Impact of Malicious Nodes on Throughput, Packets Dropped and Average Latency in MANETs

IOSR Journal of Computer Engineering (IOSR-JCE)
e-ISSN: 2278-0661,p-ISSN: 2278-8727, Volume 17, Issue 6, Ver. I (Nov – Dec. 2015), PP 55-63
www.iosrjournals.org
DOI: 10.9790/0661-17615563 www.iosrjournals.org 55 | Page
Impact of Malicious Nodes on Throughput, Packets Dropped and
Average Latency in MANETs
Md. Razibuddin Ahmed1,
Mozmin Ahmed2
McLeod Russel India Limited, Kolkata, West Bengal, India 700001
North Eastern Regional Institute of Science and Technology, Itanagar,
Arunachal Pradesh, India 791109
Abstract:- Mobile Ad-hoc Network is a decentralized wireless network[1]. Here the mobile nodes make and
break the links with the neighbouring nodes available in the radio range without actually being physically
connected. These networks are temporary and keep on changing from time to time. MANET applications are
getting importance in both civilian and military areas. MANETs can be applied in disaster communications and
used as the backup network of traditional mobile communication networks as well. Network throughput, number
of packets dropped and average latency are important parameters to evaluate the performance of wireless ad
hoc network[3]. Generally, it is difficult to achieve high throughput and low packet drop with minimum
delay[5]. In this paper, the objective is to achieve high throughput while keeping the packet drop and the
average latency under certain acceptable limits[10]. We tried to study the signature pattern of these malicious
nodes and made conclusions with the results obtained. The performance is evaluated with the following
parameters: network throughput, number of packets dropped and the average latency. We used NS2 simulator
and extracted data from the trace files[2]. Ad-hoc On Demand Distance Vector (AODV) routing protocol has
been used in our experiments[4]. Similar to our previous work, the nodes are free to move or remain static in all
the quadrants in the defined space[8].
Keywords: MANET, Malicious Node, AODV routing Protocol, Network Throughput, Packets Dropped,
Average Latency, NS2.
I. Introduction
The Ad hoc On Demand Distance Vector (AODV) algorithm enables dynamic, self-starting multi-hop
routing between participating mobile nodes. AODV can handle low, moderate and relatively high mobility rates,
as well as variety of data and traffic levels. The AODV routing protocol is designed for MANETs with
population of ten to thousand mobile nodes.
Malicious node disrupts the network activity in ad hoc networks. A node that sends out false routing
information could be a compromised node, or merely a node that has temporarily stale routing table due to
volatile physical condition. A malicious attacker can readily become a router and disrupt network operations by
its malicious activity like the black hole attack, gray hole attack, worm hole attack, sink hole attack etc[6]. We
need to identify the malicious behavior of the system and isolate the misbehaving node as quickly as possible so
that the communication through the network is not affected.
The performance and use of wireless technologies has increased tremendously, opening up avenues for
application in the less explored areas. MANET is one important field of concern, in which the mobile nodes
organize themselves in a network without the help of any predefined infrastructure. Securing MANETs is an
important part of deploying and utilizing them since, MANET is used in critical applications where data and
communication integrity is important. Existing solutions for wireless networks can be used to obtain a certain
level of such security. Nevertheless, these solutions may always be sufficient, as ad hoc network have their own
vulnerabilities which cannot be addressed by these solutions.
II. Implementing Aodv Routing Protocols
Simulation Environment:
We have created a simulation program with 20 nodes. These 20 nodes are divided into four groups with
five members in each group. Out of the five nodes in a group, one node is the source node and the other four are
the receiver nodes. In the beginning, the source node sends information to its group members lying within its
radio range. The movement of the five nodes in each group is limited to its own designated area. The receiver
nodes broadcast the message to its immediate neighbours within the range once it receives any information from
the source node or any other forwarding node acting as a router. We implemented this scenario using AODV
routing protocol and studied the performance of the network considering the parameters: network throughput,
packets dropped and average latency. The scenario is shown below:

