UAV in the Army of the future operations

NAVAL
POSTGRADUATE
SCHOOL
MONTEREY, CALIFORNIA
THESIS
Approved for public release; distribution is unlimited.
AN EXPLORATION OF UNMANNED AERIAL VEHICLES IN THE
ARMY’S FUTURE COMBAT SYSTEMS FAMILY OF SYSTEMS
by
Charles A. Sulewski
December 2005
Thesis Advisor: Thomas Lucas
Second Reader: Jeffrey B. Schamburg

THIS PAGE INTENTIONALLY LEFT BLANK

i
REPORT DOCUMENTATION PAGE Form Approved OMB No. 0704-0188
Public reporting burden for this collection of information is estimated to average 1 hour per response, including
the time for reviewing instruction, searching existing data sources, gathering and maintaining the data needed, and
completing and reviewing the collection of information. Send comments regarding this burden estimate or any
other aspect of this collection of information, including suggestions for reducing this burden, to Washington
headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite
1204, Arlington, VA 22202-4302, and to the Office of Management and Budget, Paperwork Reduction Project
(0704-0188) Washington DC 20503.
1. AGENCY USE ONLY (Leave blank) 2. REPORT DATE
December 2005
3. REPORT TYPE AND DATES COVERED
Master’s Thesis
4. TITLE AND SUBTITLE: An Exploration of Unmanned Aerial
Vehicles in the Army’s Future Combat Systems Family of Systems
6. AUTHOR(S) Charles Sulewski
5. FUNDING NUMBERS
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
Naval Postgraduate School
Monterey, CA 93943-5000
8. PERFORMING
ORGANIZATION REPORT
NUMBER
9. SPONSORING /MONITORING AGENCY NAME(S) AND ADDRESS(ES)
N/A
10. SPONSORING/MONITORING
AGENCY REPORT NUMBER
11. SUPPLEMENTARY NOTES The views expressed in this thesis are those of the author and do not reflect the official
policy or position of the Department of Defense or the U.S. Government.
12a. DISTRIBUTION / AVAILABILITY STATEMENT
Approved for public release; distribution is unlimited
12b. DISTRIBUTION CODE
13. ABSTRACT (maximum 200 words)
Unmanned aerial vehicles (UAVs) will be a critical part of the U.S. Army’s Future Force. The
Future Force will be a highly mobile, network enabled family of systems with integrated sensors and
precision munitions. The Future Force will rely heavily on UAVs to provide eyes on the battlefield.
These eyes will trigger the deployment of precision munitions by other platforms, and possibly by
UAVs themselves.
To provide insight into how the numbers and capabilities of UAVs affect a Future Force
Combined Arms Battalion’s (CAB’s) ability to secure a Northeast Asia urban objective, a simulation
was built and analyzed. 46,440 computational experiments were conducted to assess how varying the
opposing force and the numbers, tactics, and capabilities of UAVs affects the CAB’s ability to secure
the objective with minimal losses. The primary findings, over the factors and ranges examined, are:
UAVs significantly enhance the CAB’s performance; UAV capabilities and their tactics outweigh the
number of UAVs flying; battalion level UAVs, especially when armed, are critical in the opening
phases of the battle, as they facilitate the rapid attrition of enemy High Pay-off Targets; and, at least one
company level and a platoon level UAV enhances dismounts survivability later in the battle.
15. NUMBER OF
PAGES
184
14. SUBJECT TERMS Agent-based models, MANA, Project Albert, Nearly Orthogonal Latin
Hypercube, Design of Experiment, Unmanned Aerial Vehicles, UAV, FCS, Future Force, Objective
Force
16. PRICE CODE
17. SECURITY
CLASSIFICATION OF
REPORT
Unclassified
18. SECURITY
CLASSIFICATION OF THIS
PAGE
Unclassified
19. SECURITY
CLASSIFICATION OF
ABSTRACT
Unclassified
20. LIMITATION
OF ABSTRACT
UL
NSN 7540-01-280-5500 Standard Form 298 (Rev. 2-89)
Prescribed by ANSI Std. 239-18

ii

iii
Approved for public release; distribution is unlimited.
AN EXPLORATION OF UNMANNED AERIAL VEHICLES IN THE ARMY’S
FUTURE COMBAT SYSTEMS FAMILY OF SYSTEMS
Charles A. Sulewski
Captain, United States Army
B.S., United States Military Academy, 1996
Submitted in partial fulfillment of the
requirements for the degree of
MASTER OF SCIENCE IN OPERATIONS RESEARCH
from the
NAVAL POSTGRADUATE SCHOOL
December 2005
Author: Charles A. Sulewski
Approved by: Thomas Lucas
Thesis Advisor
Jeffrey B. Schamburg
Second Reader
James N. Eagle
Chairman, Department of Operations Research

iv

v
ABSTRACT
Unmanned aerial vehicles (UAVs) will be a critical part of the U.S. Army’s
Future Force. The Future Force will be a highly mobile, network enabled family of
systems with integrated sensors and precision munitions. The Future Force will rely
heavily on UAVs to provide eyes on the battlefield. These eyes will trigger the
deployment of precision munitions by other platforms, and possibly by UAVs
themselves.
To provide insight into how the numbers and capabilities of UAVs affect a Future
Force Combined Arms Battalion’s (CAB’s) ability to secure a Northeast Asia urban
objective, a simulation was built and analyzed. 46,440 computational experiments were
conducted to assess how varying the opposing force and the numbers, tactics, and
capabilities of UAVs affects the CAB’s ability to secure the objective with minimal
losses. The primary findings, over the factors and ranges examined, are: UAVs
significantly enhance the CAB’s performance; UAV capabilities and their tactics
outweigh the number of UAVs flying; battalion level UAVs, especially when armed, are
critical in the opening phases of the battle, as they facilitate the rapid attrition of enemy
High Pay-off Targets; and, at least one company level and a platoon level UAV enhances
dismounts survivability later in the battle.

vi

vii
THESIS DISCLAIMER
The reader is cautioned that the computer programs presented in this research may
not have been exercised for all cases of interest. While every effort has been made,
within the time available, to ensure that the programs are free of computational and logic
errors, they cannot be considered validated. Any application of these programs without
additional verification is at the risk of the user.

viii

ix
TABLE OF CONTENTS
I. INTRODUCTION........................................................................................................1
A. TRANSFORMATION BACKGROUND ......................................................1
B. UAVS: THE FCS FACILATER.....................................................................5
C. PROBLEM STATEMENT .............................................................................7
D. SCOPE ..............................................................................................................9
II. NORTHEAST ASIA ATTACK SCENARIO OVERVIEW..................................11
A. FCS SYSTEMS DESCRIPTION .................................................................11
1. Unmanned Aerial Vehicle - Class I, II, and III...............................12
2. Mounted Combat System (MCS) .....................................................14
3. Infantry Carrier Vehicle (ICV) ........................................................14
4. Armed Robotic Vehicle Assault Variant (ARV-A).........................15
5. Armed Robotic Vehicle Assault Variant (ARV-L) .........................15
6. Armed Robotic Vehicle - Reconnaissance Surveillance, and
Target Acquisition Variant (ARV-RSTA).......................................15
7. Reconnaissance and Surveillance Vehicle (R&SV) ........................15
8. Non-Line-of-Sight Mortor (NLOS Mortor).....................................16
9. Non-Line-of-Sight Launch System (NLOS LS)...............................16
10. Non-Line-of-Sight Cannon (NLOS Cannon)...................................16
11. Land Warrior System........................................................................16
12. Apache Attack Helicopter AH-64D..................................................17
13. JSF (Joint Strike Fighter) .................................................................17
B. RED FORCE DESCRIPTION .....................................................................17
1. BMP-3 System....................................................................................18
2. 82 Mortor System...............................................................................18
3. Dismounted Soldier............................................................................18
a. Surface-to-Air System (SA-16)...............................................18
b. Rocket Propelled Grenade System (RPG 7)...........................18
c. Anti-Tank System (AT-7)........................................................19
d. RPK-74 ....................................................................................19
4. Armored Personnel Carrier (APC) BTR-80 ...................................19
5. T-72 Tank System ..............................................................................19
C. MODEL VIGNETTE DESCRIPTION........................................................20
III. MODEL DEVELOPMENT......................................................................................27
A. AGENT-BASED SIMULATION (ABS) OVERVIEW ..............................27
B. WHY MANA?................................................................................................29
C. MODELING METHODOLOGY.................................................................31
1. Scaling: Configure Battlefield Settings...........................................31
2. Model Unit Summary ........................................................................36
a. Players .....................................................................................36
b. Weapons ..................................................................................37
c. Aggregation.............................................................................38
3. Movement Rates.................................................................................38

x
4. Personalities........................................................................................41
5. Sense and Detect.................................................................................43
a. Ground and other Air (Non UAV) Sensors ...........................43
b. UAV Sensors ...........................................................................45
4. Communication Characteristics.......................................................50
6. Weapon Characteristics ....................................................................51
a. Kinetic Weapon Modeling ......................................................52
b. Area Fire Weapon Modeling..................................................53
7. Armor and Concealment...................................................................55
D. MODEL LIMITATIONS..............................................................................56
IV. DESIGN METHODOLOGY....................................................................................59
A. DESIGN OF EXPERIMENT........................................................................59
1. Design Factors....................................................................................60
2. Measures of Effectiveness (MOE) ....................................................62
B. TOOLS AND TECHNIQUES ......................................................................63
1. DOE Software Tools ..........................................................................63
a. Spreadsheet Modeling with Excel ..........................................64
b. XML.........................................................................................64
c. Tiller© .....................................................................................64
d. Ruby Code and Scripting........................................................65
2. Analysis Software Tools (JMP Statistical Discovery Software
TM
)........................................................................................................65
3. Analysis Techniques ..........................................................................66
a. Graphical Analysis..................................................................66
b. Classification and Regression Trees (CART) ........................66
c. Multiple Regression ................................................................67
V. DATA ANALYSIS.....................................................................................................69
A. DATA COMPILATION................................................................................69
B. INITIAL OBSERVATIONS.........................................................................70
C. CLOSING OBSERVATIONS RELATED TO THESIS QUESTIONS ...77
1. Battlefield Time Hacks ......................................................................77
2. The Early Fight ..................................................................................79
a. How many Platoon, Company, and Battalion level UAVs
are needed for the FCS to secure the urban environment? ..81
b. How will armed battalion level UAVs enhance the FCS’s
ability to secure the urban environment?...............................93
e. Is it better to arm Warrior UAVs with Hellfire missiles at
the CAB level, or to use APKWS 2.75 inch guided rockets
with M151 HE warheads attached to the CL III UAVs? ......97
VI. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE STUDY .........99
A. SUMMARY OF CONCLUSIONS AND GAINED INSIGHT ..................99
1. Data Analysis Conclusions ................................................................99
2. Modeling and DOE Methodology Findings...................................101
B. RECOMMENDATIONS FOR FUTURE STUDY ...................................103

xi
APPENDIX A. MANA SPREADSHEET MODELING..................................................105
A. SCALING: CONFIGURE BATTLEFIELD SETTINGS.......................106
B. MODEL UNIT SUMMARY.......................................................................107
C. MOVEMENT RATES.................................................................................108
D. SENSE AND DETECT................................................................................109
E. PERSONALITIES.......................................................................................111
F. COMMUNICATION CHARACTERISTICS...........................................116
G. WEAPON CHARACTERISTICS..............................................................118
H. ARMOR AND CONCEALMENT .............................................................123
APPENDIX B. DOE MODELING....................................................................................125
A. DOE SPREADSHEET MODELING.........................................................125
B. TILLER ........................................................................................................129
C. RUBY SCRIPTING.....................................................................................130
APPENDIX C. ADDITIONAL DATA ANALYSIS ........................................................133
A. INITIAL OBSERVATIONS.......................................................................134
B. THE EARLY FIGHT ..................................................................................138
C. INTERACTIONS.........................................................................................147
LIST OF REFERENCES....................................................................................................149
INITIAL DISTRIBUTION LIST.......................................................................................155

xii

xiii
LIST OF FIGURES
Figure 1. Unit of Action Tree Diagram.............................................................................5
Figure 2. Future Combat Systems: Platforms ...............................................................12
Figure 3. NEA 50.2 Area of Operation Map...................................................................21
Figure 4. Combined Arms Battalion Tree Diagram........................................................22
Figure 5. NEA 50.2 MANA Screenshot..........................................................................32
Figure 6. Adjusted Average Sensor Value......................................................................44
Figure 7. UAV Sensor Probability of Detection Graph ..................................................47
Figure 8. Modeled UAV Sensor Probability of Detection Graphs..................................49
Figure 9. Carleton Function.............................................................................................54
Figure 10. Regression Tree, with MOE: Proportion of HPT Killed .................................71
Figure 11. Regression Tree, with MOE: Proportion of Dismounts Survived..................72
Figure 12. Histograms of Initial Analysis with Robust DOE ...........................................73
Figure 13. Tests of Main Effects (Stepwise Linear Regression Model Fit)......................75
Figure 14. Graphical Analysis: Battlefield Time Hack without robust DOE ..................77
Figure 15. Histograms at 450 seconds (7.5 minutes) ........................................................78
Figure 16. Histograms at 900 seconds (15 minutes) .........................................................78
Figure 17. t-Test Results Between a 15-minute and 2-hour Battle ...................................80
Figure 18. Scatterplot Matrix (Positive Correlation Between HPTs and Dismounts) ......82
Figure 19. Regression Model (Proportion of HPTs Killed at 450 seconds)......................84
Figure 20. Regression Tree (Proportion of HPTs Killed at 450 seconds).........................86
Figure 21. Regression Model (Proportion of HPTs Killed at 900 seconds)......................87
Figure 22. Regression Tree (Proportion of HPTs Killed at 900 seconds).........................88
Figure 23. Regression Model (Proportion of Dismounts Survived at 900 seconds).........91
Figure 24. Regression Tree (Proportion of Dismounts Survived at 900 seconds)............91
Figure 25. Regression Model (Interaction Measured by HPTs) .......................................94
Figure 26. Interaction Plot of CL III UAVs Armed with Munitions ................................95
Figure 27. Additional Interaction Plot...............................................................................97
Figure 28. Ruby PatchExcursion.rb Code ......................................................................130
Figure 29. Ruby Scripting Command..............................................................................130
Figure 30. Multiple Regression Output for Initial Analysis of Robust DOE..................134
Figure 31. Regression Model (Proportion of HPTs Killed at 450 seconds)....................138
Figure 32. Regression Model (Proportion of HPTs Killed at 900 seconds)....................141
Figure 33. Regression Model (Proportion of Dismounts Survived at 900 seconds).......144
Figure 34. Determine Interactions Model, MOE: Proportion of HPTs Killed................147

xiv

xv
LIST OF TABLES
Table 1. NEA 50.2 Team Disposition............................................................................23
Table 2. Red Force Disposition......................................................................................24
Table 3. Scaling Equations.............................................................................................34
Table 4. Edit Terrain Properties.....................................................................................35
Table 5. Real World Basic Movement Rates ................................................................38
Table 6. MANA Movement Speeds...............................................................................40
Table 7. Numerical Sensor Value ..................................................................................44
Table 8. FCS UAV Sensor Type....................................................................................46
Table 9. FCS UAV Sensor Type Definitions ................................................................46
Table 10. Modeled Communication Types......................................................................50
Table 11. Weapon Characteristics ...................................................................................51
Table 12. Modeled P(Kill) for Area Fire Weapons using the Carleton Function............54
Table 13. Factor and Level Description for DOE............................................................61
Table 14. Significant Factors (Proportion of HPTs Killed at 450 seconds) ....................85
Table 15. Significant Factors (Proportion of HPTs Killed at 900 seconds) ....................88
Table 16. UAV Estimates (Proportion of HPTs Killed at 450 and 900 seconds)............89
Table 17. Significant Factors (Proportion of Dismounts Survived at 900 seconds)........92
Table 18. UAV Estimates (Proportion Dismounts Survived at 900 seconds) .................93
Table 19. Table of Instruction to Modify a Skeleton study.xml File .............................131

xvi

xvii
LIST OF ACRONYMS AND ABBREVIATIONS
ABM Agent-Based Models
ABS Agent-Based Simulations
AoA Analysis of Alternatives
APKWS Armed Precision Kill Weapon System
ARV-A Armed Robotic Vehicle - Assault Variant
ARV-L Armed Robotic Vehicle - Light
ARV-RSTA Armed Robotic Vehicle - Reconnaissance, Surveillance, and Target
Acquisition Vehicle
BLOS Beyond-Line-of-Sight
CA Combined Arms
COA Course of Action
CPU Central Processing Unit
DA Department of the Army
DoD Department of Defense
FCS Future Combat Systems
GUI Graphical User Interface
GWOT Global War on Terrorism
HPT High Pay-off Target
HVT High Value Target
ICV Infantry Carrier Vehicle
LOS Line-of-Sight
MASINT Measurement and Signature Intelligence
MCS Mounted Combat System

xviii
METT-T Mission, Enemy, Troops, Terrain, and Time
MOUT Military Operations in Urbanized Terrain
NAI Named Areas of Interest
NEA Northeast Asia
NLOS Non-Line-of-Sight
OBJ Objective
RSTA Reconnaissance, Surveillance, Target Acquisition
R&SV Reconnaissance and Surveillance Vehicle
SIGINT Signals Intelligence
TA Target Acquisition
TOS Time on Station
TRAC Training and Doctrine Command Analysis Center
TRADOC Training and Doctrine Command
UA Unit of Action
UAV Unmanned Air Vehicle
UE Unit of Employment
UGV Unmanned Ground Vehicle
UH Utility Helicopter
UAMBL Unit of Action Maneuver Battle Lab
VIC Vector-in-Commander
VTOL Vertical Take-off and Landing
WSMR White Sands Missile Range
XML Extensible Markup Language

xix
ACKNOWLEDGMENTS
I would like to begin by thanking the Lord for watching over me, and providing
me with the wisdom, courage, and perseverance to not only live, but to enjoy it as well.
I would also like to thank my great thesis team. I thank Professor Thomas Lucas
for his guidance and extensive statistics, design of experiments, and write up guidance.
The journey and final product would not have been the same without your mentorship. I
would also like to thank LTC Jeffrey B. Schamburg for his military expertise and keeping
me focused and headed in the correct direction within the parameters of my work. Two
additional faculty members here at NPS contributed to overcoming obstacles along the
way in this analysis, and to these two individuals, Professors Paul and Susan Sanchez, I
say thank you. I would also like to acknowledge the talented staff at Project Albert, and
especially Dr. Gary Horne, for the opportunity to ask questions, learn insights, and find
surprises.
Finally, I would like to thank my family. Stacey, your support, love, and charm is
unbounded. Without you, my life would be incomplete. Thanks for taking extra good
care of the children during this time, and for always keeping me smiling. Thanks also to
Jessica and Anthony, the apples for each of my eyes, who always love Daddy for just
being Dad.

xx

xxi
EXECUTIVE SUMMARY
Unmanned aerial vehicles (UAVs) are playing an increasingly important role in
the Global War on Terrorism (GWOT). These roles are part of the United States
Department of Defense’s (DoD’s) greatest transformation of the armed forces since
World War II. This transformation is a holistic approach to modernize our forces’
equipment, methods, and tactics to ensure success for future conflicts.
The Army’s Future Force (formerly “Objective Force”) focuses on a lighter, more
agile force, permitting the troops to move quickly in order to seize the initiative and
finish decisively. Since conventional systems are inadequate to facilitate all of the goals
of the Army’s transformation, the Army is developing the core building block of the
Future Force—known as the Future Combat Systems (FCS) Family of Systems (FoS).
The FCS is a networked “system of systems” comprised of 18 individual system
platforms, the network, and the soldier. Unmanned Aerial Vehicles are among these
platforms.
This area of research is significant because the Army’s FCS relies heavily on
unmanned air vehicles (UAVs) and unmanned ground vehicles (UGVs) to provide eyes
on the battlefield. These eyes will trigger the deployment of precision munitions by fixed
wing Close Air Support (CAS), Beyond-Line-of-Sight (BLOS), Non-Line-of-Sight
(NLOS) weapon platforms, and possibly by UAVs themselves. The FCS UAVs are the
hunters in the sky for tomorrow’s battles.
FCS UAVs are currently broken down into classes I, II, III, and IV(a, b). This
thesis only focuses on classes I, II, and III. Class I UAVs within the FCS provide
Reconnaissance, Surveillance, and Target Acquisition (RSTA) capabilities at the platoon
level. Class II UAVs provide RSTA capabilities and target designation at the platoon and
company level. Class III UAVs provide RSTA capability, target designation,
communication relay, and mine detection at the combined arms battalion (CAB) level.
Both the CL IV(a and b) provide similar capabilities at the Unit of Action (UA) level of
the battlefield, and are outside the scope of this thesis.

xxii
This thesis applies a low-resolution model to examine the U.S Army Training and
Doctrine Command’s (TRADOC’s) tasked analysis questions regarding the effectiveness
of the FCS within an urban environment. The objective is to identify a preferred
numerical mix of class I, II, and III RSTA, and precision guided armed UAVs needed in
a combined arms battalion of the Army’s Future Force to identify, engage, and destroy
enemy targets in a specified MOUT environment.
This analysis focuses on an UA Combined Arms Battalion (CAB) attacking in a
Northeast Asia (NEA) area of operation (Refer to Figure ES1). The scenario and Blue
Force structure for the analysis is adopted from the Training and Doctrine Analysis
Center—White Sands Missile Range (TRAC-WSMR) CASTFOREM modeled vignette
NEA 50.2. The Red Force Order-of-Battle, modified slightly, represents a plausible
stronger threat. This ensures that the blue CAB does not gain complete victory with
every simulation, thus facilitating the search for outliers and surprise.
FIGURE ES1 Northeast Asia Area of Operation

xxiii
The intent is to replicate the CASTFOREM vignette as closely as possible using
an agent-based model (Map Aware Non-uniform Automata, or MANA) while exploring
future aspects of UAVs. (Note: the original CASTFOREM vignette does not include the
use of armed UAVs). This thesis studies the effectiveness of the FCS while varying the
number, capabilities, and tactics of UAVs and considering the use of armed CL III
battalion level UAVs. The primary goal is to identify a number of CL I, II, and III
UAVs, for this specific MOUT region, where UAVs enable the effective use of precision
munitions—thus enhancing the UA’s ability to fight. The analysis focuses on a critical 2-
hour window of operation where the CAB assaults onto the urban objective.
The questions scoping this thesis are as follows:
• How many Platoon, Company, and Battalion level UAVs are needed for the
FCS to secure the urban environment?
• How will armed battalion level UAVs enhance the FCS’s ability to secure the
urban environment?
• Is it better to arm Warrior UAVs with Hellfire missiles at the CAB level, or to
use APKWS 2.75 inch guided rockets with M151 HE warheads attached to
the CL III UAVs?
Applying a Nearly Orthogonal Latin Hypercube design of experiment with 258
design points provided a multitude of data. Initial observations of the data portrayed
three things:
• The enemy and terrain (two elements of mission, enemy, troops, terrain, and
time or METT-T) provide greater significance to the mission outcome than
the number and capability of UAVs deployed within the CAB at any level.
• The tactical employment, and capabilities of each UAV, provides greater
significance to the CAB’s mission accomplishment than does the actual
numbers of UAVs at each level.
• The joined platform capabilities within the FCS is so robust, that eliminating
an entire platform category, such as all the UAVs from the battle space, has

xxiv
little effect on the CAB’s ability to still maintain 95% of its Dismount
population while destroying 90% of the enemy HPTs.
The findings listed above were surprising to the author. As such, the author
evaluated several outliers portraying greater detriment to the Blue Force. These outliers
called for a slight modification to the original experimental design. Modifications
stabilized the varying environmental and enemy factors at levels providing the greatest
detriment to the Blue Force.
Upon applying the modified experimental design, the final analysis showed that
within a critical 2-hour window of the CAB’s assault on the urban terrain:
• 11 or more battalion level UAVs provide the FCS’s ability to act quickly and
decisively by bringing the biggest punch against the enemy as measured by
both the proportion of HPTs killed and the proportion of Blue Dismounts
Survived.
• The model portrays the CAB’s increased lethality against the HPTs, while
minimizing Blue Dismount deaths when adding precision munitions to CAB
UAV assets.
• The CAB needs the CL III UAV for the deep fight and preparation of the
battlefield by destroying the HPTs.
• Once the battlefield is prepared and the Dismounts arrive, then the CL I UAVs
are more significant because they provide the local situational awareness (over
the next hill) to these Dismounts.
• The APKWS missiles tend to provide more benefit to the mission
immediately upon the start of the battle.
• As the battle moves on, Hellfire missiles become more significant as
measured by the proportion of HPTs killed at 900 seconds.
• Hellfire missiles also seem to provide more application as measured by the
proportion of Blue Dismounts survived at 900 seconds. However, at 900
seconds there is already a large loss to the Red Force.

xxv
• Each tactical team benefits when deployed with between one and three
platoon level UAVs. The benefit of adding one platoon level UAV per team
increases the overall CAB survival proportion of Blue Dismounts by almost
one percent.
• Need at least one CL II UAV per tactical team. The exact number of CL II
UAVs is still unknown from this thesis.
• Lower class UAVs provide the eyes “over the next hill” for Dismounts.
Operators need to balance the tactical flight pattern in order to cover as much
ground as possible while minimally loitering over detected targets.

xxvi

1
I. INTRODUCTION
"CEDAT FORTUNA PERITIS"
(Skill is Better Than Luck)
US Army Field Artillery School
A. TRANSFORMATION BACKGROUND
Unmanned aerial vehicles (UAVs) are playing an increasingly important role in
the Global War on Terrorism (GWOT). These roles are part of the United States
Department of Defense’s (DoD’s) greatest transformation of the armed forces since
World War II. This transformation is a holistic approach to modernize our forces’
equipment, methods, and tactics to ensure success for future conflicts. Dovetailing
tomorrow’s technology with innovative tactics will enable the US Army to transform into
the next Future or Objective Force “in order to quickly and effectively respond to
situations across a full spectrum of contingencies.”1
The United States Army’s adaptation of the new force structure intends to meet
the needs of the next millennium. The vision for accomplishing this, as defined by the
senior Army leadership, is to invest in a “leap ahead” capability that will be the heart of
mounted close combat for the Army after next.2 There exists the need to blend the
capabilities of several battlefield-operating platforms, into a common System of Systems
(SoS), that will re-engineer the Army’s ability to quickly and effectively respond to
situations across a full spectrum of contingencies. Tomorrow’s threats pose complex
asymmetric situations which demands our response with an Army capable of deploying a
combat-capable brigade anywhere in the world within 96 hours, a full division in 120
hours, and five divisions on the ground within 30 days.3 Rising technology, integrated
with evolutionary tactics, will propel the US Army’s transformation in its development of
the Future Force to meet these needs.
1 Examining the Army’s Future Warrior, Force-on-Force Simulation of Candidate Technologies, Rand
Arroyo Center, 2004, p. xi.
2 Global Security.org, Future Combat Systems – Background, Retrieved 28 June 2005 from the World
Wide Web at http://www.globalsecurity.org/military/systems/ground/fcs-back.htm
3 Global Security.org, Future Combat Systems, Retrieved 1 August 2005 from the World Wide Web at
http://www.globalsecurity.org/military/systems/ground/fcs.htm

2
The Army’s Future Force (formerly “Objective Force”) focuses on a lighter, more
agile force, permitting the troops to move quickly and versatile in order to seize the
initiative and finish decisively.4 Since conventional systems are inadequate to facilitate
all of the goals of the Army’s transformation, the Army is developing the core building
block of the Future Force—known as the Future Combat Systems (FCS) Family of
Systems (FoS). The FCS is a networked “system of systems” comprised of 18 individual
system platforms, the network, and the soldier.5 These platforms are designed to operate
in concert with each other using greater quantities of precision munitions, with minimal
soldier manning. In addition, advanced communications and technologies will link
soldiers with both manned and unmanned, ground and air, platforms and sensors.
The FCS has currently progressed into the System Development and
Demonstration (SDD) Phase of its program.6 It is a living entity, with almost monthly
modifications, as new information regarding tomorrow’s technological needs unfold. As
such, it will be interesting for the reader to note the similarities and differences describing
the FCS now and from a thesis written during the Concept and Technology Development
(CTD) Phase by CPT Joseph Lindquist, June 2004, addressing degraded communications
in the Army’s Future force. Lindquist’s references provided a stepping-stone for
launching this research. Some of the source names are the same, but the publishing dates
and source descriptions have changed. In addition, Lindquist’s thesis served as a
template to follow in format, as this thesis contains similar aspects with regard to the FCS
and agent-based modeling.
As Lindquist pointed out, there exist two critical components to transform the
vision of the Future Force into a prevailing reality. The first is the requirement of high
situational understanding of the battlefield and the second is decisive tactical combat.7
Situational understanding of both friendly and enemy forces permits the commander to
4 Boeing, Future Combat Systems, Retrieved 15 November 2005 from the World Wide Web at
http://www.boeing.com/defense-space/ic/fcs/bia/about.html
5 Boeing, Future Combat Systems, Retrieved 5 August 2005 from the World Wide Web at
6 US Army Material Systems Analysis Activity (US AMSAA), FCS UA Design Concept Baseline
Description (UA-001-01-050124), 3 March 2005.
7 Naval Postgraduate School Thesis, An Analysis of Degraded Communications in the Army’s Future
Force using Agent Based Modeling, Joseph M. Lindquist, June 2004, pp. 2-3.

3
enter the fight on his conditions and seize the initiative. Decisive tactical combat refers
to sophisticated capabilities enabling mobility and long-range precision fires. This
permits the commander to safely engage and attrite the enemy at a greater distance.8 For
purposes of this research, the former focuses more on the Command, Control,
Communications, Computer, Intelligence, Surveillance, and Reconnaissance (C4ISR) and
Reconnaissance, Surveillance, Target Acquisition (RSTA) of the battlefield. One
excellent method to gain C4ISR and to perform RSTA for the FCS, while eliminating
multiple inherent flight risks to humans, is with unmanned aerial vehicles (UAVs).
Before proceeding, it is important to identify FCS features. The Army is
currently developing an Operational and Organizational plan to reorganize the current
fighting force and field this revolutionary "leap ahead" system as the centerpiece of the
Army's ground combat force between FY2015 and FY2020.9
The FCS is the catalyst for achieving the Army's transformation vision of
fielding a Future Force by the end of this decade. The Future Force will
operate as part of a joint, combined, and/or interagency team, it will be
capable of conducting rapid and decisive offensive, defensive, stability
and support operations, and be able to transition among any of these
missions without a loss of momentum. It will be lethal and survivable for
warfighting and force protection; responsive and deployable for rapid
mission tailoring and the projection required for crisis response; versatile
and agile for success across the full spectrum of operations; and
sustainable for extended regional engagement and sustained land combat.
The FCS will network fires and maneuver in direct combat, deliver direct
and indirect fires, perform intelligence, surveillance, and reconnaissance
functions, and transport Soldiers and material as the means to tactical
success.10
Over time, the FCS may actually replace the current inventory of ‘heavy’
vehicles. Vehicles such as the Abrams tank, Bradley Fighting Vehicle, and Paladin
howitzer may fade away, as the new family of manned and unmanned, ground and aerial
vehicles enter the battlefield. The ground vehicles will weigh tremendously less, each
8 US Army Training and Doctrine Command, The Army Future Force: Decisive 21st
Century
Landpower Strategically Responsive Full Spectrum Dominate. pp. 4-5.
9 Global Security.org, Future Combat Systems – Background, Retrieved 28 June 2005 from the World
Wide Web at http://www.globalsecurity.org/military/systems/ground/fcs-back.htm
10 Global Security.org, Future Combat Systems, Retrieved 3 August 2005 from the World Wide Web
at http://www.globalsecurity.org/military/systems/ground/fcs.htm

4
with the requirement of weighing less than 20 tons. Two of these smaller and lighter
vehicles must fit inside one C-130 or C-141 cargo aircraft. Though lighter, the
capabilities of each platform will increase, blending current single capabilities among
multiple platforms. The combined capabilities include Line-of-Sight (LOS) / Beyond-
Line-Of-Sight (BLOS) / Non-Line-of-Sight (NLOS) precision munitions weapon
systems, robotic C4ISR platforms, soldier Land Warrior platforms, and support
platforms. Hence, the FCS Family of Systems facilitates the response needs to the more
complex and asymmetric future fronts, with the ability to deploy a brigade size element
any where in the world, within the 96 hour time limit.11
The FCS is broken down into smaller elements; each called a Unit of Action (UA)
(Refer to Figure 1). The UA will replace a brigade size element with modularity and
agility. Within one UA, there exist three Combined Arms Battalions (CAB) comprised
of a Headquarters and Headquarters Company, one Brigade Intelligence Company, one
Communications Battalion, one NLOS Battalion, and a Forward Support Battalion.
Within a CAB, there is a Headquarters Company, two to four Infantry Companies, two to
four Mounted Combat System (MCS) companies, a Recon Troop, a Mortor Battery, and a
Reconnaissance Surveillance Target Acquisition (RSTA) Squadron. These smaller
organizations blend into smaller teams, allowing for a diverse tailorable force that moves
with speed and versatility, allowing teams of troops to conduct a variety of missions on
the future battlefield, including Military Operations in Urban Terrain (MOUT).
11 Global Security.org Future Combat Systems, Retrieved 3 August 2005 from the World Wide Web at
http://www.globalsecurity.org/military/systems/ground/fcs.htm

5
Figure 1. Unit of Action Tree Diagram12
"PRIMUS AUT NULLUS"
(First, or Not at All)
1st
Field Artillery
B. UAVS: THE FCS FACILATER
This area of research is significant because the Army’s FCS relies heavily on
unmanned air vehicles (UAVs) and unmanned ground vehicles (UGVs) to provide eyes
on the battlefield. These eyes will trigger the deployment of precision munitions by fixed
wing Close Air Support (CAS); Beyond-Line-of-Sight (BLOS) and Non-Line-of-Sight
(NLOS) weapon systems; and possibly by UAVs themselves. As of September 2004,
“some twenty types of coalition [unmanned aerial vehicles], large and small, have flown
12 Unit of Action Maneuver Battle Lab, Change 3, to TRADOC Pamphlet 525-3-90 O&O, The United
States Future Force Operational and Organizational Plan Maneuver Unit of Action (DRAFT), 30 July
2004, Fort Knox, KY 40121, section 3.2, p.18.

6
over 100,000 total flight hours in support of Operation Enduring Freedom (OEF) and
Operation Iraqi Freedom (OIF).”13 The FCS UAVs are the hunters in the sky for
tomorrow’s battles. In addition to triggering the deployment of precision munitions, they
will provide situational awareness of the engagement area, and will assist in all
communication aspects throughout the combat maneuver area and theater area of
operations.
FCS UAVs are currently broken down into classes I, II, III, and IV(a, b). Class I
UAVs within the FCS provide RSTA capabilities at the platoon level. Class II UAVs
provide RSTA capabilities and target designation at the platoon and company level.
Class III UAVs provide RSTA capability, target designation, communication relay, and
mine detection at the combined arms battalion (CAB) level. Class IVa UAVs provide
RSTA capability, target designation, communications relay, mine detection at the UA
level and supports manned/unmanned teaming operations with manned aviation. Class
IVb UAVs provide RSTA capability, target designation, communications relay, long
endurance persistent staring, and wide area surveillance for the UA.14
Currently the US Air Force is using and testing Hellfire packed Predator UAVs.
The Armed Forces is currently flying these UAVs in Afghanistan and Iraq, but little
analysis explains the full effectiveness of armed UAVs on the battlefield.15 In addition,
the Army plans to procure 11 Warrior systems, a new Extended Range Multi Purpose
(ERMP) UAV. Each system consists of 12 aircraft, five ground control stations and other
support equipment. The Warrior begins operational deployment in 2009.16 The once
reconnaissance only role is now shared with strike, force protection, and signals
collection, and, in doing so, has helped to reduce the complexity and time lag in the
sensor-to-shooter chain for a broad range of mission capabilities.17
13 Stephen Cambone, Kenneth Krieg, Peter Pace, Linton Wells, Unmanned Aircraft System Roadmap
2005-2015, Department of Defense, 4 August 2005, p. 1.
15 United States Department of Defense, Predator UAV Proves its Worth, Retrieved 10 August 2005
from the World Wide Web at http://usmilitary.about.com/cs/afweapons/a/preditor.htm
16 Greg Grant, “Army picks General Atomics for ERMP program,” Army Times, 8 Aug 2005,
Retrieved 11 October 2005 from the World Wide Web at http://www.armytimes.com/story.php?f=1-
292925-1021240.php
17 Cambone, p. 1.

7
The Warrior contains flexible payloads, with equal lethality to the Air Force's
Predator. The Army accelerated the [Extended Range Multi-Purpose UAV] ERMP
program after US commanders in Iraq “clamored for a drone that could carry Hellfire
missiles and perform the more traditional intelligence, surveillance, and reconnaissance
mission.”18 Though heavy, the Warrior can carry up to four Hellfire missiles. For lighter
payload options, an Advanced Precision Kill Weapon System (APKWS) of guided
rockets may also prove useful if attached to the current planned CL III UAV category.
The Army accelerated the ERMP with precision munitions. Contradictory, FCS planners
do not currently consider Hellfire, APKWS, or any other guided munitions, as part of any
FCS UAV. Even though the ERMP UAV posses a higher-class level then the current
planned CAB Class III UAV, planners must consider “what if questions?” What effect
occurs on the battlefield if the CAB gains control of UAV assets with Hellfire or lighter
APKWS guided rocket payloads? For this thesis, Warrior and Class III UAVs will be
similar for modeling purposes.
"CELERITAS ET ACCURATIO"
(Speed and Accuracy)
Third Field Artillery Regiment
C. PROBLEM STATEMENT
The underlying questions of this research ask how many UAVs are needed, and
how will armed UAVs affect mission performance? “Combatant Commanders are
requesting [UAVs] in even greater numbers. Our challenge is the rapid and coordinated
integration of this technology to support the joint fight.”19 This research assumes that
UAMBL’s classification and capabilities of FCS UAVs is correct, with the exception of
possibly adding precision guided missiles to the CL III UAV. The UA planning
numbers, as shown in Figure 1, per UAV class is part of the FCS MSB Update, dated 18
May 2005.20 However, in speaking with experts from UAMBL, AMSAA, and TRAC,
18 Greg Grant.
19 Department of Defense, Memorandum for Secretaries of the Military Departments, Subject:
Unmanned Aircraft Systems (UAS) Roadmap, 2005 -2015, 4 August 2005.
20 Unit of Action Maneuver Battle Lab.

8
they all agree that more research similar to this needs to be completed between now and
2015. This additional research will help validate, field, and quantify the actual number of
UAVs needed to facilitate a 24-hour operation in different environments. Continued
research will also balance the needs of the future force along with the logistics necessary
to create and support it. Advanced phases of the FCS program prompted AMSAA to
change the name of the platform description manual from the Army Future Combat
Systems Unit of Action Systems Book Version 3.0, 22 May 2003, to the FCS UA Design
Concept Baseline Description (UA-001-01-050124). Upon starting this thesis in June
2005, the 9 May 2005 publication was the most up to date manual, which supersedes
previous manuals dated 3 March 2005, and even 4 May 2005, which portrays constant
updates due to advanced breaks in research.
In addition, Jane’s Information Group, Inc. published a listing of the 59 US
made UAVs, and 114 known foreign made UAVs.21 Each year these numbers and the
capabilities of each also increase. Traditionally, surveillance UAV military users have
tended to regard them as semi-expendable battlefield assets. However, the continued
development of more sophisticated UAVs, coupled with the platform design of the FCS,
brings the need directly back for continued research.
With the collection of multiple programs, increasing UAV technologies, and
future threats, a specific need exists to identify the number of UAVs, by class type and
capabilities, needed to perform a variety of missions in different environments.22 The
Director, Headquarters United States Army Training and Doctrine Command, Futures
Center, tasked the US Army Training and Doctrine Command (TRADOC), Analysis
Center, to conduct an operational analysis of precision munitions deployed as part of the
FCS FoS.23 In an effort to assist in this essential task, this research focuses on UAV
related key battlefield and targeting factors that necessitate precision delivery of effects,
21 Kenneth Munson, Jane’s Unmanned Aerial Vehicles and Targets: Issue Twenty-Three,
(Alexandria: Jane’s Information Group Inc, 2004), p. 20.
22 Interview with Thomas Lancarich, Senior Operations Research Analyst, Chief, Scenario
Integration & Methodology Development Division, TRADOC Analysis Center-White Sands Missile
Range, New Mexico, 25 June 2005.
23 Headquarters United States Army Training and Doctrine Command (Director Futures Center),
Memorandum for U.S. Army TRADOC Analysis Center, Fort Leavenworth, KS, 9 July 2004.

