Self-Management in Cellular Networks

Drivers, Vision, Gains, Achievements & Challenges

Remco Litjens
Performance of Networks & Systems SCVT ‘12
Expertise Centre Technical Sciences November 16, 2012
TNO Eindhoven, The Netherlands
OUTLINE

!  INTRODUCTION
!  TRENDS & DRIVERS
!  VISION
!  EXPECTED GAINS
!  ACHIEVEMENTS
!  CHALLENGES
!  CONCLUDING REMARKS
OUTLINE

!  INTRODUCTION
!  TRENDS & DRIVERS
!  VISION
!  EXPECTED GAINS
!  ACHIEVEMENTS
!  CHALLENGES
!  CONCLUDING REMARKS
INTRODUCTION

!  Cellular networks are traditionally largely manually operated

!  Separation of network planning and optimization
!  (Non-)automated planning/simulation tools used to select sites,
determine antenna tilt and optimise radio parameters
!  ‘Over-abstraction’ of applied technology models
!  Manual configuration of sites
!  Radio (resource management) parameters updated weekly/monthly
!  Delayed, manual and poor handling of cell/site failures
!  Future networks will exhibit a significant degree of self-management
!  Broad attention
!  3GPP, NGMN, FP7, CELTIC, …
!  Operators, equipment/tool vendors, academia/research institutes
OUTLINE

!  INTRODUCTION
!  TRENDS & DRIVERS
!  VISION
!  EXPECTED GAINS
!  ACHIEVEMENTS
!  CHALLENGES
!  CONCLUDING REMARKS
TRENDS & DRIVERS

!  Complexity of future/contemporary wireless access networks

!  Multitude of tunable parameters with intricate dependencies
!  Multitude of radio resource management mechanisms on different
time scales
!  Complexity is needed to maximize potential of wireless access
networks
TRENDS & DRIVERS

!  Heterogeneous access networks to be cooperatively managed

!  Multi-RAT: 2G/3G/LTE/LTE-A/WiFi
!  Multi-layer: Macro/micro/pico/femto
!  Multi-vendor
!  Mode of operation
!  Network sharing, e.g.
national roaming,
hardware/spectrum sharing
!  Network ‘dynamicity’, incl.
uncoordinated femtocell
deployment
!  QoE/service-centric traffic management
TRENDS & DRIVERS

!  Higher operational frequencies

!  Multitude of cells to be managed
!  Service/traffic
!  Increasing demand for ubiquitous mobile broadband access
!  Growing suite of services with distinct characteristics &
requirements
TRENDS & DRIVERS

!  Market perspective
!  Need to reduce time-to-market of innovative services
!  Reduce operational hurdles of service introduction
!  Pressure to remain competitive
!  Reduce costs (OPEX/CAPEX)
!  Enhance resource efficiency
!  ‘Society’
!  Mobile communications becomes more
and more a critical infrastructure for
economy and society as a whole: need
for network robustness/resilience
OUTLINE

!  INTRODUCTION
!  TRENDS & DRIVERS
!  VISION
!  EXPECTED GAINS
!  ACHIEVEMENTS
!  CHALLENGES
!  CONCLUDING REMARKS
VISION

!  Self-Management for Unified

Heterogeneous Radio Access Networks
!  Service provider
!  Network operator
!  Self-management system
!  Network resources
VISION

!  Self-Management for Unified

Heterogeneous Radio Access Networks
!  Policy transformation
!  SON coordination
!  SON functions
!  Multi-RAT/layer
!  Single-RAT/layer
!  Decision support system
VISION

!  Self-Management for Unified

Heterogeneous Radio Access Networks
!  SON functions
!  Self-configuration
!  Incidental, intentional events
!  ‘Plug and play’ installation
of new base stations and features
!  Self-healing
!  Incidental, non-intentional events
!  Cell outage detection
!  Cell outage compensation
VISION