Impact of Malicious Nodes on Throughput, Packets Dropped and Average Latency in MANETs
Figure 1: Network with four groups each having one source and four destination nodes.
We further increased the group size keeping two groups with one source and nine receiving nodes in each group.
Two nodes in the scenario showing malicious behaviour in each group dropped all packets it received without
any further communication. All the nodes could freely move around in the entire simulation area. This scenario
was implemented using AODV routing protocol. Simulations were carried out to study the performance with
respect to the network throughput, packets dropped and average latency. The scenario is shown below.
Figure 2: Network with two groups each with one source, nine destinations with two malicious nodes.
The scenario was extended with 100 nodes in the network divided into four groups[9]. Each group had
one source and twenty-four receivers. Thus, the maximum group size in our simulations was of twenty-five
members. Amongst the receiver nodes, a few nodes were programmed to show malicious behavior[7]. We
study the performance of the network under various changing parameters by repeating the simulations for ten
times under each case. Each simulation was run for 200 seconds. We plotted graphs from the average results of
these ten simulations for the throughput, packet drop and delay. The main goal of our simulation was to model
the behaviour of AODV protocol without malicious nodes and with malicious nodes. The simulation space was
kept at 2000m X 800m. The mobility of the nodes varied from 1m/sec to 10m/sec.
Figure 3: Distribution of Eight Malicious Nodes

System requirements and specifications
Processor : Intel(R) Core(TM) i3 CPU M 380 @ 2.53 GHz processor
RAM : 3.00 GB
Hard Disk : 80 GB
Input Device : Standard Keyboard and Mouse
Output Device : LCD Monitor
Software Requirement
Software : NS-2.35
Operating System: Debian Linux 4
The number of malicious nodes in each group was increased from zero to ten nodes. The result of each
simulation is extracted from the trace file. We have developed perl programs to extract data for the average
latency, network throughput and packets dropped from the trace files.
III. Simulations, Results And Analysis
Table 1 : AODV Scenario Parameters.
The initial and the final node positions along with the speed of the nodes were pre-defined. Nodes
moved from their initial position to the final position in a straight line.
During our experiment, we have observed that if any vital node was defined as malicious, the system tripped
with a message and the program gets terminated. We thus had to repeat the simulation by changing the position
of the malicious node after observing the nam plots. Some of the errors noted are given below;
root@Mozmin-PC:/home/mozmin/NS/ns-allinone-2.35/TCL Files 9# num_nodes is set 100
num_nodes is set 96
warning: Please use -channel as shown in tcl/ex/wireless-mitf.tcl
INITIALIZE THE LIST xListHead
channel.cc:sendUp - Calc highestAntennaZ_ and distCST_
highestAntennaZ_ = 1.5, distCST_ = 550.0
SORTING LISTS ...DONE!
MAC_802_11: accessing MAC cache_ array out of range (src 96, dst 66, size 96)!
[suppressing additional MAC cache_ warnings]

num_nodes is set 100
root@Mozmin-PC:/home/mozmin/NS/ns-allinone-2.35/TCL Files 9# check_pktTx:Invalid MAC Control
subtype
[1]+ Exit 1 ns 100nodes-4S-24R-4M-AODV-TCP-ran-20008001.tcl
[1]+ Segmentation fault ns 100nodes-4S-24R-4M-AODV-TCP-ran-20008001.tcl
PacketQueue:: remove() couldn't find target
Direction for pkt-flow not specified; Sending pkt up the stack on default.
Direction for pkt-flow not specified; Sending pkt up the stack on default.
PacketQueue:: remove() couldn't find target
IV. Results And Discussions
Scenario: The simulation area was kept 2000m X 800m. The number of malicious node is varied from zero to
ten nodes in each group.
The throughput obtained is tabulated in Table 3 and is plotted in Figure 4 and Figure 5.
The number of packets dropped is tabulated in Table 4 and is plotted in Figure 6 and Figure 7.
The Average Latency is tabulated in Table 2 and is plotted in Figure 8.
The average latency, packet drop and network throughput under each case was extracted from the trace file and
are tabulated as follows.