9
and what acquisition force adjustments are relevant to the FCS-equipped UA and UEx
organizations for the delivery of precision munitions.
This thesis applies a low-resolution model to examine TRADOC’s tasked analysis
questions regarding the effectiveness of the FCS within an urban environment. The
objective is to identify a preferred numerical mix of class I, II, and III RSTA and
precision guided rocket packed UAVs needed in a combined arms battalion of the
Army’s Future Force to identify, engage, and destroy enemy targets in a specified MOUT
environment. This analysis output should not replace higher resolution physics-based
modeling techniques. It does however; applaud the lower resolution data process for its
delivery of quick results and analysis, while using limited resources, and possible
uncovering hidden surprises.
"NOLI ME TANGERE"
(Do Not Touch Me)
1st
Battalion (ABN), 321st
Field Artillery Regiment
The U.S. Army's Only 155mm Airborne Artillery
D. SCOPE
There exist countless questions regarding how to integrate UAVs into the Future
Force. Mission, Enemy, Troops, Terrain, and Time (METT-T) has always scoped the
battlefield. Friendly and enemy Order-of-Battle also play a key component on how to
utilize UAVs. However, this thesis will only focus, and provide insight, on one Military
Operations in Urban Terrain (MOUT) scenario.
This analysis focuses on an UA Combined Arms Battalion (CAB) attacking in a
North East Asia area of operation. The scenario and Blue Force structure for the analysis
is adopted from the Training and Doctrine Analysis Center—White Sands Missile Range
(TRAC-WSMR) CASTFOREM modeled vignette started in the Spring of 2005.24 The
Red Force Order-of-Battle, modified slightly, represents a plausible stronger threat. This
24 Thomas Lancarich, Senior Operations Research Analyst, Chief, Scenario Integration &
Methodology Development Division, TRADOC Analysis Center-White Sands Missile Range, New Mexico
North East Asia Vignettes (Vignette NEA50.2) FCS BN(-) attack vs enemy stronghold of city, May 2005.

10
ensures that the blue CAB does not gain complete victory with every simulation, thus
facilitating the search for outliers and surprise.
The intent is to replicate this vignette as closely as possible using an agent-based
model while exploring future aspects of UAVs. However, the original CASTFOREM
vignette does not include the use of armed UAVs. Lastly, there is no complete analysis
regarding data output from the CASTFOREM vignette. Therefore, this thesis will not
compare and contrast the methodology, design of experiments, or output between both
models, but will study the effectiveness of the FCS while varying the number of UAVs
and considering the use of armed CL III battalion level UAVs. The primary goal is to
identify a number of CL I, II, and III UAVs, for this specific MOUT region, where UAVs
enable the effective use of precision munitions—thus enhancing the UA’s ability to fight.
To complete this thesis within the allotted time, with limited reasonable
exploration, the following research questions scope the direction of this research:
• How many Platoon, Company, and Battalion level UAVs are needed for the
FCS to secure the urban environment?
• How will armed battalion level UAVs enhance the FCS’s ability to secure the
urban environment?
• Is it better to arm Warrior UAVs with Hellfire missiles at the CAB level, or to
use APKWS 2.75 inch guided rockets with M151 HE warheads attached to
the CL III UAVs?

11
II. NORTHEAST ASIA ATTACK SCENARIO OVERVIEW
"FESTINA LENTE"
(Make Hast Slowly)
42nd Field Artillery Regiment
The first portion of this chapter outlines the players within the scenario, while the
second portion of this chapter outlines the actual scenario studied and then modeled
within this research. The players are broken down into Blue and Red Forces. The Blue
force is comprised of a Combined Arms Battalion with Unit of Action assets as part of
the Future Combat Systems. The Red Force is the enemy. Their detailed description
follows later in this chapter. There is no Neutral (Yellow) Force modeled.
A. FCS SYSTEMS DESCRIPTION
The FCS is a networked “system of systems” comprised of 18 individual system
platforms, the network, and the soldier.25 These platforms operate in concert with each
other using greater quantities of precision munitions, with minimal soldier staffing. In
order to reduce the logistics burden on the FCS equipped UA, all FCS manned platforms
have a common core chassis, and a common set of base capabilities. Each platform will
weigh less then 20 tons in order to fly two FCS platforms inside of one C-130 cargo
aircraft. To facilitate weight requirements, counter ballistic projection, and add-on armor
capabilities substitute the full-up armor protection observed on today’s manned
platforms.26 In addition, advanced technologies will link soldiers to any combination of
manned, unmanned, air, and ground platforms or sensors.
25 Boeing, Future Combat Systems, Retrieved 5 August 2005 from the World Wide Web at
26 US Army Material Systems Analysis Activity (US AMSAA), Army Future Combat Systems Unit
of Action Systems Book Version 3.0, 22 May 2003.

12
Figure 2. Future Combat Systems: Platforms 27
The following paragraphs describe each FCS system modeled within this vignette.
Each FCS description is a direct excerpt from one of three sources. Paragraph 1 comes
directly from one of the Unit of Action Maneuver Battle Lab’s Operational Requirements
Document.28 Paragraphs 2 through 11 are direct excerpts from the FCS UA Design
Concept Baseline Description.29 Paragraphs 12 and 13 arrive directly from the World
Wide Web.
1. Unmanned Aerial Vehicle - Class I, II, and III
The Class III Unmanned Aerial Vehicle (CL III UAV) is a multifunction aerial
system capable of providing reconnaissance, security/early warning, target acquisition,
and designation for precision fires, throughout the battalion area of influence by remotely
over-watching and reporting changes in key terrain, avenues of approach and danger
27 Global Security.org Future Combat Systems, Retrieved 17 November 2005 from the World Wide
Web at http://www.globalsecurity.org/military/systems/ground/images/fcs-2005armymodernization.jpg
28 Unit of Action Maneuver Battle Lab, Change 1, to Joint Requirements Oversight Council (JROC) –
approved Future Combat Systems (FCS) Operational Requirements Document (ORD), June 2004, Fort
Knox, KY 40121, Annex E.

13
areas in open, rolling, restrictive, and urban areas. The aerial system will provide
information from operating altitude and standoff range both day/night and in adverse
weather. The aerial system should be capable of communication relay, detecting mines,
performing CBRN detection, and performing meteorological survey for the NLOS
battalion to deliver precision fires.
The UAV at the Battalion level must provide multiple capabilities, to include:
Reconnaissance and security/early warning capability for the UA during day and night;
Remotely over-watch and report changes in key terrain, avenues of approach and danger
areas in open and restrictive terrain, and urban areas; Perform target acquisition and
designation for the UA; Act as a communications (wide band) relay; Perform target area
meteorological survey; Does not require an airfield; Support CAB by performing R&S on
a minimum of three routes or nine NAIs; Enable NLOS targeting and fires.
The CL II UAV is a multifunctional aerial system capable of providing
reconnaissance, security/early warning, target acquisition, and designation for the
Infantry Company and MCS Platoon within the UA in support of LOS/BLOS and NLOS
cooperative engagements. The CL II UAV will be a vehicle-mounted system that
provides LOS enhanced dedicated imagery. This capability greatly reduces the
operational and tactical risks associated with small unit operations in all environments.
CL II UAVs provide RSTA operations under canopy, open, rolling, complex, and urban
terrain. It is carried by selected platforms and capable of autonomous flight and
navigation. The aerial system should be capable of acting as a communication relay.
The CL II UAV supports the following tasks: Provide a reconnaissance and
security/early warning capability for the UA, day or night; Remotely over-watch and
report changes in key terrain, avenues of approach and danger areas in open and
restrictive terrain, and urban areas; Perform target acquisition for the UA (LOS, BLOS
and NLOS); Perform limited communications relay; Provide teaming opportunity
between itself and other manned systems for the purpose of target acquisition, R&S;
Does not require an airfield; Capable of covering three Named Areas of Interest (NAIs).
The CL I UAV provides RSTA operations in open, rolling, complex, and urban
terrain under canopy, and in MOUT. Selected platforms and dismounted soldiers will

14
manpack the UAV. It will use autonomous flight and navigation with Vertical Take-off
and Landing (VTOL).
One system consists of two UAVs and a control interface, which displays the
information to the operator and allows human interface with the AV. The control
interface is interoperable with the dismounted soldier and the FCS Battle Command
system for mounted control. The system will provide a networked SA capability to the
UA and small unit (platoon), in all missions, securing areas, and providing RSTA.
Soldiers will employ the system and dismounted soldiers will carry it in a container that
fits within a man-packed “MOLLE pack” and protects the system from the effects of the
weather and terrain (rain, dust, etc).
The CL I UAV supports the following tasks: Provide a reconnaissance and
security/early warning capability for the UA, day or night; Remotely over-watch and
report changes in key terrain, avenues of approach and danger areas open, rolling and
restrictive terrain, and urban areas; Provide target information for the LOS/BLOS;
Provide target information for area fire munitions; Perform limited communications relay
(narrow band, short duration) in restrictive terrain within echelon; Does not require
airfields.
2. Mounted Combat System (MCS)
The Future Combat System’s (FCS) Mounted Combat System (MCS) is a manned
combat platform that provides offensive maneuver to close with and destroy enemy
forces. The MCS is a joint effort between the Army and the Defense Advanced Research
Projects Agency intended to replace the Army’s current fleet of General Dynamics M1
Abrams tanks, United Defense M2 and M3 Bradley Fighting Vehicles and other armored
vehicles.
3. Infantry Carrier Vehicle (ICV)
The ICV is the FCS Manned Combat Platform that provides the mobility for 11
personnel (two-man crew and nine-man infantry squad) on the battlefield. It is located
within the infantry platoons and companies within the CAB. The ICV delivers
dismounted forces to the close battle, supports the squad by providing self-defense
weapons support, and carries the majority of equipment freeing the individual soldier of
excess weight.

15
4. Armed Robotic Vehicle Assault Variant (ARV-A)
The ARV-A provides the Infantry platoon Reconnaissance, Surveillance, and
Target Acquisition (RSTA), direct fire and BLOS capabilities in support of maneuver and
dismounted operations. It responds to actions on contact, executing fire and maneuver
and tactical assault to ensure lethality overmatch. It supports cooperative engagements in
the full variety of terrain sets including "point and shoot" engagements by dismounted
soldiers and designation of firing missions from other platforms or dismounted elements.
ARV-A is the primary unmanned ground platform for reconnaissance and surveillance
operations and the primary unmanned ground system enabler of BLOS in the Infantry
platoon. The ARV-A RSTA mission is three-fold: Provide the sophisticated on-board
sensors; Enable the delivery of precision BLOS fires; Detect, recognize, and identify
targets with enough fidelity to support the use of LOS, BLOS and NLOS assets to
support cooperative engagement.
5. Armed Robotic Vehicle Assault Variant (ARV-L)
The ARV-L is an FCS Unmanned System, transportable by UH-60 that will
remotely provide reconnaissance capability and provide LOS/BLOS over-watching fires.
6. Armed Robotic Vehicle - Reconnaissance Surveillance, and Target
Acquisition Variant (ARV-RSTA)
The Armed Robotic Vehicle-Reconnaissance, Surveillance, and Target
Acquisition (ARV-RSTA) is the primary unmanned ground platform for reconnaissance
and surveillance operations and the primary unmanned ground system enabler of BLOS
in the MCS Company within the Unit of Action. The ARV-RSTA’s mission is three-
fold: Provide the Recon Troop Scout with sophisticated on-board sensors; Enable the
Mounted Combat System delivery of precision BLOS fires; Detect, recognize and
identify targets with enough fidelity to support the use of LOS, BLOS and NLOS assets
to support cooperative engagement.
7. Reconnaissance and Surveillance Vehicle (R&SV)
The R&SV is the FCS Manned Combat Platform that conducts streamlined
acquisition, discrimination of multiple target sets, and provides a dynamic hunter-killer
capability using on-board systems and Comanche and other UA organic, UE, Joint, and
Coalition lethal systems. It provides sophisticated on-board sensors and a suite of tools
to integrate other sensors such as MASINT, SIGINT, and EO/IR. It is employed within

16
teams of both manned and unmanned robotics sensor platforms as well as unattended
systems. Highly trained multi-functional scouts operate it. It provides sensors that will
detect, locate, track, classify, and automatically identify targets from increased standoff
ranges under all climatic conditions, day or night.
8. Non-Line-of-Sight Mortor (NLOS Mortor)
NLOS Mortors are the FCS Manned Combat Platform that provides short-range
indirect fires in support of assault battle units. It accommodates a smoothbore 120 mm
Mortar System, which can fire the full family of mortar ammunition (HE, illumination,
IR illumination, smoke, precision-guided, DPICM, training, and non-lethal).
9. Non-Line-of-Sight Launch System (NLOS LS)
NLOS LS is the FCS System that provides networked, extended-range targeting
and precision attack of armored, lightly armored, stationary, and moving targets during
day, night, obscured, and adverse weather conditions. The system’s primary purpose is
to provide responsive precision attack of High Pay-off Targets in support of the UA in
concert with other UA NLOS, external and Joint capabilities. The system also provides
“discriminating” capability via automatic target recognition and limited battle damage
assessment.
10. Non-Line-of-Sight Cannon (NLOS Cannon)
NLOS Cannon is the FCS Manned Combat Platform that provides networked,
extended-range targeting and precision attack of point and area targets in support of the
UA with a suite of munitions that include special purpose capabilities. It provides
sustained fires for close support and destructive fires for tactical standoff engagement. It
provides responsive fires in support of Combined Arms Battalions and their subordinate
units in concert with LOS, BLOS, NLOS, external, and joint capabilities. It provides
flexible support through its ability to change effects round-by-round and mission-by-
mission. It provides rapid response to calls for fire, high rate of fire, and a variety of
effects on command.
11. Land Warrior System
Existing program leveraged by FCS that provides an overwhelmingly lethal and
survivable Soldier System of Systems capable of dominance across the entire spectrum of
operations. For purposes of this model, two separate types of infantry soldiers

17
transported via the ICV model the Land Warrior. One type of modeled soldier is using
an M-16 rifle, and the other modeled soldier is using an M-249 squad automatic weapon.
12. Apache Attack Helicopter AH-64D
The AH-64D is a quick-reacting, airborne weapon system that can fight close and
deep to destroy, disrupt, or delay enemy forces. The Apache is designed to fight and
survive during the day, night, and in adverse weather conditions throughout the world.
The principal mission of the Apache is the destruction of high-payoff targets using the
HELLFIRE missile. It is also capable of employing a 30 mm M230 chain gun and Hydra
70 (2.75 inch) rockets that are lethal against a wide variety of targets. The Apache has a
full range of aircraft survivability equipment and has the ability to withstand hits from
rounds up to 23 mm in critical areas.30
13. JSF (Joint Strike Fighter)
The Joint Strike Fighter (JSF) is a multi-role fighter optimized for the air-to-
ground and close-air-support (CAS) roles, designed to affordably meet the needs of the
Air Force, Navy, Marine Corps and allies, with improved survivability, precision
engagement capability, the mobility necessary for future joint operations and the reduced
life cycle costs associated with tomorrow’s fiscal environment. JSF will benefit from
many of the same technologies developed for F-22 and will capitalize on commonality
and modularity to maximize affordability.31
B. RED FORCE DESCRIPTION
The enemy does not obtain a characterization of any traditional military echelon,
but is rather decentralized and autonomous in nature. Enemy descriptions listed in the
following paragraphs are excerpts from the Federation of American Scientists (FAS)
Military Analysis Network.32
30 FAS Military Analysis Network, AH-64 Apache, Retrieved 22 September 2005, from the World
Wide Web at http://www.fas.org/man/dod-101/sys/ac/ah-64.htm
31 FAS Military Analysis Network, Joint Strike Fighter, Retrieved 22 September 2005, from the
World Wide Web at http://www.fas.org/man/dod-101/sys/ac/jsf.htm
32Federation of American Scientists, Retrieved 22 September, from the World Wide Web at
http://www.fas.org/main/home.jsp

18
1. BMP-3 System
The BMP-3 was accepted for service in 1990 and while of a similar size to other
Infantry Fighting Vehicles (IFVs) it is more heavily armed than any previous IFV as it
mounts a 100mm 2A70 rifled gun, 30mm 2A42 cannon and a 7.62mm PKT machine
gun.33
2. 82 Mortor System
The 82 mm Mortor unit provides unique indirect fires that are organizationally
responsive to the ground maneuver commander. Military history has repeatedly
demonstrated the effectiveness of mortars. Their rapid, high-angle, plunging fires are
invaluable against dug-in enemy troops and targets in defilade, which are not vulnerable
to attack by direct fires.34
3. Dismounted Soldier
The dismounted soldier contains an array of capabilities and threats. The
following sub-paragraphs identify the weapon systems fired by the dismounted soldiers.
a. Surface-to-Air System (SA-16)
SA-16 GIMLET (Igla-1 9K310) man-portable surface-to-air missile
system, a further development from the SA-7 & SA-14 series, is an improved version of
the SA-18 GROUSE, which was introduced in 1983, three years before the SA-16.
Features added to the SA-16 include a new “seeker” and modified launcher nose cover.
The 9M313 missile of the SA-16 employs an Infrared (IR) guidance system using
proportional convergence logic, and an improved two-color seeker, presumably IR and
UV.35
b. Rocket Propelled Grenade System (RPG 7)
The RPG-7 anti-tank grenade launcher is one of the most common and
most effective infantry weapons in contemporary conflicts. It is rugged, simple and
carries a lethal punch. Whether downing US Blackhawk helicopters in Somalia, blasting
33 Zaloga, Steven J. BMP Infantry Combat Vehicle, 2nd Ed, Concord Publications, 1990, Hong Kong.
34 FAS Military Analysis Network, Mortars, Retrieved 23 September 2005, from the World Wide
Web at http://www.fas.org/man/dod-101/sys/land/mortars.htm
35 FAS Military Analysis Network, SA-16 Gimlet, Retrieved 23 September 2005, from the World
Wide Web at http://www.fas.org/man/dod-101/sys/missile/row/sa-16.htm

19
Russian tanks in Chechnya, or attacking government strong points in Angola, the RPG-7
is the weapon of choice for many infantrymen and guerrillas around the world.36
c. Anti-Tank System (AT-7)
The Russians characterize the AT-7 ATGM as a complex and light or man
portable (5-20 kg) anti-tank system. It permits long-distance carry by dismounted
infantry. Since the module is small, and fires quickly corrected by shifting its field of
view, it may also be used to engage hovering or stationary helicopters.37
d. RPK-74
The RPK-74 is a machine gun version of the AKM-74, firing the same
ammunition. Instead of the prominent muzzle brake used on the AK-74, the machine gun
has a short flash suppressor. The magazine is longer than that normally used with the
AK-74, but the magazines are interchangeable. The RPK-74 has a bipod.38
4. Armored Personnel Carrier (APC) BTR-80
The BTR-80 is a modern, lightly armored vehicle with a diesel power train. It has
been in service since the early 1980s. The BTR-80 is a lightly armored amphibious
vehicle with a collective chemical-biological-radiological (CBR) protective system.
Operated by a crew of three, the vehicle can deliver a squad of seven infantry troops on
the battlefield while provide close fire support. It can also perform reconnaissance,
combat support and patrol missions.39
5. T-72 Tank System
The T-72, is a Russian medium size tank which entered production in 1971. The
T-72 has six large road wheels and three track return rollers, which carries a 120 mm
main gun capable of firing both traditional and precision guided munitions.40
36 Lester W. Grau, For All Seasons: The Old But Effective RPG-7 Promises to Haunt the Battlefields
of Tomorrow, Foreign Military Studies Office, Fort Leavenworth, KS Retrieved 23 September 2005 from
the World Wide Web at http://www.g2mil.com/RPG.htm
37 FAS Military Analysis Network, AT-7, Retrieved 23 September 2005, from the World Wide Web at
http://www.fas.org/man/dod-101/sys/land/row/at-7.htm
38Retrieved 23 September 2005 from the World Wide Web at
http://www.sovietarmy.com/small_arms/rpk-74.html
39 FAS Military Analysis Network, BTR-80, Retrieved 11 October 2005, from the World Wide Web at
http://www.fas.org/man/dod-101/sys/land/row/btr-80.htm
40 FAS Military Analysis Network, T-72, Retrieved 23 September 2005, from the World Wide Web at
http://www.fas.org/man/dod-101/sys/land/row/t72tank.htm

20
C. MODEL VIGNETTE DESCRIPTION
TRAC-WSMR provided the initial vignette, Northeast Asia (NEA) 50.2, for the
basis of this research. The nomenclature NEA 50.2 identifies the specific vignette
modeled within CASTFOREM at TRAC-WSMR. NEA 50.2 grew from the NEA 50
scenario modeled within VIC at TRAC-Leavenworth. NEA 50.1 is the same scenario but
modeled with CASTFOREM. The difference between NEA 50.1 and NEA 50.2 lays
within the Blue force Structure. NEA’s 50.1 Blue Force is a traditional Brigade Combat
Team (BCT). NEA’s 50.2 Blue Force is a Combined Arms Battalion (CAB), as part of a
Unit of Action (UA), from the Army’s Future Combat Systems.
The use of the model, Map Aware Non-Uniform Automata (MANA), replicates
the CASTFOREM NEA 50.2 vignette. The following chapter provides an overview of
MANA. The initial scenario models an 18-hour battle, starting from the initial Start
Position (SP), followed by the Order of March towards the Release Point (RP), and
finishes with the attack of an urban location. However, the scope of this thesis focuses on
modeling a critical 2-hour window of the NEA 50.2 scenario using MANA. This critical
2-hour window models the overwhelming mission and goal of the CAB to clear and
secure OBJ DALLAS within an urban terrain (OBJ TEXAS) in a timely manner (See
Figure 3).

21
Figure 3. NEA 50.2 Area of Operation Map
Control of this key terrain is extremely important because follow on units from
the Southeast will need to use OBJ TEXAS as part of a main supply and logistics route in
order to continue another advance towards the capitol city located Northwest of OBJ
TEXAS.41 The terrain surrounding the urban area is quite mountainous and covered with
varying dense vegetation. Along the avenue of approach is a river. The FCS platforms
are tested in their ability to negotiate all obstacles providing protection to the forces in the
city as well as the FCS’s ability to use LOS, BLOS, NLOS weapons in a completely
networked manner to clear and ultimately secure the city. The city itself provides
varying buildings and urban obstacles that may hamper the FCS’s ability to clear and
41 Brigade and Below Scenario (BBS) slide show, March 2005, provided by Mr. Tom Loncarich,
TRAC-WSMR during office visit 25 June 2005.

22
secure the area in a timely manner. Not modeled in this vignette is a BCT arriving from
the East to secure the denser part of the city easterly of OBJ DALLAS.
Figure 4 outlines the Blue Force Combined Arms Battalion (CAB) disposition.
The CAB, with additional UA assets, is blended into four teams; A, B, C, and D, as
shown in Table 1. Each team has a specific mission. Team A provides reinforcing fire
and support from a position West of OBJ TEXAS. Teams C and D will cross the river to
the North and advance onto OBJ EL PASO and OBJ DALLAS. Team B secures OBJ
HOUSTON and provides over-watching fires as Team C secures OBJ EL PASO and
allows a passing of lines from Team D to secure OBJ DALLAS.
CAB
Figure 4. Combined Arms Battalion Tree Diagram

23
ICV
MCS
R&SV
NLOS-M
ARV-RSTA
ARV-Assault
ARV-Light
CL
I
UAV
CLII
UAV
CL
III
UAV
NLOS-C
NLOS-LS
JSF
Air
Strike
Force
AH
64
D
Team A
MCS PLT 1 3 1
MCS PLT 2 3 1
INF PLT 3 5 1 1 2
HQ 1 3
Team B
MCS PLT 1 3 1
MCS PLT 2 3 1
INF PLT 3 5 1 1 2
HQ 1 3
Team C
INF PLT 1 5 1 1 2
INF PLT 2 5 1 1 2
MCS PLT 3 3 1
MTR SEC 2
HQ 1 3
Team D
INF PLT 1 5 1 1 2
INF PLT 2 5 1 1 2
MCS PLT 3 3 1
MTR SEC 2
HQ 1 3
REC TRP
REC PLT 1 3 1 6
REC PLT 2 3 1 6
REC PLT 3 3 1 6
UAV SEC 12
MTR BAT (-)
MTR PLT 4
UA Supporting Assets
UA NLOS A BAT (+)
NLOS PLT 1 3 6
NLOS PLT 2 3 6
Air Assets 48
6
Table 1. NEA 50.2 Team Disposition

24
Table 2 outlines the enemy force disposition. In order to maintain an unclassified
thesis, the true enemy (Red Force Order of Battle) from the original vignette will remain
unidentified. However, within the limits of an unclassified disclaimer, a traditional
military echelon does not characterize the enemy, but the enemy is rather decentralized
and autonomous in nature. Each enemy soldier and platform has 100% strength and
capabilities. A generality of the enemy from the original vignette is as follows:
The Operational Environment that the Threat would assume, from what I
believe our Threat Experts would tell you, is that few armored vehicles
would be isolated in any one urban area. They would be in small groups,
platoon size or less, and would be scattered throughout the entire terrain
area in hidden positions. They would move only short distances to avoid
detection from aerial sensors, and would be used only when it was felt
they would be at an advantage in an isolated situation.
-Tom Loncarich, Senior Operations
Research Analyst (TRAC-WSMR)
The author modeled this type of enemy, but assumed greater numbers with more
aggressiveness and lethality. Tom Loncarich noted that the disposition of the modeled
Red Force assumed for the MANA scenario is rather, “more high-end, aggressive threat
excursion. Perhaps possible, but not probable.” Since this research includes the use of
“Data Farming” tools intended to unleash possibility and surprise, and the ability to use
an exhaustive and thorough Design of Experiments exists, then there presents a need to
model a flexible and challenging enemy Order of Battle in order to identify any “what if”
or “worst case” plausible outcomes.
Asset Quantity
Red BMP-3 6
Red 82 Mortors 6
Red SA-16 Infantryman 5
Red RPG-7 8
Red AT-7 5
Red Scout 5
Red RPK-74 6
Red AK-M Infantryman 80
Red SVD 3
Red APC 6
Red T-72 6
Table 2. Red Force Disposition

25
The Red Force uses the urban area as a hide position in order to attack the Blue
Force when advantageous. The Red Force mission within the urban area is to defend and
deny US and allies access to important avenues of approach, in order to help protect the
regime from intervention by US and combined forces.42
42 Brigade and Below Scenario (BBS) slide show, March 2005.

26

27
III. MODEL DEVELOPMENT
"SIC ITUR AD ASTRA"
(This Is the Way to the Stars)
102D FIELD ARTILLERY REGIMENT
The purpose of this chapter is three fold. First, it provides the reader with an
understanding of the model. Second, it provides a methodology for developing an
advanced simulation technique. Some readers may consider the second point of most
interest as it provides systematic directions, explaining the author’s methodology to
develop the critical values within this scenario. Considering George Box’s quote that “all
models are wrong, some are useful,” the last part of this chapter outlines limitations
within the modeling environment and the techniques the author used to develop a useful
scenario within the model.
Looking back, nobody really knows when humans first introduced simulation to
represent warrior battle maneuvers. Possibly, a polished stone represented the first “toy
soldier” and a flat piece of dirt represented his battle space. Historians accredit Sun Tzu,
the Chinese general and military philosopher, as inventing the first simulation, or war-
game, known as Wei Hai (meaning “encirclement”) about five thousand years ago.43
Though initially titled as a game, it truly offered a primitive simulation process that
replicated a battle as many times as the player desired, training a military mindset in the
art of war. Improved simulation techniques continued to emerge through the years.
A. AGENT-BASED SIMULATION (ABS) OVERVIEW
The Department of Defense (DoD) incorporates simulation modeling techniques
to support decision makers. Primarily, DoD simulation models encompass high-
resolution, complex, and resource intensive modeling procedures.
The scenario generation process for our high-resolution simulations is
man-hour intensive and requires detailed knowledge of the simulation’s
underlying data and operating assumptions. Often times, the analyst is
43 Peter P. Perla, The Art of War-Gamming, United States Naval Institute, Annapolis, Maryland, 1990,
p. 15.

28
limited to a small set of simulation runs due to the simulation’s
complexity, scenario development constraints, and the decision maker’s
timeline. Consequently, they may only obtain a limited view of possible
outcomes.44
For example, to replicate a howitzer firing a projectile in a high-resolution model, the
analyst must know more information then just the classical ‘trajectory in a vacuum’
physics problem. Instead, the analyst must take into account interior, exterior, and
terminal ballistics. Each includes, but is not limited to, factors such as projectile square
weight, propellant temperature, propellant moisture, muzzle velocity variation, and tube
wear effecting interior ballistics, as well as meteorological atmospheric conditions such
as air temperature, air moisture, wind direction, wind speed, and the rotation of the Earth
effecting exterior ballistics. These examples only name a few factors that the analyst
could consider when modeling the howitzer firing the projectile. This process then
repeats for every other howitzer in the battery, positioned at different locations, and any
other munitions also fired. As such, a simulation requiring multiple munitions, from
several platforms demands significant computing ability just to provide the decision
maker with useful insights required for his decision.
As a result, an innovative class of simulation, known as agent-based simulation
(ABS), emerged as a low-resolution simulation to compliment, and augment, previously
established more computationally intensive physics-based simulation models. The role
of ABS should not replace high-resolution models. However, the author maintains that
over the past few years, ABS increasingly proves useful to the DoD in primarily two
areas. The first is to use ABS up front in an exploratory analysis, in order to gain quick
insight and narrow the focus of seemingly endless possibilities of factors, parameters, and
variables in order to expedite building high-resolution physics-based simulations.45 This
saves time and money on the front end of a simulation project. The second is to use ABS
in order to offset timely resource intensive key battlefield objectives that otherwise
require excessive recourses in physics-based models. Here the analyst switches back and
forth between two models in order to gain advanced scenario insight.
44 Lloyd Brown, Thomas Cioppa, and Thomas Lucas, “Agent-Based Simulations Supporting Military
Analysis,” Phalnex, April 2004.
45 Brown, Cioppa, and Lucas.

29
Insight, surprise, and outliers all hail from analysis. ABS offers quick scenario
generation, fast run times, rapid data turn around, and permits the analyst to consider
many alternatives in a short amount of time. ABS complements and augments physics-
based models permitting analysts to examine the problem over a greater range of
plausible possibilities, while helping to fix the aforementioned quantities.
B. WHY MANA?
The author chose Map Aware Non-Uniform Automata (MANA) as the agent-
based simulation-modeling tool to support this research. MANA’s individual agent and
squad situation awareness (SA) aptitude, coupled with its networked communication
parameters supports use of this tool to replicate the NEA 50.2 scenario.
FCS are networked via a C4ISR architecture including networked
communications, network operations, sensors, Battle Command system,
training, and both manned and unmanned reconnaissance and
surveillance (R&S) capabilities that will enable levels of SA and
synchronized operations heretofore unachievable.46
New Zealand’s Defense Technology Agency (DTA), initially developed MANA,
and has continuously updated the model as needed. As a general notation, the MANA
User Handbook provides direct annotation for the following paragraphs.47
The reader must first appreciate the meaning of MANA. Concurring with
Lindquist’s dissection48 of each word constructing the acronym MANA, we have:
• Map Aware — Agents are aware of and respond to, not only their local
surroundings and terrain, but also a collective registry of recorded battlefield activities.
• Non-Uniform — Not all agents move and behave in the same way (much like
soldiers, sailors or airmen).
46 Unit of Action Manuever Battle Lab, TRADOC Pam 525-3-90, Future Force Operational and
Organization Plan, Maneuver Unit Action, with Change 3, Fort Knox, KY, 30 July 2004.
47 Galligan, David P., Mark A. Anderson, Michael K. Lauren, Map Aware, Non-Uniform Automata
version 3.0, New Zealand Defense Technology, July 2004.
48 Lindquist, p.27.

30
• Automata — Agents can react independently to events, using their own
“personalities.” Personalities, in general, are propensities that guide an agent’s actions to
move.
Fundamentally, analysts use MANA for two reasons. The first is because the
behavior of the entities within a combat model (both friend and foe) adds possibilities to
the analysis of the possible outcomes. The second is because analysts have limited time
to determine particular force mixes and each side’s combat effectiveness necessary for
programming into higher resolution models.
The behavior of troops in any given scenario plays an important role in
simulations. However, as is the weather, human nature is mathematically intangible, and
often overlooked by analysts. MANA, as with other ABMs, contains entities controlled
by decision-making algorithms. Hence, agents representing military units make their
own decisions, as opposed to the modeler explicitly determining their behavior in
advance.
To differentiate MANA from highly detailed models also using agents, analysts
sometimes refer to MANA as an Agent Based Distillation (ABD), which reflects the
intention to model only the essence of a problem. MANA falls into a subset of these
models, called cellular automaton (CA) models. CA models have their origin in physics
and biology. The famous Ising model of magnetic spin alignment is an example of such
a model in physics, while Conway’s “Game of Life” is an example of a CA model
designed to explore biological ideas. MANA and other CA models encompass complex
adaptive systems (CAS) properties because entities react to their surrounding. Agents’
decisions, actions, and reactions alter as agents switch among their state conditions.
Some properties exhibited in MANA include:
• Local interactions among agents emerge into a “global” behavior
• Agents interact with each other in non-linear ways, and “adapt” to their local
environment
• The influence of situational awareness when deciding an action
• The importance of sensors and how to use them to best advantage

31
MANA users may sit down and obtain a good understanding of the model within
a few short hours, while completing their first scenario soon after. MANA offers a
simple to use graphical user interface (GUI), including drop down window capabilities
much like many Window based applications. As a reminder, the preceding information
came primarily from the MANA User Handbook.
C. MODELING METHODOLOGY
This section describes detailed information used to create the scenario within the
MANA model. In turn, it provides the reader a methodology to facilitate the model
development process implemented within this simulation technique. The reader wishing
more detail may consider viewing each corresponding section within Appendix A,
SPREADSHEET MODELING to the section headings within this chapter prior to
advancing to each new section. Each appendix shows a snapshot of modeling
spreadsheets built with Excel. Spreadsheet modeling describes the approach
implemented to transform real world data into scaled MANA values.
1. Scaling: Configure Battlefield Settings
Scaling the scenario is the most important step, as it also parallels as the first step.
The model’s output becomes useless if the scenario fails proper scaling. Part of the
conclusions, and lessons learned section of this thesis, describes in more detail the trials
and errors associated with scaling. In addition, Appendix A provides the screen shots of
the spreadsheet modeling referenced throughout this chapter. Spreadsheet modeling
assisted in the entire scaling and model development of this scenario. CAPT Mike
Babilot, United States Marine Corps, developed a baseline spreadsheet, which the author
incorporated within this work.49 A modified and upgraded version of the baseline
spreadsheet fits this scenario, and may assist in a wider array of future scenario
applications. The intent of Appendix A is two fold. First, it provides the reader with the
input values assigned to each modeling entity within MANA, such that the reader can
replicate the scenario by inputting each value into a MANA version 3.0.39, or newer,
49 Naval Postgraduate School Thesis, Comparison of a Distributed Operations Force to a Traditional
Force in Urban Combat, Michael Babilot, September 2005.

32
simulation model. Second, it provides a graphical representation of the modeling
methodology.
As humans, we typically express distances in feet, miles, kilometers; time in
seconds, minutes, hours; and velocities in feet per second, miles per hour, or kilometers
per hour. In essence, we think of a distance and time. MANA provides distance and time
in grids (or pixels) and time steps. The user defines the resolution settings for each
MANA scenario as any rectangle between the values of 1 square and 1000 square grid
matrix. As such, the user also defines the relationship of MANA grids to real world
distances. One pixel may represent any metric of length. Possible examples include a
centimeter, foot, kilometer, or even 5 miles. The model is a stochastic simulation,
allowing the user to define each time step as a second, minute, hour, 5 hours or any other
time metric.
Three parameters molded together, properly scale any simulation scenario. The
first labels the model terrain distance. The second represents the total time the scenario
runs with respect to real world time. The third defines the velocity at which agents travel
along the terrain. This scenario encompasses a 500 by 500 square grid resolution
representing a 2.6 by 2.6 kilometer terrain piece upon the Earth’s surface (Figure 5).
Figure 5. NEA 50.2 MANA Screenshot

33
The full scenario lasts for 7200 time steps, which represents the critical 2-hours of
real time to secure the urban objective. Thus, each time step corresponds to one second.
Calculations stemming from these two parameters yield the correct MANA speed in
which each agent travels. Immediately one might ask why the maximum resolution of
1000 square grids does not scale the scenario. The answer lies in the velocity at which
each agent travels.
The model itself limits agent’s velocities. Optimally, an agent should travel with
a velocity not exceeding one grid per each time step. Here, a value x, represents the
agent’s velocity, such that in one time step, the agent advances to the next grid with a
probability of x over 100. Therefore, the ratio 0/100 describes a stationary agent while
100/100 describes an agent’s ability to advance one grid with 100% probability per time
step. As such, 200/100 described the agent’s ability to advance two grids with a
probability of one. Ultimately, agents appear to move at different velocities.
MANA limits the ratio to not exceed greater then 1000/100. As the numerator
grows past 100, certain side effects occur. The MANA User Guide describes these side
effects in greater detail. However, one side effect increases the possibility of two agents
passing right by each other without detection of the other agent. This side effect actually
represents possible real world occurrences, and the author accepts it within the scenario.
Combining the equations shown in Table 3 balances the distance, duration, and
velocity—yielding a 500 square grid resolution.
Given the battle lasts for 2 hours, and the terrain encompasses 2.6 square
kilometers, experimentation with associated values for time step, second, and grid, led to
a feasible scaling for this specific scenario. Notice an increase of time steps per second
provides unrealistic characteristics allowing each agent to have multiple capabilities per
second. In real life, a second reflects a short amount of time, limiting a soldier’s
cognitive and reaction process. Inverting the relationship with an increase of seconds per
each time step, or setting the resolution above 500 grids, dramatically amplifies the
converted MANA movement ratios towards 1000/100, and increases more side effects.
The feasible scaled values assume a compromise between extremes. Notice each air
movement speed may result with a failed probability to detect other agents within

34
proximity. However, this possible failure indicatively represents air assets flying rapidly
at high altitudes.
2.6 KM 1000 meters meters
5.2
500 grid 1 KM grid
• =
60min 60sec 1 timestep
2 hours 7200 timesteps
1 hour 1min 1sec
• • • =
General speed conversions of tactical speeds modeled in this scenario conversion
Dismounts 1.6 km * 1 hour * 1 min * 1 sec * 500 grids = 0.09 grids * 100 = 8.547008547 9
1 hours 60 min 60 sec 1 steps 2.6 km 1 step
Ground Vehicles 16 km * 1 hour * 1 min * 1 sec * 500 grids = 0.85 grids * 100 = 85.47008547 85
UAV CL I 60 km * 1 hour * 1 min * 1 sec * 500 grids = 3.21 grids * 100 = 320.5128205 321
UAV CL II and Helo 80 km * 1 hour * 1 min * 1 sec * 500 grids = 4.27 grids * 100 = 427.3504274 427
UAV CL III 140 km * 1 hour * 1 min * 1 sec * 500 grids = 7.48 grids * 100 = 747.8632479 748
CAS 300 km * 1 hour * 1 min * 1 sec * 500 grids = 16 grids * 100 = 1602.564103 1000
Air
Ground
rounded mana input / 100
Table 3. Scaling Equations

35
Table 4 edits the terrain properties, represented by colors, within the model. The
user defines each color with the Red-Green-Blue (RGB) schematic found in most
paintbrush applications. The user assigns a name to each color. Each color represents an
associated going, cover, and concealment value. Going and movement speed are
synonymous. Cover provides protection from bullets, and concealment shields them
from other’s visibility. The color affects each agent’s movement speed, as well as their
cover and concealment from others, for each time step while traveling within that terrain
color. For this scenario, each value estimates percentages of speed, cover, and
concealment when traveling through similar terrain and vegetation features as
experienced by the author. For example, the color defining a Wall prevents an agent
from going through it, while providing 100% cover and concealment. In contrast, the
color defining a Road permits an agent to travel an average rate of 90% of its maximum
speed, and provides zero cover and concealment.
Table 4. Edit Terrain Properties

36
Refer to Appendix A, section “Configure Battlefield Settings” to view the
remaining input values associated with the “Configure Battlefield Settings” portion of the
Model. Each spreadsheet screenshot correlates to an associated series of main menu tabs
located within the GUI of the MANA application. All appendices include the necessary
values needed for entry to build this scenario.
2. Model Unit Summary
Chapter II outlined both the Blue and Red players modeled in this scenario. This
section discusses in detail how to model each player in MANA. Appendix A, section
“Model Unit Summary,” is a tablature format of multiple inputs from the General,
Ranges, and Weapons GUI tabs within the MANA application. Though other sections in
Appendix A describe these three tabs in detail, fundamental rules and assumptions
established to build this scenario lay within this specific section. Following in each
paragraph is a description for each table column. Refer to the actual table in “Model Unit
Summary,” for each associated value.
a. Players
Unit Type / Squad: Each group of real world players has an assigned
squad value within the model. Squads fall into two categories, Red or Blue, followed by
the traditional name for that specific player. There are 33 squads built in this scenario.
Squads one through 11 are Red Force units and squads 12 through 33 are Blue Force
units.
Start # - End #: Each squad has a number for record keeping. Most
squads have identical start and end numbers. However, each of the four maneuver teams,
A, B, C, and D, has identical UAV squads assets. As such, the scenario has four squads
for each of the Class I and Class II UAVs, resulting in different start and end numbers.
# Type Squads: Following from the preceding bullet, this column
identifies the number of squads built in the scenario to represent the real world player.
Thirty-three squads represent the real world players.
# Agents: Within each squad, there may be multiple agents. Each icon on
the battlefield map defines a separate agent.

