J. Software Engineering & Applications, 2009, 2:50-54
Published Online April 2009 in SciRes (www.SciRP.org/journal/jsea)
Research on Survivability of Mobile Ad-hoc Network
Yuan Zhou1, Chunhe Xia1, Haiquan Wang1, Jianzhong Qi1
1Key Laboratory of Beijing Network Technology, School of Computer Science and Engineering, Beihang University, Beijing, China.
Received November 26th, 2008; revised January 7th, 2009; accepted February 16th, 2009.
ABSTRACT
In this paper, we analyze the survivability of Mobile Ad Hoc Network systemically and give a detailed description of the
survivability issues related to the MANET. We begin our work with analyzing the requirements of survivability of ad
hoc network, and then we classify the impacts that affect survivability into three categories: dynamic topology, faults
and attacks. The impacts of these factors are analyzed individually. A simulation environment for the MANET towards
survivability is designed and implemented as well. Experiments that under the requirements and the impacts we de-
clared are done based on this environment.
Keywords: Ad hoc, Survivability, GTNeTS, Simulation
1. Introduction
A mobile ad-hoc network (MANET) is a kind of
self-organized wireless network with a collection of mo-
bile hosts that is multi-hop instead of using any fixed
infrastructure or centralized management [1,2,3]. There
are a lot of potential applications of MANET in both ci-
vilian and military areas. For instance, it can be applied
in disaster communications, used as the back-up network
of traditional mobile communication networks and used
to build tactical network, etc. With the characteristic of
MANET like: dynamic topology, no infrastructure or
trust institution and limited power resources, research on
survivability of MANET is becoming an academic hot-
spot.
The concept of survivability was first presented by
Barnes in 1993. After that, Knight, K.Sullivan, R. J. Elli-
son and many research institutes have done lots of effort
on issues of survivability. To be brief, survivability is the
capability of a system to provide essential services under
attacks, failures or accidents. At present, many researches
of MANET survivability put emphasis on designing new
survivable routing protocols for specific problems [4,5].
However, there is few integrated and systematic analysis
in the MANET survivability area [6]. On the other hand,
research institutions usually use simulation approaches
for experimental verification on MANET. However, the
existing MANET simulators are mainly constructed for
the purpose of experimenting performance or protocol of
MANET [7]. Attributions related to MANET survivabil-
ity are absent in the simulator structure. Additionally,
threats like faults and attacks can not to be brought in to
those simulators easily to verify MANET survivability
issues.
For the above reasons, this paper presents a systematic
analysis of MANET towards survivability. It describes
how various impacts affect the survivability of MANET
in details. Meanwhile, a simulation environment is designed
and implemented for MANET towards survivability,
which incorporates not only the attributions related to
MANET survivability but also fault events and attack
events against MANET. Users can draw many kinds of
scenarios of MANET to test survivability issues under
this simulation environment.
2. Analysis of Survivability of MANET
2.1 Definition of Survivability
Barnes presented the conception of “Survivability” for
the first time in 1993, but until now, there is no all-
acceptant definition of Survivability. The most influ-
encing definition is presented by a research group of
CMU/SEI, who define survivability as the capability of
a system to fulfill its mission in a timely manner, in the
presence of attacks, failures, or accidents [8]. Many
other researchers define survivability from different
perspective: In the area of software engineering,
Deutsch defined surviva- bility as the degree to which
essential functions are still available even though some
part of the system is down [9]. Ellison et al. introduced
the following definition: survivability is the ability of a
network computing system to provide essential services
in the presence of attacks and failures, and recover full
services in a timely manner [10]. Researchers like Moi-
tra [11], Jha [12], and Wilson [13] also defined the sur-
vivability similarly.
This work was supported by three projects: the National 863 Projec
t
-Research on high level description of network survivability model and
its validation simulation platform under Grant No.2007AA01Z407, The
Co-Funding Project of Beijing Municipal education Commission unde
r
Grant No.JD100060630 and National Foundation Research Project.

Research on Survivability of Mobile Ad-hoc Network 51
The definitions above well describe the meaning of
survivability in natural language, but attributions towards
survivability are not embodied, especially those of the
MANET. Compared with the descriptive definition, a
formal description can give a more rigorous definition of
survivability of MANET.
2.2 Definition of Survivability of MANET
After analyzing the definitions listed above, we find that
the essence of survivability is the ability of providing
essential system-wide service. Papers [6,14,15] presented
that the essential service a mobile ad hoc network must
provide is the communication service. That is to say, in
the context of MANET, the essential service is primarily
pivotal to a fundamental requirement: establishing a
connection between any two nodes in an ad hoc network
at any instant. So the survivability issue depends on how
well an ad hoc network demands the requirement. On the
other hand, the threads mobile ad hoc network must face
to are as follows:
1) Dynamic topology influence. Nodes in an MANET
can move arbitrarily. Consequently, network topology
may change rapidly and randomly.