Table 2: Average Latency.
We have tabulated the average latency, drop, and throughput in the simulation time of 200 seconds against the
number of malicious nodes. The graph is shown in Figure 9. The values on the y-axis are plotted on a
logarithmic scale.
V. Conclusion
From the plots, we have observed that as we increase the number of malicious nodes from 0 to 40, the
Average Packets Dropped is lowest with 4 malicious nodes in the system and Network Throughput is highest
with 0 malicious nodes in the system. The average end to end delay of the system is lowest with 4 malicious
nodes in the system.
Figure 4 and Figure 5 shows the network throughput as the number of malicious nodes was varied from
0 to 40 numbers in a network of 100 mobile nodes. The network throughput gradually increases with the
simulation time giving highest throughput of 208699 bits/sec at 200th second when the number of malicious
node was 0. This can be seen from the Table 3, the network throughput was 178676 bits/sec with 40 malicious
node which is comparatively low.
Figure 6 and Figure 7 shows the number of control and data packets dropped. Number of packet
dropped varies with the number of malicious nodes. 0 malicious node means there is no malicious node in the
network. As the number of malicious node increased, we have observed that network suffers from higher packet
drop compared to when the number of malicious node was kept low. With 0 malicious node, the drop was 6223
packets and with 40 malicious nodes the drop was 10506 packets.
The average end to end delay as seen in Figure 8, increases gradually as the number of malicious node
increases from 0 to 40 in the network. The lowest end to end delay recorded was 0.6849 seconds with 0
malicious nodes and the highest end to end delay recorded was 1.0593 seconds with 40 malicious nodes during
the simulations.
Future Work
We intend to carry out simulations with other routing protocols like DSR and TORA. Other existing
performance metrics shall be studied through various simulations. As Ad hoc networks are open to both external
and internal attacks due to lack of any centralized security system, we will try to make a study on Black Hole
and Gray Hole attacks. These attacks are required to be analyzed on other existing MANET routing protocols.
We will try to device methods of detection of selfish nodes in the Ad hoc networks and improve the
performance of the network.

Figure 4: Network Throughput against time with
malicious nodes varying from zero to ten in each
group.
Figure 5: Network Throughput against number of
malicious node with malicious nodes varying from
zero to ten in each group.
Figure 6: Number of Packets Dropped against time
with malicious nodes varying from zero to ten in each
group.

Figure 7: Number of Packets Dropped against number of
malicious nodes with malicious nodes varying from zero
to ten in each group.
Figure 8: Average latency against number of
malicious node with malicious nodes varying from
zero to ten in each group.
Figure 9: Average Latency, Packets Dropped and
Network Throughput against number of malicious
nodes.

References
[1]. Wenye Wang, Xinbing Wang, and Arne A. Nilsson, “Energy-Efficient Bandwidth Allocation in Wireless Networks: Algorithms,
Analysis, and simulations”.
[2]. The ns Manual (formerly ns Notes and Documentation). The Network Simulator, http://www.isi.edu/nsnam/ns
[3]. Mohammad Siraj & Soumen Kanrar, “Performance of Modeling wireless networks in realistic environment”.
[4]. Kurose, Ross, “How To Misuse Aodv: A Case Study Of Insider Attacks Against Ad- Hoc Routing Protocols”.
[5]. Brian Russell, “Maximizing Throughput in a SimulatedWireless Environment”.
[6]. Bounpadith Kannhavong, Hidehisa Nakayama, Yoshiaka Nemoto, Nei Kato, and Abbas Jamalipour, “A survey of Routing Attacks
in Mobile Ad Hoc Networks”.
[7]. Elmurod A. Talipov, “Adding Malicious Node to AODV”.
[8]. Mozmin Ahmed and Dr. Md. Anwar Hussain, “Understanding Vulnerability of Adhoc Networks Under Malicious Node Attacks”.
[9]. Mozmin Ahmed and Dr. Md. Anwar Hussain, “Effect of Malicious Node Attacks Under Practical Adhoc Network”.
[10]. Baruch Awerbuch, David Holmer, and Herbert Rubens , “High Throughput Route Selection in
[11]. Multi-rate Ad Hoc Wireless Networks”.
APPENDIX
Table 3: Network Throughput with malicious nodes varying from zero to ten in each group.

Table 4: Number of Packets Dropped with malicious nodes varying from zero to ten in each group

Impact of Malicious Nodes on Throughput, Packets Dropped and Average Latency in MANETs

More Related Content

What's hot (16)

Viewers also liked (12)

Similar to Impact of Malicious Nodes on Throughput, Packets Dropped and Average Latency in MANETs (20)

More from iosrjce (20)

Recently uploaded (20)

Impact of Malicious Nodes on Throughput, Packets Dropped and Average Latency in MANETs