37
Moving Parts: Moving Parts is the total number of agents per each type of
squad. It is the product of the # Type Squads and # agents. The running tally of the
number of moving parts within the scenario facilitated aggregation in order to minimize
the run time of the scenario.
Squad Class: Each squad has an assigned class value. Red Force squad
class values range from one to three, and Blue Force squad class values range from 100
to 210. Class values limit the types of munitions fired from enemy classes. Squad Class
tightly weaves with Squad Threat Level, as well as each Target Classification value. The
Squad Class restricts, for example, a Blue Infantrymen firing a M16 rifle at a Red T72
tank, but authorizes a NLOS cannon system to fire its primary weapon at the same Red
T72 tank.
Squad Threat Level: In addition to the Squad Class, a Squad Threat Level
also designates each squad. The threat level simulates the Maneuver Commander’s
Guidance and limits the number of munitions fired from a particular squad. For example,
the Blue NLOS Cannon Platoon has authorization to shoot at a Red AK-M Infantrymen,
but it would be an expensive choice of munitions to fire at a single target. However,
threat levels of multiple agents are added together to create a cumulative group threat
level within a specified radius. Now, if an abundant number of infantrymen are located
within a specified blast radius, then they form a group. Thus, the cannon system will fire
the same projectile at this group target.
b. Weapons
Weapons: A general assumption is that all squads have, at most, two
weapon systems. This includes the primary weapon classifying a specific platform, and
an alternate weapon also found on that platform. In addition, each different kinetic
energy (LOS) weapon fires only one type of bullet. However, two different target effects
simulate the use of each area fire (NLOS or BLOS) weapon system. As such, a third
weapon added to all squads armed with NLOS or BLOS weapons works around the
model’s limitations. Weapon 3 simulates different effects the same projectile fired from
Weapon 1 has against hardened targets. Weapon 1 simulates projectile effects against
soft targets, where as Weapon 3 simulates projectile effects against hard targets. A later
section covers specific weapon modeling characteristics within the scenario.

38
Priority Target Class vs. Non Target Class: Classifies the use of weapons
fired at only specific enemy targets, and in an order of priority.
Min and Max Threat Levels: Offers a specified threat level window that
particular weapon systems are able to fire at enemy targets. This coincides with the
example detailed in Squad Threat Level regarding firing upon a group target in lieu of a
single target.
c. Aggregation
There exist three columns for aggregation. Two of these columns
primarily provide bookkeeping to count the number of squads and agents per side, and to
limit the number of icons present on the map. However, the aggregation value of “1 icon
to X number of real world objects” also doubles as the number of hits required to kill a
specific agent within each squad. This simulated ‘one hit one kill’ for all agents within
the simulation.
3. Movement Rates
As pointed out earlier, scaling the scenario is a critical part in modeling. Table 5
displays initial movement rates. Due to limitations with the model, or assumptions made,
changes occurred to each platform’s basic movement rates noted in Table 5. These
changes reflect different speeds the agent travels at in different state conditions.
Dismounts 1.6 km
1 hour
Ground Vehicles 16 km
1 hour
UAV CL I 60 km
1 hour
UAV CL II and Helo 80 km
1 hour
UAV CL III 140 km
1 hour
CAS 300 km
1 hour
Ground
Air
Table 5. Real World Basic Movement Rates50 51
51 “Unopposed Movement Rates” in FM 90-31, Chapter 4, Table IV-5.

39
Table 5 splits the movement rates into two basic categories: Ground, and Air.
There are four different subcategories, or values, identifying the air category.
Assumptions include that the different atmospheric conditions are negligible on the air
movement speeds. As such, the UAV and Helo converted movement values remain the
same for the remainder of this scenario. Notice in Table 3, the converted CAS movement
value exceeds the MANA limit, 1000. Instead of using the maximum value of 1000 to
represent CAS movement, its speed is set to zero. The CAS icon is set to the side of the
battlefield. The placement assumes the CAS is flying too fast, and at too great of an
altitude, to be effected by the enemy surface-to-air missiles. The CAS has two state
changes, active (Default) and passive (Taken Shot (Pri) ). In the Default state, the CAS
fires upon acquired targets. Upon firing its weapon, it enters a passive or Taken Shot
(Pri) state for 60 time-steps, simulating a racetrack flight route returning it to the same
launch position for future targets.
There are two different subcategories for each ground asset: Dismounted and
Ground Vehicle. Each category has different movement values depending on the squad
state. Table 6, from Appendix A, section “Movement Rates,” identifies the final possible
converted movement rates for each state change within each subcategory of ground
assets.

40
Ground Vehicle Different
State Value Settings
% of
Adjusted
Movement
Speed
MANA
Input
Speed
100% 1.20 120
10% 0.12 12
0% - 0
50% 0.60 60
60% 0.72 72
100% 1.20 120
150% 1.80 180
0% - 0
1% 0.01 1
Default movement Rate
Reach Final Waypoint
Run Start (if applied)
Taken Shot (for primary or secondary)
Shot At
(jugement call based on platforms
ability to fire at 0, 50%, 60% or full speed)
Reach Waypoint
Dismounted Different
% of
Adjusted
Movement
Speed
MANA
Input
Speed
100% 0.09 9
0% - 0
100% 0.09 9
60% 0.05 5
0% - 0
100% 0.09 9
Refuled by Anyone
Taken Shot Blue
Taken Shot Red
Default movement Rate Blue
Default movement Rate Red
Table 6. MANA Movement Speeds
Table 6 shows the final model values inputted in MANA after manipulating the
base movement rates in the movement calculator spreadsheet. The movement calculator
spreadsheet annotated in Appendix A begins with each of the researched basic movement
speeds of 1.6 kmph and 16 kmph for both dismounted and ground vehicles respectively.
Research showed a difference in tactical speeds in a restricted area verses a platform’s
maximum speed, and the author wanted to incorporate both into this scenario.
There exist two ideas behind incorporation the movement calculator. The first
idea defines a platform’s tactical speed as 100% of its movement speed, while defining
its maximum speed as 550% of its tactical speed. The maximum speed of all the FCS
ground vehicles is roughly 90 kmph, thus 550% of 15 kmph equals 88 kmph.52 For

41
simplicity, Red Force ground vehicles have the same movement abilities. Also for
simplicity, assume that a dismounted soldier sprints at about 9 kmph when in a combat
uniform, which is roughly 550% of its tactical speed (1.6 kmph * 5.5 = 8.8 kmph). There
are times when a platform, or a soldier, travel at speeds in between the tactical and
maximum speeds. In the scenario, 200% and 400% rates of the tactical speed represent
these in between speeds.
Secondly, each platform moves at different speeds depending on its combat load.
Based on prior experience of personal timed road marches while carrying combat
equipment, four adjustment factors affect each of the base movement rates. The factor
values affecting ground vehicles is 1 if unencumbered, 0.95 for a light combat load, 0.85
for a full combat load, and 0.75 for a heavy combat load. These values represent both the
strain on an engine as well as a slower safety speed when carrying increased cargo. The
factor values affecting dismounted troops are 1 if unencumbered, 0.7 for a light combat
load, 0.5 for a full combat load, and 0.2 for a heavy load. These values represent a
soldier’s physical inability to travel at the same speed when carrying increased loads.
Babilot designed this movement calculator53 for use within various applications.
For this scenario, assume that the soldier in the urban terrain would spend most of his
time walking or jogging while carrying a light to full combat load; and a ground vehicle
will spend most of its time traveling at either its tactical speed or twice that speed, while
again carrying a light to full combat load. Since some of the FCS ground vehicles are
robotic in nature, a combat load refers to its fuel, add on armor, and ballistics. In each
category, an average of each of these four values determines the adjusted speed. Lastly,
in order to simulate the agent’s reaction in different states, multiply the adjusted speed by
a certain percentage annotated in the second column of Table 6, resulting in the final
input values annotated in the last column of Table 6.
4. Personalities
The premise of ABS is the agent’s ability to act or react due to its goals and
situational awareness. MANA permits each of the agents within a squad to have one of
three categories of situational awareness: Agent Situational Awareness (SA), Squad SA,
53 Babilot.

42
and Inorganic SA. These categories are important to note here, because they help
formulate modeling different sensor, detection, communication, and weapon capabilities.
Agent SA—Response of an agent to information that it receives only from its
current local surrounding that is defined by its Sensor and Detection Ranges found within
its own SA map.
Squad SA—Response of an agent to information on other agents’ (only within the
squad) local surroundings defined by their Sensor and Detection Ranges found within
their SA map.
Inorganic SA—Response of agent to information on other agents’ (only within
the squad) inorganic SA map. Entities are places on the inorganic SA map via
communication properties among each squad.54
Appendix A, section “Personalities and Ranges,” shows each weighted value
entered into MANA for each state a squad enters. This includes the associated values
needed for entry within each Agent SA, Squad SA, and Inorganic SA field. Left to the
reader is to familiarize himself with the MANA handbook to understand each weighted
value. Operational experience, coupled with designer’s intentions for each platform,
dictate the value setting chosen for each squad’s personality traits. Setting these
personality values last makes the agents move with closer resemblance to how they
would in real life. The author claims that these settings are best applied after
mathematically determining the other parameter settings for each squad’s sensor,
detection, communication, and weapon capabilities. An increased value of a squad’s
desire to go towards the next waypoint simulates the squad’s tactical decision to maintain
a designated march route, where as an increased value of the squad’s desire to go towards
the enemy simulates the squad’s tactical decision to aggress the enemy. Opposite values
have the reverse effect upon each agent. The “Personalities and Ranges” section
summarizes into one large chart much of the inputted values discussed in the following
paragraphs.
54 Galligan, p.28.

43
5. Sense and Detect
This section describes the methodology used to model each squad’s sensor
capabilities. Appendix A, section “Sense and Detect,” portrays the numeric approach
used to set values within MANA. There are two categories: UAV Sensors, and Ground
and other Air (Non UAV) Sensors. For clarity purposes of the technique used, the
discussion of the latter precludes the former.
a. Ground and other Air (Non UAV) Sensors
An assumption made, is that all platform sensor range capabilities fall into
one of six categories: Short, Short-Medium, Medium, Medium-Long, Long, and Extra
Long; which corresponds to 150 meters or less, 200 meters or less, 250 meters or less,
350 meters or less, 500 meters or less, and 1300 meters or less. MANA’s runtime
increases dramatically depending on increased agent’s sensor ranges coupled with the
total number of agents in a scenario. Since this scenario has 280 total agents within the
squads, there existed a need to reduce the sensor ranges. As such, we assume a scaled
down distance of real world sensor ranges to minimize runtime. This scaled down
distance simulates possible degraded sensor capabilities within an urban terrain. Based
on the scenario and terrain, this had little, if any, influence on the results.
A matrix consisting of rows depicting each squad, and columns depicting
each type of sensor is part of Appendix A, section “Sense and Detect, Ground and other
Air (non UAV) Platforms.” There are 18 columns in this matrix. The first three columns
represent whether a squad has short, medium, or long-range antenna capabilities.
Columns four through 18 characterize each of the possible sensor capabilities outlined in
the FCS UA Design Concept Baseline Description.55 The value 1 in each row/column
intersection indicates that the squad modeled has that type of sensor capability. Using the
formula in Figure 6, a weighted adjusted value between 1 and 3.6, numerically describes
each squad’s sensor capability.

44
Sensor Types
1 ( ) 2 ( ) 3 ( ) Adjusted Average Value
15
short medium long
• + • + • + =
∑
Figure 6. Adjusted Average Sensor Value
The following example explains the formula in figure 6: The MCS has
two of the 15 possible sensor types listed in the FCS UA Design Concept Baseline. In
addition, the overall sensor range capability of the MCS has a medium range associated to
it.56 Each of the values short, medium, and long is binary and has the assigned value of
“1” only if it describes that platform's capability. Therefore, MCS’s Adjusted Average
(sensor) Value is characterized by the following values: (short) = 0, (medium) = 1, (long)
= 0, and the sum of the Sensor Types equal to 2. Substituting these values into Figure 6,
the MCS Adjusted Average Value = 2.13.
Each weighted Adjusted Average Value falls within one of the six sensor
range categories (Numerical Value) shown in Table 7. Using these categories, each
squad corresponds to a predetermined table value found in Appendix A, section “Sense
and Detect, Ground and other Air (non UAV) Platforms.” These predetermined table
values convert real world metrics to MANA units and depict the squad’s modeled
distance and probability of detection at each distance.
Range
Short
Medium
Long
Short-Medium
Medium-Long
Extra Long
Numerical Value
>3
= 1
= 2
= 3
1<x<2
2<x<3
Table 7. Numerical Sensor Value

45
Each table’s distance is monotonically increasing, while the probability of
detection is monotonically decreasing. This represents most ground and traditional air
assets with simplistic sensors: However, this is not generally true for UAV sensor
ranges.
b. UAV Sensors
Generally for UAV sensors that were modeled in this research, the UAV
sensors’ probability of detection increases at greater ranges, up to a certain distance.
Then the probability decreases. Notice in Appendix A, section “Sense and Detect,” each
UAV’s adjusted average value depicted in the chart with 18 columns, is greater then the
value of three. Hence, the algorithm annotated in Figure 6 could not be used alone to
depict the increased UAV sensor ranges.
Due to a UAVs complex set of sensor capabilities, each class of UAVs fly
at a specific height while pointing their sensors at an optimal angle towards the ground.
Aviators call this angle, the field of view57. A 90-degree field of view, pointing straight
at the ground, as well as a 0-degree field of view, pointing straight at the horizon,
provides minimal footprints on the ground causing limited detection abilities. Instead, an
optimal angle obtained optimizes the sensor footprint on the ground. The footprint is the
piece of the earth that the UAV sensor performs a sweep width. Different UAVs have
different sensor footprint capabilities.
MANA limits each squad with only one sensor and detection range.
However, each class of FCS UAVs has multiple sensors, as noted in Table 8, generated
from the FCS Design Concept Baseline.58 Refer to Table 9 for definitions of each sensor
type with respect to UAVs only. Again, the procedure alone outlined in paragraph a
above, is insufficient for modeling UAVs. Added to the procedure is a need to create
three additional subclasses within the category, Extra Long, which specify the greater
sensor capabilities of the platoon, company, and battalion level UAVs. All UAV classes
yielded an adjusted average numerical value greater then three, and require a modeling
57 Department of the Navy, Office of the Chief of Naval Operations, Integration of Unmanned
Vehicles into Maritime Missions, TM 3-22-5-SW, chap. 2, p. 4.

46
table, which monotonically increases in both range and probability of detection, similar to
the graph in Figure 7.
AITR
EO
IR
TD
CM
RADAR
Warning
Plum
Dect
Standoff
Chem
Det
SIGNINT
Combat
ID
UAV CL I x x x
UAV CL II x x x
UAV CL III x x x x x x x x x
Table 8. FCS UAV Sensor Type
Table 9. FCS UAV Sensor Type Definitions 59

47
Figure 7. UAV Sensor Probability of Detection Graph60
The coverage factor is an adjusted weighted value comprised of four
factors: UAV speed, sensor sweep width (footprint), time on station (TOS), and size of
area patrolled. The coverage factor is directly proportional to its speed, sweep width, and
TOS, while inversely proportional to the size of the patrolled area.61 The base scenario
assumes maintaining the speed, TOS, and size of patrolled area constant for each
modeled UAV, leaving only the sweep width affecting the probability of detection.
Therefore, each modeled UAV’s probability of detection is solely dependent upon the
length of the sweep width (measured in meters on the ground), or in MANA terms, the
sensor range in grids. Hence, the idea behind modeling each of the UAV sensor
capabilities is to replicate the curve in Figure 7 for each class of UAVs flying at a
specified height, with an optimal field of view, yielding the greatest sweep width
(footprint) on the ground. The graphs in Figure 8 each depict this intent while assuming
the following characteristics for each UAV modeled.
60 Department of the Navy, chap. 2, p. 4.

48
CL I UAV has a 350 ft footprint, which converts to 21 MANA grids. To
obtain this size footprint in real life, the UAV must fly at 500 ft while using a 30-degree
field of view.62
CL II UAV has a 650 ft footprint (38 MANA grids). To obtain this, the
UAV must fly at 1000 ft while using a 30-degree field of view.63
CL III UAV has a 2500 ft footprint (147 MANA grids). To obtain this,
the UAV must fly at 2500 ft while using a 45-degree field of view.64 This is actually 500
ft higher then the recommended window of 1000 – 2000 ft for the FCS CL III UAV 65 66;
however, the only value of concern needed for input into MANA is the width of the
footprint (sensor range).
MANA’s battlefield is only two-dimensional, and in the model, the UAVs
are actually flying at the ground level. In order to simulate the UAV, and all other air
assets flying in this scenario, the scenario has the “Terrain Affects Going” turned off for
all airborne squads. This eliminates the modeled terrain from affecting the speed of the
squads as noted in the Terrain and Battlefield section of this chapter, making the flying
height of each air platform negligible. Refer to Appendix A, section “Sense and Detect,”
for the spreadsheet model behind each graph in Figure 8.67
66 Presentation to the CSA on the FCS Brigade Combat Team Operational & Organizational Plan, by
US Army Futures Center, TRADOC, 7 October 2005.
67 The methodology used to model each squad’s sensor capabilities is adopted by combining lecturer
material from OA3602 Search Theory and Detection, Naval Postgraduate School and the references noted
in footnotes 57, 58, and 66.

49
P(det) of UAV Class I Flying at 500 Ft
Using 30 Degree Field of Veiw With a
350 Ft Foot Print
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 50 100 150
Meters on the Ground
P(det)
P(det) of UAV Class II Flying at 1000 Ft
Using 30 Degree Field of View with a
650 Ft Foot Print
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 50 100 150 200 250
P
(det)
P(det) of UAV Class III Flying at 2500
Ft Using 45 Degree Field of View
with a 2500 Ft Foot Print
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 200 400 600 800 1000
P
(det)
Figure 8. Modeled UAV Sensor Probability of Detection Graphs

50
4. Communication Characteristics
This scenario assumes that each squad uses one of eight communications devices
annotated in Table 10.
Device Type Notes
Cellphone or
equivalent VHF Limited Reliability
Basic Radio or
equivalent UHF LOS
Personal Role
Radio (PRR)
or equivalent UHF
Intra-Team
Communications
PRC 148 or
equivalent VHF/UHF
Platoon – Squad – Team
C2 - CAS Control
JTRS
Cluster(8
channel) or
equivalent Digitial
Future Internet
Networked Protocal
System (Joint Tactical
Radio System)
JTRS
Cluster(4
channel) or
equivalent Digitial
Future Internet
Networked Protocal
Radio System)
JTRS Cluster
5 SFF-D-E-G
or equivalent Digitial
Future Internet
Networked Protocal
Radio System)
PRC 117 or
equivalent
VHF /
UHF /
Satellite
Communi
cations
Squad – Plat – HHQ
CAS/Fires Control (OTH -
Digital)
Table 10. Modeled Communication Types
Appendix A, section “Communication Characteristics” explains in detail each
communication devise assigned to each squad. Each device modeled encompasses
specific parameters outlined in Appendix A. Each relates to its signal transmission range;
outgoing message capacity; outgoing message buffer size; latency of message to reach
receiving squad; reliability of devise to send a transmission; if sent, the message accuracy
in which it is received, maximum length of time a message sent remains in queue; level
of confidence the receiver has in the message; and deliverability conformation.

51
6. Weapon Characteristics
The scenario assumes there is a maximum of only two weapon systems per squad,
falling into two categories, Kinetic or Area Fire. Kinetic (LOS) weapons are those
similar to a rifle or a traditional tank, where as Area Fire (BLOS or NLOS) weapons are
those similar to an indirect artillery system. Table 11 from Appendix A, section
“Weapon Characteristics” provides detailed information of each weapon built in this
scenario including the weapon name, minimum effective range, maximum effective
range, maximum weapon range, blast shot radius, maximum number of targets each
weapon can engage in one minute, and the weapon’s basic load of carried rounds. Each
value converts into values entered into MANA.
Table 11. Weapon Characteristics

52
Depending on whether the weapon is kinetically or aerially modeled depends
directly on limitations within the model, and hence calls for separate spreadsheet
modeling techniques. Refer to Appendix A to identify the modeling technique applied
for each weapon.
a. Kinetic Weapon Modeling
Each kinetic weapon assigned the probabilities of 1.0, 0.5, and 0, to the
minimum effective range, maximum effective range, and maximum weapon range,
respectively. The maximum effective range is the “the distance from a weapon system at
which a 50 percent probability of target hit is expected.”68 From this definition, the
scenario assumes the other two hit probabilities, facilitating the graphing function that
yields the probability of hit dependent upon each weapon system’s range to target.
Rather then formulating a piecewise linear regression connecting each of the weapon’s
three data points, a more exhaustive graphical smoothing spline maps the probability of
hit for each meter, starting at 0 meters, and increases to each maximum weapon range. A
smoothing spline is an excellent way to get an idea of the shape of the expected value of
the distribution of y across x. A spline may vary in smoothness (or flexibility) according
to a user-defined lambda, a tuning parameter within the spline formula.69 For
consistence, the scenario assumes a very stiff lambda equal to 1,000,000 for each kinetic
weapon modeled. Three data points per weapon system entered into a spline formula
provided by JMP IN software resulted in a smooth distribution of hit probability across
meters.
Since the distribution is a smooth approximation that best fits the three
initial data points, some fitted values annotated in Appendix A, section “Raw Spline
Data,” exceed the numerical probability limits of 1.0 and 0. Importing each string of
values into Excel and using a series of “if, then statements,” any value outside the limit
becomes 0 or 1.0. Nested inside are additional “if, then statements” ensuring that all
approximated values adhere to the original weapon minimum and maximum limits. For
example, the Guided Hellfire arms at the minimum effective range of 500 meters; it has a
68 “Operational Terms and Graphics” in FM 101-5-1, chap. 1, p. m.
69 JMP Start Statistics, A Guide to Statistics and Data Analysis using JMP and JMP IN Software,
Third Edition, (SAS Institute Inc. 2005) p. 245.

53
maximum effective range of 7000 meters, and a maximum launch range of 8000
meters.70 Refer to Appendix A, section “Raw Spline Data” to observer that the spline
technique estimated values starting at zero and continued past 8000, where as the “Spline
Look-up Table” used the series of nested “if, then statement” to replicate minimum
arming distances for modeling purposes within MANA. The following Excel coding
script is an example of the cell codes within the “Spline Look-up Table.”
=IF('Raw Spline Data'!$A5<500,0,IF('Raw Spline
Data'!$A5>8000,0,IF('Raw Spline Data'!$R5<0,0,IF('Raw Spline
Data'!$R5>=1,1,'Raw Spline Data'!$R5))))
Using the same lambda to estimate each weapon’s “best fit” did inflate
each weapon’s maximum effective range. However, the scenario assumes this point mute
since the inflation is identical for all kinetic weapon systems. An additional assumption
regarding the LOS kinetic energy weapons is that they cannot travel through walls to
engage targets. However, the Hellfire, APKWS, LOCAAS, SA-16 guided rockets, and
the AT-12 stabber do not track traditional ballistic trajectories. Since the model limits
ballistics to follow straight paths, the scenario does assume these munitions modeled as
kinetic energy systems, to travel through walls to engage targets. This modeling
assumption simulates their precision guidance characteristics.
b. Area Fire Weapon Modeling
The scenario models area fire weapons much simpler. Assumptions
include that all area fire weapons can fire through walls to engage a target, simulating the
“lobbing effect” of indirect fire. This assumption holds true for both traditional
munitions, as well as precision guided munitions modeled. A third weapon system added
to each squad simulates the difference in effects that the same projectile has against both
soft and hard targets. As noted earlier, the third weapon system truly replicates the
primary weapon system (Weapon 1) when fired against hardened targets.
70 Global Security.org, Hellfire, Getting the Most from a Lethal Weapon System, referenced 7 October
2005 on the World Wide Web at http://www.globalsecurity.org/military/library/news/1998/01/1helfire.pdf

54
The Carleton Function, Figure 9, where r is the blast radius and b is a
coefficient identifying the lethality to the target, determines the probability of hit for each
area fire weapon. For this model, p(hit) = p(kill).
Figure 9. Carleton Function71
The blast (shot) radius in Appendix A, section “Weapon Specifications,” is the maximum
effective range for each projectile. The maximum blast radius has p(kill) = 0.5 when
applying an appropriate b coefficient for each light (soft) target noted in Table 12. The
model assumes a direct hit with a p(kill) = 1, and that the same weapon system has half
the effects on heavy (hardened) targets at the maximum blast radius. Selecting an
appropriate b coefficient models these assumptions and provides various p(hit) values for
different targets located with the corresponding blast radii annotated in Table 12.
Platform Target Type b
NLOS M real world range 0 20 40 60
MANA units 0 4 8 12
light target 51 1 0.925988 0.735228 0.500553
heavy target 36 1 0.856997 0.539408 0.249352
NLOS C/LS real world range 0 16.66667 33.33333 50
MANA units 0 3 6 10
light target 43 1 0.927636 0.740476 0.508627
heavy target 30 1 0.856997 0.539408 0.249352
guided xm36 real world range 0 5 10 15
MANA units 0 1 2 3
light target 13 1 0.928705 0.743893 0.513924
heavy target 9 1 0.856997 0.539408 0.249352
guided 82mm real world range 0 5 10 15
MANA units 0 1 2 3
light target 13 1 0.928705 0.743893 0.513924
heavy target 9 1 0.856997 0.539408 0.249352
Table 12. Modeled P(Kill) for Area Fire Weapons using the Carleton Function
71 Thomas Lucas, OA4655 Combat Modeling, Naval Postgraduate School, lecture presentation:
Entity-level Attrition: Some Phit and Pkill Algorithms.
2
2
-
2
p(hit) =
r
b
e
⎛ ⎞
⎜ ⎟
⎜ ⎟
⎜ ⎟
⎜ ⎟
⎝ ⎠

55
7. Armor and Concealment
Weighted values of each system’s platform capabilities models both the squad’s
armor and concealment MANA values. There existed a need to link FCS platform
defensive capabilities together in order to model each squad’s armor and concealment.
The Armor Thickness is a weighted average of possible capabilities classified
within each category described by the FCS UA Design Concept Baseline:72 Ballistic
protection, active measures, passive measures, threat warning receivers, countermine
abilities, and additional body armor. Refer to Appendix A, section “Armor and
Concealment,” to observe each of the possible capabilities within each category.
Summing the capabilities of each platform and dividing by the total number of
capabilities yields an average numerical value associated per squad. Seventy-five percent
of each averaged numerical value is the final weighted value defined in MANA. The
weighted value compliments the penetration value of each modeled weapon system. For
example, the value 75 annotates the armor value for an MCS vehicle. As such, only
weapons modeled with penetration values of 75, or greater, can kill the MCS. A close
look at the scenario reviews that an AK-M rifle cannot kill the MCS, whereas the AT-
Stabber can. The scenario assumes the Red Forces to have similar capabilities among
similar platforms in order to obtain a robust scenario.
Caveats to the algorithm in place include the author’s decision to model the
NLOS Cannon and Launch systems, CAS, and Apache squads to all have an armor value
of 100. A value of 100 makes each of these squads invincible to any other weapon
system. This simulates the CAS and Apache’s flying at altitudes greater then the SA-16
missile can engage. This also simulates the NLOS systems’ positions at greater distances
then actually portrayed on the scenario map. Model limitations dictate current positions
of the NLOS systems.
The squad concealment rate represents the signature management capability of
each platform. Each platform has a level 0, 1 or 2 signature management capability as

56
defined by the FCS UA Design Concept Baseline.73 In addition, the author included a
binary value, 0 or 1, to represent if there exists a human in the loop decision to position
the platform, or squad, in a concealed manner, rather than exposed in the open.
Multiplying 10 to the sum of each row in the Concealment table, Appendix A, section
“Armor and Concealment,” yields the MANA input value for each squad.
Red Force squads assume similar capabilities to maintain a robust scenario. In
addition, an x in the last row of the table identifies the author’s assumption to model the
squad with a different concealment rate. This serves for two reasons. First, it speeds up
computer run time by disabling enemy squad’s acquisition of air and NLOS assets on
their SA map, since these squads are invincible. Second, it provides the sniper and UAVs
greater concealment to represent real world occurrences.
D. MODEL LIMITATIONS
MANA version 3.0.39 presented several unique challenges to work around, or to
simply accept as limitations. This research uncovered a “bug” which prompted an
accelerated distribution of version 3.1.1 from New Zealand’s Defense Technology
Agency. The “bug” allows the agent the ability to engage targets through walls with the
use of their non-precision modeled kinetic energy weapons. This only occurred if the
agent acquired a target thru their inorganic situational awareness map. However, even a
direct hit, failed to kill the target. In essence, the “bug” lowered the agent’s ammunition
count, without posing harm to the target. However, this reflects what may occur in real
battles. A soldier may request a second soldier among their squad to provide suppressive
fires towards a particular building. The purpose of these fires may be to cover the first
soldier’s movement to better position him to engage a target. It is in this case that the
target is not harmed by the suppressive fires provided by the second soldier.
Ironically, the scenario settings specific to this research caused the newer version
of MANA to execute with a slower computer run time. As such, the author accepted this
“bug,” and continued with version 3.0.39 declaring the “bug” as a simulation providing
suppressive fires. Observing the simulation shows that suppressive fires do not harm

57
each target. However, each agent soon repositioned himself and the detection of the
targets drifted from their inorganic situational awareness map to their personal agent
situation awareness map. Once this occurred, the agent’s weapons killed each target.
All modeled UAVs encompass a 360-degree sweep width around their platform
even with the careful modeling considerations outlined earlier. This limitation in MANA
gives the UAVs an increased ability to detect other agents, where as in real life, their
sweep width only protrudes in one direction from the nose of the UAV. This limitation
was mitigated by bounding the maximum sensor range for each UAV class where the
p(det) approached the value one, as annotated in the predetermined table values
converting real world metrics to MANA units in Appendix A, section “Sense and
Detect.”
Modeling Hellfire, APKWS, LOCAAS, SA-16 guided rockets, and the AT-12
stabber, as kinetic energy weapons allows each to travel through walls with desired
effects upon the target. The reader should not confuse this technique used with the “bug”
discussed above. The author modeled these weapons as kinetic energy instead of area
fire weapons because all agents within a squad fire an area fire weapon simultaneously at
the same target, which would have resulted in an additional waste of precision guided
munitions all targeted upon the same object. The downfall is that each of these precision
munitions kinetically modeled incur a p(kill) = 1 for the entire blast radius, which is not
necessarily representative of real life. This limitation is mitigated by only firing precision
munitions against targets having threat level values within the boundary limits annotated
in Appendix A, section, “Model Unit Summary.” This simulates only firing precision
guided munitions against intended targets as authorized by a maneuver commander on
the battlefield with the specific intent to destroy (not neutralize or suppress) each target.
As noted earlier, the author scaled down each platform sensor range to increase
the simulation run time. The same holds true for each maximum range modeled as a
kinetic energy weapon. Appendix A, section “Weapon Characteristics,” provides
converted valued needed for input out to 500 grids, or the entire battlefield length of 2.6
kilometers. The author experienced an agonizing sluggish run time as each agent
searched the entire battlefield for targets. Shortening the maximum range to 96 grids

58
(500 meters) for each kinetic energy modeled weapon improved the simulation run time
without significantly changing the results.
The last major workaround built within the scenario included two inactive and
invisible “ghost” blue-dismounted squads with prepositioned locations on the battlefield.
Once the Blue Force ICV drove within the specified distance of 20 grids (approximately
100 meters on the ground) the “ghost” agents changed states into active visible blue force
dismounts. The downfall is that within one time step (equal to one second) the dismounts
obtained a position equivalent to 100 meters on the ground. Again, the author judged this
as acceptable for modeling purposes as it replicates the quick dispersion of infantrymen
in securing a perimeter. In addition, this too had little, if any, consequences on the
results.

59
IV. DESIGN METHODOLOGY
"NOUS SOUTIENDRONS"
(We will support)
42nd Field Artillery Brigade
This chapter outlines the design of experiment (DOE) which supports, and
bridges, the model development to the data analysis. Factors applied to help answer
thesis questions are included within the DOE. This chapter also describes each measure
of effectiveness (MOE) chosen to scope and quantify the analysis conclusions based upon
the DOE. A brief mention of the tools and techniques supporting the UAV exploration
follows at the last part of the chapter.
A. DESIGN OF EXPERIMENT
An effective design of experiment (DOE) supports the simulation model that
provides the data output for analysts to perform supporting work in the decision-making
process. As mentioned earlier, and as a product of Project Albert, Data Farming
provides a method to grow an abundance of data points for further exploration. The
initial DOE chosen to support this analysis was a Nearly-Orthogonal Latin Hypercube
(NOLH). The NOLH design efficiently searches the high-dimensional input space
defined by an intricate response surface. The NOLH has the following characteristics74:
• Approximate orthogonality of all input factors
• A collection of experimental cases representative of the subset of points in the
hypercube of explanatory variables (space filling)
• Ability to examine 20, or more, variables efficiently
• The flexibility to analyze and estimate multiple effects, interactions and
thresholds
• Requires minimal a priori assumptions on the response
74 Cioppa, Thomas M., Efficient Nearly Orthogonal and Space-Filling Experimental Designs for
High-Dimensional Complex Models, (PhD. Dissertation, Operations Research Department, Naval
Postgraduate School, Monterey, CA), 2002.

60
• Easy design generation
• An ability to gracefully handle premature experiment termination
Refer to Cioppa’s dissertation for additional information regarding a NOLH.
Specific to the final study, a crossed robust NOLH DOE with 20 nearly
uncorrelated input factors yielded 258 design points and paved the way towards the data
analysis. The reader may appreciate the following example identifying one benefit for
choosing such a design. A simple grid design consisting of 20 factors observed at only
two levels each, requires 220
(or 1,048,576) design points. Design points and data runs
are synonyms. If each run lasted only one computer minute, then it would still take 1.99
CPU years to finish running a single replication of the entire full design. Under the same
conditions, 258 design points takes only 4.3 hours using a single computer.
A crossed design captures the single NOLH, with 129 design points, stacked on
top of another NOLH with an additional 129 design points, while varying only one factor
different between the two stacks. The remaining factors and each of their levels maintain
the same values. A robust design captures both controllable and uncontrollable factors.
Uncontrolled factors are synonymous with noise factors. This better reflects real world
occurrences since it captures both controlled and uncontrolled situational entities.
1. Design Factors
Several assumptions mentioned within the Model Development chapter of this
thesis double as design factors. Since the FCS is a futuristic entity with some unknowns,
each factor selected for the DOE supports a modeling assumption or addresses a thesis
question. Selection of both controlled and noise factors ensured evaluating a robust
design. Each controlled factor specifies UAV values, and each noise factor portrays
uncontrolled elements such as environmental conditions, and enemy force sizes. Table
13 portrays the 20 nearly uncorrelated factors chosen for this design, respective levels,
and factor explanations.
Factors numbered four and five outlined in Table 13 reveal the necessity for the
crossed design. For this thesis, one battalion level UAV cannot carry both Warrior and
APKWS missiles at the same time. The thesis explores the benefits of one missile type
against the other by attaching only one type of missile per UAV for 129 runs each.

61
Keeping the remaining factors the same and substituting the Warrior missiles for
APKWS missiles systematically, builds the crossed design and doubles the number of
design points (runs) to 258.
Factor
Number
Potential Decision
(Controlled) Factors
Applied to
each Squad
# in MANA Low Level High Level
1 Number of UAVs CL I per team 20,21,22,23 0 6
2
Number of UAVs CL II per
team 24,25,26,27 0 6
3 Number of UAVs CL III 28 0 16
4
Number of Hellfire missiles in
UAV Warrior 28 0 4
5
Number of APKWS missiles in
UAV CL III 28 0 8
6
Sensor range and P(det) UAV
CL I 20,21,22,23 0 2
7
CL II 24,25,26,27 0 2
8
CL III 28 0 2
9
Agents desire to go after
enemy UAV CL I and II
20,21,22,23,
24,25,26,27 0 20
10
Agents desire to go to next way
point UAV CL I and II
20,21,22,23,
24,25,26,27 0 20
11
enemy UAV CL III 28 0 20
12
point UAV CL III 28 0 20
13 UAV CL I flying speed 20,21,22,23 60 80
14 UAV CL II flying speed 24,25,26,27 80 100
15 UAV CL III flying speed 28, 80 140
Potential Noise
(Uncontrolled) Factors
16
Number of initial enemy high
pay off targets
1,2,3,6,10,
11 1 12
17
Map editor city cover and
concealment all 1% 100%
18
Map editor inside building
cover and concealment all 1% 100%
19
Communication Reliability due
to inclement weather 20-28 0.75 1
20 UAV Concealment 20-28 0 0.9
Density of obstacles and darkness within the urban
location
Density of walls or other obstacles and darkness within
the buildings
The UAV communication links to ground elements are
greatly hindered in inclement weather such as rain
UAVs concealed by low cloud cover
The equivalent ground speed of this type of UAV
Initial number of enemy high pay-off targets
Tactical flight pattern of the UAV to fly towards a
detected target
Tactical flight pattern of the UAV to fly upon its
intended path
The P(det) at a given sensor range for this type of UAV
Tactical flight pattern of the UAV to fly towards, and
circle (or possible) hover over a detected target
Tactical flight pattern of the UAV to fly upon its
intended path
Explenation: Appriviate titles are listed as the Decision
and Noise Factors for programing purposes
Number of CL I UAVs per each A, B, C, and D teams
Number of CL II UAVs per each A, B, C, and D teams
Number of battalion level UAVs (This includes Warrior
UAVs or CL III UAVs)
The number of precision guided missiles upon a
battalion level UAV
The number of precision guided missiles upon a
battalion level UAV
Table 13. Factor and Level Description for DOE
This next portion follows the example listed in the preceding paragraph regarding
the time saving benefit of the NOLH DOE. Applying these 20 factors to a full factorial
design, and evaluating incremented levels between the low and high level of each,
combined with a six minute computer runtime for each design point, results in 6.9E48
CPU years to complete one iteration of the whole design. The crossed NOLH DOE
limited the number of design points, or runs, to again only 258. By lowering the number

62
of design points, and using a cluster set of 12 computers to share all the runs, the number
of computing hours lowered dramatically. The decreased total time allotted an additional
29 iterations per design point, enabling a “large sample” of 30 observations per point.
Even with 30 iterations per design point, the total number of computing hours cumulated
to only 2.68 CPU days per computer, resulting with 7740 rows and 102 columns of raw
data ready for analysis scoped by the measures of effectiveness. This process repeated
six times, evaluating different time-hacks within the battle. In total, the final production
runs consisted of 46,440 simulated battles.
2. Measures of Effectiveness (MOE)
Measures of effectiveness (MOEs) scope the analysis. An MOE is specific to the
success or failure of the military mission. While the thesis concentrates on UAVs, recall
that the UAV, and other FCS platforms, are only supporters of combat soldiers. One of
the Army’s mottos, “Mission first, people always” helped narrow the focus of the MOEs
for this thesis.
Recall that the CAB’s mission is to secure the urban area, OBJ Dallas. Though
the UAVs, and precision munitions platforms are an intricate part of the mission
accomplishment, much of the FCS is robotic in nature, and the only way to effectively
secure the urban area is with the dismounted infantry. This suggests looking at ways to
measure mission accomplishment through the success or failure of the infantry. An 80%
survival proportion of the Blue Dismounts at their final waypoint at the end of a 2-hour
battle portrays seizing the objective for this analysis. The CAB’s ability to fire precision
munitions against Red Force High Pay-off Targets (HPTs) directly affects the ability of
the CAB to accomplish their mission. Scouting platforms, such as the UAVs, provide the
TA for the use of precision munitions. For this analysis, the HPTs are the Red Force
entities precluding the Blue Force in delivering infantry to the close fight, thus obscuring
the specific mission to secure the objective. The HPTs include the SA-16 agents trying to
destroy the Blue UAVs and other air assets. Other HPTs are the BMP-3, 82 mm mortars,
scouts, APC, and T72 platforms, who deliver firepower to the deep fight, intended to
minimize the CAB’s penetration and delivery of dismounts to the close fight. To
accomplish the mission, the Blue Force has a desire to preserve their High Value Targets
(HVTs).

63
In this model, the HVTs are the Blue Force platforms that if destroyed by the
enemy will fail to protect the dismounts prior to arriving to the close fight. This effect
ultimately causes deaths among the Dismounts and failure to their mission. People
always, reflects the sacred desire to minimize dismounted deaths, for without the
dismounted infantry, the Blue Force would never secure the urban area. TRAC-
Monterey approved the following MOEs,75 chosen for this analysis in this order of
importance:
• Proportion of Blue Dismounts (Infantry) survived
• Proportion of Red High Pay-off Targets (HPTs) killed
Note: For this thesis, the Blue Dismounts (Infantry) only refer to those soldiers
who dismount from an ICV with the specific mission to secure the urban objective while
on foot. The ICV driver, who remains inside the ICV, as well as other soldiers who
remain inside other platforms such as an MCS, are not included in the calculations as
measured by the first MOE.
B. TOOLS AND TECHNIQUES
Visual observation of the MANA model provides a certain degree of value;
however, the purpose of MANA is essentially to “explore the greatest range of possible
outcomes with the least set-up time.”76 This section describes the tools and techniques
used to complement MANA’s quick build up approach and to create a valuable DOE
resulting in a quick, vast, and effective data analysis.
1. DOE Software Tools
The tools bridging MANA to the analysis include spreadsheet modeling with
Excel; Tiller©; XML; and Ruby scripting. As described in the Model Development
chapter of this thesis, the author maintains that spreadsheet modeling provides an
organized method to perform the thought process, while simultaneously cataloging
important modeling parameters.
75 Jeffrey Schamburg, LTC, Director, TRADOC Analysis Center – Monterey, Naval Postgraduate
School, Monterey California.
76 Galligan, p. 2.