!  Self-Management for Unified

Heterogeneous Radio Access Networks
!  SON functions
!  Self-optimization
!  Measurements
!  Performance indicators
!  Network, traffic, mobility,
propagation conditions
!  Optimize periodicity, accuracy
!  Automatic tuning
!  Smart algorithms process measurements into parameter
adjustments, e.g. tilt, power, handover/scheduling parameters, …
VISION

!  Self-Management for Unified

Heterogeneous Radio Access Networks
!  Policy transformation
!  Heading harmonization
!  Transform high-level operator objective
to SON function specific objectives
!  Example - Optimization weights between call dropping, ping-
pong handovers and radio link failure in handover optimization
!  SON coordination
!  Tailing harmonization
!  Detect/resolve conflicts between SON functions
!  Conflicting actions, conflicting performance effects, …
VISION

!  Self-Management for Unified

Heterogeneous Radio Access Networks
!  Decision support system
!  Intelligent selection of required hardware upgrades
!  Advice on migration of sites to other technologies or frequency
bands
!  Estimation of the resource cost of performance to assist the (re)
negotiation of SLAs
OUTLINE

!  INTRODUCTION
!  TRENDS & DRIVERS
!  VISION
!  EXPECTED GAINS
!  ACHIEVEMENTS
!  CHALLENGES
!  CONCLUDING REMARKS
EXPECTED GAINS

!  OPEX reductions …
!  Primary objective!
!  Less human involvement in
!  Network planning/optimization
!  Performance monitoring, drive testing
!  Troubleshooting
!  About 25% of OPEX is related to
network operations
!  x00 million € savings potential per
network
EXPECTED GAINS

!  … and/or CAPEX reductions …

!  Via delayed capacity expansions
!  Smart eNodeBs may however be more expensive
!  … and/or performance enhancements
!  Enhanced service availability (robustness/resilience), service quality
IMPACT OF 'SELF-OPTIMISATION' IMPACT OF 'SELF-HEALING'

WITH SELF-OPTIMIISATION:
LESS QUALITY DEGRADATION

QUALITY
SERVICE QUALITY
WITH SELF-HEALING:
QUICK RECOVERY
SERVICE QUALITY

TRAFFIC LOAD
TO TOLERABLE LEVEL

LOCAL SERVICE
DELAYED
WITHOUT SELF-OPTIMISATION: INVESTMENTS
MORE QUALITY DEGRADATION

LOCAL
WITHOUT SELF-HEALING:
DRAMATIC DROP
TO INTOLERABLE LEVEL
REQUIRED SERVICE QUALITY

0 20 40 60 80 100 0 20 40 60 80 100
TIME SITE TIME
FAILURE
OUTLINE

!  INTRODUCTION
!  TRENDS & DRIVERS
!  VISION
!  EXPECTED GAINS
!  ACHIEVEMENTS
!  CHALLENGES
!  CONCLUDING REMARKS
ACHIEVEMENTS

!  Existing solutions primarily target

!  Self-configuration
!  Stand-alone self-optimization functions
!  Single-RAT, single-layer scenarios
!  Examples
!  Smooth automated integration of new sites
!  Neighbor cell list optimization
!  Handover parameter optimization
!  Cell outage compensation
!  Self-optimization of admission control thresholds
!  Load balancing
!  Interference coordination
!  Automatic on/off-switching of sites
ACHIEVEMENTS

!  Smooth automated integration of new sites

!  Minimize disruptive effects
!  Observation phase
!  Data collection
!  Off-line tilt/power
optimization
!  Pre-computed trajectory
!  Integration phase
!  Gradual activation
!  Power 0 à P*
!  Tilt 90o à T*
!  Adoption in regular optimization cycles
 Start

ACHIEVEMENTS
Collect HPIs

Observation No No
time >
Optimisation
Opt. intervall
Yes
Yes HPI_PP < threshold No
HPI_HOF < threshold

!  Handover parameter optimization HPI_RLF < HPI_RLF <

Yes threshold No Yes threshold No

Decrease HPI Decrease TTT Increase TTT Increase HPI

thresholds & Hysteresis & Hysteresis thresholds

Reset observ. Reset observ.