2) Faults that may happen in nodes and links. For in-
stance, a node can shutdown itself for certain reasons,
and a link may be influenced by an obstacle.
3) Every layer of ad hoc network architecture may be
under attack. The Wormhole, Blackhole attacks aim at
the network layer, jamming and eavesdropping attacks
aim at the physical layer, and SYN flooding attacks aims
at the transport layer, to name a few.
With the above analysis we can define the survivability
of a mobile ad hoc network as, the ability of establishing
a communication service between any two nodes in the
network at any instant under the impact of dynamic to-
pology, faults and attacks.
2.3 Detailed Description of Survivability of MANET
Based on the definition of survivability of MANET
above, the system of MANET towards survivability is-
sues can be abstracted into three members: mobile ad hoc
network, the impacts which can affect the survivability of
the mobile ad hoc network, and the essential service that
a mobile ad hoc network must provide. It can be de-
scribed as follows. Then, we will analyze them in detail
SurvivalAdhoc={ADHOC,SERVICE,IMPACT} (1)
The mobile ad hoc network is composed of a number
of nodes with certain attributions related to the surviv-
ability and directed links. We consider the node obtaining
following attributions: IP address, location, radio range,
state (transmit, receive, idle and down), power, protocol
type and, mobility type. All these attributions are related
to the survivability of MANET. It can also be described
as follows:
ADHOC::={NODE,LINK} (2)
NODE::={node1,node2,…noden} (3)
LINK::={<nodei,nodej>|nodei,nodej
∈ NODE} (4)
node=(ip,id,location,R,state,power,
protocoltype,mobilitytype) (5)
The essential service of a mobile ad hoc network is to
establish a communication service between any two
nodes in the network at any instant. The ad hoc network
establishes a connection by two steps. First, connections
between any two adjacent nodes are provided by the link
layer and physical layer, and then such connections are
extended by the network layer form one-hop to multi-
hops. That is to say: for an ad hoc network with N nodes,
there need to be a directional path (Nodei, Nodej….Nodel,
Nodm,) between any two nodes. And the path must meet
the following requirements:
1) The distance between any two nodes in neighbors,
Nodei and Nodei+1, is less than the transmission range of
Nodei.
2) Any node in this path must possess a routing entry
that sets the target node as the destination node, and the
next hop is the succeeding node.
The impacts affecting the ad hoc network survivability
are dynamic topology, faults and attacks. The essence of
all these three factors is to destroy the path between two
nodes so that the network is disconnected. Thus, the fac-
tors are described as follows:
:: { |()()
(), }
I
MPACTaDynamicToplogy aFaulta
Attack aaActions
=
∨
∨∈ (6)
The factor of dynamic topology is caused by the ac-
tions such as mobility of nodes. It can result in the loca-
tion changed of the node, and lead to the distance re-
quirement of link layer connection dissatisfied. The in-
fluence of dynamic topology is that the new distance of
two nods in neighbor may be larger than the transmission
range of the first node, so that the path is destroyed. If
there is no other path between the destination node and
target node, they can’t afford the communication service.
Faults are considered to include node faults and link
faults [14]. Node faults are the actions that cause the
state of the nodes changed to down. If the state of a
node is changed to down, the transmission range will be
changed to 0, which will make the path dissatisfy the
requirement. Link faults are introduced by obstacles
between nodes, or by signals fading effect which may
influence the transmission range of nodes. It can be de-
scribed as follows:
()() (),
F
aultaNodeaLinkaaActions→∨ ∈ (7)
Attacks of MANET are usually classified by network
layers. The attack aiming at the application layer is more
or less the same as the wired network. Its main object is
to disrupt the application service, and worm is an in-
stance. Attacks to the transportation layer are the actions
to disrupt the transportation layer protocols, like SYN
flooding. The attacks to the network layer are the actions
that destroy the routing and forwarding processes in ad
hoc network [16]. The attacks to link layer are the actions

52 Research on Survivability of Mobile Ad-hoc Network
which destroy the connection of adjacent nodes. The at-
tacks to physical layer are actions to jam, intercept or to
eavesdrop the wireless channel. The description of at-
tacks to network survivability is:
()() ()
() ()()
Attack aPhysicalAttack aLinkAttack a
NetworkAttack aTranspAttack aApplicationAttack a
→∨
∨∨∨
(8)
3. Simulation Environment
Existing simulators are not well-equipped to serve our
purpose. Hence, we design a survivability-based simulation
environment of MANET to test how those factors influ-
ence the network.
3.1 Characteristics of Simulation Environment
While designing the simulation environment, we concern
the following factors:
·Towards Survivability: The users can configure the pa-
rameters freely, such as node amount, mobility model, radio
range, and power of node, which may affect MANET sur-
vivability. Fault events and attack events can be added to
simulate scenarios towards survivability thoroughly.