64
a. Spreadsheet Modeling with Excel
Appendix B, section “DOE Spreadsheet Modeling” outlines the crossed
NOLH DOE. There exist three spreadsheet models. The first is the factor description
and is similar to that of Table 13. It outlines both the controlled and noise factors
creating the robust design. The second spreadsheet is a NOLH coded spreadsheet for 17-
22 factors detailing the factor levels used at each of the 129 design points.77 The third
spreadsheet is a design file and looks very similar to the second. This file adds an
additional nine correlated factors. These are correlated to each of the UAV P(det)
factors. The correlation represents the modeled monotonic increase in the P(det) incurred
at extended ranges, rather then just studying a single “cookie-cutter” sensor range. The
design file incorporates the final crossed NOLH DOE with 258 design points. The
process dovetails both the design file and the Ruby scripting procedures annotated in the
following paragraphs.
b. XML
Though MANA offers an easily viewed GUI to input data values, analysts
may also build MANA scenarios and edit them using the Extensible Markup Language
(XML), as all MANA databases are stored and transmitted in XML. XML offers a
simple and very flexible text format device derived from SGML (ISO 8879).
Technicians originally designed SGML to meet the challenges of large-scale electronic
publishing; XML also plays an increasingly important role in the exchange of a wide
variety of data on the internet.78 Storing scenarios in XML permits the analyst to
transmit scenario files quite rapidly over the internet to perform Data Farming
techniques. This process occurs with agencies such as the Maui High Performance
Computing Center (MHPCC) and enables thousands of design points to run over a
networked cluster of computers in a short amount of time.
c. Tiller©
The Tiller, Version 0.7.0.0, Copyright 2004 Referentia Systems
Incorporated, is a product developed in support of Project Albert and the Marine Corps
77 NOLH 17-22 Factors, coded by Professor Susan Sanchez, Naval Postgraduate School, Monterey,
California.
78 W3C, Extensible Markup Language, referenced 18 October 2005 from the World Wide Web at
http://www.w3.org/XML/

65
Warfighting Laboratory. Its primary purpose is to prepare model XML scenarios for
Data Farming. It provides DOE options such as the Random Latin Hypercube, coded by
Professor Paul Sanchez, Naval Postgraduate School, and a Nearly Orthogonal Latin
Hypercube, coded by Professor Susan Sanchez, Naval Postgraduate School. The final
output of the Tiller is a usable study.xml file containing the chosen DOE for running at
any computer cluster facility.
The Tiller application may be used alone to process the DOE, or as
performed in this thesis, may be used in conjunction with an object-oriented
programming language, such as Ruby, to modify the XML. XML modifications lockstep
the additional nine correlated factors within this design. In addition, it quickly links the
multiple squads depicting the same factor values as annotated from the design. Though
the Tiller is useful, the author found the application rather lengthy when applying all 20
factors, at each level, for each squad, and for each set of pre-analysis DOE iterations
performed. Instead, the author used the Tiller to build a skeleton study.xml file once, and
then performed further XML manipulation solely with the rapid process of Ruby
Scripting. Appendix B, section “Tiller,” outlines the Tiller GUI.
d. Ruby Code and Scripting
Ruby is a reflective, object-oriented programming language. It combines
syntax inspired by Ada and Perl with Smalltalk-like object-oriented features, and also
shares some features with Python, Lisp, Dylan and CLU. Ruby is a single-pass
interpreted language. Programmers describe Ruby as behaving intuitively, or as the
programmer assumes it should, not as expected by the computer itself.79
Refer to Appendix B, section “Ruby Scripting,” to observe the Ruby code
and scripting process written by Paul Sanchez that modified the skeleton Tiller study.xml
file for all DOE iterations performed.
2. Analysis Software Tools (JMP Statistical Discovery Software TM
)
JMP Statistical Discovery Software™ contains the software features used for the
Data Analysis portion of this thesis. The Data Analysis is included in the next chapter of
this thesis.
79 Wikipedia.org, Ruby Programming Language, referenced 18 October 2005 on the World Wide Web
at http://en.wikipedia.org/wiki/Ruby_programming_language

66
The author chose JMP as the tool to support the majority of the Data Analysis
because JMP provides interactive graphical and desktop statistics. JMP excels at helping
analysts uncover relationships and outliers within the data. This unveils valuable
discoveries, unleashes surprises, and supports better decision-making. It joins statistics
with graphics, and the flexibility to see the data from all angles to discover these
relationships and outliers.80
3. Analysis Techniques
Most large databases yield the flexibility to perform a wide array of data analysis
techniques. Though this analysis applies statistical tests, the core analysis focuses
primarily on three techniques: Graphical Analysis, Multiple Regression, and
Classification and Regression Trees.
a. Graphical Analysis
Graphical analysis provides a visual method to sift and explore through
data sets to find unexpected relationships. Statistical experts describe exploratory
analysis as data-driven hypothesis generation in search of structures that may indicate
deeper relationships between cases or variables.81 The output graphs from this analysis
will assist military decision makers by providing UAV insights without requiring the
decision maker to read the entire thesis.
b. Classification and Regression Trees (CART)
The CART (Classification and Regression Trees) algorithm is a widely
used statistical procedure for producing classification and regression models with a tree-
based structure. The principle behind building tree models is to identify significant
factors. This is done by partitioning the space spanned by the factors to minimize the
score of variance (or impurity) of response data at each branch node. Depending on the
particular score chosen, high purity occurs when the majority of points in each cell of the
partition are similar. This is a recursive process and repeats as many times as necessary
so that each end branch defines a separate node.82 83 The regression tree yields a
80 JMP, The Statistical Discovery Software, referenced 18 October 2005 on the World Wide Web at
http://www.jmp.com/product/jmp5_brochure.pdf
81 Hand, David, Heikki Mannila, and Padhraic Smyth, Principles of Data Mining, (MIT Press,
Cambridge, Massachusetts, 2001), p. 53.
82 Montgomery, Douglas, Elizabeth Peck, and Geoffrey Vining, Introduction to Linear Regression
Analysis, Third Edition, (John Wiley and Sons, Inc, 2001), p. 516.

67
continuous output. Classification trees, however, are the product of a discrete categorical
output based on a hierarchy of univariate binary decisions.84 The CART algorithm will
classify significant UAV factors into classes complimented by further regression
analysis.
c. Multiple Regression
A general regression analysis is a statistical process that investigates the
relationship between two or more variables (factors) related in a nondeterministic
fashion. Regression itself means coming or going back. The objective in multiple
regression is to build a probabilistic model that relates a dependent variable y to more
than one independent or predictor variables. Then the predicted values of each variable
are “pulled back in” towards the mean.85 The actual y values in a sample differ from the
predicted values. The errors or residuals denoted by e, are the differences between the
observed and predicted values, hopefully possessing a normal distribution with constant
variances.86 The regression analysis is practical for gaining insight on which predictor
variables (design factors) have the greatest significance towards the success of the FCS
CAB mission, as measured by the previously mentioned MOEs. Regression analysis is
also useful in identifying interactions between input variables.
83 Hand, pp. 145,343.
84 Hand, p. 147.
85 Devore, Jay L., Probability and Statistics for Engineering and the Sciences, Sixth Edition,
(Brooks/Cole, 2004), pp. 497,587.
86 Devore, p. 587.

68

69
V. DATA ANALYSIS
"CONJUNCTI STAMUS"
(United We Stand)
27th Field Artillery Regiment
This chapter contains the significant results of the data analysis drawn for
conclusions. Within the chapter, there are three sections: Data Compilation, Initial
Observations, and Closing Observations Related to Thesis Questions. Each section
paints the iterative process identifying significant findings. The closing observations
section outlines each thesis question, the measure of effectiveness addressing the
question, and the significant observations and findings pertaining to each question.
(Note: Dismounts and Infantrymen are synonymous throughout the analysis)
A. DATA COMPILATION
Receiving a multitude of data consisting of over 46640 data runs, with 102
variables each, begs the question, what now? This is raw data. Analysis of the raw data
could be an endless process. In addition, since MANA is stochastic in nature,
heteroscedasticity, or variance of the variability, can be quite prevalent within the raw
data. On one hand, ignoring it may bias the standard errors and p values. On the other
hand, its effect, though not detrimental, possibly weakens an analysis. In an attempt to
minimize, and apply better-suited models without losing core information, the aggregated
means of each of the replicated 30-design points builds a single measure of centrality
used for analysis procedures.87 The benefit of aggregating the means becomes lucid after
viewing Figure 12 in the next section.
For simplicity, this analysis concentrates on the multiple means, or averages, of
the outcomes. Though this technique delivers possibly an inflated R2
value (measuring
how well the regression line approximates real data points), it compliments the analyst’s
ability to identify otherwise unforeseen significant factors when Data Farming.
87 Lindquist, p. 59.

70
B. INITIAL OBSERVATIONS
Applying the robust crossed NOLH DOE outlined in Chapter IV of this thesis, the
initial analysis presented surprising results. The overwhelming flavor of the results
suggested that the noise (uncontrollable) factors included within the robust experimental
design were more significant than that of the actual number of UAVs assigned within the
CAB. The regression trees shown in Figures 10 and 11 identify enemy and terrain factors
as having greater significance than that of the number of UAVs assigned within the CAB.
Specifically, we observe the city and building density (modeled as cover and
concealment) and the initial number of enemy HPTs possessing higher significance.
There are 258 observations within each tree. Initial observations also show that the Blue
Force predominately achieves their objective while maintaining most of their Infantry
and annihilating most of the enemy HPTs. The trees show 0.9 as the mean for the
proportion of HPTs killed and 0.95 as the mean for the proportion of the surviving Blue
Dismounts. Notice in Figure 10, the first significant split occurring at the factor labeled
“City Cover and Concealment,” depicts a vast difference among the number of
observations and its respective mean—much more so than that of each subsequent
branch. Though the “number of CL I UAVs” factor does appear in Figure 10, suggesting
its significance, it does so only once and on the third split. In addition, numerous splits of
“Building Cover and Concealment” suggest possibly a non-linear relationship.
Figure 10 shows multiple paths that span out as branches of the tree. One path is
as follows. There are 258 total observations. Recall that each observation is an
aggregated mean of 30 replications. The overall mean is 0.90 as measured by the
proportion of HPTs killed. The first split occurs on the parameter City Cover and
Concealment. Of these observations, 236 occur when the parameter value is less then
0.92, indicating a slightly less dense city environment comprised of perhaps walls,
obstacles, and rubble. Among the 236 observations, only eight occur when the Building
Cover and Concealment parameter exceeds 0.97, indicating a denser environment within
the buildings. When the Building Cover and Concealment is less dense, as in this split at
0.97, then the Blue force performs better, as seen by a mean of 0.91 over 0.80 from the
other eight grouped observations. Finally, of the 228 observations, 198 occur when the
initial number (of each type) of HPTs at the beginning of the battle is equal to three or

71
more. From the initial robust DOE, we observe that the proportion of HPTs killed is
inversely proportional to the initial number of HPTs on the battlefield, suggesting that the
Blue Force is not as capable against a larger enemy, nor when fighting in a denser city.
Observe in Figure 10, the mean is highest among a smaller sample (only 30 observations)
in which the number (of each type) of enemy HPTs is less then three, and when the fight
occurs in a less dense city and building environment.
Figure 10. Regression Tree, with MOE: Proportion of HPT Killed
The next split would occur at this candidate because
it has the next largest Sum of Squares. At this next
split, there is also the largest delta of impurity among
parameters.
The “Candidates” are
the remaining
parameters where
additional splits
may occur.
Initial observations as measured by
the proportion of HPTs Killed identifies
only one controllable Blue Force
factor. The Blue Force has no control
over all the other “noise” factors
shown in this regression tree.
The mean
proportion
of HPTs Killed
increases as
there are less
initial HPTs at
the beginning of
the battle. This
suggests that
The Blue Force
does better
against a smaller
Enemy.
There are 236 observations when the
City Cover and Concealment
parameter is less then 0.92. When
this occurs, the mean proportion of HPTs
Killed increases by 1% from 0.90 to 0.91.
Among the 236 observations,
228 occur when the Building
Cover and Concealment para-
meter is less then 0.97, and
only 8 observations occur at an
equal or greater parameter value.
There are 258 total observations.
Each observation is an
aggregate of 30 replications.
The overall mean is 0.90
The next split would occur at this candidate because
it has the next largest Sum of Squares. At this next
split, there is also the largest delta of impurity among
parameters.
The “Candidates” are
the remaining
parameters where
additional splits
may occur.
Initial observations as measured by
the proportion of HPTs Killed identifies
only one controllable Blue Force
factor. The Blue Force has no control
over all the other “noise” factors
shown in this regression tree.
The mean
proportion
of HPTs Killed
increases as
there are less
initial HPTs at
the beginning of
the battle. This
suggests that
The Blue Force
does better
against a smaller
Enemy.
There are 236 observations when the
City Cover and Concealment
parameter is less then 0.92. When
this occurs, the mean proportion of HPTs
Killed increases by 1% from 0.90 to 0.91.
Among the 236 observations,
228 occur when the Building
Cover and Concealment para-
meter is less then 0.97, and
only 8 observations occur at an
equal or greater parameter value.
There are 258 total observations.
Each observation is an
aggregate of 30 replications.
The overall mean is 0.90

72
Figure 11. Regression Tree, with MOE: Proportion of Dismounts Survived
Furthermore, the initial analysis suggests that the Blue Force is overwhelming in
this scenario, and that changing the levels of each factor, to include the number of UAVs,
has little effect on the overall outcome. Again, the Blue Force predominately maintained
almost all of its infantry, while almost destroying the enemy’s entire supply of HPTs.
Figure 12 shows two histograms and their associated box plots, quantiles, and
moments information. The histogram (bar chart) represents a frequency distribution
predicting the number of observations occurring at each of the recorded proportions. The
proportion scales from zero to one. The box plot graphically represents the numerical
information listed in the quantiles and moments portions of the figure. Quantiles are the
points at which various percentages of the total sample are above or below, and moments
combine the individual data points to form descriptions of the entire data set.88 The
median is the horizontal line in the center location of the box. In both, the right edge of
the box is much closer to the median then is the left edge, indicating a very substantial
skew in the middle half of the data.89 The whiskers protruding from each box represent
the observations outside the quartiles, and the single dots represent possible outliers. The
furthest dots from the mean are then extreme outliers. The box itself represents the
interquartile range, and symbolizes observations ranging from the 25th
to the 75th
88 Sall, p. 118.
89 Devore, p. 41.
Blue Force’s only
controllable
factor, all others
are noise factors.
Blue Force’s only
controllable
factor, all others
are noise factors.

73
percentiles of the collected data. Refer to the key within Figure 12 for additional
information regarding the observations.
(Key)
Figure 12. Histograms of Initial Analysis with Robust DOE 90
Figure 12 contains 258 observations in each plot. Each histogram portrays a
skewed advantage towards Blue Dismounted Infantrymen surviving, and the annihilation
of Red HPTs. Each histogram illustrates two extreme outliers as measured by the
established MOEs. The histogram on the bottom portrays two observations reflecting an
unacceptable survival level of Blue Dismounts at only 60%. The histogram on the top
90 JMP IN, JMP 5.5.2 Help Command, SAS Institute Inc, 2004.

74
portrays two observations reflecting 65% of enemy HPTs killed, in relation to its mean at
90%. Recall that each of these data points is an aggregation of 30 original observations
averaged about each point. The 30 replications yield similar observations due to initial
battlefield settings determined from the experimental design. Therefore, each outlier is
not a single observation, but rather the mean of 30 observations. This identifies
something significant causing a possible spread of 30 undesirable outcomes affecting the
mission. As suggested previously, aggregating the means brought forth an insight
otherwise difficult to observe. These outliers implored the author to determine the initial
parameter settings that caused such undesirable mission results.
Examining the model and data output simultaneously identified a generality
among each of these specific outliers. It revealed that the initial parameter levels for
several of the noise factors were higher in each of these 30 replications then that of other
data runs. The most dominant of these noise factors contributing to mission detriment, as
measured by the MOEs, is a denser city environment coupled with a greater number of
initial enemy HPTs. In essence, a value closer to “1” for both the city and the building
cover and concealment parameters within the model yielded a denser city with perhaps
more obstacles that offered greater protection to the enemy from the Blue Force.
A fitted model developed through a stepwise regression and labeling each of the
MOEs as the y variable resulted with a summary of fit and parameter estimates
complimenting the regression tree analysis. Setting y as the proportion of Blue
Dismounts surviving, and examining all 20 factors, without interactions, resulted in a
fitted model with R2
equal to 0.42. This R2
suggests that the fit to the real data points is
lower then desired. However, Figure 13 maintains that the noise factors are more
significant then the others as measured by their high F-ratios. This measurement is with
respect to the proportion of Blue Infantrymen surviving. Appendix C, “Initial
Observations,” holds the entire model as determined by the multiple regression process.
The entire output, as well as similar results for the Red HPTs killed, is within this
appendix. The F-ratios portrayed from multiple regression also suggest the significance
of having armed battalion level UAVs. In addition, UAV tactical capabilities such as
speed, sensor range, and employment to fly towards the enemy targets are more
significant then that of the specific number of UAVs assigned within the CAB.

75
Figure 13. Tests of Main Effects (Stepwise Linear Regression Model Fit)
An interesting note is that performing a multiple regression with interactions
between factors raised the R2
to 0.80, suggesting an improved fitted model. With
interactions applied to the model, the Effect Test output, similar to Figure 13, is too large
for the main body of the thesis. The output for this model is located in Appendix C,
“Initial Observations.” This improved model was similar to the first in that the most
significant factors are those that are uncontrolled by the Blue Force.
Identifying this generality resulted in modifying the DOE, and setting the
parameter levels for the final observations within the data analysis. Changes to the DOE
included eliminating the various levels of each of the three noise factors already
discussed and setting their levels to stable values which provide a greater amount of
detriment to the CAB’s ability to complete its overall mission. Similar insight on some
of the other outliers portrayed in Figure 12 led the author to stabilize the two remaining
noise factors: Communication Reliability due to inclement weather and UAV
Concealment due to various cloud cover. The enemy, terrain, and weather predominately
outweighed any controlled factors within the DOE. Stabilizing the level of each the noise
factors at values that posed a stronger threat against the Blue Force, eliminated the
robustness of the design. Eliminating the robustness at this stage parallels the
Intelligence community’s process in providing the enemy’s most capable course of action
(COA) during a war-gaming design exercise. This action permitted the author to
concentrate the remaining analysis on controllable Blue Force factors. This follows suit
with the Operations community building friendly COAs. The observations obtained
through the initial regression analysis set each of the noise parameter levels for all the
Weapons
added to UAVs
are key to
to mission
success

76
remaining data runs. The stable levels for each noise parameter are as follows: 12
platforms for each type of HPT, 0.85 for the Map Editor City Cover and Concealment,
0.95 for Map Editor Building Cover and Concealment, 100 for Communication
Reliability, and 90 for UAV Concealment due to cloud cover.
The fitted model determined by the process of multiple regression identifies the
number of UAVs flying at each level. For both the initial and closing observations
sections of this chapter, the model is in the form:

77
C. CLOSING OBSERVATIONS RELATED TO THESIS QUESTIONS
The iterative process detailing the data analysis identified the need to stabilize all
the noise factors (minus communications) at levels stressing to the Blue Force.
Simultaneous efforts also raised an inquiry to question if different time hacks on the
battlefield provide any insight to answering the thesis-based questions.
1. Battlefield Time Hacks
Recall that the CASTFORM NEA 50.2 vignette is an 18-hour battle, and that this
research focuses only on a 2-hour window. Within the 2-hours, what time is most
critical? Stabilizing the noise levels, and performing six additional iterations of the battle
(running each simulation for the first 7.5, 15, 30, 60, 90, 120 minutes) shows that the
battle damage asymptotes as time increases. Figure 14 depicts the asymptotic curves
suggesting that the Blue Force kills most of the Red HPTs early in the fight—fifty
percent within the first 450 seconds (7.5 minutes) and sixty-five percent within the first
900 seconds (15 minutes) of the battle. A more important observation reveals a 5% loss
in Blue Dismounts within the first 15 minutes. The percentage increases until the end of
the first hour (3600 seconds) where it tapers off to 25% (75% strength of initial force).
These observations focused the remaining analysis toward the initial part of the battle.
Proportion of Red HPT Killed
Plotted over Time (seconds)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 900 1800 2700 3600 4500 5400 6300 7200
Time
Mean
Proportion
Killed
Proportion of Blue Dismount Infantrymen Survived
Plotted over Time (seconds)
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 900 1800 2700 3600 4500 5400 6300 7200
Time
Mean
Proportion
Alive
Figure 14. Graphical Analysis: Battlefield Time Hack without robust DOE
Note: The Blue Infantry normally do not dismount from their ICV until roughly
600 seconds into the battle. Recall that this simulation is a stochastic (not a time) driven
event. Therefore, the time varies occasionally as reflected in Figure 15. Figure 15 shows
a possible Blue Dismount killed by the 450th
second.

78
Figure 15. Histograms at 450 seconds (7.5 minutes)
Figure 16. Histograms at 900 seconds (15 minutes)

79
2. The Early Fight
Prior to analyzing the early fight, a t-test identifies the significance of the first 15
minutes in comparison to the entire 2-hour fight. The 15-minute mark chosen for the t-
test ensures the infantry’s delivery to the close fight (Figure 16). The null hypothesis is
that the means are equal when comparing the 15-minute and a 2-hour battle. Recall the
15-minute battle observations result from a DOE depicting a stronger enemy, where as
the 2-hour battle observations result from a DOE encompassing a variety of noise factor
levels.
The reader should not compare the two t-tests depicted in Figure 17 to each
other. Each graph and corresponding t-test represents different entities. Recall the
MOEs are the proportion of HPTs killed, and the proportion of Blue Dismounts that
survived. As such, each t-test speaks volumes on their own accord, as outlined in the
following paragraphs.
There exists a significant difference between the means when comparing the
proportion of Red HPTs killed. This significance is proved by the two-sided P-value
(Prob > |t|) equal to “0” as shown in the top half of Figure 17. A smaller P-value
suggests more contradiction to the null hypothesis,91 thus identifying a significant
difference between the means. Figure 17 shows expected results benefiting Blue’s fight
as measured by the first MOE, proportion of Red HPTs killed. Contrary, the same figure
also portrays what should be dreadful results to the military reader as measured by the
second MOE, proportion of Blue Dismounts survived.
There is not as much significant difference between the means when comparing
the proportion of surviving Blue Dismounts; however, the variances are clearly different.
The two-sided P-value is equal to 0.16, and the single sided Prob < t is equal to 0.92.
Therefore, there does not exist enough evidence to reject the null hypothesis, and for all
practical purposes, the means are the same. The author claims that this initial 5% loss of
infantry during the first 15 minutes of combat is detrimental to the mission. Recall that
this is the same 5% loss occurring at the end of a 2-hour fight with a more random
enemy, as posed by the robust DOE. This raises the author’s eyebrow and suggests that
91 Devore, p. 347.

80
military leaders should devise a system minimizing casualties within the first 15-minutes
of a fight when up against a strong enemy.
Figure 17. t-Test Results Between a 15-minute and 2-hour Battle

81
Left to the reader is the option to perform an additional t-test identifying similar
results when comparing the 7.5-minute mark to the entire 2-hour battle. The analysis
format for the remainder of this chapter mirrors the order of the thesis questions outlined
in Chapter I.
a. How many Platoon, Company, and Battalion level UAVs are
needed for the FCS to secure the urban environment?
Securing the urban area is binary, either the Blue Force did, or it did not.
TRAC-Monterey defines securing the urban environment for this scenario as the Blue
Force Dismounts reaching their final waypoint with 80% of their initial strength
remaining. Recall that the initial analysis of the robust DOE showed the mean proportion
of Blue Dismounts surviving at 0.95. This is different from the secondary analysis of the
Blue Dismounted strength when the changed DOE reflected a 25% loss at the end of the
same 2-hour duration. Since the iterative process drove the analysis to concentrate on the
initial part of the battle, the Blue Dismounts do not have enough time to reach their final
waypoint at either of the 7.5 or 15-minute time hacks. Therefore, the question asking if
the Dismounts reached their final waypoint is not addressed within the context of this
analysis. Instead, the question asks what needs to occur early in the fight in order to
minimize Infantry deaths (less then 20%) by the end of the 2-hour duration. The answer
is to minimize the HPTs prior to the Infantry’s arrival to their dismounted checkpoint.
The scatterplot in Figure 18 supports the claim in minimizing HPTs. The
covariance matrix, also in Figure 18, depicts how strong the two output MOEs relate to
one another. The proportion provides reason for the small values appearing within the
covariance matrix. According to 95% of observations (depicted by the oval shape), there
is a positive correlation (about 0.4) between the proportion of surviving Blue Infantry and
the proportion of Red HPTs killed. This positive correlation supports the observations
gleaned when viewing the simulation model. There is a lower survival rate of Blue
Dismounts when the Red Force has more HPTs alive on the battlefield.

82
Figure 18. Scatterplot Matrix (Positive Correlation Between HPTs and Dismounts)

83
The top portion of Figure 19 portrays the model fit with each observation
positioned along the line of fit within an ideal manner. The line of fit is the centered
straight-line protruding at a 45-degree angle. The line of fit shows where the actual
response and the predicted response are equal. The distance between the line of fit and
each observation is the residual, or error (e), for that point. The horizontal dashed line
identifies the mean 0.92 The adjusted R2
for this model is 0.89. The closer the adjusted
R2
is to 1.0 implies a better fitted model for its data. This adjusted R2
suggests a good
fitted model for this data.
The middle portion of Figure 19 is a diagnostic plot (a basic plot that
assesses the validity and usefulness of a model, also known as a residual plot). The
residual (e) is on the vertical axis, and the MOE is on the horizontal axis. The points
follow a random distribution about 0 implying constant variances, free of
heteroscedasticity (explained earlier in the chapter). This is observed from the absence of
any unusual or distinct pattern of points, thus providing a good visual assessment of
model effectiveness.93 The bottom portion of Figure 19 is another diagnostic plot useful
for visualizing the extent to which the residuals are normally distributed. The histogram
of the residuals appears to have a normal distribution. The appearance of a normal
distribution is reinforced by the diagonal straight line shown in the Normal Quantile Plot.
This kind of plot is also called a quantile-quantile plot, or Q-Q plot. The Q-Q plot also
shows Lilliefors confidence bounds, reference lines, and a probability scale.94
Refer to Appendix C, “Early Fight,” to observe the full model for Figure
19. Also refer to Table 14 to observe the most significant factors and interactions
yielding the greatest effects within this regression. Since the Sum of Squares for each are
all quite small (due to measuring proportions), and the model is quite large, the author
lists F-values greater then 25.0 in order to identify the significant factors. Table 14
outlines these factors. Refer to the appendix to review the remaining significant factors.
92 Sall, p. 314.
93 Devore, pp. 557-559.
94 JMP IN, JMP 5.5.2 Help Command, SAS Institute Inc, 2004.

84
Figure 19. Regression Model (Proportion of HPTs Killed at 450 seconds)

85
Single Factor F-ratio
# CL I 145.72
# CL II 25.70
# CL III 676.50
# Hellfire on Warrior 188.91
# APKWS on CL III 260.05
CL I and II Desire to Enemy 26.68
CL I and II Desire to next waypoint 82.61
Interaction of Factors
# CL III and Hellfire on Warrior 43.77
# CL III and # APKWS on CL III 133.60
CLI and II Desire to Enemy and CL I and II Desire to next waypoint 46.43
Quadratic
# CL I and # CL I 52.60
# APKWS on CLIII and # APKWS on CL III 37.93
Table 14. Significant Factors (Proportion of HPTs Killed at 450 seconds)
Table 14 (extracted from Appendix C) shows that the most significant
factor, as measured by the MOE proportion of HPTs killed, is the number of CL III
UAVs. Recall these are battalion level UAVs. The F-ratio for each UAV class identifies
their significance in the early fight to prepare the battlefield for the infantry’s arrival. In
addition, the interaction of battalion UAVs with APKWS weapons is also very valuable,
as measured by the same MOE. A partition of factors shown in the regression tree
(Figure 20) coupled with the parameter estimates outlined in the full model (found in
Appendix C, “Early Fight,”) helps identify the number of the UAVs needed to facilitate
the early destruction of Red HPTs. As found in the initial analysis of the robust DOE, we
find that the tactical employment of the UAVs is extremely important. Tactical
employment refers to the UAV operator’s decision to fly the UAV along the intended
flight path verses loitering over detected targets. This is seen from both the single factors
and the interaction of factors labeled in Table 14. Observe the significance of UAVs
flying towards the enemy verses towards their intended flight route, and their interaction.

86
Figure 20. Regression Tree (Proportion of HPTs Killed at 450 seconds)
The Regression Tree compliments the fitted regression model by showing
an increase in the purity of the model at the first split by identifying the number of
battalion level UAVs. The proximity of the means upon the first split is closer than
expected, but the means do clearly show the benefit of having more then 11 CL III UAVs
during the early fight. The larger means on the right side of the regression tree identify
this benefit. The second split, across both paths, shows that armed battalion level UAVs
are significant. The proximity of the means among each split suggests that perhaps about
three or four APKWS missiles will have the same increased affect on the battlefield. The
third split identifies the significance of platoon level UAVs. Since the means are rather
close, we can conclude that roughly three-platoon level UAVs among each team facilitate
the CAB’s mission. Recall from the scenario, that there are four tactical teams within the
CAB. Team A, B, C, and D.
Performing the same analysis on this MOE at 900 seconds resulted in a
stepwise fitted model with an adjusted R2
value at 0.82. This value is slightly lower then
the regression model developed at 450 seconds, but still quite high, and a good fit.
Figure 21 paints the predicted by actual plot of the model. Again, the observations fall
quite symmetric about the line of fit. The residual plot is distributed without any distinct
pattern, and reinforces the validity of this model. The histogram and the Q-Q plot
suggest a normal distribution of the residuals.

87

88
Table 15 identifies the significant factors of the regression model with an
F-ratio above 25.0. To observe the entire model, with the parameter and estimate effects,
refer to Appendix C, section “Early Fight.” The importance of the extracted F-ratios
portrayed in both Tables 14 and 15 lays in the similarity of significant factors. The
battalion level UAV remains as the single most important factor as measured by the
proportion of HPTs killed. Though not as significant, both company and platoon level
UAVs are important. Noticeable again, precision munitions attached to battalion level
UAVs are quite significant, as is the tactical employment of the UAVs. The interaction
suggests the need for the UAVs to follow their flight plan as well as sometimes
continuing in their scoping operations of detected enemy targets. Figure 22 again helps
determine the quantifiable number of UAVs needed to assist the Blue Force in obtaining
their mission to secure the urban area by depleting the Red HPTs.
# CL I 47.93
# CL II 54.41
# CL III 324.89
CLI and II Desire to Enemy and CL I and II Desire to next waypoint 27.06
Table 15. Significant Factors (Proportion of HPTs Killed at 900 seconds)
Figure 22. Regression Tree (Proportion of HPTs Killed at 900 seconds)

89
A consistency between Figures 20 and 22 shows that battalion level UAVs
bring the most punch to the battlefield in order to maximize the proportion of Red HPTs
killed. Though the means are relatively close, the right side of the regression tree does
again yield higher means in the destruction of HPTs when deploying more then 11 CL III
UAVs. The significance of having at least one platoon level UAV per team becomes
apparent again. Since the means are relatively close among each split, the CAB may
launch less then 11 CL III UAVs if deemed necessary after performing a cost benefit
analysis (outside the scope of this thesis). The presence of CL III UAVs appearing twice
in the regression tree suggests a non-linear fit, thus supporting the quadratic stepwise
regression model performed and displayed in Appendix C.
Though the CL III UAV seems to deliver the greatest punch to the battle
as measured by the regression trees and F-ratios, the military never depends on one asset
alone. On both the 450 and 900-second regression trees, notice the absence of CL II
UAVs. Table 14 possibly explains their absence by showing that even though the CL II
UAVs are significant as determined by their F-ratio, they are not as significant to the
model when applying this particular MOE. However, the parameter estimates for both
regression models does support the significance of CL II UAV presence as outlined in
Table 16 (extracted from Appendix C, section “The Early Fight.”)
Each estimate in Table 16 is positive, annotating a positive effect on
increasing the number of HPTs killed. An increase of one UAV within each class in turn
increases the proportion of HPTs killed by their respective estimates outlined in Table 16.
For example, given an increase of one CL III UAV from 11 to 12, provides almost a
0.5% increase in the proportion of Red HPTs killed within the first 450 seconds.
Parameter Estimate Parameter Estimate
# CL I UAVs 0.0055 # CL I UAVs 0.0032
# CL II UAVs 0.0023 # CL II UAVs 0.0034
# CL III UAVs 0.0045 # CL III UAVs 0.0032
450 Seconds 900 Seconds
Table 16. UAV Estimates (Proportion of HPTs Killed at 450 and 900 seconds)

90
Thus far, mostly one MOE, proportion of Red HPTs killed, has provided
insight to answering the thesis question. This next section performs the same analysis
techniques already described, but by applying the MOE proportion of Blue Dismounts
survived. This section, shortened for brevity, only examines the 900-second time-hack as
the stochastic simulation predominantly maintains a later arrival of Dismounts to the
close fight than that at the 450-second time-hack.
Figure 23, again portrays the regression fitted model with each
observation falling along the line of fit. The R2
in this model is 0.61, and the adjusted R2
for this model is slightly lower, only 0.53. This adjusted R2
is not as high as seen in the
past, but it is not laughable either. The model, significant factors, and parameter
estimates provide continued insight into our questions as measured with the MOE,
proportion of Blue Dismounts survived. Appendix C, “The Early Fight,” contains the
entire model.
The regression tree in Figure 24 compliments this entire model, proposing
that the CL I and II UAV traveling to the next waypoint is key to maintain a higher
survival proportion of Blue Dismounts. This suggests that the UAV operators play a
critical part in providing the eyes for the fight. Both the CL I, and CL II, UAV has
excellent sensor capabilities, that when flown routinely provides battlefield signature
patterns resulting in keeping Dismounts alive. The first split minimizing the impurity
occurs with a factor level of 15. This means on a scale between zero and 20, that there is
a stronger desire for the operators to fly the UAVs along the intended flight route. The
delta between the means about each split continues to be minimal. The mean for both (#
CL I UAV >=1) and (# CL I UAV < 3) is about 0.95, suggesting the significance in
having between one and three platoon size UAVs per team. This observation supports
the same number lower bound of CL I UAVs determined when applying the previous
MOE. The remaining splits identify tactical measures when deploying the UAVs as
having greater significance then other factors. These factors are not present within the
tree when looking at the MOE proportion of Blue Dismounts survived.

91
Figure 23. Regression Model (Proportion of Dismounts Survived at 900 seconds)
Figure 24. Regression Tree (Proportion of Dismounts Survived at 900 seconds)
The absence of the number of CL II and III UAVs within the tree in
Figure 24 is possibly explained by the impact of killing a large quantity of HPTs within
the first 450 seconds of the battle and prior to the arrival of the Dismounts. This
observation again supports the importance of preparing the battlefield for the Infantry’s

92
arrival. Thus, the CAB needs the CL III UAV for the deep fight and preparation of the
battlefield by destroying the HPTs. Once the Dismounts arrive, the CL I UAV is more
significant, as shown by Table 17, because it provides the local situational awareness
(over the next hill) to these Dismounts.
In addition, Table 17 extracted from the full regression model in Appendix
C, “The Early Fight,” has very few significant factors with F-ratios greater then 25.0.
The author listed the examples outlined in Table 17 because of their interesting values.
Supporting the corresponding regression tree, the most significant factor as measured by
its F-ratio, is the tactical employment of the CL I and II UAVs towards their next
waypoint. This supports the need of the smaller UAVs by the Dismounts to use them for
local situational awareness, covering as much territory as possible. Completely opposite
to this finding is the appearance in the small amount of significance of the CL I and II
UAVs aggressive flight pattern circling detected enemy targets. This suggests that
operators should fly both the CL I and II UAVs according to their flight pattern, even
after detecting an enemy target. There is little need for loitering, or hovering over an
established target with these UAV classes for the MOE proportion of Blue Dismounts
survived.
# CL I 26.61
# CL II 5.73
# CL III 3.20
CL I and II Desire to Enemy 0.04
# CL I and CL I and II Desire to next waypoint 22.28
# CL II and # APKWS on CL III 13.16
# CL III and # Hellfire on Warrior 24.68
# CL III and # APKWS on CL III 14.58
Quadratic
# CL II and # CL II 9.11
# CL III and # CL III 4.98
Table 17. Significant Factors (Proportion of Dismounts Survived at 900 seconds)
The parameter estimates outlined in Table 18, extracted from the full
model, identify the significance of adding one additional UAV per class at 900 seconds
into the battle. Adding an additional platoon UAV to each team increases the proportion

93
of surviving Blue Dismounts by almost one percent. Comparing this observation with the
interaction of factors outlined in Table 17, and the regression tree in Figure 24, suggests
the significance of the scouting abilities of the platoon level UAV. This is even stronger
as it continues along its flight pattern. Increasing the number of platoon UAVs from one
to three may save the proportion of Infantry lives by two percent.
Parameter Estimate
# CL I UAVs 0.9470
# CL II UAVs 0.0022
# CL III UAVs -0.0010
900 Seconds
Table 18. UAV Estimates (Proportion Dismounts Survived at 900 seconds)
The negative valued estimate corresponding to the number of battalion
level UAVs suggests that an increase in CL III UAVs may not preserve additional lives
once the battle reaches 900 seconds. This may call for a shift in prioritizing Blue Force
assets. There is a continued trend showing that success in the opening stages of the battle
paves the battlefield for the Infantry’s arrival. Once the battlefield is prepared, there is
less necessity for this battalion level asset.
b. How will armed battalion level UAVs enhance the FCS’s ability
to secure the urban environment?
Continued analysis, using two smaller models with four factors apiece
helped establish the effect of armed UAVs as measured by the two established MOEs.
Performing a stepwise regression and only selecting variables pertaining to CL III UAVs
and types of missiles associated with each resulted in a model that easily identifies
interactions among these specific variables. The actual versus predicted plot in Figure 25
portrays similar characteristics found in the larger model detailed in the previous section.
The R2
is smaller (0.51) in this model as expected since eliminating the majority of the
factors cannot add to the accuracy of the model.

94
Figure 25. Regression Model (Interaction Measured by HPTs)
This process leads to a more important fact outlined in Figure 26, that the
non-parallel lines clearly identifies significant interactions between the number of
battalion level UAVs and armed battalion level UAVs. There are two added variables
“mean UAV with Hellfire Missiles,” and the “mean proportion of payload.” These
additional columns (variables) added to the raw data are a measuring device used to assist
in Data Mining procedures. The bottom left cell of Figure 26 shows two lines labeled as
“0” and “1.” The “0” represents unarmed UAVs, and the “1” identifies armed UAVs.
Following the x-axis, from left to right, we observer that the mean proportion of HPTs
killed (y-axis) climbs much higher with an increased number of armed UAVs over that of
unarmed UAVs. The entirety of this smaller model appears in Appendix C,
“Interactions,” and supports the observations portrayed by each of the Figures and Tables
of the previous section.
In an interaction plot, the y-axes are the response, and each small plot
shows the effect of two factors on the response. One factor (associated
with the column of the matrix of plots) is on the x-axis. This factor’s effect
shows as the slope of the lines in the plot. The other factor becomes
multiple prediction profiles (lines) as it varies from low to high. This
factor shows its effect on the response as the vertical separation of the
profile lines. If there is an interaction, then the slopes are different for the
different profile lines.95
95 Sall, p. 421.

95
Figure 26. Interaction Plot of CL III UAVs Armed with Munitions
When studying the previous section’s Tables and Figures, notice the
slightly decreased F-ratio as well as the decreased parameter estimates of the CL I and III
UAVs when comparing the 450-second regression model to the 900-second regression
model (Refer to Tables 14, 15, and 16). Observing the simulation model reminds the
reader that this vignette does not model the entire battle, and that the vignette does not
simulate a lead up to all the military units arriving at their attack position. Rather, the
vignette opens with each asset already in its attack position. The scenario has a 2.6 by
2.6 square kilometer battlefield. Observing the scenario in the “play” mode reveals that
each of the CL III armed UAVs, detect, classify, and almost immediately fire upon Red
HPTs at the beginning of each run. Therefore, as the battle continues, the big punch
depleting the enemy force up front, possibly leaves less need for the CL III UAVs at the
end of the battle. The proportion of Red HPTs killed over time performs this
measurement. The similarities among Tables 14 and 15 identify a significant effect in
killing Red HPTs when deploying armed UAVs.
Note: The lines of a cell in the
interaction plot are dotted when
there is no corresponding
interaction term in the model.
Non-parallel lines indicate a
significant interaction between the
# of battalion level UAVs and armed
battalion level UAVs.
The bottom left cell of Figure 26 shows
two lines labeled as “0” and “1.” The “0”
represents unarmed UAVs, and
the “1” identifies armed UAVs.
Following the x-axis, from left to right,
we observer that the mean
proportion of HPTs killed (y-axis)
climbs much higher with an
increased number of armed UAVs
over that of unarmed UAVs.
The bottom left cell of Figure 26 shows
two lines labeled as “0” and “1.” The “0”
represents unarmed UAVs, and
the “1” identifies armed UAVs.
Following the x-axis, from left to right,
we observer that the mean
proportion of HPTs killed (y-axis)
climbs much higher with an
increased number of armed UAVs
over that of unarmed UAVs.