time time

HO optimisation Legend

SeNB TeNB Hysteresis

Time-to-
Trigger
Connected
Cell
UE1
Interfering
Cell

User (UE)
Received signal strength

SeNB TeNB
TTT
Hysteresis
Offset

Hysteresis eNodeB

Handover
area

Distance
ACHIEVEMENTS

!  Cell outage compensation

Control  parameters  

Operator  policy:  
Coverage,  QoS  

Compensa-on  

-70
2000
-80
1500
-90
1000
-100

500 -110

0 -120

-500

-1000
-130

-140 Detec-on   Measurements  

-150

Coverage/QoS  
-1500
-160
-2000
-170
-2000 -1500 -1000 -500 0 500 1000 1500 2000 2500

es-ma-on  
ACHIEVEMENTS

!  Cell outage compensation

!  Mitigate local degradation of coverage; keep QoS at acceptable level
!  Determine set of compensating sites
!  Adjust control parameters
!  Antenna downtilt
!  UL/DL powers
!  …
ACHIEVEMENTS

!  Cell outage compensation

ACHIEVEMENTS

!  Cell outage compensation

Failed Cell
Compensating Cell
Operational Cell

Covered Area
(a)
Coverage Hole (b)
+/- 80% compensated
Compensated Coverage Hole

Figure: (a) Outage situation with no compensation, and (b)

snapshot of situation with largely converged compensation.
ACHIEVEMENTS

!  Load balancing
!  Detect load asymmetry between neighboring cells
!  Adjust handover control parameters to balance loads
! Example scenario: moving hot spot
non-regular network; hotspot is moving from cell 27 to 13 scenario 3
2000
Base Station 200 reference
shotspot route
23 with load balancing
1500 26 25 180
22

1000 24 160
27

n unsatisfied users in network

5 4 20 19 140
500
28 8
29 7 6
1 120
0 9 2 21
104
Y [m]

3 100
11 10
-500 30 17
16 80
12 14 13
-1000
18 60
32
31 15
-1500 40 35
35 34
33 20
-2000
36
0
-2000 -1500 -1000 -500 0 500 1000 1500 0 200 400 600 800 1000 1200 1400 1600 1800
X [m]
t [s]
OUTLINE

!  INTRODUCTION
!  TRENDS & DRIVERS
!  VISION
!  EXPECTED GAINS
!  ACHIEVEMENTS
!  CHALLENGES
!  CONCLUDING REMARKS
CHALLENGES

!  Development of SON functions

!  Use cases
!  Self-optimisation of ‘multiflow’, CoMP and ICIC features
!  Automatic traffic steering in a multi-layer/RAT network
!  Dynamic spectrum management in a multi-layer/RAT network
!  Automated tuning of Active Antenna Systems
!  Intrinsic challenges
!  Algorithm development
!  Multi-objective optimization
!  Data gathering: performance vs signalling cost
!  Timely detection for timely response
!  Dealing with delayed feedback
!  Effect of control actions vs natural variations
CHALLENGES

!  Development of unified self-management system

!  Policy transformation
!  Dependencies
!  SON coordination
!  Guard function
! Autognostics
!  Conflict handling
!  Intricate parameter
dependencies
!  Prevent oscillations
CHALLENGES

!  Development of unified self-management system

!  Decision support systems
!  Gradual introduction
!  Build confidence
!  On/off functions
!  Degree of freedom
!  Manual approvals
!  Architectural embedding
!  Measurements
!  Protocols
!  Interfaces
OUTLINE

!  INTRODUCTION
!  TRENDS & DRIVERS
!  VISION
!  EXPECTED GAINS
!  ACHIEVEMENTS
!  CHALLENGES
!  CONCLUDING REMARKS
CONCLUDING REMARKS

!  Self-management is the key approach to

!  … reduce O/CAPEX
!  … cost-effective provisioning of high-quality services
!  … reduce time-to-market of new features, services
!  Key aspects
!  Policy transformation
!  SON coordination
!  SON functions
!  Decision support system
!  Key challenges
!  Effectiveness, reliability, stability
!  Measurements, interfaces, protocols, architectures
QUESTIONS?