96
Recall that the analysis of surviving Blue Dismounts at 900 seconds into
the battle revealed less need for CL III UAVs at that particular time of the battle (Table
17). However, the significant interactions among “Hellfire missiles on Warrior” and “CL
III UAVs,” and that of “APKWS missiles on CL III UAVs” and “CL III UAVs” in the
full model suggests that providing armed UAVs under the CAB’s control proves
beneficial to the survival of Blue Dismounts. In addition, performing similar analysis,
applying a standard least squares analysis reinforces the interaction of specific factors as
outlined in Figure 27. The interactions identified within multiple cells of Figure 27
reveal that armed UAVs (denoted by “1”) help the mission. With respect to this MOE,
armed UAVs increase the survival proportion of Blue Dismounts (y-axis) and unarmed
UAVs lowers the number of Blue Dismounts surviving when reading each x-axis from
left to right.

97
Figure 27. Additional Interaction Plot
e. Is it better to arm Warrior UAVs with Hellfire missiles at the
CAB level, or to use APKWS 2.75 inch guided rockets with M151
HE warheads attached to the CL III UAVs?
Noticeably, armed UAVs appear significant to mission accomplishment as
measured by both MOEs. The question of which type of missile is better to use is not
quite as clear. What appears evident is that both types of missiles do materialize as
significant depending upon the application. The higher F-ratios in Table 14 identify the
APKWS missiles more significant then Hellfire missiles as measured by the proportion of
HPTs killed at 450 seconds. This holds true for all the single factors, interaction of these
factors, and their quadratic effects as well. Therefore, the APKWS missiles tend to
provide more benefit to the mission immediately upon the start of the battle. As the
battle moves on, Hellfire missiles become more significant. This is explained possibly

98
because APKWS is better to use in denser urban locations in order to minimize
unintentional destruction of nearby buildings. As the battle starts in this scenario, the
APKWS missiles engage HPTs masked by urban buildings and obstacles at a rapid rate.
As the battle continues, the HPTs are destroyed while the UAVs have fired their entire
payload. With Hellfire missiles, the UAVs fired at a steadier rate and at targets possibly
less hidden. Many of the same hidden HPTs in the urban environment were possibly
destroyed by other FCS platforms later in the scenario. The Hellfire missiles possibly
maintained their significance later in the battle due to their steady rate of fire toward the
remaining HPTs.
The regression tree in Figure 22 identifies Hellfire missiles as having
greater significance then that of the APKWS as measured by the proportion of HPTs
killed later in the battle at 900 seconds. Table 17 again identifies Hellfire missiles and
their interaction terms as having greater significance as measured by the proportion of
Blue Dismounts survived at 900 seconds. Looking at each of the interaction plots for
both MOEs, the proportion of payload is clearly significant for the battalion level UAVs.
A closer look at the percentiles of the means in the interaction plots for each appears
negligible when trying to determine a winner.

99
VI. CONCLUSIONS AND RECOMMENDATIONS FOR FUTURE
STUDY
"PER ANGUSTA AD AUGUSTA"
(Through Difficulties To Things Of Honor)
218TH FIELD ARTILLERY REGIMENT
This chapter contains a summary of conclusions and gained insight from the data
analysis. Following the summary of conclusions and gained insight section of this
chapter are some recommendations for future study.
A. SUMMARY OF CONCLUSIONS AND GAINED INSIGHT
The summary of conclusions and gained insight has two sections: Data Analysis
Conclusions and Modeling and DOE Methodology Findings. The division separates
aspects of the entire research that may have varying weighted values depending on the
reader.
1. Data Analysis Conclusions
The underlying questions of this research ask how many UAVs are needed, and
how will armed UAVs affect mission performance? Initial observations portrayed three
things:
• The enemy and terrain (two elements of METT-T) provide greater
significance to the mission outcome than the number and capability of UAVs
at any level.
• The tactical employment, and capabilities of each type of UAV, provides
greater significance to the CAB’s mission accomplishment than does the
actual numbers of UAVs at each level.
• The joined platform capabilities within the FCS is so robust, that eliminating
an entire platform category, such as all the UAVs from the battle space, has
little effect on the CAB’s ability to still maintain 95% of its Dismount
population while destroying 90% of the enemy HPTs.

100
Identifying outliers and modifying the parameter levels within the DOE to reflect
a very strong enemy steered the final analysis. This change to the parameter values
portrayed an enemy situation greater than four times the strength of the original
CASTFOREM Red Force order of battle.
Final analysis, employing a strong Red Force order of battle, and a dense urban
terrain environment showed that:
• 11 or more battalion level UAVs provide the FCS’s ability to act quickly and
decisively by bringing the biggest punch against the enemy as measured by
both the proportion of HPTs killed and the proportion of Blue Dismounts
Survived.
• The model portrays the CAB’s increased lethality against the HPTs, while
minimizing Blue Dismount deaths when adding precision munitions to CAB
UAV assets.
• The CAB needs the CL III UAV for the deep fight and preparation of the
battlefield by destroying the HPTs.
• Once the battlefield is prepared and the Dismounts arrive, then the CL I UAVs
are more significant because they provide the local situational awareness (over
the next hill) to these Dismounts.
• The APKWS missiles tend to provide more benefit to the mission
immediately upon the start of the battle.
• As the battle moves on, Hellfire missiles become more significant as
measured by the proportion of HPTs killed at 900 seconds.
• Hellfire missiles also seem to provide more application as measured by the
proportion of Blue Dismounts survived at 900 seconds. However, at 900
seconds there is already a large loss to the Red Force.
• Each tactical team benefits when deployed with between one and three
platoon level UAVs. The benefit of adding one platoon level UAV per team

101
increases the overall CAB survival proportion of Blue Dismounts by almost
one percent.
• Need at least one CL II UAV per tactical team. The exact number of CL II
UAVs is still unknown from this thesis.
• Lower class UAVs provide the eyes “over the next hill” for Dismounts.
Operators need to balance the tactical flight pattern in order to cover as much
ground as possible while minimally loitering over detected targets.
The quantitative values identifying the number of UAVs needed are for those
currently flying within a critical 2-hour window. A logistician still needs to determine
how many UAVs are needed in reserve due to maintenance schedules and recovery
assets.
The thesis and analysis determined an abundance of outcomes. The data analysis
responds quantifiably to the questions posed within this research. These answers afford
UAV insight to the operational analysis and the military community. However, this
section would be incomplete if the research failed to mention the insight drawn from both
the modeling and DOE methodologies. The ABS community benefits from the advance
techniques outlined within each of these methodologies.
2. Modeling and DOE Methodology Findings
Paramount to all modeling conclusions is the need to catalog ABM vignettes and
detailed methodologies outlining the parameter values used within each scenario. At the
October 2005 Military Operations Research Society (MORS) Workshop, Agent-Based
Models and Other Analytic Tools in Support of Stability Operations, the author
established the importance of such cataloging. Models, including MANA, are not widely
accepted beyond the research community. This is possibly because decision makers are
not aware of the vast scenarios already built by such models. An easily assessable library
consisting of MANA scenarios and parameter methodologies may assist in fostering this
needed acceptance.
Spreadsheet modeling offers a perfect way to capture modeling methodologies.
Spreadsheet modeling provides quick set up, flexibility, and an effortlessness cataloging
capability of each scaled parameter. The scaling is important since the operator defines

102
each MANA battlefield parameter. Again, cataloging efforts yield decision makers with
a history of scenarios, while offering analysts references to adopt similar aspects into
their own models. This also fosters the ability to build ABM vignettes in even a quicker
amount of time, without losing accuracy.
Accuracy and resolution are two different entities. The MANA run time can
become extremely slow if the operator defines the model with too much resolution. An
example of this is providing agents with sensor and weapon capabilities across the entire
terrain map. This modeling approach may not be of best interest to the modeler even if
the real life scaling permits it. The 2.6 by 2.6 square kilometer battle space of this
scenario is small enough for certain platforms to potentially range the entire playing field.
However, maximizing their sensor and weapon ranges slows the model run time almost
to a halt. The modeler should consider the terrain and environment prior to setting an
agent’s maximum range. In this scenario, certain line of sight platforms can sense and
engage targets past 2.6 kilometers in a desert. However, the mountains and MOUT
terrain of this scenario precludes most line of sight weapons to at most 500 meters or less.
Shortening the maximum weapon engagement range to only 500 meters (96 pixels)
decreased the run time to a desired speed for analysis purposes without losing accuracy.
The author found the Tiller application as an excellent tool to build a DOE with
minimal factors. The large number of factors combined with their correlated and
lockstep association to each multiple MANA squads having the same characteristics
called for additional programming using object-oriented programming. The author
recommends that the Project Albert staff adds the programming code used in this thesis to
the Tiller application. Professor Paul Sanchez, Naval Postgraduate School, is the author,
and point of contact for this code. This code will facilitate the Tiller application of larger
experimental designs.
While the author believes this as a beneficial exploration, discoveries must remain
within the context of its domain, agent-based simulation. Generally, ABS is an
exploratory tool yielding analysis based from low-resolution model output. The author
maintains that the modeled scenario is free of major flaws and modeling errors.

103
However, the conclusions drawn are from only one modeled vignette, and research
addressing additional vignettes will assist in the final development of the entire FCS.
B. RECOMMENDATIONS FOR FUTURE STUDY
As the research unfolded, a multitude of tangential and parallel topics came into
light for future study. One particular area of study is to compare and contrast the data
analysis output from this thesis to conclusions drawn from the original CASTFOREM
vignette at TRAC-WSMR. Though the CASTFOREM vignette did not model armed
UAVs, the 20 factors chosen within the DOE provide a multitude of data and analysis
output outlined by each of the full regression models in Appendix C and Chapter V. A
comparison of each simulation model about identical vignettes may bridge the process of
validating and verifying agent-based simulations (ABS) for future DOD use in planning
and analysis operational phases. In addition, future study of the same vignette modeled
in other agent-based models could provide insight to the ABS community as a whole.
This analysis drew from a CAB(-) asset. Due to limiting the number of agents
within the scenario, the author omitted the modeling of all Unmanned Ground Vehicles
(UGVs), certain command and control platforms, and all logistic platforms. The FCS is
very robotic in nature, and further study on each of the robotic platforms may provide
additional insight prior to fielding. Possibly the simplest of any follow-on study, may be
to perform an analysis of UGVs in lieu of UAVs by changing the parameters and
capabilities of all UAV modeled agents to represent that of UGVs in the MANA model.
Additionally, the existing modeled CAB(-) may be lifted out of this scenario and
placed in a completely new vignette representing a different tactical environment to see if
the same CAB is capable of performing a wide array of tactical missions. The procedure
is simple to perform by obtaining a digital version of the XML code from the author, or
by following the spreadsheet modeling techniques in Appendix A outlining all modeled
parameters. Slight changes may be necessary if the vignette scaling is different or to
change routes of march.
Concluded is the necessity to prepare the battlefield for the Infantry’s arrival.
This begs the question of what tactical deployment procedures and assets can better

104
prepare the MOUT battle space for the arrival of dismounts, such that their survival is
closer to 100 percent. Also concluded, is the benefit of battalion level CL III UAVs (or
Warrior UAVs under battalion control) carrying and deploying precision munitions. The
idea of armed UAVs changes the weight and payload balance requirements of each UAV.
An additional analysis of the balance between munitions, sensors, and fuel can establish
future building requirement of the FCS UAVs.
Similarly, there was a 5% loss of Blue Dismounts occurring at the end of a 2-hour
fight with a more random enemy, as posed by the robust DOE. There was the same 5%
loss within the first 15 minutes of a fight when posed against a stronger enemy. This
raises the author’s eyebrow and suggests that military leaders should devise a system
minimizing casualties within the initial stages of a fight when up against a strong enemy
situation.
Though at least one CL II UAVs per team is deemed significant in the
conclusions, there is an absence regarding the overall estimate of the number of company
level UAVs needed within a CAB. A nonlinear optimization model, using the parameter
estimates and the regression models in Appendix C may provide additional insight and
identify this exact number of company level UAVs. This nonlinear optimization problem
will also confirm the number of platoon and battalion level UAVs determined in this
thesis.
This research concluded that between one and three CL I, at least one CL II, and
11 CL III UAVs improve mission performance in this scenario. A cost-benefit-
estimation analysis on the regression models in Appendix C would help to identify the
trade-offs between applying different combinations of UAVs and other FCS platforms
within this and other operational settings.

105
APPENDIX A. MANA SPREADSHEET MODELING
The appendix provides the reader with the modeling methodology details used to
facilitate the model development process implemented within this simulation technique.
Each part of this appendix shows a snapshot of modeling spreadsheets built with Excel.
Spreadsheet modeling describes the approach implemented to transform real world data
into scaled MANA parameters. The spreadsheet modeling also offers a cataloging
approach to capture everything needed to replicate the scenario, or to adopt future
scenarios as well with minimal changes to the scaling process.

106
A. SCALING: CONFIGURE BATTLEFIELD SETTINGS
MAP SCALE
X Y JUSTIFICATION
500 500
square of
7070 4545 2600 meters Max
7330 4805 2600 meters 2600
Speedier - fog of war
Speedier - fog of war
1.00 (grids) prevents unecessary clutter of id locations
2 meters
LOS Mode Advanced
Real World Elevation Range: Min = 0 Max = 255
Terrain Effect Range 1 (grids) affects speed of model - highter = slower
5 meters
Move Selection
Best Move Precision
Move Precision 200
Multiple Agents in Cell
X Diagonal Motion Correction
X Navigate Obstacles
Squad Moves Together
X Going affects speed and Terrain affects LOS
Calculations
2 120 7200 1 7,200 60
1 second per 1 step
2600 2600 6760000
5.2 500 500 250000
2.8846154
general speed
conversion sec 1 steps 1 grids 500 km 2.6 Can't model CAS at 1000, so assume stationary
Assume Helo travels only at 60 knots for model
inf mech uav I uav II uav III cas helo
1.6 16 60 80 140 300 140
General speed conversions conversion
Dismounts 1.6 km 1 hour 1 min 1 sec 500 grids = 0.08547 grids 100 = 8.547008547 9
Ground Vehicles 16 km 1 hour 1 min 1 sec 500 grids = 0.854701 grids 100 = 85.47008547 85
UAV CL I 60 km 1 hour 1 min 1 sec 500 grids = 3.205128 grids 100 = 320.5128205 321
UAV CL II and Helo 80 km 1 hour 1 min 1 sec 500 grids = 4.273504 grids 100 = 427.3504274 427
UAV CL III 140 km 1 hour 1 min 1 sec 500 grids = 7.478632 grids 100 = 747.8632479 748
CAS 300 km 1 hour 1 min 1 sec 500 grids = 16.02564 grids 100 = 1602.564103 1000
mana input / 100
CONFIGURE BATTLEFIELD SETTINGS
Manage New Contact By:
ALGORITHM TAB - SAME FOR ALL UNITS
Steps in
Scenario
Minutes In
Scenario
Seconds
In
Scenario steps/sec
G
enral
M
ovem
ent
Settings
Agent Location
Underlying Contact ID
Stephen Algorithm
Hours In Scenario steps/min
meters per grid
square
Total Grid
Squares in
Grid
Squares
on X axis
Grid
Squares
on Y axis
m on X axis of
Terrain Map
m on Y
axis of
Terrain
Total m2
in Map

107
B. MODEL UNIT SUMMARY
Start
#
End
#
UNIT
TYPE
/
Squad
#
Type
Squads
#
agents
moving
parts
Squad
Class
Squad
Threat
Level
weapon
1
(light
for
Area
Fire)
Min
Threat
Level
Max
Threat
Level
Min
Threat
Level
Max
Threat
Level
weapon
2
Min
Threat
Level
Max
Threat
Level
total
entites
total
agents
1
-
1
Red
BMP-3
1
6
6
3
200
2A-42
/30
mm
130
160
100
100
800
Guided
2A-70M100mm
tube
firing
AT12
guided
stabber
100
100
800
1
to
2
2
-
2
Red
82
Mortors
1
6
6
2
100
82
mm
Mortar
100
all
but
100
30
800
100
30
800
ak
m/47
rifle
130160170140
3
800
1
to
2
3
-
3
Red
SA-16
Infantryman
1
5
5
1
3
Guided
SA-16
Surface
to
Air
Missle
200
all
but
200
900
900
1
to
1
4
-
4
Red
RPG-7
1
8
8
1
3
anti
tank
grenade
launcher
160
100
100
800
1
to
2
5
-
5
Red
AT-7
1
5
5
1
3
anti
tank
missle
160
100
100
800
1
to
1
6
-
6
Red
Scout
1
5
5
1
1
ak
m/47
rifle
3
99
1
to
1
7
-
7
Red
RPK-74
1
6
6
1
3
rpk
74
light
machine
gun
100
140
120
160
130
3
99
1
to
1
8
-
8
Red
AK-M
Infantryman
1
80
80
1
1
ak
m
/
47
rifle
100
140
120
160
130
3
99
1
to
1
9
-
9
Red
SVD
1
3
3
1
1
SVD
7.62
sniper
100
140
120
160
130
3
99
1
to
1
10
-
10
Red
APC
1
6
6
2
100
2A-42
/30
mm
130
100
100
800
rpk
74
light
machine
gun
140160130120
3
800
1
to
2
11
-
11
Red
T72
1
6
6
3
200
2A-46
/125mm
160
130
100
100
800
rpk
74
light
machine
gun
140160130120
3
800
1
to
2
12
-
12
Blue
NLOS
Mortor
Sec
1
4
4
140
100
120
mm
BLOS
guided
munition
1
2
3
30
800
2
3
1
30
800
xm307
25mm
1
2,3
3
200
1
to
2
13
-
13
Blue
NLOS
Cannon
Plt
1
2
2
170
na
155
mm
std
1
2
3
30
800
3
2
1
30
800
155
mm
guided
(heavy
targets
only)
1
30
800
1
to
3
14
-
14
Blue
NLOS
LS
Plt
1
2
2
170
na
payload
assit
mod
(PAM)
1
2
3
30
800
3
2
1
30
800
1
to
6
15
-
15
Blue
ICV
Platoon
1
6
6
120
100
MK44
30
mm
2
1
3
100
M240B
7.62mm
1
1
99
1
to
5
16
-
16
Blue
MCS
Platoon
1
6
6
160
200
Guided
xm36
120mm
1
2
3
30
800
3
2
1
30
800
xm307
25mm
1
2,3
3
200
1
to
3
17
-
17
Blue
ARV-A
1
6
6
130
100
MK44
30
mm
2
1
30
100
M240B
7.62mm
2
3
1
99
1
to
1
18
-
18
Blue
ARV-A(L)
1
6
6
130
100
xm307
25
mm
1
2
3
200
Javelin
Anti
Tank
Missle
3
1
2
100
200
1
to
1
19
-
19
Blue
ARV-RSTA
1
6
6
130
100
xm307
25
mm
1
2
3
200
1
to
1
20
-
23
Blue
UAV
CL
1
4
3
12
200
900
1
to
1
24
-
27
Blue
UAV
CL
2
4
3
12
200
900
1
to
1
28
-
28
Blue
UAV
CL
3
1
12
12
200
900
Guided
Hellfire
3
2
1
100
800
Guided
AKPWS
1
2
3
6
200
1
to
1
29
-
29
Blue
R&SV
1
3
3
150
100
xm307
25
mm
1
2
3
200
1
to
3
30
-
30
Blue
Infantryman
1
54
54
100
3
m16
1
3
1
99
1
to
3
31
-
31
Blue
MachineGunner
M240b
1
10
10
100
3
m240B
7.62mm
1
2
3
1
99
1
to
1
32
-
32
Blue
CAS
1
1
1
210
900
m230
/
30
mm
3
2
1
3
800
Guided
LOCAAS
3
2
1
100
800
1
to
48
33
-
33
Blue
Apache
1
2
2
210
900
m230
/
30
mm
3
2
1
3
800
Guided
Hellfire
3
2
1
100
800
1
to
3
totals
33
262
280
notes:
Aggregation
is
depedent
upon
platoon
sizes.
IE:
1
icon
of
a
NLOS
Mortar
Section
represents
2
real
world
Motor
Tubes
Weapon
3
is
a
subclass
of
Weapon
1.
Weapon
3
represents
a
different
type
of
projectile
fired
from
the
same
tube
of
weapon
1.
22
144
11
Priority
Target
Class
for
Weapon
1
Non
Target
Class
136
Aggregation:
1
icon(agent)
to
X
real
objects
Priority
Target
Class
for
Weapon
2
Non
Target
Class
Priority
Target
Class
for
Weapon
3
Non
Target
Class
Target
Classifications
HVY
Target
Projectile
Classifications
Squad
Classifications
Target
Classificatons
AGGREGATION
PLAYERS
WEAPON
1
Weapon
3
(Subclass
of
Weapon
1)
WEAPON
2

108
C. MOVEMENT RATES
MOVEMENT CALCULATOR FOR ALL GROUND VEHICLES
Base Movement Rate (kmph) 16 16000 (meters per hour) 2.75
tacticle
100%
increase
200%
increase max
100% 200% 400% 550%
Adjustment Factor tacticle 100% increase 200% increase max
Adjustment
Factor tacticle
100%
increase
200%
increase max
Unencumbered 1.00 267 533 1,067 1,467 Unencumbered 875 1,749 3,499 4,811
Light Combat Load 0.98 261 523 1,045 1,437 Light Combat Load 857 1,714 3,429 4,714
Full Combat Load 0.89 237 475 949 1,305 Full Combat Load 778 1,557 3,114 4,281
Heavy Load 0.78 208 416 832 1,144 Heavy Load 682 1,364 2,729 3,752
tacticle
100%
increase
200%
increase max tacticle
100%
increase
200%
increase max
Unencumbered 4.4 8.9 17.8 24.4 Unencumbered 14.6 29.2 58.3 80.2
Light Combat Load 4.4 8.7 17.4 24.0 Light Combat Load 14.3 28.6 57.1 78.6
Full Combat Load 4.0 7.9 15.8 21.8 Full Combat Load 13.0 25.9 51.9 71.4
Heavy Load 3.5 6.9 13.9 19.1 Heavy Load 11.4 22.7 45.5 62.5
tacticle
100%
increase
200%
increase max Dismounted Infantry
Unencumbered 0.9 1.7 3.4 4.7
Light Combat Load 0.8 1.7 3.4 4.6
Full Combat Load 0.8 1.5 3.0 4.2
Heavy Load 0.7 1.3 2.7 3.7
1.20
3.28 feet = 1 meter
Notes: Picked Restricted movement rates due to traveling through urban area
Scenario occurs at day in combat, and mounted vehicles have scensor devices that allow traveling at optimal speeds
Ground Vehicle Different
% of
Adjusted
Movement
Speed
MANA
Input
Speed
100% 1.20 120
10% 0.12 12
0% - 0
50% 0.60 60
60% 0.72 72
100% 1.20 120
150% 1.80 180 ROUND(DXX*10,1)*10
0% - 0
1% 0.01 1
Relative movement to tacticle speed
Default movement Rate
Run Start (if applied)
Taken Shot (for primary or secondary)
Shot At
(jugement call based on platforms
ability to fire at 0, 50%, 60% or full speed)
Reach Waypoint
Adjusted Speed = Target Zone (average rate)
Adapted From FM90-31 - Ch4
Armored/Mechanized Infantry Movement Rates: Ideal Terrain (grids per step)
Armored/Mechanized Movement Rates: Ideal Terrain (meters per min)
Armored/Mechanized Movement Rates: Ideal Terrain (meters per sec)
Armored/Mechanized Movment Rates: Ideal Terrain (feet per min)
Armored/Mechanized Movement Rates: Ideal Terrain (feet per sec)
MOVEMENT CALCULATOR FOR DISMOUNTS
Base Movement Rate (kmph) 1.6 1600 (meters per hour) 8.8
rounded
Walk Jog Run Sprint
100% 200% 400% 550%
Adjustment Factor Walk Jog Run Sprint
Adjustment
Factor Walk Jog Run Sprint
Unencumbered 1.00 27 53 107 147 Unencumbered 87 175 350 481
Light Combat Load 0.90 24 48 96 132 Light Combat Load 79 157 315 433
Full Combat Load 0.50 13 27 53 73 Full Combat Load 44 87 175 241
Heavy Load 0.30 8 16 32 44 Heavy Load 26 52 105 144
Walk Jog Run Sprint Walk Jog Run Sprint
Unencumbered 0.4 0.9 1.8 2.4 Unencumbered 1.5 2.9 5.8 8.0
Light Combat Load 0.4 0.8 1.6 2.2 Light Combat Load 1.3 2.6 5.2 7.2
Full Combat Load 0.2 0.4 0.9 1.2 Full Combat Load 0.7 1.5 2.9 4.0
Heavy Load 0.1 0.3 0.5 0.7 Heavy Load 0.4 0.9 1.7 2.4
Walk Jog Run Sprint Dismounted Infantry
Unencumbered 0.1 0.2 0.3 0.5
Light Combat Load 0.1 0.2 0.3 0.4
Full Combat Load 0.0 0.1 0.2 0.2
Heavy Load 0.0 0.1 0.1 0.1
0.09 AVERAGE(C22:D23)
3.28 feet = 1 meter
Notes: Picked Restricted movement rates due to traveling through urban area
Scenario occurs at day in combat, but assuming night speads because of enemy hide positions, and traveling in dark city allies
Dismounted Different
% of
Adjusted
Movement
Speed
MANA
Input
Speed
100% 0.09 9
0% - 0
100% 0.09 9
60% 0.05 5
0% - 0 ROUND(DXX*10,1)*10
100% 0.09 9
Default movement Rate Blue
Adjusted Speed = Target Zone (average rate)
Relative movement to walking speed
Default movement Rate Red
Adapted From FM90-31 - Ch4
Model Dismounted Infantry Movement Rates: Ideal Terrain (grids per step)
Dismounted Infantry Movement Rates: Ideal Terrain (meters per min)
Dismounted Infantry Movement Rates: Ideal Terrain (meters per sec)
Dismounted Infantry Movement Rates: Ideal Terrain (feet per min)
Dismounted Infantry Movement Rates: Ideal Terrain (feet per sec)
Refueled by Anyone
Taken Shot Blue
Taken Shot Red

109
D. SENSE AND DETECT
UAV Platforms
Intent: Replicate the Liklihood of Detection graph from TM 3-22-5-SW for each UAV classes I, II, and III
Integration of Unmanned Vehicles into Maritime Missions
TM 3-22-5-SW
Department of the Navy, Office of the Chief of Naval Operations
p 2-4
1 foot = 0.3048 meters
Predetermined Table Values Converting Real World Metrics to MANA Units
Meters Grids
UAV CL I flying at 500 ft 106.7 21 Meters 13.34 26.68 53.35 106.7 Grid 3 5 10 21
350 ft foot print with a 30 degree field of view flying at 500 ft P(det) 0.2 0.5 0.8 1 P(det) 0.2 0.5 0.8 1.0
Meters Grids
UAV CL II flying at 1000 ft 198.2 38 Meters 24.77 49.54 99.09 198.2 Grid 5 10 19 38
650 ft foot print with a 30 degree field of view flying at 1000 feet P(det) 0.2 0.5 0.8 1 P(det) 0.2 0.5 0.8 1.0
Meters Grids
CL III flying at 2500 ft 762.2 147 Meters 95.27 190.5 381.1 762.2 Grid 18 37 73 147
2500 ft foot print with a 45 degree field of view flying at 2500 ft P(det) 0.2 0.5 0.8 1 P(det) 0.2 0.5 0.8 1.0
Classify (MANA INPUT)
P(det) of UAV Class I Flying at 500 Ft
Using 30 Degree Field of Veiw With a 350
Ft Foot Print
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 20 40 60 80 100 120
P(det)
P(det) of UAV Class II Flying at 1000 Ft
650 Ft Foot Print
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 50 100 150 200 250
P
(det)
P(det) of UAV Class III Flying at 2500 Ft
2500 Ft Foot Print
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 200 400 600 800 1000
P
(det)

110
Ground and Other Air (non UAV) Platforms
Range
Meters Grids
Short 150 29 Meters 100 125 150 Grid 19 24 29
P(det) 0.9 0.8 0.7 P(det) 0.9 0.8 0.7
Meters Grids
Medium 250 48 Meters 150 200 250 Grid 29 38 48
P(det) 0.9 0.8 0.7 P(det) 0.9 0.8 0.7
Meters Grids
Long 500 96 Meters 300 400 500 Grid 58 77 96
P(det) 0.9 0.8 0.7 P(det) 0.9 0.8 0.7
Meters Grids
Short-Medium 200 38 Meters 150 175 200 Grid 29 34 38
P(det) 0.9 0.8 0.7 P(det) 0.9 0.8 0.7
Medium-Long 350 67 Meters 250 300 350 Grid 48 58 67
P(det) 0.9 0.8 0.7 P(det) 0.9 0.8 0.7
Extra Long 1300 250 Meters 700 900 1100 1300 Grid 135 173 212 250
P(det) 0.9 0.8 0.7 0.6 P(det) 0.9 0.8 0.7 0.6
short
medium
long
multi
function
ka
band
radar
AITR
acoustic
emmiter
mapping
remote
EO
IR
TD
CM
RADAR
Warning
Plum
Dect
Standoff
Chem
Det
SIGNINT
Combat
ID
mast
sensor
Red BMP-3 1 1 2.06667
Red 82 Mortors 1 1
Red SA-16 Infantryman 1 1
Red RPG-7 1 2
Red AT-7 1 2
Red Scout 1 1 3.06667
Red RPK-74 1 2
Red AK-M Infantryman 1 1
Red SVD 1 2
Red APC 1 1 1 1.13333
Red T72 1 1 1 2.13333
Blue NLOS Mortor Sec 1 1 1 1.13333
Blue NLOS Cannon Plt 1 1 1 1.13333
Blue NLOS LS Plt 1 1
Blue ICV Platoon 1 1 1 2.13333
Blue MCS Platoon 1 1 1 2.13333
Blue ARV-A 1 1 1 1 2.2
Blue ARV-A(L) 1 1 1 2.13333
Blue ARV-RSTA 1 1 1 1 1 1 2.33333
Blue UAV CL 1 1 1 1 1 3.2
Blue UAV CL 2 1 1 1 1 3.2
Blue UAV CL 3 1 1 1 1 1 1 1 1 1 1 3.6
Blue R&SV 1 1 1 1 1 1 3.33333
Blue Infantryman 1 1
Blue MachineGunner M240b 1 1
Blue CAS 1 1 1 3.13333
Blue Apache 1 1 1 3.13333
column 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
Sensor type based off of C4ISR
Adjusted
Average
Value
per
Squad
= 1
= 2
2<x<3
= 3
1<x<2
Numerical Value
<3

111
E. PERSONALITIES
Agent
SA
State
Inorganic
SA
Ranges
Squad
SA
ID
Name
of
Agent
Summary
Justification
Enemies
Combat
Enemy
Threat
1
EnemyThreat
2
EnemyThreat
3
Ideal
Enemy
En.
Class
Uninjured
Friends
Injured
Friends
Cluster
Neutrals
Next
Waypoint
Advance
Alt.
Waypoint
Easy
Going
Cover
Concealment
Line
Center
Enemy
Threat
1
EnemyThreat
2
EnemyThreat
3
Squad
Friends
Other
Friends
Neutrals
Unknowns
Enemy
Threat
1
EnemyThreat
2
EnemyThreat
3
Friends
Neutrals
Unknowns
Icon
Allegiance
Threat
Agent
Class
Movement
Speed
No
Hits
to
kill
Stealth
Armour
Thickness
Waypoint
Radius
Sensor
Class
Range
Sensor
Detect
Range
Fuel
Usage
Rate
Refuel
Trigger
Rate
Prob
Refuel
Enemy
Prob
Refuel
Friend
Prob
Refuel
Neutral
1-1
Red
BMP-3
State
1
Default
State
Start
state
and
Default
fallback
state
10
10
5
5
10
10
10
31
2
200
3
120
2
30
43
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
10
10
5
5
10
31
2
200
3
60
2
30
43
5
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
10
10
5
5
10
31
2
200
3
120
2
30
43
2
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
10
10
5
5
10
31
2
200
3
180
2
30
43
2
X
X
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
39
2
100
2
2
20
16
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
39
2
100
2
2
20
16
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
39
2
100
2
2
20
16
X
X
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
-10
1
1
38
2
3
1
8
1
20
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
1
1
38
2
3
1
5
1
20
X
X
no
selection
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
-10
-10
10
1
1
-10
-10
10
10
27
2
3
1
8
2
10
5
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
-10
-10
10
1
1
-10
-10
10
10
37
2
3
1
5
2
10
5
5
X
X
no
selection
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
10
1
1
10
10
37
2
3
1
8
1
10
5
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
10
1
1
10
10
37
2
3
1
5
1
10
5
5
X
X
no
selection
no
selection
no
selection
no
selection
no
selection
3-3
Red
82
Mortars Red
RPG-7 Red
AT-7
1-1 2-2 4-4 5-5
Red
SA-16
Infantryman
Red
BMP-3

112
1
Default
State
Start
state
and
Default
fallback
state
40
2
1
1
0
1
60
5
0
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
40
2
3
1
0
1
60
5
0
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
40
2
1
1
0
1
60
5
0
X
X
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
10
-10
-10
1
1
10
10
27
2
3
1
8
1
10
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
10
-10
-10
1
1
10
10
27
2
3
1
5
1
10
5
X
X
no
selection
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
20
10
10
30
30
30
10
10
10
26
2
1
1
8
1
10
11
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
30
30
5
5
5
5
5
5
26
2
1
1
5
1
10
11
X
X
no
selection
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
10
10
10
28
2
1
1
1
10
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
147
2
3
1
1
10
5
X
X
no
selection
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
5
10
10
10
10
10
10
10
10
10
30
2
100
2
120
2
20
43
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
5
10
10
10
10
10
10
10
10
10
30
2
100
2
0
2
20
43
5
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
5
10
10
10
10
10
10
10
10
10
30
2
100
2
60
2
20
43
5
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
10
10
10
10
10
10
10
10
10
10
30
2
100
2
180
2
20
43
5
X
X
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
10
20
5
5
10
10
10
10
10
10
32
2
200
3
120
2
20
38
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
10
20
5
5
10
10
10
10
10
10
32
2
200
3
0
2
20
38
5
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
10
20
5
5
10
10
10
10
10
10
32
2
200
3
60
2
20
38
5
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
10
20
5
5
10
10
10
10
10
10
32
2
200
3
180
2
20
38
5
X
X
no
selection
no
selection
no
selection
6-6 8-8
7-7
Red
AK-M
Infantryman
Red
Scout Red
RPK-74
11-11
Red
T72
9-9
Red
SVD
10-10
Red
APC

113
ID
Name
of
Agent
Description
Summary
time
in
state
Enemies
Combat
Enemy
Threat
1
EnemyThreat
2
EnemyThreat
3
Ideal
Enemy
En.
Class
Uninjured
Friends
Injured
Friends
Cluster
Neutrals
Next
Waypoint
Advance
Alt.
Waypoint
Easy
Going
Cover
Concealment
Line
Center
Enemy
Threat
1
EnemyThreat
2
EnemyThreat
3
Squad
Friends
Other
Friends
Neutrals
Unknowns
Enemy
Threat
1
EnemyThreat
2
EnemyThreat
3
Friends
Neutrals
Unknowns
Icon
Allegiance
Threat
Agent
Class
Movement
Speed
No
Hits
to
kill
Stealth
Armour
Thickness
Waypoint
Radius
Sensor
Class
Range
Sensor
Detect
Range
Fuel
Usage
Rate
Refuel
Trigger
Range
Prob
Refuel
Enemy
Prob
Refuel
Friend
Prob
Refuel
Neutral
1
Default
State
Start
state
and
Default
fallback
state
0
-10
20
10
10
14
1
3
140
120
2
20
75
1
X
X
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
100
-10
20
10
10
14
1
3
140
12
2
20
75
1
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
30
-10
20
10
10
14
1
3
140
0
2
20
75
1
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
6
-10
20
10
10
14
1
3
140
72
2
20
75
1
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
50
-10
20
10
10
14
1
3
140
180
2
20
75
1
X
X
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
7200
-10
10
10
10
10
10
-20
10
10
10
14
1
3
140
12
1
20
75
1
X
X
36
Run
Start
Squad
state
at
the
beginning
of
a
run
(can
be
used
as
delay)
50
-10
10
10
14
1
3
140
2
20
75
1
X
X
no
selection
0
0
no
selection
1
Default
State
Start
state
and
Default
fallback
state
0
13
1
3
170
0
100
100
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
15
13
1
3
170
0
100
100
X
X
no
selection
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
0
15
1
3
170
0
100
100
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
60
15
1
3
170
0
100
100
X
X
no
selection
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
0
-10
20
5
1
3
120
120
5
20
70
1
X
X
25
100
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
100
-10
20
10
10
5
1
3
120
12
5
20
70
1
X
X
25
100
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
6
-10
20
10
10
5
1
3
120
0
5
20
70
1
X
X
25
100
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
6
-10
20
10
10
5
1
3
120
72
5
20
70
1
X
X
25
100
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
50
-10
20
10
10
5
1
3
120
180
5
20
70
1
X
X
25
100
19
Injured
Agent
state
when
injured
(shot
at
and
hit)
7200
-10
20
10
10
5
1
3
120
12
3
20
70
1
X
X
25
100
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
7200
30
-10
10
10
30
30
30
-20
30
30
30
5
1
3
120
0
5
20
70
1
X
X
25
100
36
Run
Start
Squad
state
at
the
beginning
of
a
run
(can
be
used
as
delay)
200
5
1
3
120
0
5
20
70
1
X
X
25
100
no
selection
1
Default
State
Start
state
and
Default
fallback
state
10
30
10
10
7
1
3
160
120
3
20
75
1
X
X
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
100
10
30
10
10
7
1
3
160
12
3
20
75
1
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
15
10
30
10
10
7
1
3
160
0
3
20
75
1
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
6
10
30
10
10
7
1
3
160
72
3
20
75
1
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
50
10
30
10
10
7
1
3
160
180
3
20
75
1
X
X
19
Injured
Agent
state
when
injured
(shot
at
and
hit)
7200
10
30
10
10
7
1
3
160
12
2
20
75
1
X
X
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
7200
30
10
10
10
30
30
30
-20
30
30
30
7
1
3
160
0
3
20
75
1
X
X
36
Run
Start
Squad
state
at
the
beginning
of
a
run
(can
be
used
as
delay)
200
7
1
3
160
0
X
X
1
Default
State
Start
state
and
Default
fallback
state
20
6
1
2
130
120
2
30
54
1
X
X
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
100
20
6
1
2
130
12
2
30
54
1
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
6
20
6
1
2
130
0
2
30
54
1
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
6
20
6
1
2
130
72
2
30
54
1
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
50
20
6
1
2
130
180
2
30
54
1
X
X
19
Injured
Agent
state
when
injured
(shot
at
and
hit)
7200
20
6
1
2
130
12
1
30
54
1
X
X
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
7200
30
30
30
30
-20
30
30
30
6
1
2
130
0
2
30
54
1
X
X
no
selection
130
0
X
X
15-15
Blue
ARV-A
16-16
Blue
ICV
Platoon
Blue
NLOS
LS
Plt Blue
MCS
Platoon
State
HO
-
Motor
BTRY
H1
-
C
H2
-
D
17-17
14-14
Blue
NLOS
Mortar
Sec
13-13
12-12
Blue
NLOS
Cannon
Plt

114
1
Default
State
Start
state
and
Default
fallback
state
20
6
1
2
130
120
2
20
32
1
X
X
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
20
6
1
2
130
12
2
20
32
1
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
20
6
1
2
130
0
2
20
32
1
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
20
6
1
2
130
72
2
20
32
1
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
20
6
1
2
130
180
2
20
32
1
X
X
19
Injured
Agent
state
when
injured
(shot
at
and
hit)
20
6
1
2
130
12
1
20
32
1
X
X
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
30
20
6
1
2
130
0
2
20
32
1
X
X
36
Run
Start
Squad
state
at
the
beginning
of
a
run
(can
be
used
as
delay)
-10
6
1
2
130
0
2
20
32
1
X
X
no
selection
1
Default
State
Start
state
and
Default
fallback
state
-20
20
4
1
2
130
120
2
30
54
5
X
X
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
-20
20
4
1
2
130
12
2
30
54
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
-20
20
4
1
2
130
0
2
30
54
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
4
1
2
130
2
30
54
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
-20
20
4
1
2
130
180
2
30
54
X
X
19
Injured
Agent
state
when
injured
(shot
at
and
hit)
-20
20
4
1
2
130
12
1
30
54
X
X
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
30
-20
-10
4
1
2
130
0
2
30
54
X
X
no
selection
no
selection
no
selection
1
Default
State
Start
state
and
Default
fallback
state
5
30
-10
12
1
0
200
321
1
90
11
5
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
1
Default
State
Start
state
and
Default
fallback
state
5
30
-10
112
1
0
0
427
1
90
11
5
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
18-18
Blue
ARV-A(L)
19-19
Blue
ARV-RSTA
20-23
Blue
UAV
CL
1
24-27
Blue
UAV
CL
2

115
1
Default
State
Start
state
and
Default
fallback
state
30
-10
9
1
0
200
427
100
100
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
1
Default
State
Start
state
and
Default
fallback
state
-10
20
-10
-11
-1
4
1
2
150
120
3
30
70
5
X
X
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
100
-10
20
-10
-11
-1
4
1
2
150
12
3
30
70
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
6
-10
20
-10
-11
-1
4
1
2
150
0
3
30
70
5
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
6
0
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
50
-10
20
-10
-11
-1
4
1
2
150
180
3
30
70
5
X
X
19
Injured
Agent
state
when
injured
(shot
at
and
hit)
7200
-10
20
-10
-11
-1
4
1
2
150
12
2
30
70
5
X
X
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
7200
-10
20
-10
4
1
2
150
0
3
30
70
5
X
X
no
selection
X
X
no
selection
1
Default
State
Start
state
and
Default
fallback
state
100
9
16
5
X
X
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
5.4
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
5.4
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
5.4
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
14
X
X
19
Injured
Agent
state
when
injured
(shot
at
and
hit)
4.5
X
X
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
5.4
X
X
36
Run
Start
Squad
state
at
the
beginning
of
a
run
(can
be
used
as
delay)
0
X
X
no
selection
1
Default
State
Start
state
and
Default
fallback
state
100
9
16
5
X
X
2
Reach
Waypoint
Agent
state
when
any
waypoint
is
reached
100
5
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
100
5
X
X
4
Taken
Shot
(Sec)
Agent
state
when
agent
has
fired
its
secondary
weapon
at
an
enemy
(may
not
have
hit
target)
100
5
X
X
5
Shot
At
(Pri)
Agent
state
when
shot
at
by
an
enemy’s
primary
weapon
(may
not
have
been
hit)
100
14
X
X
19
Injured
Agent
state
when
injured
(shot
at
and
hit)
100
5
X
X
35
Reach
Final
Waypoint
Agent
state
when
final
waypoint
is
reached
100
5
X
X
36
Run
Start
Squad
state
at
the
beginning
of
a
run
(can
be
used
as
delay)
no
selection
X
X
1
Default
State
Start
state
and
Default
fallback
state
10
1
3
210
0
48
100
16
0
X
X
3
Taken
Shot
(Pri)
Agent
state
when
agent
has
fired
its
primary
weapon
at
an
enemy
(may
not
have
hit
target)
60
10
1
3
210
0
48
100
100
0
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
X
X
no
selection
no
selection
X
X
31-31
Blue
MachineGunner
M240b
30-30
Blue
Infantryman
29-29
Blue
R&SV
28-28
Blue
UAV
CL
3
32-32
Blue
CAS

116
F. COMMUNICATION CHARACTERISTICS
Item # Device Type Notes Range(meters)
range
(model_grids)
Capacity
(msgs/sec)
capacity
(model_steps)
Queue
Buffer Size
Latency
(sec)
latency
(model) Self Reliab. 100 MxAge
Rank
Filter Include
Delivery
(Guarant
eed of F-
N-F)
1
Cellphone or
equivalent VHF Limited Reliability 2,000 385 1 1 2 10 10 120 70 100 30 High SETC F-N-F
2
Basic Radio
or equivalent UHF LOS 50 10 1 1 2 10 10 120 70 100 30 High SET F-N-F
3
Personal Role
Radio (PRR)
or equivalent UHF
Intra-Team
Communications 500 96 1 1 2 10 10 120 93 100 30 High SNETC F-N-F
4
PRC 148 or
equivalent VHF/UHF
Platoon – Squad – Team
C2 - CAS Control 6,500 500 1 1 2 10 10 120 93 100 30 High SNETC F-N-F
5
JTRS
Cluster(8
channel) or
equivalent Digitial
Future Internet
Networked Protocal
Radio System) 50,000 500 8 8 16 10 10 120 93 100 30 High SNETC F-N-F
6
JTRS
Cluster(4
channel) or
equivalent Digitial
Future Internet
Networked Protocal
Radio System) 50,000 500 4 4 8 10 10 120 93 100 30 High SNETC F-N-F
7
JTRS Cluster
5 SFF-D-E-G
or equivalent Digitial
Future Internet
Networked Protocal
Radio System) 50,000 500 5 5 10 10 10 120 98 100 30 High F-N-F
8
PRC 117 or
equivalent
VHF /
UHF /
Satellite
Communi
cations
Squad – Plat – HHQ
CAS/Fires Control (OTH -
Digital) 11,500 500 1 2 2 10 10 120 93 100 30 High SNETC F-N-F
notes: call waiting
not used in my model
Number Squad With Radio Capabilities or Similiarities to:
1 Red BMP-3 6
2 Red 82 Mortors 4
3 Red SA-16 Infantryman 1
4 Red RPG-7 1
5 Red AT-7 4
6 Red Scout 4
7 Red RPK-74 2
8 Red AK-M Infantryman 4
9 Red SVD 1
10 Red APC 6
11 Red T72 6
12 Blue NLOS Mortor Sec 5
13 Blue NLOS Cannon Plt 5
14 Blue NLOS LS Plt 7
15 Blue ICV Platoon 5
16 Blue MCS Platoon 5
17 Blue ARV-A 6
18 Blue ARV-A(L) 6
19 Blue ARV-RSTA 6
20 Blue UAV CL 1 7
21 Blue UAV CL 2 7
22 Blue UAV CL 3 7
23 Blue R&SV 5
24 Blue Infantryman 3
25 Blue MachineGunner M24 3
26 Blue CAS 8
27 Blue Apache 8
Notes:
Blue Force Radio reference from FCS UA Design Concept Baseline Descriptions UA-001-01-050124
Blue Force CAS and Apache referenced from pilots currently stationed at Naval Postgraduate School academic year 2005
Red Force Radio designed to be equivalent to Blue Force capabilities
1 transmission
at a time
FOR
MY
MODEL
THESE
COMMS
ARE
ESSENTIALLY
THE
SAME
(CANT
DO
INTERIOR/EXTERIOR
EFFECTS)
time to make
call every 2 min max hold time
unitl decide to
call back
BLUE FORCE Agent Memory 30 seconds
Intra-Squad Comms Delay - min link Rank low
Squad Threat Persistence 30 Inorganic Threat Persistance 30
Fuse Unknowns No Fuse Unknowns on Inorg map No
Fuse Time - Fuse Time -
Fuse Radius - Fuse Radius -
Outbound Comm Link X
Type Type # Range Capacity Buffer Latency Self Reliab. Acc. MxAge Rank Filter Include Delivery
Blue NLOS Mortor Sec 12 #N/A n 5
Blue NLOS Cannon Plt 13 #N/A n 5
Blue NLOS LS Plt 14 #N/A n 7
Blue ICV Platoon 15 12 Blue NLOS Mortor Sec y 5 JTRS Cluster(8 channel) or equivalent 500 8 16 10 120 93 100 30 High SNETC F-N-F
Blue ICV Platoon 15 13 Blue NLOS Cannon Plt y 5 JTRS Cluster(8 channel) or equivalent 500 8 16 10 120 93 100 30 High SNETC F-N-F
Blue ICV Platoon 15 16 Blue MCS Platoon y 5 JTRS Cluster(8 channel) or equivalent 500 8 16 10 120 93 100 30 High SNETC F-N-F
Blue MCS Platoon 16 14 Blue NLOS LS Plt y 5 JTRS Cluster(8 channel) or equivalent 500 8 16 10 120 93 100 30 High SNETC F-N-F
Blue MCS Platoon 16 23 Blue UAV CL 1 y 5 JTRS Cluster(8 channel) or equivalent 500 8 16 10 120 93 100 30 High SNETC F-N-F
Blue ARV-A 17 15 Blue ICV Platoon y 6 JTRS Cluster(4 channel) or equivalent 500 4 8 10 120 93 100 30 High SNETC F-N-F
Blue ARV-A 17 24 Blue UAV CL 2 y 6 JTRS Cluster(4 channel) or equivalent 500 4 8 10 120 93 100 30 High SNETC F-N-F
Blue ARV-A(L) 18 16 Blue MCS Platoon y 6 JTRS Cluster(4 channel) or equivalent 500 4 8 10 120 93 100 30 High SNETC F-N-F
Blue ARV-RSTA 19 16 Blue MCS Platoon y 6 JTRS Cluster(4 channel) or equivalent 500 4 8 10 120 93 100 30 High SNETC F-N-F
Blue UAV CL 1 20 15 Blue ICV Platoon y 7 JTRS Cluster 5 SFF-D-E-G or equivalent 500 5 10 10 120 98 100 30 High 0 F-N-F
Blue UAV CL 2 24 16 Blue MCS Platoon y 3 Personal Role Radio (PRR) or equivalent 96 1 2 10 120 93 100 30 High SNETC F-N-F
Blue UAV CL 3 28 23 Blue UAV CL 1 y 8 PRC 117 or equivalent 500 2 2 10 120 93 100 30 High SNETC F-N-F
Blue R&SV 29 14 Blue NLOS LS Plt y 8 PRC 117 or equivalent 500 2 2 10 120 93 100 30 High SNETC F-N-F
Blue R&SV 29 26 Blue UAV CL 2 y 8 PRC 117 or equivalent 500 2 2 10 120 93 100 30 High SNETC F-N-F
Blue R&SV 29 27 Blue UAV CL 2 y 8 PRC 117 or equivalent 500 2 2 10 120 93 100 30 High SNETC F-N-F
Blue Infantryman 30 15 Blue ICV Platoon y 8 PRC 117 or equivalent 500 2 2 10 120 93 100 30 High SNETC F-N-F
Blue MachineGunner M24 31 15 Blue ICV Platoon y 8 PRC 117 or equivalent 500 2 2 10 120 93 100 30 High SNETC F-N-F
Blue CAS 32 #N/A n 8
Blue Apache 33 #N/A n 8
add to Latency an additional 20 seconds to all NLOS Cannon and NLOS Launch Systems take into account time of flight and another 10 seconds for computation procedures
add to Latency an additional 45 seconds to all NLOS Mortars Latencey to take into account time of flight and antother 10 seconds for computational procedures
From
Squad
To
Squad
LINK
(Y/N) DEVICE

117
RED FORCE Agent Memory 30 seconds
Intra-Squad Comms Delay min link Rank low
Squad Threat Persistence 30 Inorganic Threat Persistance 30
Fuse Unknowns No Fuse Unknowns on Inorg map No
Fuse Time - Fuse Time -
Fuse Radius - Fuse Radius -
Outbound Comm Link X
Type Type # Range Capacity Buffer Latency Self Reliab. Acc. MxAge Rank Filter Include Delivery
Red BMP-3 1 #N/A n 6
Red 82 Mortors 2 #N/A n 4
Red SA-16 Infantryman 3 #N/A n 1
Red RPG-7 4 #N/A n 1
Red AT-7 5 11 Red T72 y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red Scout 6 1 Red BMP-3 y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red Scout 6 2 Red 82 Mortors y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red Scout 6 4 Red RPG-7 y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red Scout 6 5 Red AT-7 y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red Scout 6 9 Red SVD y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red Scout 6 11 Red T72 y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red RPK-74 7 #N/A n 2
Red AK-M Infantryman 8 1 Red BMP-3 y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red AK-M Infantryman 8 2 Red 82 Mortors y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red AK-M Infantryman 8 39 Red T72 y 4 PRC 148 or equivalent 500 1 2 10 120 93 100 30 High SNETC F-N-F
Red SVD 9 #N/A n 1
Red APC 10 2 Red 82 Mortors y 6 JTRS Cluster(4 channel) or equivalent 500 4 8 10 120 93 100 30 High SNETC F-N-F
Red APC 10 #N/A n 6
Red T72 11 #N/A n 6
add to Latency an additional 45 seconds to all Mortars Latencey to take into account time of flight and antother 10 seconds for computational procedures
From
Squad To Squad
LINK
(Y/N) DEVICE

118
G. WEAPON CHARACTERISTICS
Max Terrain Dimension 2600 Meters 5.200 Meters per grid
# CELLs in maximum dimension 500 # GRIDS 5.689 Yds per grid
Steps per Minute 60 Steps 17.066 Feet per grid
Steps per Second 1 Steps
TABLE A
Platform Weapon
Min
Effective
Range (m)
Max
Effective
Range (m)
Max
weapon
Range
Shot
Radius (m)
Max
Targets/
min
Carried
Rounds
{High
Rate of
Fire /
min}
Blue NLOS Mortor Sec 120 mm BLOS guided munition 500 12000 15000 60 2 62 24
xm307 25mm 1 450 2000 1 10 300 250
Blue NLOS Cannon Plt 155 mm std 500 30000 30000 50 4 24 10
155 mm guided (heavy targets only) 500 30000 30000 50 4 24
10
Blue NLOS LS Plt payload assit mod (PAM) 500 40000 40000 50 1 15 1
Blue ICV Platoon MK44 30 mm 1 2000 6000 1 10 320 400
M240B 7.62mm 1 1800 3725 1 10 1200 200
Blue MCS Platoon Guided xm36 120mm 40 2000 4000 15 4 27 4
xm307 25mm 1 450 2000 1 10 300 250
Blue ARV-A MK44 30 mm 1 2000 6000 1 10 320 400
M240B 7.62mm 1 1800 3725 1 10 1200 200
Blue ARV-A(L) xm307 25 mm 1 450 2000 1 10 300 250
Javelin Anti Tank Missle 75 2000 2000 5 2 2 2
Blue ARV-RSTA xm307 25 mm 1 450 2000 1 10 300 250
Blue UAV CL 3 Guided Hellfire 500 7000 8000 30 16 4 16
APKWS 500 6000 6500 10 4 6 4
Blue R&SV xm307 25 mm 1 450 2000 1 10 300 250
Blue Infantryman m16 1 550 3600 1 10 1260 16
Blue MachineGunner M24m240B 7.62mm 1 1800 3725 1 10 1200 200
Blue CAS m230 / 30 mm 1 1830 6000 1 10 1200 625
Guided LOCAAS 100 100000 100000 50 1 16 1
Blue Apache m230 / 30 mm 1 1830 6000 1 10 1200 625
Guided Hellfire 500 7000 8000 30 16 16 16
Red BMP-3 2A-42 /30 mm 1 4000 unk 5 4 500 15
Guided 2A-70M100mm tube firing
AT12 guided stabber 100 5500 unk 15 4 50
3
Red 82 Mortors 82 mm Mortar 1000 4000 4000 15 4 65 10
ak m/47 rifle 1 300 1000 1 10 240 600
Red SA-16 Infantryman Guided SA-16 Surface to Air Missle 500 3500 5000 5 2 2
2
Red RPG-7 anti tank grenade launcher 50 500 920 5 6 6 6
Red AT-7 anti tank missle 40 500 1000 5 2 2 2
Red Scout ak m/47 rifle 1 300 1000 1 10 240 600
Red RPK-74 rpk 74 light machine gun 1 450 2500 1 10 1000 150
Red AK-M Infantryman ak m / 47 rifle 1 300 1000 1 10 240 600
Red SVD SVD 7.62 sniper 1 1300 3800 1 1 10 30
Red APC 2A-42 /30 mm 1 300 2500 1 10 240 100
rpk 74 light machine gun 1 450 2500 1 10 1000 150
Red T72 2A-46 /125mm 50 2120 10000 15 4 60 8
rpk 74 light machine gun 1 450 2500 1 10 1000 150
1 0.5 0
Maximum effective range is the maximum range within which a weapon is effective against its intended target. interpret to be 50% kill rate
Weapon Specs

119
TABLE B
Weapon
Effects in
Grid
Range
Pkill at
Max Grid
Range
Grid Shot
Radius
engagmnt/s
tep
Targets /
100
time in shot
taken state
Blue NLOS Mortor Sec 120 mm BLOS guided munition 500 1 12 0.03 100 30
0 xm307 25mm 87 0 0 0.17 100 6
Blue NLOS Cannon Plt 155 mm std 500 1 10 0.07 100 15
0 155 mm guided (heavy targets only)
500 1 10 0.07 100 15
Blue NLOS LS Plt payload assit mod (PAM) 500 1 10 0.02 100 60
Blue ICV Platoon MK44 30 mm 385 1 0 0.17 100 6
0 M240B 7.62mm 346 1 0 0.17 100 6
Blue MCS Platoon Guided xm36 120mm 385 1 3 0.07 100 15
0 xm307 25mm 87 0 0 0.17 100 6
Blue ARV-A MK44 30 mm 385 1 0 0.17 100 6
0 M240B 7.62mm 346 1 0 0.17 100 6
Blue ARV-A(L) xm307 25 mm 87 0 0 0.17 100 6
0 Javelin Anti Tank Missle 385 1 1 0.03 100 30
Blue ARV-RSTA xm307 25 mm 87 0 0 0.17 100 6
Blue UAV CL 3 Guided Hellfire 500 1 6 0.27 100 4
APKWS 500 1 2 0.07 100 15
Blue R&SV xm307 25 mm 87 0 0 0.17 100 6
Blue Infantryman m16 106 1 0 0.17 100 6
Blue MachineGunner
M240b
m240B 7.62mm
346 1 0 0.17 100 6
Blue CAS m230 / 30 mm 352 1 0 0.17 100 6
0 Guided LOCAAS 500 1 10 0.02 100 60
Blue Apache m230 / 30 mm 352 1 0 0.17 100 6
0 Guided Hellfire 500 1 6 0.27 100 4
Red BMP-3 2A-42 /30 mm 500 1 1 0.07 100 15
0
AT12 guided stabber 500 1 3 0.07 100 15
Red 82 Mortors 82 mm Mortar 500 1 3 0.07 100 15
0 ak m/47 rifle 58 0 0 0.17 100 6
Red SA-16 Infantryman Guided SA-16 Surface to Air Missle
500 1 1 0.03 100 30
Red RPG-7 anti tank grenade launcher 96 1 1 0.10 100 10
Red AT-7 anti tank missle 96 1 1 0.03 100 30
Red Scout ak m/47 rifle 58 0 0 0.17 100 6
Red RPK-74 rpk 74 light machine gun 87 0 0 0.17 100 6
Red AK-M Infantryman ak m / 47 rifle 58 0 0 0.17 100 6
Red SVD SVD 7.62 sniper 250 1 0 0.02 100 60
Red APC 2A-42 /30 mm 58 0 0 0.17 100 6
rpk 74 light machine gun 87 0 0 0.17 100 6
Red T72 2A-46 /125mm 408 1 3 0.07 100 15
0 rpk 74 light machine gun 87 0 0 0.17 100 6
RANGE PROFILE FOR MAP (MANA conversion for Kinetic Weapon Factors only)
TABLE C max req 2600
MANA values if modeled as Kinetic Energy Weapon
Weapon
Real World 0 25 50 300 450 501 750 1000 1500 2000 2600
GRID 0 5 10 58 87 96 144 192 288 385 500
Blue NLOS Mortor Sec 120 mm BLOS guided munition 0.00 0.00 0.00 0.00 0.00 1.00 1.00 1.00 1.00 1.00 1.00
0 xm307 25mm 0.92 0.92 0.89 0.78 0.72 0.68 0.53 0.40 0.17 0.00 0.00
Blue NLOS Cannon Plt 155 mm std 0.00 0.00 0.00 0.00 0.00 1.00 1.00 0.99 0.98 0.98 0.96
0 155 mm guided (heavy targets only)
0.00 0.00 0.00 0.00 0.00 1.00 1.00 0.99 0.98 0.98 0.96
Blue NLOS LS Plt payload assit mod (PAM) 0.00 0.00 0.00 0.00 0.00 1.00 1.00 0.99 0.99 0.98 0.97
Blue ICV Platoon MK44 30 mm 1.00 1.00 0.98 0.92 0.88 0.87 0.80 0.74 0.62 0.51 0.39
0 M240B 7.62mm 1.00 1.00 0.99 0.92 0.87 0.86 0.79 0.73 0.58 0.46 0.29
Blue MCS Platoon Guided xm36 120mm 0.00 0.00 1.00 0.93 0.89 0.88 0.82 0.76 0.63 0.51 0.35
0 xm307 25mm 0.92 0.92 0.89 0.78 0.72 0.68 0.53 0.40 0.17 0.00 0.00
Blue ARV-A MK44 30 mm 1.00 1.00 0.98 0.92 0.88 0.87 0.80 0.74 0.62 0.51 0.39
0 M240B 7.62mm 1.00 1.00 0.99 0.92 0.87 0.86 0.79 0.73 0.58 0.46 0.29
Blue ARV-A(L) xm307 25 mm 0.92 0.92 0.89 0.78 0.72 0.68 0.53 0.40 0.17 0.00 0.00
0 Javelin Anti Tank Missle 0.00 0.00 0.00 0.94 0.90 0.89 0.82 0.77 0.63 0.51 0.00
Blue ARV-RSTA xm307 25 mm 0.92 0.92 0.89 0.78 0.72 0.68 0.53 0.40 0.17 0.00 0.00
Blue UAV CL 3 Guided Hellfire 0.00 0.00 0.00 0.00 0.00 1.00 1.00 1.00 1.00 1.00 1.00
APKWS
Blue R&SV xm307 25 mm 0.92 0.92 0.89 0.78 0.72 0.68 0.53 0.40 0.17 0.00 0.00
Blue Infantryman m16 0.94 0.94 0.91 0.77 0.70 0.67 0.54 0.43 0.23 0.11 0.03
Blue MachineGunner
M240b
m240B 7.62mm
1.00 1.00 0.99 0.92 0.87 0.86 0.79 0.73 0.58 0.46 0.29
Blue CAS m230 / 30 mm 1.00 1.00 0.98 0.91 0.87 0.85 0.78 0.72 0.59 0.48 0.36
0 Guided LOCAAS 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.99 0.99 0.99
Blue Apache m230 / 30 mm 1.00 1.00 0.98 0.91 0.87 0.85 0.78 0.72 0.59 0.48 0.36
0 Guided Hellfire 0.00 0.00 0.00 0.00 0.00 1.00 1.00 1.00 1.00 1.00 1.00
Red BMP-3 2A-42 /30 mm 1.00 1.00 0.99 0.96 0.94 0.94 0.91 0.88 0.81 0.75 0.68
0
AT12 guided stabber 0.00 0.00 0.00 0.98 0.97 0.96 0.94 0.92 0.87 0.83 0.77
Red 82 Mortors 82 mm Mortar 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.92 0.84 0.73
0 ak m/47 rifle 0.92 0.92 0.89 0.73 0.56 0.50 0.23 0.00 0.00 0.00 0.00
Red SA-16 Infantryman Guided SA-16 Surface to Air Missle
0.00 0.00 0.00 0.00 0.00 1.00 0.97 0.94 0.87 0.80 0.69
Red RPG-7 anti tank grenade launcher 0.00 0.00 1.00 0.71 0.54 0.48 0.20 0.00 0.00 0.00 0.00
Red AT-7 anti tank missle 0.00 0.00 0.98 0.73 0.58 0.52 0.26 0.02 0.00 0.00 0.00
Red Scout ak m/47 rifle 0.92 0.92 0.89 0.73 0.56 0.50 0.23 0.00 0.00 0.00 0.00
Red RPK-74 rpk 74 light machine gun 0.91 0.91 0.89 0.78 0.72 0.69 0.54 0.43 0.23 0.10 0.00
Red AK-M Infantryman ak m / 47 rifle 0.92 0.92 0.89 0.73 0.56 0.50 0.23 0.00 0.00 0.00 0.00
Red SVD SVD 7.62 sniper 0.99 0.99 0.97 0.88 0.82 0.80 0.71 0.63 0.47 0.34 0.20
Red APC 2A-42 /30 mm 0.92 0.92 0.89 0.73 0.56 0.50 0.23 0.00 0.00 0.00 0.00
0 rpk 74 light machine gun 0.91 0.91 0.89 0.78 0.72 0.69 0.54 0.43 0.23 0.10 0.00
Red T72 2A-46 /125mm 0.00 0.00 1.00 0.93 0.90 0.88 0.82 0.76 0.64 0.54 0.41
0 rpk 74 light machine gun 0.91 0.91 0.89 0.78 0.72 0.69 0.54 0.43 0.23 0.10 0.00
Note: Table C reflects flexible data values, for simplified changes to the model if needed. Weapons finally modeled as Area Fire weapons are reflected in Table D.
Note: simply highlight last column and expand to right
to cover additional distance or change Real world
values to desired values

120
TABLE D
Area Fire Weapon Data Determined by Real World Blast Radius and Pk is determined by Carleton Function
MANA values if modeled as Area Fire Weapon
Platform Target Type b
NLOS M real world range 0 20 40 60
MANA units 0 4 8 12
light target 51 1 0.925988 0.735228 0.500553
heavy target 36 1 0.856997 0.539408 0.249352
NLOS C/LS real world range 0 16.66667 33.33333 50
MANA units 0 3 6 10
light target 43 1 0.927636 0.740476 0.508627
heavy target 30 1 0.856997 0.539408 0.249352
guided xm36 real world range 0 5 10 15
MANA units 0 1 2 3
light target 13 1 0.928705 0.743893 0.513924
heavy target 9 1 0.856997 0.539408 0.249352
guided 82mm real world range 0 5 10 15
MANA units 0 1 2 3
light target 13 1 0.928705 0.743893 0.513924
heavy target 9 1 0.856997 0.539408 0.249352 Carleton Function
2
2
-
2
p(hit) =
r
b
e
⎛ ⎞
⎜ ⎟
⎜ ⎟
⎜ ⎟
⎜ ⎟
⎝ ⎠

121
Raw Spline Data
Note: This is only a portion of the whole table, and does not show maximum
ranges.
120
mm
BLOS
guided
munition
Spline
Predictor
for
120
mm
xm307
25mm
Spline
Predictor
xm307
155
mm
Spline
Predictor
for
155
payload
assit
mod
(PAM)
Spline
Predictor
for
PAM
MK44
30
mm
Spline
Predictor
for
mk44
M240B
7.62mm
Spline
Predictor
for
M240
Guided
xm36
120mm
Spline
Predictor
for
xm36
Javelin
Anti
Tank
Missle
Spline
Predictor
for
Javelin
Guided
Hellfire
Spline
Predictor
for
Hellfire
m16
Spline
Predictor
for
m16
m230
/
30
mm
Spline
Predictor
for
m230
Guided
LOCAAS
Spline
Predictor
for
LOCAAS
2A-42
/30
mm
Spline
Predictor
for
2A-42
Guided
2A-70M100mm
tube
firing
AT12
gu
Spline
Predictor
for
2A-70
82
mm
Mortar
Spline
Predictor
for
82
mm
Mortar
Guided
SA-16
Surface
to
Air
Missle
Spline
Predictor
for
SA-16
anti
tank
grenade
launcher
Spline
Predictor
for
anti
tank
grenade
anti
tank
missle
Spline
Predictor
for
anti
tank
missle
rpk
74
light
machine
gun
Spline
Predictor
for
rpk
74
ak
m
/
47
rifle
Spline
Predictor
for
ak
m
SVD
7.62
sniper
Spline
Predictor
for
SVD
2A-46
/125mm
Spline
Predictor
for
2A-46
500
1.0002
1
0.915585
500
1
500
1
1
0.997232
1
0.999336
40
0.999836
75
1
500
1.002204
1
0.935484
1
0.996305
100
1
1
1
100
1
1000
1
500
1.003111
50
1.005046
40
0.993619
1
0.910425
1
0.919151
1
0.991039
50
0.997654
12000
0.4980
450
0.717736
30000
0.5
40000
0.5
2000
0.508303
1800
0.502569
2000
0.500649
2000
0.5
7000
0.466936
550
0.652257
1830
0.51063
100000
0.5
4000
0.5
5500
0.5
4000
0.5
3500
0.481336
500
0.479093
500
0.524502
450
0.718387
300
0.730765
1300
0.527234
2120
0.505925
15000
0.0008
2000
-0.024453
6000
-0.001383
3725
-0.000621
4000
-0.000161
8000
0.014328
3600
-0.011613
6000
-0.001621
5000
0.006221
920
0.005407
1000
-0.00587
2500
-0.019619
1000
-0.034534
3800
-0.004656
10000
-0.000616
0
0.9978
0
0.916057
0
1.008475
0
1.006329
0
0.997496
0
0.999616
0
1.010069
0
1.019481
0
0.959522
0
0.936035
0
0.996593
0
1.000501
0
1.000125
0
1.009259
0
1.166667
0
1.065419
0
1.063394
0
1.034509
0
0.910886
0
0.919795
0
0.991424
0
1.010441
40
0.9980
40
0.897207
40
1.007797
40
1.005823
40
0.986912
40
0.988415
40
0.999836
40
1.009091
40
0.962866
40
0.913997
40
0.985076
40
1.0003
40
0.995124
40
1.005556
40
1.16
40
1.060335
40
1.016716
40
0.993619
40
0.892479
40
0.894077
40
0.976047
40
1.000212
50
0.9980
50
0.892501
50
1.007627
50
1.005696
50
0.984266
50
0.985615
50
0.997278
50
1.006494
50
0.963705
50
0.908493
50
0.982197
50
1.00025
50
0.993873
50
1.00463
50
1.158333
50
1.059068
50
1.005046
50
0.983397
50
0.887884
50
0.887654
50
0.972203
50
0.997654
75
0.9981
75
0.880758
75
1.007203
75
1.00538
75
0.977651
75
0.978614
75
0.990882
75
1
75
0.965806
75
0.894748
75
0.975001
75
1.000125
75
0.990748
75
1.002315
75
1.154167
75
1.055907
75
0.975873
75
0.95784
75
0.87642
75
0.871618
75
0.962597
75
0.99126
100
0.9983
100
0.869058
100
1.00678
100
1.005063
100
0.971039
100
0.971614
100
0.984486
100
0.993506
100
0.967915
100
0.881036
100
0.967807
100
1
100
0.987622
100
1
100
1.15
100
1.052754
100
0.946699
100
0.932286
100
0.865002
100
0.855622
100
0.952995
100
0.984866
125
0.9984
125
0.857414
125
1.006356
125
1.004747
125
0.964428
125
0.964615
125
0.978091
125
0.987013
125
0.970029
125
0.867368
125
0.960615
125
0.999875
125
0.984496
125
0.997685
125
1.145833
125
1.049611
125
0.917523
125
0.906734
125
0.853643
125
0.83968
125
0.943399
125
0.978474
150
0.9985
150
0.845841
150
1.005932
150
1.00443
150
0.957819
150
0.957616
150
0.971695
150
0.980519
150
0.972148
150
0.853754
150
0.953426
150
0.99975
150
0.98137
150
0.99537
150
1.141667
150
1.046475
150
0.888344
150
0.881185
150
0.84236
150
0.823807
150
0.933811
150
0.972082
175
0.9986
175
0.834354
175
1.005508
175
1.004114
175
0.951213
175
0.950617
175
0.9653
175
0.974026
175
0.974273
175
0.840205
175
0.94624
175
0.999625
175
0.978245
175
0.993056
175
1.1375
175
1.043347
175
0.859162
175
0.855642
175
0.831168
175
0.808015
175
0.924232
175
0.965692
200
0.9987
200
0.822967
200
1.005085
200
1.003797
200
0.944611
200
0.94362
200
0.958905
200
0.967532
200
0.976403
200
0.826732
200
0.93906
200
0.999499
200
0.975119
200
0.990741
200
1.133333
200
1.040225
200
0.829976
200
0.830104
200
0.820081
200
0.792319
200
0.914663
200
0.959304
225
0.9989
225
0.811694
225
1.004661
225
1.003481
225
0.938012
225
0.936623
225
0.95251
225
0.961039
225
0.978537
225
0.813346
225
0.931884
225
0.999374
225
0.971993
225
0.988426
225
1.129167
225
1.03711
225
0.800784
225
0.804573
225
0.809116
225
0.776732
225
0.905107
225
0.952918
250
0.9990
250
0.800549
250
1.004237
250
1.003165
250
0.931417
250
0.929627
250
0.946115
250
0.954545
250
0.980676
250
0.800059
250
0.924714
250
0.999249
250
0.968867
250
0.986111
250
1.125
250
1.034001
250
0.771587
250
0.779051
250
0.798287
250
0.761268
250
0.895564
250
0.946536
275
0.9991
275
0.789547
275
1.003814
275
1.002848
275
0.924827
275
0.922633
275
0.93972
275
0.948052
275
0.982818
275
0.786881
275
0.91755
275
0.999124
275
0.965741
275
0.983796
275
1.120833
275
1.030897
275
0.742382
275
0.753537
275
0.787609
275
0.745941
275
0.886036
275
0.940156
300
0.9992
300
0.778702
300
1.00339
300
1.002532
300
0.918243
300
0.91564
300
0.933326
300
0.941558
300
0.984964
300
0.773823
300
0.910393
300
0.998999
300
0.962616
300
0.981481
300
1.116667
300
1.027797
300
0.71317
300
0.728033
300
0.777099
300
0.730765
300
0.876525
300
0.93378
325
0.9993
325
0.768029
325
1.002966
325
1.002215
325
0.911663
325
0.908648
325
0.926932
325
0.935065
325
0.987112
325
0.760896
325
0.903243
325
0.998874
325
0.95949
325
0.979167
325
1.1125
325
1.024702
325
0.68395
325
0.702541
325
0.76677
325
0.701981
325
0.867033
325
0.927408
350
0.9995
350
0.757541
350
1.002542
350
1.001899
350
0.90509
350
0.901657
350
0.920538
350
0.928571
350
0.989263
350
0.748111
350
0.896102
350
0.998749
350
0.956364
350
0.976852
350
1.108333
350
1.021611
350
0.65472
350
0.677061
350
0.756638
350
0.673356
350
0.85756
350
0.92104
375
0.9996
375
0.747254
375
1.002119
375
1.001582
375
0.898524
375
0.894668
375
0.914145
375
0.922078
375
0.991417
375
0.735479
375
0.888969
375
0.998624
375
0.953238
375
0.974537
375
1.104167
375
1.018523
375
0.625479
375
0.651595
375
0.746719
375
0.644883
375
0.848108
375
0.914677
400
0.9997
400
0.737181
400
1.001695
400
1.001266
400
0.891964
400
0.887681
400
0.907752
400
0.915584
400
0.993572
400
0.72301
400
0.881846
400
0.998498
400
0.950113
400
0.972222
400
1.1
400
1.015437
400
0.596228
400
0.626143
400
0.737027
400
0.616557
400
0.838679
400
0.90832
425
0.9998
425
0.727337
425
1.001271
425
1.000949
425
0.885412
425
0.880696
425
0.901359
425
0.909091
425
0.995729
425
0.710717
425
0.874732
425
0.998373
425
0.946987
425
0.969907
425
1.095833
425
1.012354
425
0.566964
425
0.600707
425
0.727578
425
0.588371
425
0.829274
425
0.901967
450
1.0000
450
0.717736
450
1.000847
450
1.000633
450
0.878868
450
0.873712
450
0.894967
450
0.902597
450
0.997887
450
0.69861
450
0.86763
450
0.998248
450
0.943861
450
0.967593
450
1.091667
450
1.009272
450
0.537688
450
0.575287
450
0.718387
450
0.560321
450
0.819895
450
0.895621
475
1.0001
475
0.700572
475
1.000424
475
1.000316
475
0.872333
475
0.866731
475
0.888575
475
0.896104
475
1.000045
475
0.686699
475
0.860538
475
0.998123
475
0.940735
475
0.965278
475
1.0875
475
1.006191
475
0.508398
475
0.549885
475
0.702055
475
0.532399
475
0.810543
475
0.889281
500
1.0002
500
0.683661
500
1
500
1
500
0.865806
500
0.859751
500
0.882184
500
0.88961
500
1.002204
500
0.674996
500
0.853459
500
0.997998
500
0.937609
500
0.962963
500
1.083333
500
1.003111
500
0.479093
500
0.524502
500
0.685993
500
0.504601
500
0.80122
500
0.882948
525
1.0003
525
0.666999
525
0.999576
525
0.999684
525
0.859289
525
0.852774
525
0.875793
525
0.883117
525
1.004363
525
0.663512
525
0.846392
525
0.997873
525
0.934484
525
0.960648
525
1.079167
525
1.00003
525
0.450976
525
0.497861
525
0.670197
525
0.47692
525
0.791928
525
0.876623
550
1.0004
550
0.650583
550
0.999153
550
0.999367
550
0.852782
550
0.845799
550
0.869403
550
0.876623
550
1.006522
550
0.652257
550
0.839338
550
0.997748
550
0.931358
550
0.958333
550
1.075
550
0.99695
550
0.422845
550
0.471238
550
0.654664
550
0.44935
550
0.782668
550
0.870304
575
1.0006
575
0.634407
575
0.998729
575
0.999051
575
0.846285
575
0.838827
575
0.863014
575
0.87013
575
1.00868
575
0.637151
575
0.832298
575
0.997623
575
0.928232
575
0.956019
575
1.070833
575
0.993868
575
0.3947
575
0.444633
575
0.639392
575
0.421886
575
0.77344
575
0.863994
600
1.0007
600
0.618468
600
0.998305
600
0.998734
600
0.839798
600
0.831858
600
0.856624
600
0.863636
600
1.010837
600
0.622283
600
0.825273
600
0.997497
600
0.925106
600
0.953704
600
1.066667
600
0.990784
600
0.366543
600
0.418045
600
0.624376
600
0.394522
600
0.764248
600
0.857692
625
1.0008
625
0.602762
625
0.997881
625
0.998418
625
0.833323
625
0.824891
625
0.850236
625
0.857143
625
1.012992
625
0.607651
625
0.818262
625
0.997372
625
0.92198
625
0.951389
625
1.0625
625
0.987699
625
0.338373
625
0.391473
625
0.609613
625
0.367251
625
0.755093
625
0.851399
650
1.0009
650
0.587284
650
0.997458
650
0.998101
650
0.82686
650
0.817926
650
0.843848
650
0.850649
650
1.015145
650
0.593254
650
0.811267
650
0.997247
650
0.918855
650
0.949074
650
1.058333
650
0.98461
650
0.310193
650
0.364916
650
0.5951
650
0.340069
650
0.745975
650
0.845115
675
1.0011
675
0.57203
675
0.997034
675
0.997785
675
0.820409
675
0.810965
675
0.837461
675
0.844156
675
1.017296
675
0.57909
675
0.804289
675
0.997122
675
0.915729
675
0.946759
675
1.054167
675
0.981519
675
0.282003
675
0.338373
675
0.580834
675
0.312968
675
0.736897
675
0.83884
700
1.0012
700
0.556996
700
0.99661
700
0.997468
700
0.813971
700
0.804007
700
0.831075
700
0.837662
700
1.019445
700
0.565157
700
0.797327
700
0.996997
700
0.912603
700
0.944444
700
1.05
700
0.978424
700
0.253804
700
0.311843
700
0.56681
700
0.285944
700
0.72786
700
0.832576
725
1.0013
725
0.542178
725
0.996186
725
0.997152
725
0.807545
725
0.797052
725
0.824689
725
0.831169
725
1.02159
725
0.551453
725
0.790383
725
0.996872
725
0.909477
725
0.94213
725
1.045833
725
0.975325
725
0.225597
725
0.285325
725
0.553027
725
0.25899
725
0.718865
725
0.826322
750
1.0014
750
0.527573
750
0.995763
750
0.996835
750
0.801134
750
0.7901
750
0.818304
750
0.824675
750
1.023733
750
0.537975
750
0.783457
750
0.996747
750
0.906352
750
0.939815
750
1.041667
750
0.972221
750
0.197383
750
0.258817
750
0.539481
750
0.232101
750
0.709914
750
0.820079

122
Spline Look-Up Table
Note: This is only a portion of the whole table, and does not show maximum
ranges.
REAL
WORLD
(meters)
GRID
120
mm
BLOS
guided
munition
xm307
25mm
155
mm
xm307
25mm
payload
assit
mod
(PAM)
MK44
30
mm
M240B
7.62mm
Guided
xm36
120mm
xm307
25mm
MK44
30
mm
M240B
7.62mm
xm307
25
mm
Javelin
Anti
Tank
Missle
xm307
25
mm
Guided
Hellfire
xm307
25
mm
m16
m240B
7.62mm
m230
/
30
mm
Guided
LOCAAS
m230
/
30
mm
Guided
Hellfire
2A-42
/30
mm
Guided
2A-70M100mm
tube
firing
AT12
guided
stabber
82
mm
Mortar
Guided
SA-16
Surface
to
Air
Missle
anti
tank
grenade
launcher
anti
tank
missle
rpk
74
light
machine
gun
ak
m
/
47
rifle
SVD
7.62
sniper
ak
47
rifle
2A-46
/125mm
0
-
0.0000
0.9161
0.0000
0.9161
0.0000
0.9975
0.9996
0.0000
0.9161
0.9975
0.9996
0.9161
0.0000
0.9161
0.0000
0.9161
0.9360
0.9996
0.9966
1.0000
0.9966
0.0000
1.0000
0.0000
0.0000
0.0000
0.0000
0.0000
0.9109
0.9198
0.9914
0.9198
0.0000
40
8
0.0000
0.8972
0.0000
0.8972
0.0000
0.9869
0.9884
0.9998
0.8972
0.9869
0.9884
0.8972
0.0000
0.8972
0.0000
0.8972
0.9140
0.9884
0.9851
1.0000
0.9851
0.0000
0.9951
0.0000
0.0000
0.0000
0.0000
0.9936
0.8925
0.8941
0.9760
0.8941
0.0000
50
10
0.0000
0.8925
0.0000
0.8925
0.0000
0.9843
0.9856
0.9973
0.8925
0.9843
0.9856
0.8925
0.0000
0.8925
0.0000
0.8925
0.9085
0.9856
0.9822
1.0000
0.9822
0.0000
0.9939
0.0000
0.0000
0.0000
1.0000
0.9834
0.8879
0.8877
0.9722
0.8877
0.9977
75
14
0.0000
0.8808
0.0000
0.8808
0.0000
0.9777
0.9786
0.9909
0.8808
0.9777
0.9786
0.8808
1.0000
0.8808
0.0000
0.8808
0.8947
0.9786
0.9750
1.0000
0.9750
0.0000
0.9907
0.0000
0.0000
0.0000
0.9759
0.9578
0.8764
0.8716
0.9626
0.8716
0.9913
100
19
0.0000
0.8691
0.0000
0.8691
0.0000
0.9710
0.9716
0.9845
0.8691
0.9710
0.9716
0.8691
0.9935
0.8691
0.0000
0.8691
0.8810
0.9716
0.9678
1.0000
0.9678
0.0000
0.9876
1.0000
0.0000
0.0000
0.9467
0.9323
0.8650
0.8556
0.9530
0.8556
0.9849
125
24
0.0000
0.8574
0.0000
0.8574
0.0000
0.9644
0.9646
0.9781
0.8574
0.9644
0.9646
0.8574
0.9870
0.8574
0.0000
0.8574
0.8674
0.9646
0.9606
0.9999
0.9606
0.0000
0.9845
0.9977
0.0000
0.0000
0.9175
0.9067
0.8536
0.8397
0.9434
0.8397
0.9785
150
29
0.0000
0.8458
0.0000
0.8458
0.0000
0.9578
0.9576
0.9717
0.8458
0.9578
0.9576
0.8458
0.9805
0.8458
0.0000
0.8458
0.8538
0.9576
0.9534
0.9997
0.9534
0.0000
0.9814
0.9954
0.0000
0.0000
0.8883
0.8812
0.8424
0.8238
0.9338
0.8238
0.9721
175
34
0.0000
0.8344
0.0000
0.8344
0.0000
0.9512
0.9506
0.9653
0.8344
0.9512
0.9506
0.8344
0.9740
0.8344
0.0000
0.8344
0.8402
0.9506
0.9462
0.9996
0.9462
0.0000
0.9782
0.9931
0.0000
0.0000
0.8592
0.8556
0.8312
0.8080
0.9242
0.8080
0.9657
200
38
0.0000
0.8230
0.0000
0.8230
0.0000
0.9446
0.9436
0.9589
0.8230
0.9446
0.9436
0.8230
0.9675
0.8230
0.0000
0.8230
0.8267
0.9436
0.9391
0.9995
0.9391
0.0000
0.9751
0.9907
0.0000
0.0000
0.8300
0.8301
0.8201
0.7923
0.9147
0.7923
0.9593
225
43
0.0000
0.8117
0.0000
0.8117
0.0000
0.9380
0.9366
0.9525
0.8117
0.9380
0.9366
0.8117
0.9610
0.8117
0.0000
0.8117
0.8133
0.9366
0.9319
0.9994
0.9319
0.0000
0.9720
0.9884
0.0000
0.0000
0.8008
0.8046
0.8091
0.7767
0.9051
0.7767
0.9529
250
48
0.0000
0.8005
0.0000
0.8005
0.0000
0.9314
0.9296
0.9461
0.8005
0.9314
0.9296
0.8005
0.9545
0.8005
0.0000
0.8005
0.8001
0.9296
0.9247
0.9992
0.9247
0.0000
0.9689
0.9861
0.0000
0.0000
0.7716
0.7791
0.7983
0.7613
0.8956
0.7613
0.9465
275
53
0.0000
0.7895
0.0000
0.7895
0.0000
0.9248
0.9226
0.9397
0.7895
0.9248
0.9226
0.7895
0.9481
0.7895
0.0000
0.7895
0.7869
0.9226
0.9175
0.9991
0.9175
0.0000
0.9657
0.9838
0.0000
0.0000
0.7424
0.7535
0.7876
0.7459
0.8860
0.7459
0.9402
300
58
0.0000
0.7787
0.0000
0.7787
0.0000
0.9182
0.9156
0.9333
0.7787
0.9182
0.9156
0.7787
0.9416
0.7787
0.0000
0.7787
0.7738
0.9156
0.9104
0.9990
0.9104
0.0000
0.9626
0.9815
0.0000
0.0000
0.7132
0.7280
0.7771
0.7308
0.8765
0.7308
0.9338
325
63
0.0000
0.7680
0.0000
0.7680
0.0000
0.9117
0.9086
0.9269
0.7680
0.9117
0.9086
0.7680
0.9351
0.7680
0.0000
0.7680
0.7609
0.9086
0.9032
0.9989
0.9032
0.0000
0.9595
0.9792
0.0000
0.0000
0.6839
0.7025
0.7668
0.7020
0.8670
0.7020
0.9274
350
67
0.0000
0.7575
0.0000
0.7575
0.0000
0.9051
0.9017
0.9205
0.7575
0.9051
0.9017
0.7575
0.9286
0.7575
0.0000
0.7575
0.7481
0.9017
0.8961
0.9987
0.8961
0.0000
0.9564
0.9769
0.0000
0.0000
0.6547
0.6771
0.7566
0.6734
0.8576
0.6734
0.9210
375
72
0.0000
0.7473
0.0000
0.7473
0.0000
0.8985
0.8947
0.9141
0.7473
0.8985
0.8947
0.7473
0.9221
0.7473
0.0000
0.7473
0.7355
0.8947
0.8890
0.9986
0.8890
0.0000
0.9532
0.9745
0.0000
0.0000
0.6255
0.6516
0.7467
0.6449
0.8481
0.6449
0.9147
400
77
0.0000
0.7372
0.0000
0.7372
0.0000
0.8920
0.8877
0.9078
0.7372
0.8920
0.8877
0.7372
0.9156
0.7372
0.0000
0.7372
0.7230
0.8877
0.8818
0.9985
0.8818
0.0000
0.9501
0.9722
0.0000
0.0000
0.5962
0.6261
0.7370
0.6166
0.8387
0.6166
0.9083
425
82
0.0000
0.7273
0.0000
0.7273
0.0000
0.8854
0.8807
0.9014
0.7273
0.8854
0.8807
0.7273
0.9091
0.7273
0.0000
0.7273
0.7107
0.8807
0.8747
0.9984
0.8747
0.0000
0.9470
0.9699
0.0000
0.0000
0.5670
0.6007
0.7276
0.5884
0.8293
0.5884
0.9020
450
87
0.0000
0.7177
0.0000
0.7177
0.0000
0.8789
0.8737
0.8950
0.7177
0.8789
0.8737
0.7177
0.9026
0.7177
0.0000
0.7177
0.6986
0.8737
0.8676
0.9982
0.8676
0.0000
0.9439
0.9676
0.0000
0.0000
0.5377
0.5753
0.7184
0.5603
0.8199
0.5603
0.8956
475
91
0.0000
0.7006
0.0000
0.7006
0.0000
0.8723
0.8667
0.8886
0.7006
0.8723
0.8667
0.7006
0.8961
0.7006
0.0000
0.7006
0.6867
0.8667
0.8605
0.9981
0.8605
0.0000
0.9407
0.9653
0.0000
0.0000
0.5084
0.5499
0.7021
0.5324
0.8105
0.5324
0.8893
500
96
1.0000
0.6837
1.0000
0.6837
1.0000
0.8658
0.8598
0.8822
0.6837
0.8658
0.8598
0.6837
0.8896
0.6837
1.0000
0.6837
0.6750
0.8598
0.8535
0.9980
0.8535
1.0000
0.9376
0.9630
0.0000
1.0000
0.4791
0.5245
0.6860
0.5046
0.8012
0.5046
0.8829
525
101
1.0000
0.6670
0.9996
0.6670
0.9997
0.8593
0.8528
0.8758
0.6670
0.8593
0.8528
0.6670
0.8831
0.6670
1.0000
0.6670
0.6635
0.8528
0.8464
0.9979
0.8464
1.0000
0.9345
0.9606
0.0000
1.0000
0.4510
0.4979
0.6702
0.4769
0.7919
0.4769
0.8766
550
106
1.0000
0.6506
0.9992
0.6506
0.9994
0.8528
0.8458
0.8694
0.6506
0.8528
0.8458
0.6506
0.8766
0.6506
1.0000
0.6506
0.6523
0.8458
0.8393
0.9977
0.8393
1.0000
0.9314
0.9583
0.0000
0.9969
0.4228
0.4712
0.6547
0.4493
0.7827
0.4493
0.8703
575
111
1.0000
0.6344
0.9987
0.6344
0.9991
0.8463
0.8388
0.8630
0.6344
0.8463
0.8388
0.6344
0.8701
0.6344
1.0000
0.6344
0.6372
0.8388
0.8323
0.9976
0.8323
1.0000
0.9282
0.9560
0.0000
0.9939
0.3947
0.4446
0.6394
0.4219
0.7734
0.4219
0.8640
600
115
1.0000
0.6185
0.9983
0.6185
0.9987
0.8398
0.8319
0.8566
0.6185
0.8398
0.8319
0.6185
0.8636
0.6185
1.0000
0.6185
0.6223
0.8319
0.8253
0.9975
0.8253
1.0000
0.9251
0.9537
0.0000
0.9908
0.3665
0.4180
0.6244
0.3945
0.7642
0.3945
0.8577
625
120
1.0000
0.6028
0.9979
0.6028
0.9984
0.8333
0.8249
0.8502
0.6028
0.8333
0.8249
0.6028
0.8571
0.6028
1.0000
0.6028
0.6077
0.8249
0.8183
0.9974
0.8183
1.0000
0.9220
0.9514
0.0000
0.9877
0.3384
0.3915
0.6096
0.3673
0.7551
0.3673
0.8514
650
125
1.0000
0.5873
0.9975
0.5873
0.9981
0.8269
0.8179
0.8438
0.5873
0.8269
0.8179
0.5873
0.8506
0.5873
1.0000
0.5873
0.5933
0.8179
0.8113
0.9972
0.8113
1.0000
0.9189
0.9491
0.0000
0.9846
0.3102
0.3649
0.5951
0.3401
0.7460
0.3401
0.8451

123
H. ARMOR AND CONCEALMENT
ballistic
protection
active
measures
passive
measures
threat
warning
receivers
countermine
body
armor
MANA
Value
=
75%
of
the
proportion
value
to
the
max
value
Human
In
The
Loop
ie
using
terrrain
or
cammo
None
level
1
level
2
changed
%
for
modeling
purposes
MANA
VALUE
Red BMP-3 1 2 2 1 1 1 43 1 2 30
Red 82 Mortors 0 1 0 1 0 1 16 1 1 20
Red SA-16 Infantryman 0 0 0 0 0 1 5 1 10
Red RPG-7 0 0 0 0 0 1 5 1 10
Red AT-7 0 0 0 0 0 1 5 1 10
Red Scout 0 0 0 0 0 1 5 1 2 x 60
Red RPK-74 0 0 0 0 0 1 5 1 10
Red AK-M Infantryman 0 0 0 1 0 1 11 1 10
Red SVD 0 0 0 0 0 1 5 1 10
Red APC 2 1 2 2 1 0 43 1 1 20
Red T72 1 2 2 1 1 1 43 1 1 20
Blue NLOS Mortor Sec 4 2 3 3 1 1 75 1 1 20
Blue NLOS Cannon Plt 4 1 3 3 1 1 100 1 1 x 100
Blue NLOS LS Plt 0 0 0 0 0 1 100 1 0 x 100
Blue ICV Platoon 3 1 3 3 2 1 70 1 1 20
Blue MCS Platoon 4 1 3 3 2 1 75 1 1 20
Blue ARV-A 3 2 2 1 1 1 54 1 2 30
Blue ARV-A(L) 2 1 0 1 1 1 32 1 1 20
Blue ARV-RSTA 3 2 2 1 1 1 54 1 2 30
Blue UAV CL 1 0 0 0 0 0 0 0 0 x 90
Blue UAV CL 2 0 0 0 0 0 0 0 0 x 90
Blue UAV CL 3 3 2 3 3 1 1 70 0 2 x 90
Blue R&SV 0 0 0 1 0 2 16 1 10
Blue Infantryman 0 0 0 1 0 2 16 1 10
Blue MachineGunner M240b 0 1 2 0 0 0 16 1 0 10
Blue CAS 0 1 2 1 0 0 100 1 0 x 100
Blue Apache 0 1 2 1 0 0 100 100
Auto Cannon
Integrate
APS CBRN LWR
AT Mine
Protection
HMG
Smoke
Greanades EMP MWR (UV)
AO Mine
Protection
HE Frag
Smart Top
Attack
Fixed
Wavelength
Laser
NBC
Warning
Internal
Critical
Component
Ballistic Prot
Top Attack EM Armor
Fire
Extinguishers
JCAD Chem
Point Det
14.5mm all
around
LVOSS
Smoke
Dispensing
Fire
Suppression
152 mm HE
Frag Local SA ERA
Categories HITL and Signature Management
consisting of these individual capabilites
Armor Thickness Concealment

124

125
APPENDIX B. DOE MODELING
A. DOE SPREADSHEET MODELING
This appendix outlines the crossed NOLH DOE. There exist three spreadsheet
models within this appendix. The first is the factor description and is similar to that of
Table 13. It outlines both the controlled and uncontrolled noise factors creating the
robust design. The second spreadsheet is a NOLH coded spreadsheet for 17-22 factors
detailing the factor levels used at each of the 129 design points.96 The third spreadsheet
is a design file, similar to the second, but adds the additional 9 correlated factors to each
of the UAV p(det) factors, but at extended ranges. The design file is the final crossed
NOLH DOE with 258 design points.
Factor
Number Potential Controlled Factors
Effecting
Modeled
Squad Units Low Level High Level Mana factor Mana Low Mana High
1 number of UAVs CL I per team 20,21,22,23 0 6 UAV CL I 0 6
2 number of UAVs CL II per team 24,25,26,27 0 6 UAV CL II 0 6
3 number of UAVs CL III 28 0 16 UAV CL III 0 16
4
number of Hellfire missiles in
UAV Warrior 28 0 4 Rounds 0 4
5
number of APKWS missiles in
UAV CL III 28 0 8 Rounds 0 8
6 sensor range P(det) UAV CL I 20,21,22,23 0% 2%
Sensor
Cababilities 0 2000
7 sensor range P(det) UAV CL II 24,25,26,27 0% 2%
Sensor
Cababilities 0 2000
8 sensor range P(det) UAV CL III 28 0% 2%
Sensor
Cababilities 0 2000
9
enemy UAV CL I and II
20,21,22,23,
24,25,26,27 0 20
Agent SA
Enemies 0 20
10
point UAV CL I and II
20,21,22,23,
24,25,26,27 0 20
Agent SA
Next Way
Point 0 20
11
enemy UAV CL III 28 0 20
Agent SA
Next Way
Point 0 20
12
point UAV CL III 28 0 20
Agent SA
Next Way
Point 0 20
13 UAV CL I flying speed (kmph) 20,21,22,23 60 80 speed 261 427
14 UAV CL II flying speed (kmph) 24,25,26,27 80 100 speed 427 534
15 UAV CL III flying speed (kmph) 28, 80 140 speed 427 748
Potential Noise Factors
16
number of initial enemy high
pay off targets
1,2,3,6,10,
11 1 12 No. of agents 1 12
17
map editor city cover and
concealment all 1% 100% all 0.01 1
18
map editor inside building cover
and concealment all 1% 100% all 0.01 1
19
Communication Reliabilty due
to inclement weather 20-28 0.75 1 reliabilty 75 100
20 UAV Concealment 20-28 0 0.9 concealment 0 90
Model Values Converted MANA Values
96 NOLH 17-22 Factors, coded by Professor Susan Sanchez, Naval Postgraduate School, Monterey,
California.

126
low level 0 0 0 0 0 0 0 0 0 0 0 0 261 427 427 1 0 0 75 0
high level 6 6 16 0 8 2 2 2 20 20 20 20 374 534 748 12 1 1 100 90
decimals 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 3 3 0
factor name
number
of UAVs
CL I per
team
number
of UAVs
CL II per
team
number
of UAVs
CL III
number
of
missiles
in UAV
CL III
number
of
APKWS
missiles
in UAV
CL III
sensor
range
and
P(det)
UAV CL I
sensor
range
and
P(det)
UAV CL II
sensor
range
and
P(det)
UAV CL
III
Agents
desire to
go after
enemy
UAV CL I
and II
Agents
desire to
go to
next way
point
UAV CL I
and II
Agents
desire to
go after
enemy
UAV CL
III
Agents
desire to
go to
next way
point
UAV CL
III
UAV CL I
flying
speed
UAV CL II
flying
speed
UAV CL
III flying
speed
number
of initial
enemy
high pay
off
targets
map
editor
city
cover
and
conceal
ment
map
editor
inside
building
cover
and
conceal
ment
Commun
ication
Reliabilty
due to
inclemen
t weather
UAV
Conceal
ment
1 1 3 6 0 3 1 1 2 11 19 15 15 329 534 725 8 0.844 0.539 97.461 86
2 5 2 7 0 1 1 1 0 10 8 18 12 331 528 660 10 0.906 0.578 90.234 65
3 3 5 0 0 3 0 2 1 12 11 9 3 316 519 745 8 0.555 0.148 95.898 59
4 4 5 5 0 4 2 0 1 5 9 8 8 299 532 643 12 0.609 0.359 89.648 70
5 0 2 9 0 1 1 1 2 20 13 2 20 357 455 570 6 0.875 0.672 93.164 58
6 4 3 11 0 3 1 1 0 5 2 4 20 332 479 527 5 0.922 0.875 93.945 66
7 2 6 12 0 1 0 0 2 19 15 14 2 261 439 577 4 0.914 0.344 98.438 61
8 3 4 14 0 3 2 2 0 6 4 10 2 280 457 535 6 0.469 0.094 99.219 52
9 0 0 4 0 2 2 1 1 15 17 15 5 328 455 730 10 0.781 0.977 83.203 2
10 6 0 4 0 1 0 0 1 4 3 10 3 340 440 720 6 0.727 0.695 80.469 33
11 0 6 8 0 3 0 2 0 12 17 3 11 292 451 668 11 0.719 0.211 82.422 0
12 6 6 4 0 1 2 1 2 6 6 5 13 289 474 698 7 0.586 0.195 79.883 6
13 3 2 15 0 3 1 0 1 13 18 0 4 363 505 530 2 0.766 0.828 86.523 43
14 5 1 15 0 1 0 2 1 4 8 0 6 356 527 542 3 0.672 0.484 87.305 22
15 2 3 16 0 4 0 0 0 20 16 17 15 287 483 522 2 0.625 0.109 78.516 41
16 5 5 16 0 1 1 2 1 5 8 19 12 298 490 510 2 0.984 0.375 89.258 25
17 1 1 3 0 0 1 1 1 4 12 13 12 284 533 440 10 0.07 0.898 95.703 6
18 6 2 4 0 3 0 0 1 15 1 12 14 310 493 452 8 0.477 0.773 98.828 27
19 1 5 5 0 2 0 1 1 5 18 6 10 341 530 447 8 0.055 0.414 84.57 28
20 4 6 6 0 1 2 1 0 11 5 1 9 335 492 432 8 0.359 0.047 87.891 21
21 2 1 13 0 4 1 1 2 2 13 3 20 263 475 690 2 0.109 0.664 99.805 34
22 4 2 10 0 3 0 2 1 19 11 1 12 262 440 603 5 0.172 0.617 96.875 5
23 1 4 12 0 3 1 0 2 1 18 16 1 363 470 683 7 0.164 0.25 95.313 35
24 4 4 16 0 2 1 2 1 20 8 13 3 366 458 638 4 0.023 0.445 91.016 19
25 2 1 7 0 3 2 1 1 9 16 11 9 291 438 442 10 0.336 0.914 78.32 83
26 6 0 7 0 1 1 1 2 13 6 13 7 264 448 547 11 0.148 0.938 84.375 89
27 1 4 4 0 0 0 1 1 6 17 9 14 318 445 550 7 0.492 0.133 80.664 78
28 6 6 6 0 1 2 1 1 20 1 6 16 330 430 590 9 0.211 0.313 85.938 75
29 2 3 12 0 3 1 0 1 6 9 1 1 265 499 673 2 0.438 0.883 76.953 76
30 4 3 16 0 4 1 1 2 11 7 6 3 302 502 748 1 0.352 0.836 82.813 49
31 3 4 11 0 1 1 0 0 3 17 19 16 322 496 608 2 0.18 0.477 88.086 70
32 3 4 15 0 1 1 1 2 16 0 15 14 338 501 663 4 0.398 0.297 75.977 53
33 0 2 5 0 4 1 2 1 18 10 18 15 325 518 635 2 0.039 0.398 75 80
34 5 3 2 0 4 1 1 0 7 10 18 19 315 497 693 3 0.25 0.32 84.18 54
35 2 6 1 0 6 0 2 2 11 8 5 7 277 495 615 3 0.313 0.547 81.25 89
36 3 5 3 0 5 2 0 0 8 16 5 2 282 491 655 5 0.422 0.594 77.734 56
37 0 2 10 0 6 2 1 1 12 0 4 15 345 452 575 7 0.305 0.07 85.547 74
38 5 1 8 0 8 1 1 1 2 16 3 15 360 450 545 8 0.008 0.234 83.008 77
39 2 6 15 0 6 0 0 2 14 2 12 8 309 476 475 7 0.453 0.969 81.445 82
40 5 5 14 0 8 2 2 1 3 15 13 10 276 449 470 8 0.227 0.711 83.789 68
41 2 3 1 0 6 1 1 0 17 5 17 1 367 450 688 1 0.063 0.563 98.242 32
42 5 2 7 0 6 1 0 1 4 14 18 4 342 435 593 3 0.102 0.281 96.289 23
43 3 4 3 0 6 1 2 0 17 7 7 16 281 484 738 4 0.133 0.859 99.609 40
44 4 5 2 0 5 1 0 2 6 13 2 17 288 480 618 1 0.242 0.82 94.727 17
45 1 0 8 0 7 1 1 1 18 7 4 1 362 516 467 8 0.195 0.039 88.281 42
46 5 3 10 0 5 0 1 1 8 17 7 3 324 524 517 9 0.383 0.188 94.141 44
47 0 5 11 0 6 0 0 0 17 0 12 9 270 529 490 9 0.258 0.727 86.328 3
48 3 5 13 0 8 2 1 2 9 15 12 13 327 517 472 9 0.297 0.781 91.602 27
49 3 2 3 0 6 2 2 2 8 5 20 9 287 501 460 2 0.969 0.203 76.367 9
50 5 1 2 0 6 1 0 1 19 16 11 12 266 515 497 1 0.734 0.492 81.641 26
51 2 3 5 0 4 0 2 1 2 6 4 0 333 489 562 3 0.633 0.734 79.492 51
52 5 5 2 0 6 1 1 1 15 11 1 6 370 520 565 3 0.859 0.758 82.617 44
53 2 1 10 0 8 2 0 1 7 2 7 13 280 486 695 10 0.57 0.008 83.594 11
54 5 1 15 0 7 0 1 1 10 19 9 18 295 461 670 6 0.656 0 77.344 8
55 1 4 9 0 8 0 0 1 4 1 20 7 334 447 623 10 0.813 0.844 85.156 13
56 5 5 13 0 8 1 1 0 13 20 10 2 312 467 680 11 0.711 0.945 84.961 30
57 1 1 5 0 5 1 2 0 2 8 12 1 285 454 457 2 0.953 0.352 99.414 79
58 4 2 2 0 5 0 1 2 13 11 11 6 267 465 465 4 0.797 0.016 97.852 72
59 1 5 7 0 8 1 1 0 3 6 4 16 361 435 595 4 1 0.57 92.969 60
60 4 4 7 0 6 2 1 1 19 19 5 15 339 473 525 6 0.883 0.609 97.07 51
61 2 2 14 0 4 2 1 0 10 6 6 3 271 525 675 9 0.484 0.078 94.922 72
62 3 1 10 0 6 1 2 1 13 10 3 9 283 498 740 12 0.68 0.258 88.867 75
63 2 4 14 0 6 1 1 0 3 2 14 18 321 531 713 7 0.594 0.633 96.094 53
64 5 3 11 0 7 1 2 1 19 15 17 10 349 508 620 8 0.539 0.531 92.773 86
65 3 3 8 0 4 1 1 1 10 10 10 10 318 481 588 7 0.5 0.5 87.5 45
66 5 3 10 0 5 1 1 0 9 1 5 5 306 427 450 5 0.156 0.461 77.539 4
67 1 4 9 0 7 1 1 2 10 12 2 8 304 433 515 3 0.094 0.422 84.766 25
68 3 1 16 0 5 2 1 1 8 9 11 17 319 442 430 5 0.445 0.852 79.102 31
69 2 1 11 0 5 0 2 1 15 11 12 13 336 429 532 1 0.391 0.641 85.352 20
70 6 4 7 0 7 1 1 0 0 7 18 0 278 506 605 7 0.125 0.328 81.836 32
71 2 3 5 0 5 1 1 2 15 18 16 0 303 482 648 8 0.078 0.125 81.055 24
72 4 0 4 0 7 2 2 0 1 5 6 18 374 522 598 9 0.086 0.656 76.563 29
73 3 2 2 0 5 0 0 2 14 16 10 18 355 504 640 7 0.531 0.906 75.781 38
74 6 6 12 0 6 0 1 1 5 3 5 15 307 506 445 3 0.219 0.023 91.797 88
75 0 6 13 0 7 2 2 1 16 17 10 17 295 521 455 7 0.273 0.305 94.531 57
76 6 0 8 0 6 2 0 2 8 3 17 9 343 510 507 2 0.281 0.789 92.578 90
77 0 0 12 0 7 0 1 0 14 14 15 7 346 487 477 6 0.414 0.805 95.117 84
78 3 4 1 0 5 1 2 1 7 2 20 16 272 456 645 11 0.234 0.172 88.477 47
79 2 5 1 0 7 2 0 1 16 12 20 14 279 434 633 10 0.328 0.516 87.695 68
80 4 3 0 0 4 2 2 2 0 4 3 5 348 478 653 11 0.375 0.891 96.484 49
81 1 2 0 0 7 1 0 1 15 12 1 8 337 471 665 11 0.016 0.625 85.742 65
82 5 5 13 0 8 1 1 1 16 8 8 8 351 428 735 3 0.93 0.102 79.297 84
83 0 4 12 0 5 2 2 1 5 19 8 6 325 468 723 5 0.523 0.227 76.172 63
84 5 1 12 0 6 2 1 1 15 2 14 10 294 431 728 5 0.945 0.586 90.43 62
85 2 0 10 0 7 0 1 2 9 15 19 11 300 469 743 5 0.641 0.953 87.109 69
86 4 5 3 0 4 1 1 0 18 7 17 0 372 486 485 11 0.891 0.336 75.195 56
87 2 4 6 0 5 2 0 1 1 9 19 8 373 521 572 8 0.828 0.383 78.125 85
88 5 2 4 0 5 1 2 0 19 3 4 19 272 491 492 6 0.836 0.75 79.688 55
89 2 2 1 0 7 1 0 1 0 12 7 18 269 503 537 9 0.977 0.555 83.984 71
90 4 5 9 0 5 0 1 1 11 4 9 11 344 523 733 3 0.664 0.086 96.68 7
91 0 6 9 0 7 1 1 0 8 14 7 13 371 513 628 2 0.852 0.063 90.625 1
92 5 2 12 0 8 2 1 1 14 3 11 6 317 516 625 6 0.508 0.867 94.336 12
93 0 0 10 0 7 1 1 1 0 19 14 4 305 531 585 4 0.789 0.688 89.063 15
94 4 3 4 0 5 1 2 1 14 11 19 19 370 462 502 11 0.563 0.117 98.047 14
95 2 3 0 0 4 1 1 0 9 13 14 17 333 459 427 12 0.648 0.164 92.188 41
96 3 2 5 0 7 1 2 2 18 3 1 4 313 465 567 11 0.82 0.523 86.914 20
97 3 2 1 0 7 1 1 0 4 20 5 6 297 460 512 9 0.602 0.703 99.023 37
98 6 4 11 0 4 1 0 1 2 10 2 5 310 443 540 11 0.961 0.602 100 10
99 1 3 14 0 4 1 1 2 13 10 2 1 320 464 482 10 0.75 0.68 90.82 36
100 4 0 15 0 2 2 0 0 9 12 15 13 358 466 560 10 0.688 0.453 93.75 1
101 3 1 13 0 3 0 2 2 12 4 15 18 353 470 520 8 0.578 0.406 97.266 34
102 6 4 6 0 2 0 1 1 8 20 16 5 290 509 600 6 0.695 0.93 89.453 16
103 1 5 8 0 0 1 1 1 18 4 17 5 275 511 630 5 0.992 0.766 91.992 13
104 4 0 1 0 2 2 2 0 6 18 8 12 326 485 700 6 0.547 0.031 93.555 8
105 1 1 2 0 1 0 0 1 17 5 7 10 359 512 705 5 0.773 0.289 91.211 23
106 4 3 15 0 2 1 1 2 3 15 3 19 268 511 487 12 0.938 0.438 76.758 58
107 1 4 9 0 2 1 2 1 16 6 2 16 293 526 582 10 0.898 0.719 78.711 67
108 3 2 13 0 2 1 0 2 3 13 13 4 354 477 437 9 0.867 0.141 75.391 50
109 2 1 14 0 3 1 2 0 14 7 18 3 347 481 557 12 0.758 0.18 80.273 73
110 5 6 8 0 1 1 1 1 2 13 16 19 273 445 708 5 0.805 0.961 86.719 48
111 1 3 6 0 3 2 1 1 12 3 13 17 311 437 658 4 0.617 0.813 80.859 46
112 6 1 5 0 2 2 2 2 3 20 8 11 365 432 685 4 0.742 0.273 88.672 87
113 3 1 3 0 0 0 1 0 11 5 8 7 308 444 703 4 0.703 0.219 83.398 63
114 3 4 13 0 2 0 0 0 12 15 0 11 348 460 715 11 0.031 0.797 98.633 81
115 1 5 14 0 2 1 2 1 1 4 9 8 369 446 678 12 0.266 0.508 93.359 64
116 4 3 11 0 4 2 0 1 18 14 16 20 302 472 613 10 0.367 0.266 95.508 39
117 1 1 14 0 2 1 1 1 5 9 19 14 265 441 610 10 0.141 0.242 92.383 46
118 4 5 7 0 0 0 2 1 13 18 13 7 355 475 480 3 0.43 0.992 91.406 79
119 1 5 1 0 1 2 1 2 10 1 11 2 340 500 505 7 0.344 1 97.656 82
120 5 2 8 0 0 2 2 1 16 19 0 13 301 514 552 3 0.188 0.156 89.844 77
121 1 1 3 0 0 1 1 2 7 0 10 18 323 494 495 2 0.289 0.055 90.039 60
122 5 5 11 0 3 1 0 2 18 13 8 19 350 507 718 11 0.047 0.648 75.586 11
123 2 4 15 0 3 2 1 0 7 9 9 14 368 496 710 9 0.203 0.984 77.148 18
124 5 1 9 0 0 1 1 2 17 14 16 4 274 526 580 9 0 0.43 82.031 30
125 2 2 9 0 2 0 1 1 1 1 15 5 296 488 650 7 0.117 0.391 77.93 39
126 4 4 2 0 4 0 1 2 10 14 14 17 364 436 500 4 0.516 0.922 80.078 18
127 3 5 6 0 2 1 0 1 7 10 18 11 352 463 435 1 0.32 0.742 86.133 15
128 4 2 3 0 2 1 1 2 17 18 6 2 314 430 462 6 0.406 0.367 78.906 37
129 1 3 6 0 1 1 0 1 1 5 3 10 286 453 555 5 0.461 0.469 82.227 4

127
First Half of Crossed NOLH DOE (Hellfire Portion)
The crossed design is 258 rows in length. The first 129 rows vary the number of Hellfire
missiles from zero to four, while keeping the number of APKWS missiles at zero. The
second 129 rows vary the number of APKWS missiles from zero to eight, while keeping
the number of Hellfire Missiles at zero. The full design is too long to show on a single
page. This first chart is only the first 129 rows of the entire DOE. The chart on the
following page is only the second 129 rows of the entire DOE.
number of
UAVs CL I
per team
number of
UAVs CL II
per team
number of
UAVs CL III
number of
Hellfire in
UAV CL III
number of
APKWS in
UAV CL III
sensor
P(det) pt 0
UAV CL I
sensor
P(det) pt 2
UAV CL I
sensor
P(det) pt 3
UAV CL I
sensor
P(det) pt 4
UAV CL I
sensor
P(det) pt 0
UAV CL II
sensor
P(det) pt 2
UAV CL II
sensor
P(det) pt 3
UAV CL II
sensor
P(det) pt 4
UAV CL II
sensor
P(det) pt 0
UAV CL III
sensor
P(det) pt 2
UAV CL III
sensor
P(det) pt 3
UAV CL III
sensor
P(det) pt 4
UAV CL III
Agents
desire to
go after
enemy
UAV CL I
and II
Agents
desire to
go to next
way point
UAV CL I
and II
Agents
desire to
go after
enemy
UAV CL III
Agents
desire to
go to next
way point
UAV CL III
UAV CL I
flying
speed
UAV CL II
flying
speed
UAV CL III
flying
speed
number of
initial
enemy
high pay
off targets
map editor
city cover
and
concealme
nt
map editor
inside
building
cover and
concealme
nt
Communic
ation
Reliabilty
due to
inclement
weather
UAV
Concelm
ent
1 3 6 2 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 11 19 15 15 329 534 725 8 0.844 0.539 97.461 86
5 2 7 2 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 10 8 18 12 331 528 660 10 0.906 0.578 90.234 65
3 5 0 1 0 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 12 11 9 3 316 519 745 8 0.555 0.148 95.898 59
4 5 5 1 0 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 5 9 8 8 299 532 643 12 0.609 0.359 89.648 70
0 2 9 1 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 20 13 2 20 357 455 570 6 0.875 0.672 93.164 58
4 3 11 0 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 5 2 4 20 332 479 527 5 0.922 0.875 93.945 66
2 6 12 1 0 0 3000 6000 8000 0 3000 6000 8000 2000 5000 8000 10000 19 15 14 2 261 439 577 4 0.914 0.344 98.438 61
3 4 14 0 0 2000 5000 8000 10000 2000 5000 8000 10000 0 3000 6000 8000 6 4 10 2 280 457 535 6 0.469 0.094 99.219 52
0 0 4 1 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 15 17 15 5 328 455 730 10 0.781 0.977 83.203 2
6 0 4 1 0 0 3000 6000 8000 0 3000 6000 8000 1000 4000 7000 9000 4 3 10 3 340 440 720 6 0.727 0.695 80.469 33
0 6 8 1 0 0 3000 6000 8000 2000 5000 8000 10000 0 3000 6000 8000 12 17 3 11 292 451 668 11 0.719 0.211 82.422 0
6 6 4 2 0 2000 5000 8000 10000 1000 4000 7000 9000 2000 5000 8000 10000 6 6 5 13 289 474 698 7 0.586 0.195 79.883 6
3 2 15 1 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 13 18 0 4 363 505 530 2 0.766 0.828 86.523 43
5 1 15 1 0 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 4 8 0 6 356 527 542 3 0.672 0.484 87.305 22
2 3 16 1 0 0 3000 6000 8000 0 3000 6000 8000 0 3000 6000 8000 20 16 17 15 287 483 522 2 0.625 0.109 78.516 41
5 5 16 0 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 5 8 19 12 298 490 510 2 0.984 0.375 89.258 25
1 1 3 3 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 4 12 13 12 284 533 440 10 0.07 0.898 95.703 6
6 2 4 3 0 0 3000 6000 8000 0 3000 6000 8000 1000 4000 7000 9000 15 1 12 14 310 493 452 8 0.477 0.773 98.828 27
1 5 5 2 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 5 18 6 10 341 530 447 8 0.055 0.414 84.57 28
4 6 6 3 0 2000 5000 8000 10000 1000 4000 7000 9000 0 3000 6000 8000 11 5 1 9 335 492 432 8 0.359 0.047 87.891 21
2 1 13 4 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 2 13 3 20 263 475 690 2 0.109 0.664 99.805 34
4 2 10 4 0 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 19 11 1 12 262 440 603 5 0.172 0.617 96.875 5
1 4 12 4 0 1000 4000 7000 9000 0 3000 6000 8000 2000 5000 8000 10000 1 18 16 1 363 470 683 7 0.164 0.25 95.313 35
4 4 16 4 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 20 8 13 3 366 458 638 4 0.023 0.445 91.016 19
2 1 7 2 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 9 16 11 9 291 438 442 10 0.336 0.914 78.32 83
6 0 7 2 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 13 6 13 7 264 448 547 11 0.148 0.938 84.375 89
1 4 4 4 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 6 17 9 14 318 445 550 7 0.492 0.133 80.664 78
6 6 6 3 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 20 1 6 16 330 430 590 9 0.211 0.313 85.938 75
2 3 12 2 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 6 9 1 1 265 499 673 2 0.438 0.883 76.953 76
4 3 16 3 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 11 7 6 3 302 502 748 1 0.352 0.836 82.813 49
3 4 11 3 0 1000 4000 7000 9000 0 3000 6000 8000 0 3000 6000 8000 3 17 19 16 322 496 608 2 0.18 0.477 88.086 70
3 4 15 4 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 16 0 15 14 338 501 663 4 0.398 0.297 75.977 53
0 2 5 1 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 18 10 18 15 325 518 635 2 0.039 0.398 75 80
5 3 2 0 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 7 10 18 19 315 497 693 3 0.25 0.32 84.18 54
2 6 1 2 0 0 3000 6000 8000 2000 5000 8000 10000 2000 5000 8000 10000 11 8 5 7 277 495 615 3 0.313 0.547 81.25 89
3 5 3 2 0 2000 5000 8000 10000 0 3000 6000 8000 0 3000 6000 8000 8 16 5 2 282 491 655 5 0.422 0.594 77.734 56
0 2 10 0 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 12 0 4 15 345 452 575 7 0.305 0.07 85.547 74
5 1 8 2 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 2 16 3 15 360 450 545 8 0.008 0.234 83.008 77
2 6 15 1 0 0 3000 6000 8000 0 3000 6000 8000 2000 5000 8000 10000 14 2 12 8 309 476 475 7 0.453 0.969 81.445 82
5 5 14 1 0 2000 5000 8000 10000 2000 5000 8000 10000 1000 4000 7000 9000 3 15 13 10 276 449 470 8 0.227 0.711 83.789 68
2 3 1 1 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 17 5 17 1 367 450 688 1 0.063 0.563 98.242 32
5 2 7 0 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 4 14 18 4 342 435 593 3 0.102 0.281 96.289 23
3 4 3 1 0 1000 4000 7000 9000 2000 5000 8000 10000 0 3000 6000 8000 17 7 7 16 281 484 738 4 0.133 0.859 99.609 40
4 5 2 1 0 1000 4000 7000 9000 0 3000 6000 8000 2000 5000 8000 10000 6 13 2 17 288 480 618 1 0.242 0.82 94.727 17
1 0 8 2 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 18 7 4 1 362 516 467 8 0.195 0.039 88.281 42
5 3 10 0 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 8 17 7 3 324 524 517 9 0.383 0.188 94.141 44
0 5 11 2 0 0 3000 6000 8000 0 3000 6000 8000 0 3000 6000 8000 17 0 12 9 270 529 490 9 0.258 0.727 86.328 3
3 5 13 1 0 2000 5000 8000 10000 1000 4000 7000 9000 2000 5000 8000 10000 9 15 12 13 327 517 472 9 0.297 0.781 91.602 27
3 2 3 4 0 2000 5000 8000 10000 2000 5000 8000 10000 2000 5000 8000 10000 8 5 20 9 287 501 460 2 0.969 0.203 76.367 9
5 1 2 2 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 19 16 11 12 266 515 497 1 0.734 0.492 81.641 26
2 3 5 3 0 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 2 6 4 0 333 489 562 3 0.633 0.734 79.492 51
5 5 2 4 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 15 11 1 6 370 520 565 3 0.859 0.758 82.617 44
2 1 10 2 0 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 7 2 7 13 280 486 695 10 0.57 0.008 83.594 11
5 1 15 3 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 10 19 9 18 295 461 670 6 0.656 0 77.344 8
1 4 9 3 0 0 3000 6000 8000 0 3000 6000 8000 1000 4000 7000 9000 4 1 20 7 334 447 623 10 0.813 0.844 85.156 13
5 5 13 3 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 13 20 10 2 312 467 680 11 0.711 0.945 84.961 30
1 1 5 3 0 1000 4000 7000 9000 2000 5000 8000 10000 0 3000 6000 8000 2 8 12 1 285 454 457 2 0.953 0.352 99.414 79
4 2 2 3 0 0 3000 6000 8000 1000 4000 7000 9000 2000 5000 8000 10000 13 11 11 6 267 465 465 4 0.797 0.016 97.852 72
1 5 7 4 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 3 6 4 16 361 435 595 4 1 0.57 92.969 60
4 4 7 3 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 19 19 5 15 339 473 525 6 0.883 0.609 97.07 51
2 2 14 3 0 2000 5000 8000 10000 1000 4000 7000 9000 0 3000 6000 8000 10 6 6 3 271 525 675 9 0.484 0.078 94.922 72
3 1 10 2 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 13 10 3 9 283 498 740 12 0.68 0.258 88.867 75
2 4 14 4 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 3 2 14 18 321 531 713 7 0.594 0.633 96.094 53
5 3 11 3 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 19 15 17 10 349 508 620 8 0.539 0.531 92.773 86
3 3 8 2 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 10 10 10 10 318 481 588 7 0.5 0.5 87.5 45
5 3 10 2 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 9 1 5 5 306 427 450 5 0.156 0.461 77.539 4
1 4 9 2 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 10 12 2 8 304 433 515 3 0.094 0.422 84.766 25
3 1 16 3 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 8 9 11 17 319 442 430 5 0.445 0.852 79.102 31
2 1 11 3 0 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 15 11 12 13 336 429 532 1 0.391 0.641 85.352 20
6 4 7 3 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 0 7 18 0 278 506 605 7 0.125 0.328 81.836 32
2 3 5 4 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 15 18 16 0 303 482 648 8 0.078 0.125 81.055 24
4 0 4 3 0 2000 5000 8000 10000 2000 5000 8000 10000 0 3000 6000 8000 1 5 6 18 374 522 598 9 0.086 0.656 76.563 29
3 2 2 4 0 0 3000 6000 8000 0 3000 6000 8000 2000 5000 8000 10000 14 16 10 18 355 504 640 7 0.531 0.906 75.781 38
6 6 12 3 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 5 3 5 15 307 506 445 3 0.219 0.023 91.797 88
0 6 13 3 0 2000 5000 8000 10000 2000 5000 8000 10000 1000 4000 7000 9000 16 17 10 17 295 521 455 7 0.273 0.305 94.531 57
6 0 8 3 0 2000 5000 8000 10000 0 3000 6000 8000 2000 5000 8000 10000 8 3 17 9 343 510 507 2 0.281 0.789 92.578 90
0 0 12 2 0 0 3000 6000 8000 1000 4000 7000 9000 0 3000 6000 8000 14 14 15 7 346 487 477 6 0.414 0.805 95.117 84
3 4 1 3 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 7 2 20 16 272 456 645 11 0.234 0.172 88.477 47
2 5 1 3 0 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 16 12 20 14 279 434 633 10 0.328 0.516 87.695 68
4 3 0 3 0 2000 5000 8000 10000 2000 5000 8000 10000 2000 5000 8000 10000 0 4 3 5 348 478 653 11 0.375 0.891 96.484 49
1 2 0 4 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 15 12 1 8 337 471 665 11 0.016 0.625 85.742 65
5 5 13 1 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 16 8 8 8 351 428 735 3 0.93 0.102 79.297 84
0 4 12 1 0 2000 5000 8000 10000 2000 5000 8000 10000 1000 4000 7000 9000 5 19 8 6 325 468 723 5 0.523 0.227 76.172 63
5 1 12 2 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 15 2 14 10 294 431 728 5 0.945 0.586 90.43 62
2 0 10 1 0 0 3000 6000 8000 1000 4000 7000 9000 2000 5000 8000 10000 9 15 19 11 300 469 743 5 0.641 0.953 87.109 69
4 5 3 0 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 18 7 17 0 372 486 485 11 0.891 0.336 75.195 56
2 4 6 0 0 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 1 9 19 8 373 521 572 8 0.828 0.383 78.125 85
5 2 4 0 0 1000 4000 7000 9000 2000 5000 8000 10000 0 3000 6000 8000 19 3 4 19 272 491 492 6 0.836 0.75 79.688 55
2 2 1 0 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 0 12 7 18 269 503 537 9 0.977 0.555 83.984 71
4 5 9 2 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 11 4 9 11 344 523 733 3 0.664 0.086 96.68 7
0 6 9 2 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 8 14 7 13 371 513 628 2 0.852 0.063 90.625 1
5 2 12 0 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 14 3 11 6 317 516 625 6 0.508 0.867 94.336 12
0 0 10 1 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 0 19 14 4 305 531 585 4 0.789 0.688 89.063 15
4 3 4 2 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 14 11 19 19 370 462 502 11 0.563 0.117 98.047 14
2 3 0 1 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 9 13 14 17 333 459 427 12 0.648 0.164 92.188 41
3 2 5 1 0 1000 4000 7000 9000 2000 5000 8000 10000 2000 5000 8000 10000 18 3 1 4 313 465 567 11 0.82 0.523 86.914 20
3 2 1 0 0 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 4 20 5 6 297 460 512 9 0.602 0.703 99.023 37
6 4 11 3 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 2 10 2 5 310 443 540 11 0.961 0.602 100 10
1 3 14 4 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 13 10 2 1 320 464 482 10 0.75 0.68 90.82 36
4 0 15 2 0 2000 5000 8000 10000 0 3000 6000 8000 0 3000 6000 8000 9 12 15 13 358 466 560 10 0.688 0.453 93.75 1
3 1 13 2 0 0 3000 6000 8000 2000 5000 8000 10000 2000 5000 8000 10000 12 4 15 18 353 470 520 8 0.578 0.406 97.266 34
6 4 6 4 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 8 20 16 5 290 509 600 6 0.695 0.93 89.453 16
1 5 8 2 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 18 4 17 5 275 511 630 5 0.992 0.766 91.992 13
4 0 1 3 0 2000 5000 8000 10000 2000 5000 8000 10000 0 3000 6000 8000 6 18 8 12 326 485 700 6 0.547 0.031 93.555 8
1 1 2 3 0 0 3000 6000 8000 0 3000 6000 8000 1000 4000 7000 9000 17 5 7 10 359 512 705 5 0.773 0.289 91.211 23
4 3 15 3 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 3 15 3 19 268 511 487 12 0.938 0.438 76.758 58
1 4 9 4 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 16 6 2 16 293 526 582 10 0.898 0.719 78.711 67
3 2 13 3 0 1000 4000 7000 9000 0 3000 6000 8000 2000 5000 8000 10000 3 13 13 4 354 477 437 9 0.867 0.141 75.391 50
2 1 14 3 0 1000 4000 7000 9000 2000 5000 8000 10000 0 3000 6000 8000 14 7 18 3 347 481 557 12 0.758 0.18 80.273 73
5 6 8 2 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 2 13 16 19 273 445 708 5 0.805 0.961 86.719 48
1 3 6 4 0 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 12 3 13 17 311 437 658 4 0.617 0.813 80.859 46
6 1 5 2 0 2000 5000 8000 10000 2000 5000 8000 10000 2000 5000 8000 10000 3 20 8 11 365 432 685 4 0.742 0.273 88.672 87
3 1 3 3 0 0 3000 6000 8000 1000 4000 7000 9000 0 3000 6000 8000 11 5 8 7 308 444 703 4 0.703 0.219 83.398 63
3 4 13 0 0 0 3000 6000 8000 0 3000 6000 8000 0 3000 6000 8000 12 15 0 11 348 460 715 11 0.031 0.797 98.633 81
1 5 14 2 0 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 1 4 9 8 369 446 678 12 0.266 0.508 93.359 64
4 3 11 1 0 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 18 14 16 20 302 472 613 10 0.367 0.266 95.508 39
1 1 14 1 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 5 9 19 14 265 441 610 10 0.141 0.242 92.383 46
4 5 7 2 0 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 13 18 13 7 355 475 480 3 0.43 0.992 91.406 79
1 5 1 2 0 2000 5000 8000 10000 1000 4000 7000 9000 2000 5000 8000 10000 10 1 11 2 340 500 505 7 0.344 1 97.656 82
5 2 8 1 0 2000 5000 8000 10000 2000 5000 8000 10000 1000 4000 7000 9000 16 19 0 13 301 514 552 3 0.188 0.156 89.844 77
1 1 3 1 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 7 0 10 18 323 494 495 2 0.289 0.055 90.039 60
5 5 11 1 0 1000 4000 7000 9000 0 3000 6000 8000 2000 5000 8000 10000 18 13 8 19 350 507 718 11 0.047 0.648 75.586 11
2 4 15 1 0 2000 5000 8000 10000 1000 4000 7000 9000 0 3000 6000 8000 7 9 9 14 368 496 710 9 0.203 0.984 77.148 18
5 1 9 0 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 17 14 16 4 274 526 580 9 0 0.43 82.031 30
2 2 9 1 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 1 1 15 5 296 488 650 7 0.117 0.391 77.93 39
4 4 2 1 0 0 3000 6000 8000 1000 4000 7000 9000 2000 5000 8000 10000 10 14 14 17 364 436 500 4 0.516 0.922 80.078 18
3 5 6 2 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 7 10 18 11 352 463 435 1 0.32 0.742 86.133 15
4 2 3 0 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 17 18 6 2 314 430 462 6 0.406 0.367 78.906 37
1 3 6 1 0 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 1 5 3 10 286 453 555 5 0.461 0.469 82.227 4

128
Second Half of Crossed NOLH DOE (APKWS Portion)
1 3 6 0 3 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 11 19 15 15 329 534 725 8 0.844 0.539 97.461 86
5 2 7 0 1 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 10 8 18 12 331 528 660 10 0.906 0.578 90.234 65
3 5 0 0 3 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 12 11 9 3 316 519 745 8 0.555 0.148 95.898 59
4 5 5 0 4 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 5 9 8 8 299 532 643 12 0.609 0.359 89.648 70
0 2 9 0 1 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 20 13 2 20 357 455 570 6 0.875 0.672 93.164 58
4 3 11 0 3 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 5 2 4 20 332 479 527 5 0.922 0.875 93.945 66
2 6 12 0 1 0 3000 6000 8000 0 3000 6000 8000 2000 5000 8000 10000 19 15 14 2 261 439 577 4 0.914 0.344 98.438 61
3 4 14 0 3 2000 5000 8000 10000 2000 5000 8000 10000 0 3000 6000 8000 6 4 10 2 280 457 535 6 0.469 0.094 99.219 52
0 0 4 0 2 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 15 17 15 5 328 455 730 10 0.781 0.977 83.203 2
6 0 4 0 1 0 3000 6000 8000 0 3000 6000 8000 1000 4000 7000 9000 4 3 10 3 340 440 720 6 0.727 0.695 80.469 33
0 6 8 0 3 0 3000 6000 8000 2000 5000 8000 10000 0 3000 6000 8000 12 17 3 11 292 451 668 11 0.719 0.211 82.422 0
6 6 4 0 1 2000 5000 8000 10000 1000 4000 7000 9000 2000 5000 8000 10000 6 6 5 13 289 474 698 7 0.586 0.195 79.883 6
3 2 15 0 3 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 13 18 0 4 363 505 530 2 0.766 0.828 86.523 43
5 1 15 0 1 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 4 8 0 6 356 527 542 3 0.672 0.484 87.305 22
2 3 16 0 4 0 3000 6000 8000 0 3000 6000 8000 0 3000 6000 8000 20 16 17 15 287 483 522 2 0.625 0.109 78.516 41
5 5 16 0 1 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 5 8 19 12 298 490 510 2 0.984 0.375 89.258 25
1 1 3 0 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 4 12 13 12 284 533 440 10 0.07 0.898 95.703 6
6 2 4 0 3 0 3000 6000 8000 0 3000 6000 8000 1000 4000 7000 9000 15 1 12 14 310 493 452 8 0.477 0.773 98.828 27
1 5 5 0 2 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 5 18 6 10 341 530 447 8 0.055 0.414 84.57 28
4 6 6 0 1 2000 5000 8000 10000 1000 4000 7000 9000 0 3000 6000 8000 11 5 1 9 335 492 432 8 0.359 0.047 87.891 21
2 1 13 0 4 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 2 13 3 20 263 475 690 2 0.109 0.664 99.805 34
4 2 10 0 3 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 19 11 1 12 262 440 603 5 0.172 0.617 96.875 5
1 4 12 0 3 1000 4000 7000 9000 0 3000 6000 8000 2000 5000 8000 10000 1 18 16 1 363 470 683 7 0.164 0.25 95.313 35
4 4 16 0 2 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 20 8 13 3 366 458 638 4 0.023 0.445 91.016 19
2 1 7 0 3 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 9 16 11 9 291 438 442 10 0.336 0.914 78.32 83
6 0 7 0 1 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 13 6 13 7 264 448 547 11 0.148 0.938 84.375 89
1 4 4 0 0 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 6 17 9 14 318 445 550 7 0.492 0.133 80.664 78
6 6 6 0 1 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 20 1 6 16 330 430 590 9 0.211 0.313 85.938 75
2 3 12 0 3 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 6 9 1 1 265 499 673 2 0.438 0.883 76.953 76
4 3 16 0 4 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 11 7 6 3 302 502 748 1 0.352 0.836 82.813 49
3 4 11 0 1 1000 4000 7000 9000 0 3000 6000 8000 0 3000 6000 8000 3 17 19 16 322 496 608 2 0.18 0.477 88.086 70
3 4 15 0 1 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 16 0 15 14 338 501 663 4 0.398 0.297 75.977 53
0 2 5 0 4 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 18 10 18 15 325 518 635 2 0.039 0.398 75 80
5 3 2 0 4 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 7 10 18 19 315 497 693 3 0.25 0.32 84.18 54
2 6 1 0 6 0 3000 6000 8000 2000 5000 8000 10000 2000 5000 8000 10000 11 8 5 7 277 495 615 3 0.313 0.547 81.25 89
3 5 3 0 5 2000 5000 8000 10000 0 3000 6000 8000 0 3000 6000 8000 8 16 5 2 282 491 655 5 0.422 0.594 77.734 56
0 2 10 0 6 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 12 0 4 15 345 452 575 7 0.305 0.07 85.547 74
5 1 8 0 8 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 2 16 3 15 360 450 545 8 0.008 0.234 83.008 77
2 6 15 0 6 0 3000 6000 8000 0 3000 6000 8000 2000 5000 8000 10000 14 2 12 8 309 476 475 7 0.453 0.969 81.445 82
5 5 14 0 8 2000 5000 8000 10000 2000 5000 8000 10000 1000 4000 7000 9000 3 15 13 10 276 449 470 8 0.227 0.711 83.789 68
2 3 1 0 6 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 17 5 17 1 367 450 688 1 0.063 0.563 98.242 32
5 2 7 0 6 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 4 14 18 4 342 435 593 3 0.102 0.281 96.289 23
3 4 3 0 6 1000 4000 7000 9000 2000 5000 8000 10000 0 3000 6000 8000 17 7 7 16 281 484 738 4 0.133 0.859 99.609 40
4 5 2 0 5 1000 4000 7000 9000 0 3000 6000 8000 2000 5000 8000 10000 6 13 2 17 288 480 618 1 0.242 0.82 94.727 17
1 0 8 0 7 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 18 7 4 1 362 516 467 8 0.195 0.039 88.281 42
5 3 10 0 5 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 8 17 7 3 324 524 517 9 0.383 0.188 94.141 44
0 5 11 0 6 0 3000 6000 8000 0 3000 6000 8000 0 3000 6000 8000 17 0 12 9 270 529 490 9 0.258 0.727 86.328 3
3 5 13 0 8 2000 5000 8000 10000 1000 4000 7000 9000 2000 5000 8000 10000 9 15 12 13 327 517 472 9 0.297 0.781 91.602 27
3 2 3 0 6 2000 5000 8000 10000 2000 5000 8000 10000 2000 5000 8000 10000 8 5 20 9 287 501 460 2 0.969 0.203 76.367 9
5 1 2 0 6 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 19 16 11 12 266 515 497 1 0.734 0.492 81.641 26
2 3 5 0 4 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 2 6 4 0 333 489 562 3 0.633 0.734 79.492 51
5 5 2 0 6 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 15 11 1 6 370 520 565 3 0.859 0.758 82.617 44
2 1 10 0 8 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 7 2 7 13 280 486 695 10 0.57 0.008 83.594 11
5 1 15 0 7 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 10 19 9 18 295 461 670 6 0.656 0 77.344 8
1 4 9 0 8 0 3000 6000 8000 0 3000 6000 8000 1000 4000 7000 9000 4 1 20 7 334 447 623 10 0.813 0.844 85.156 13
5 5 13 0 8 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 13 20 10 2 312 467 680 11 0.711 0.945 84.961 30
1 1 5 0 5 1000 4000 7000 9000 2000 5000 8000 10000 0 3000 6000 8000 2 8 12 1 285 454 457 2 0.953 0.352 99.414 79
4 2 2 0 5 0 3000 6000 8000 1000 4000 7000 9000 2000 5000 8000 10000 13 11 11 6 267 465 465 4 0.797 0.016 97.852 72
1 5 7 0 8 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 3 6 4 16 361 435 595 4 1 0.57 92.969 60
4 4 7 0 6 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 19 19 5 15 339 473 525 6 0.883 0.609 97.07 51
2 2 14 0 4 2000 5000 8000 10000 1000 4000 7000 9000 0 3000 6000 8000 10 6 6 3 271 525 675 9 0.484 0.078 94.922 72
3 1 10 0 6 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 13 10 3 9 283 498 740 12 0.68 0.258 88.867 75
2 4 14 0 6 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 3 2 14 18 321 531 713 7 0.594 0.633 96.094 53
5 3 11 0 7 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 19 15 17 10 349 508 620 8 0.539 0.531 92.773 86
3 3 8 0 4 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 10 10 10 10 318 481 588 7 0.5 0.5 87.5 45
5 3 10 0 5 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 9 1 5 5 306 427 450 5 0.156 0.461 77.539 4
1 4 9 0 7 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 10 12 2 8 304 433 515 3 0.094 0.422 84.766 25
3 1 16 0 5 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 8 9 11 17 319 442 430 5 0.445 0.852 79.102 31
2 1 11 0 5 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 15 11 12 13 336 429 532 1 0.391 0.641 85.352 20
6 4 7 0 7 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 0 7 18 0 278 506 605 7 0.125 0.328 81.836 32
2 3 5 0 5 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 15 18 16 0 303 482 648 8 0.078 0.125 81.055 24
4 0 4 0 7 2000 5000 8000 10000 2000 5000 8000 10000 0 3000 6000 8000 1 5 6 18 374 522 598 9 0.086 0.656 76.563 29
3 2 2 0 5 0 3000 6000 8000 0 3000 6000 8000 2000 5000 8000 10000 14 16 10 18 355 504 640 7 0.531 0.906 75.781 38
6 6 12 0 6 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 5 3 5 15 307 506 445 3 0.219 0.023 91.797 88
0 6 13 0 7 2000 5000 8000 10000 2000 5000 8000 10000 1000 4000 7000 9000 16 17 10 17 295 521 455 7 0.273 0.305 94.531 57
6 0 8 0 6 2000 5000 8000 10000 0 3000 6000 8000 2000 5000 8000 10000 8 3 17 9 343 510 507 2 0.281 0.789 92.578 90
0 0 12 0 7 0 3000 6000 8000 1000 4000 7000 9000 0 3000 6000 8000 14 14 15 7 346 487 477 6 0.414 0.805 95.117 84
3 4 1 0 5 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 7 2 20 16 272 456 645 11 0.234 0.172 88.477 47
2 5 1 0 7 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 16 12 20 14 279 434 633 10 0.328 0.516 87.695 68
4 3 0 0 4 2000 5000 8000 10000 2000 5000 8000 10000 2000 5000 8000 10000 0 4 3 5 348 478 653 11 0.375 0.891 96.484 49
1 2 0 0 7 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 15 12 1 8 337 471 665 11 0.016 0.625 85.742 65
5 5 13 0 8 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 16 8 8 8 351 428 735 3 0.93 0.102 79.297 84
0 4 12 0 5 2000 5000 8000 10000 2000 5000 8000 10000 1000 4000 7000 9000 5 19 8 6 325 468 723 5 0.523 0.227 76.172 63
5 1 12 0 6 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 15 2 14 10 294 431 728 5 0.945 0.586 90.43 62
2 0 10 0 7 0 3000 6000 8000 1000 4000 7000 9000 2000 5000 8000 10000 9 15 19 11 300 469 743 5 0.641 0.953 87.109 69
4 5 3 0 4 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 18 7 17 0 372 486 485 11 0.891 0.336 75.195 56
2 4 6 0 5 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 1 9 19 8 373 521 572 8 0.828 0.383 78.125 85
5 2 4 0 5 1000 4000 7000 9000 2000 5000 8000 10000 0 3000 6000 8000 19 3 4 19 272 491 492 6 0.836 0.75 79.688 55
2 2 1 0 7 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 0 12 7 18 269 503 537 9 0.977 0.555 83.984 71
4 5 9 0 5 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 11 4 9 11 344 523 733 3 0.664 0.086 96.68 7
0 6 9 0 7 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 8 14 7 13 371 513 628 2 0.852 0.063 90.625 1
5 2 12 0 8 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 14 3 11 6 317 516 625 6 0.508 0.867 94.336 12
0 0 10 0 7 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 0 19 14 4 305 531 585 4 0.789 0.688 89.063 15
4 3 4 0 5 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 14 11 19 19 370 462 502 11 0.563 0.117 98.047 14
2 3 0 0 4 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 9 13 14 17 333 459 427 12 0.648 0.164 92.188 41
3 2 5 0 7 1000 4000 7000 9000 2000 5000 8000 10000 2000 5000 8000 10000 18 3 1 4 313 465 567 11 0.82 0.523 86.914 20
3 2 1 0 7 1000 4000 7000 9000 1000 4000 7000 9000 0 3000 6000 8000 4 20 5 6 297 460 512 9 0.602 0.703 99.023 37
6 4 11 0 4 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 2 10 2 5 310 443 540 11 0.961 0.602 100 10
1 3 14 0 4 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 13 10 2 1 320 464 482 10 0.75 0.68 90.82 36
4 0 15 0 2 2000 5000 8000 10000 0 3000 6000 8000 0 3000 6000 8000 9 12 15 13 358 466 560 10 0.688 0.453 93.75 1
3 1 13 0 3 0 3000 6000 8000 2000 5000 8000 10000 2000 5000 8000 10000 12 4 15 18 353 470 520 8 0.578 0.406 97.266 34
6 4 6 0 2 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 8 20 16 5 290 509 600 6 0.695 0.93 89.453 16
1 5 8 0 0 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 18 4 17 5 275 511 630 5 0.992 0.766 91.992 13
4 0 1 0 2 2000 5000 8000 10000 2000 5000 8000 10000 0 3000 6000 8000 6 18 8 12 326 485 700 6 0.547 0.031 93.555 8
1 1 2 0 1 0 3000 6000 8000 0 3000 6000 8000 1000 4000 7000 9000 17 5 7 10 359 512 705 5 0.773 0.289 91.211 23
4 3 15 0 2 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 3 15 3 19 268 511 487 12 0.938 0.438 76.758 58
1 4 9 0 2 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 16 6 2 16 293 526 582 10 0.898 0.719 78.711 67
3 2 13 0 2 1000 4000 7000 9000 0 3000 6000 8000 2000 5000 8000 10000 3 13 13 4 354 477 437 9 0.867 0.141 75.391 50
2 1 14 0 3 1000 4000 7000 9000 2000 5000 8000 10000 0 3000 6000 8000 14 7 18 3 347 481 557 12 0.758 0.18 80.273 73
5 6 8 0 1 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 2 13 16 19 273 445 708 5 0.805 0.961 86.719 48
1 3 6 0 3 2000 5000 8000 10000 1000 4000 7000 9000 1000 4000 7000 9000 12 3 13 17 311 437 658 4 0.617 0.813 80.859 46
6 1 5 0 2 2000 5000 8000 10000 2000 5000 8000 10000 2000 5000 8000 10000 3 20 8 11 365 432 685 4 0.742 0.273 88.672 87
3 1 3 0 0 0 3000 6000 8000 1000 4000 7000 9000 0 3000 6000 8000 11 5 8 7 308 444 703 4 0.703 0.219 83.398 63
3 4 13 0 2 0 3000 6000 8000 0 3000 6000 8000 0 3000 6000 8000 12 15 0 11 348 460 715 11 0.031 0.797 98.633 81
1 5 14 0 2 1000 4000 7000 9000 2000 5000 8000 10000 1000 4000 7000 9000 1 4 9 8 369 446 678 12 0.266 0.508 93.359 64
4 3 11 0 4 2000 5000 8000 10000 0 3000 6000 8000 1000 4000 7000 9000 18 14 16 20 302 472 613 10 0.367 0.266 95.508 39
1 1 14 0 2 1000 4000 7000 9000 1000 4000 7000 9000 1000 4000 7000 9000 5 9 19 14 265 441 610 10 0.141 0.242 92.383 46
4 5 7 0 0 0 3000 6000 8000 2000 5000 8000 10000 1000 4000 7000 9000 13 18 13 7 355 475 480 3 0.43 0.992 91.406 79
1 5 1 0 1 2000 5000 8000 10000 1000 4000 7000 9000 2000 5000 8000 10000 10 1 11 2 340 500 505 7 0.344 1 97.656 82
5 2 8 0 0 2000 5000 8000 10000 2000 5000 8000 10000 1000 4000 7000 9000 16 19 0 13 301 514 552 3 0.188 0.156 89.844 77
1 1 3 0 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 7 0 10 18 323 494 495 2 0.289 0.055 90.039 60
5 5 11 0 3 1000 4000 7000 9000 0 3000 6000 8000 2000 5000 8000 10000 18 13 8 19 350 507 718 11 0.047 0.648 75.586 11
2 4 15 0 3 2000 5000 8000 10000 1000 4000 7000 9000 0 3000 6000 8000 7 9 9 14 368 496 710 9 0.203 0.984 77.148 18
5 1 9 0 0 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 17 14 16 4 274 526 580 9 0 0.43 82.031 30
2 2 9 0 2 0 3000 6000 8000 1000 4000 7000 9000 1000 4000 7000 9000 1 1 15 5 296 488 650 7 0.117 0.391 77.93 39
4 4 2 0 4 0 3000 6000 8000 1000 4000 7000 9000 2000 5000 8000 10000 10 14 14 17 364 436 500 4 0.516 0.922 80.078 18
3 5 6 0 2 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 7 10 18 11 352 463 435 1 0.32 0.742 86.133 15
4 2 3 0 2 1000 4000 7000 9000 1000 4000 7000 9000 2000 5000 8000 10000 17 18 6 2 314 430 462 6 0.406 0.367 78.906 37
1 3 6 0 1 1000 4000 7000 9000 0 3000 6000 8000 1000 4000 7000 9000 1 5 3 10 286 453 555 5 0.461 0.469 82.227 4

129
B. TILLER
The Tiller, Version 0.7.0.0, Copyright 2004 Referentia Systems Incorporated, is a
product developed in support of Project Albert and the Marine Corps Warfighting
Laboratory. Its primary purpose is to prepare model XML scenarios for Data Farming.
In addition, it provides DOE options such as the Random Latin Hypercube coded by
Professor Paul Sanchez, Naval Postgraduate School, and a Nearly Orthogonal Latin
Hypercube coded by Professor Susan Sanchez, Naval Postgraduate School. The final
output of the Tiller is a usable study.xml file containing the chosen DOE for running at
any computer cluster facility. To choose factors for Data Farming, first select specific
squad values from the Scenario Information window. Second, drag and drop these
specific values into the Scenario Variables to be Data Farmed window. The author used
the Tiller to build a skeleton study.xml file once, and performed further XML
manipulation solely with the rapid process of Ruby Scripting.

130
C. RUBY SCRIPTING
Figure 28 identifies the PatchExcurision.rb Ruby code written by Paul Sanchez
that modifies the skeleton Tiller study.xml file for all DOE iterations performed. A
Notepad application provides simple viewing of the code. Figure 29 identifies the
scripting typed by a user within a Command Prompt Window to execute the
PatchExcursion.rb Ruby code. Table 19 identifies all the steps the user needs to execute
to modify a skeleton Tiller study.xml file for use by the MHPCC.
Figure 28. Ruby PatchExcursion.rb Code 97
Figure 29. Ruby Scripting Command
97 PatchExcursion.rb, coded by Professor Paul Sanchez, Naval Postgraduate School, Monterey,
California.

131
1. Open the Tiller, and ensure Ruby is loaded onto the running PC.
2. Browse to File/Open/Scenario File (The MANA basecase.xml file
scenario location).
3. To create a skeleton study.xlm file, double click on the appropriate
factor within each squad (platform) from the “Scenario Information”
window. Each factor will then appear in the “Scenario Variables to be
Data Farmed” window. Else drag and drop from one window to the other.
Once all factors are selected, double click on the submit button, and a
study.xml file will be saved automatically in the same directory as the
basecase.xml file.
4. Create a designfile.csv from the crossed NOLH DOE with 258 design
points, and save the .csv file in the same location as study.xml file created
by the Tiller. (The intent is to create columns consisting of the factor
name and the values for each design point, or excursion, directly below
each column heading name.)
5. Write and then save a copy of PatchExcursion.rb in the same folder as
the skeleton study.xml file created by the Tiller (Refer to Figure 28).
6. Open a command window.
7. Change the directory within the command window to the same as that
of the folder that contains a copy of PathExcursion.rb, study.xml, and
designfile.csv.
8. Write the scripting code outlined in Figure 29 and press enter. (At this
time, the ruby code reads the designfile.csv containing the DOE and
merges each design point into the skeleton file created by the tiller.)
9. The outstudy.xml file automatically appears in the same directory.
10. Rename the outstudy.xml file to study.xml overwriting the old
study.xml. This is necessary because the original study.xml file is only a
skeleton file, and does not include the complete DOE. The outstudy.xml
includes the completed DOE—but has the wrong name. See step 12.
11. Recreate a Zip folder of the current working directory.
12. Submit an email to MHPCC at isaac@mhpcc.hpc.mil attaching the Zip
file and wait. The computer cluster searches the zip folder for the specific
file names outlined within this table. The zip folder must contain the
basecase.xml, terrain.bmp, and elevation.bmp from the ABS, and the DOE
scripted within the study.xml.
Table 19. Table of Instruction to Modify a Skeleton study.xml File

132

133
APPENDIX C. ADDITIONAL DATA ANALYSIS
The purpose of this appendix is streamline the Data Analysis chapter of this
thesis. Figures follow in the same order as outlined in Chapter V. The fitted models
determined by means of multiple regression help identify the number of UAVs (or any
other parameter outlined within the DOE). In each instance, the model is in the form:

134
A. INITIAL OBSERVATIONS
Figure 30. Multiple Regression Output for Initial Analysis of Robust DOE
(Note: This page contains Multiple Regression Models without Interactions,
to view the Multiple Regression Model with Interactions mentioned in the Initial
Observations section of Chapter V, refer to the next three pages.)

135
Multiple Regression Model with Interactions, as mentioned in the Initial Observations
section of Chapter V: MOE - Proportion of Blue Dismounts Survived
(Note: An interesting note is that performing a multiple regression with interactions
between factors raised the R2
to 0.80, suggesting an improved fitted model from that
portrayed in Figure 13 (or Figure 30 in Appendix A). With interactions applied to the
model, the Effect Test output, similar to Figure 13, was too large for the main body of the
thesis. The output for this model is located here in Appendix C, “Initial Observations.”
This improved model was similar to the first in that the most significant factors are those
that are uncontrolled by the Blue Force. Refer to the next two pages, to view the
Parameter Estimates, and the Effects Test supporting this improved model with an
increased R2
= 0.80.)

136
(Parameter Estimates for Multiple Regression Model with Interactions
MOE - Proportion of Blue Dismounts Survived)

137
(Effect Tests for Multiple Regression Model with Interactions
MOE - Proportion of Blue Dismounts Survived)

138
B. THE EARLY FIGHT
(Note: Refer to the next two pages to view the Parameter Estimates and the Effects Test)

139
(Proportion of HPTs Killed at 450 seconds)

140

141

142

143

144
Figure 33. Regression Model (Proportion of Dismounts Survived at 900 seconds)

145
(Proportion of Dismounts Survived at 900 seconds)

146
(Proportion of Dismounts Survived at 900 seconds)

147
C. INTERACTIONS
Figure 34. Determine Interactions Model, MOE: Proportion of HPTs Killed
(Note: Refer to the next page for the Effects Test and the Interaction Profiles)

149
LIST OF REFERENCES
Boeing, Future Combat Systems, Retrieved 15 November 2005 from the World Wide
Web at http://www.boeing.com/defense-space/ic/fcs/bia/about.html
Boeing, Future Combat Systems, Retrieved 5 August 2005 from the World Wide Web at
Boeing, Future Combat Systems, Retrieved 5 August 2005 from the World Wide Web at
Brigade and Below Scenario (BBS) slide show, March 2005, provided by Mr. Tom
Loncarich, TRAC-WSMR during office visit 25 June 2005.
Cioppa, Thomas M., Efficient Nearly Orthogonal and Space-Filling Experimental
Designs for High-Dimensional Complex Models, (PhD. Dissertation, Operations
Research Department, Naval Postgraduate School, Monterey, CA), 2002.
Department of Defense, Memorandum for Secretaries of the Military Departments,
Subject: Unmanned Aircraft Systems (UAS) Roadmap, 2005 -2015, 4 August 2005.
Department of the Navy, Office of the Chief of Naval Operations, Integration of
Unmanned Vehicles into Maritime Missions, TM 3-22-5-SW, chap. 2, p. 4.
Devore, Jay L., Probability and Statistics for Engineering and the Sciences, Sixth
Edition, (Brooks/Cole, 2004), pp. 497 587.
Examining the Army’s Future Warrior, Force-on-Force Simulation of Candidate
Technologies, Rand Arroyo Center, 2004, p. xi.
FAS Military Analysis Network, AH-64 Apache, Retrieved 22 September 2005, from the
World Wide Web at http://www.fas.org/man/dod-101/sys/ac/ah-64.htm
FAS Military Analysis Network, AT-7, Retrieved 23 September 2005, from the World
Wide Web at http://www.fas.org/man/dod-101/sys/land/row/at-7.htm

150
FAS Military Analysis Network, BTR-80, Retrieved 11 October 2005, from the World
Wide Web at http://www.fas.org/man/dod-101/sys/land/row/btr-80.htm
FAS Military Analysis Network, Joint Strike Fighter, Retrieved 22 September 2005,
from the World Wide Web at http://www.fas.org/man/dod-101/sys/ac/jsf.htm
FAS Military Analysis Network, Mortars, Retrieved 23 September 2005, from the World
Wide Web at http://www.fas.org/man/dod-101/sys/land/mortars.htm
FAS Military Analysis Network, SA-16 Gimlet, Retrieved 23 September 2005, from the
World Wide Web at http://www.fas.org/man/dod-101/sys/missile/row/sa-16.htm
FAS Military Analysis Network, T-72, Retrieved 23 September 2005, from the World
Wide Web at http://www.fas.org/man/dod-101/sys/land/row/t72tank.htm
Federation of American Scientists, Retrieved 22 September, from the World Wide Web
at http://www.fas.org/main/home.jsp .
Galligan, David P., Mark A. Anderson, Michael K. Lauren, Map Aware, Non-Uniform
Automata version 3.0, New Zealand Defense Technology, July 2004.
Global Security.org Future Combat Systems, Retrieved 17 November 2005 from the
World Wide Web at http://www.globalsecurity.org/military/systems/ground/images/fcs-
2005armymodernization.jpg
Global Security.org Future Combat Systems, Retrieved 3 August 2005 from the World
Wide Web at http://www.globalsecurity.org/military/systems/ground/fcs.htm
Global Security.org, Future Combat Systems – Background, Retrieved 28 June 2005 from
the World Wide Web at http://www.globalsecurity.org/military/systems/ground/fcs-
back.htm
Global Security.org, Future Combat Systems, Retrieved 1 August 2005 from the World
Wide Web at http://www.globalsecurity.org/military/systems/ground/fcs.htm

151
Global Security.org, Hellfire, Getting the Most from a Lethal Weapon System, Retrieved
7 October 2005 on the World Wide Web at
http://www.globalsecurity.org/military/library/news/1998/01/1helfire.pdf
Grant, Greg, “Army picks General Atomics for ERMP program,” Army Times, 8 Aug
2005, Retrieved 11 October 2005 from the World Wide Web at
http://www.armytimes.com/story.php?f=1-292925-1021240.php
Hand, David, Heikki Mannila, and Padhraic Smyth, Principles of Data Mining, (MIT
Press, Cambridge, Massachusetts, 2001), p. 53.
Headquarters United States Army Training and Doctrine Command (Director Futures
Center), Memorandum for U.S. Army TRADOC Analysis Center, Fort Leavenworth, KS,
9 July 2004.
Interview with Thomas Lancarich, Senior Operations Research Analyst, Chief, Scenario
Integration & Methodology Development Division, TRADOC Analysis Center-White
Sands Missile Range, New Mexico, 25 June 2005.
Interview with Jeffrey Schamburg, LTC, Director, TRADOC Analysis Center-Monterey,
Naval Postgraduate School, Monterey California, 15 August 2005.
JMP IN, JMP 5.5.2 Help Command, SAS Institute Inc, 2004.
JMP Start Statistics, A Guide to Statistics and Data Analysis using JMP and JMP IN
Software, Third Edition, (SAS Institute Inc. 2005) p. 245.
JMP, The Statistical Discovery Software, referenced 18 October 2005 on the World Wide
Web at http://www.jmp.com/product/jmp5_brochure.pdf
Kenneth Munson, Jane’s Unmanned Aerial Vehicles and Targets: Issue Twenty-Three,
(Alexandria: Jane’s Information Group Inc, 2004), p. 20.
Lester W. Grau, For All Seasons: The Old But Effective RPG-7 Promises to Haunt the
Battlefields of Tomorrow, Foreign Military Studies Office, Fort Leavenworth, KS
Retrieved 23 September 2005 from the World Wide Web at
http://www.g2mil.com/RPG.htm

152
Lloyd Brown, Thomas Cioppa, and Thomas Lucas, “Agent-Based Simulations
Supporting Military Analysis,” Phalnex, April 2004.
Lucas, Thomas, OA4655 Combat Modeling, Naval Postgraduate School, lecture
presentation: Entity-level Attrition: Some Phit and Pkill Algorithms.
Montgomery, Douglas, Elizabeth Peck, and Geoffrey Vining, Introduction to Linear
Regression Analysis, Third Edition, (John Wiley and Sons, Inc, 2001), p. 516.
Naval Postgraduate School, NOLH 17-22 Factors, coded by Professor Susan Sanchez,
Monterey, California. Retrieved 15 August 2005 from the World Wide Web at
http://nps.blackboard.com/webapps/portal/frameset.jsp?tab=courses&url=/bin/common/c
ourse.pl?course_id=_3888_1
Naval Postgraduate School, PatchExcursion.rb, coded by Professor Paul Sanchez,
Monterey, California.
Naval Postgraduate School Thesis, An Analysis of Degraded Communications in the
Army’s Future Force using Agent Based Modeling, Joseph M. Lindquist, June 2004, pp.
2-3.
Naval Postgraduate School Thesis, Comparison of a Distributed Operations Force to a
Traditional Force in Urban Combat, Michael Babilot, September 2005, Excel
Spreadsheet.
Peter P. Perla, The Art of War-Gamming, United States Naval Institute, Annapolis,
Maryland, 1990, p. 15.
Presentation to the CSA on the FCS Brigade Combat Team Operational &
Organizational Plan, by US Army Futures Center, TRADOC, 7 October 2005.
Retrieved 23 September 2005 from the World Wide Web at
http://www.sovietarmy.com/small_arms/rpk-74.html
Stephen Cambone, Kenneth Krieg, Peter Pace, Linton Wells, Unmanned Aircraft System
Roadmap 2005-2015, Department of Defense, 4 August 2005, p. 1.

153
Thomas Lancarich, Senior Operations Research Analyst, Chief, Scenario Integration &
Methodology Development Division, TRADOC Analysis Center-White Sands Missile
Range, New Mexico North East Asia Vignettes (Vignette NEA50.2) FCS BN(-) attack vs
enemy stronghold of city, May 2005.
Unit of Action Maneuver Battle Lab, Change 1, to Joint Requirements Oversight Council
(JROC) – approved Future Combat Systems (FCS) Operational Requirements Document
(ORD), June 2004, Fort Knox, KY 40121, Annex E.
Unit of Action Maneuver Battle Lab, Change 3, to TRADOC Pamphlet 525-3-90 O&O,
The United States Future Force Operational and Organizational Plan Maneuver Unit of
Action (DRAFT), 30 July 2004, Fort Knox, KY 40121, section 3.2, p.18.
Unit of Action Manuever Battle Lab, TRADOC Pam 525-3-90, Future Force Operational
and Organization Plan, Maneuver Unit Action, with Change 3, Fort Knox, KY, 30 July
2004.
United States Department of Defense, Predator UAV Proves its Worth, Retrieved 10
August 2005 from the World Wide Web at
http://usmilitary.about.com/cs/afweapons/a/preditor.htm
US Army Material Systems Analysis Activity (US AMSAA), Army Future Combat
Systems Unit of Action Systems Book Version 3.0, 22 May 2003.
US Army Material Systems Analysis Activity (US AMSAA), FCS UA Design Concept
Baseline Description (UA-001-01-050124), 3 March 2005.
US Army Training and Doctrine Command, The Army Future Force: Decisive 21st
Century Landpower Strategically Responsive Full Spectrum Dominate. pp. 4-5.
W3C, Extensible Markup Language, referenced 18 October 2005 from the World Wide
Web at http://www.w3.org/XML/
Wikipedia.org, Ruby Programming Language, referenced 18 October 2005 on the World
Wide Web at http://en.wikipedia.org/wiki/Ruby_programming_language
Zaloga, Steven J. BMP Infantry Combat Vehicle, 2nd Ed, Concord Publications, 1990,
Hong Kong.

154

155
INITIAL DISTRIBUTION LIST
1. Defense Technical Information Center
Ft. Belvoir, Virginia
2. Dudley Knox Library
Monterey, California
3. Professor Thomas Lucas
Monterey, CA
4. LTC Jeffrey Schamburg
TRAC-MTRY
Monterey, CA
5. Dr. Gary Horne
USMC Project Albert
Quantico, VA
6. COL Gary Krahn
Department of Mathematical Sciences
United States Military Academy
West Point, NY
7. Joseph Lindquist
Department of Mathematical Sciences
United States Military Academy
West Point, NY
8. Charles A. Sulewski
Monterey, CA
9. Thomas Loncarich
TRAC-WSMR
White Sands Missile Range, NM
10. Dr. Lucy Nowell
Advanced Research & Development Activity
Ft. Meade, MD

156
11. Dr. Francis Hughes
Defense Intelligence Agency
Bolling AFB, Washington, DC
12. Heather A. McMonagle
National Security Agency
Ft. Meade, MD
13. Dawn M. Dentato
Somers High School
Somers, NY
14. Michael Landin
FDIC Analysis Division
Ft. Sill, OK

UAV in the Army of the future operations

More Related Content

Similar to UAV in the Army of the future operations

Recently uploaded

UAV in the Army of the future operations