·Expandability: Users can easily expand the simulation
environment by adding new protocols of various layers,
modules, attack events and fault events. For this reason,
MANET towards survivability description language is
designed and a language translator is implemented for it.
Users can add new protocols, functions, faults and attacks
through appending new keywords.
·Introduce Fault-event and Attack-event: Based on the
definition and description of MANET survivability, faults
and attacks are described, modeled and simulated to
support MANET survivability verification.
·Data Acquisition Interface: One of the most important
purposes of simulation is collecting data for further
analysis from simulation process. Thus, data acquisition
interface is incorporated in the simulation environment
and users are allowed to configure what kind of result to
be collected.
Based on the analysis above, simulation environment
includes description language of MANET towards sur-
vivability and its interpreter. Then, we choose Georgia
Tech Network Simulator (GTNetS) [17] as the underly-
ing network simulation platform.
3.2 Simulation Environment Architecture
The system architecture is as shown in Figure 1:
·Graphic Configuration Interface: MANET properties,
attack-events and fault-events information are configured
by users through graphic user interface (GUI) and then be
saved in the configuration file.
·Event interpretation: it analyzes and interpreters the
configuration file, executes corresponding events by
calling functions from event class library. The event class
library consists of fault class library and attack class li-
brary, each of which is built up by specific classes. Cur-
rently, there are node fault classes, link fault classes in
fault class library. And the attack class library contains
energy-consuming attack class, signal interference attack
class, black hole attack class, SYNflooding attack class
and Worn attack class. All these classes are used by re-
ceiving parameters from the interpreter and calling sup-
port functions from GTNetS.
Figure 1. Simulation environment architecture

Research on Survivability of Mobile Ad-hoc Network 53
·Network configuration interpretation: it interprets the
network configure file and makes function calls from
GTNetS to construct simulation of MANET. All these
network properties are supported: energy-consumption
modal, AODV and DSR routing protocol, Random
Waypoint, Random Walk and Random Direction mobil-
ity models, data transfer service, node information and
topology information.
·Data Collection Module: it displays the simulation
process graphically. After that, it parses the collected data
that users care about, including total data transferred
amount, average connectivity node-pairs and so on, then
saves data into the result data file.
4. Experiments on MANET Survivability
Analysis
4.1 Related Definition
For the simulation test of the definition of MANET sur-
vivability, the capability to provide connectivity service
under various threats, terms that used in this paper are as
follows:
Average Connectivity Efficiency [6] (E): It means the
ratio of connected node pairs to all the node pairs within
an N-node MANET in a certain period. In such MANET,
each link between two nodes is directed, which means
node pair (i,j) is different from node pair (j,i). The value
of (E) reflects the capability of a MANET to provide
connectivity service at a certain time. It reaches 100%
when all the node pairs in the network can be connected.
(%)
(. )*100
1
*
Average ConnectivityEfficiency
T
no ofconnectednodepairs
i
T NumberOfNodepairs
=
∑
=
−
(9)
4.2 Experiment of Influences of Dynamic Topology
We choose number of node, radio range of node and mo-
bility speed of node as the variables in this experiment.
We would like to figure out that how the dynamic topol-
ogy influences the survivability of MANET.
We carry out the experiment in a 1000×1000 bounded
squared topology, and Parameters used in the test of the
influences on MANET survivability of dynamic topology
are as follows:
·Simulation time:300
·Node amount:20-80
·Radio range of each node:50-250
·Mobile way of node: Random Walk, 10-20
·Routing protocol of each node :AODV
·Bandwidth of each link:1Mb
From the results we can easily find out that when the
radio range increases, the connectivity efficiency increases.
Howeverm, its effect to the connectivity efficiency is not
obvious when the radio range invreases to a certain
level.As a result, when choosing the radio range, we
should consider its effort to the connectivity efficiency
and its cost to the power consumption.
When the moving speed of node increases, the topol-
ogy is more instable, which means the node will be easier
to move out of other nodes’ transmission range, so that
the connectivity efficiency decreases.
A larger number of nodes mean a higher node density.
So from the results we can see in a bounded area, large
number of nodes will contribute to the connectivity effi-
ciency.
4.3 Experiment of Influence of Faults and Attacks
There are node faults and link faults. When node faults
happen, nodes can shut down themselves randomly.
When link faults happen, it can be emerging obstacles
that affect the links. This paper assumes that only one
fault happens at one time, and ignores the mobility attri-
bution of the nodes, so that the topology is stable when
dealing with the influences of faults.
Figure 2. Average connectivity efficiency vs. transmission
range for different number of node when speed =10
Figure 3. Average connectivity efficiency vs. transmission
range for different number of node when speed =15
Figure 4. Average connectivity efficiency vs. transmission
range for different number of node when speed =20
Speed=10
N=20
N=30
N=50
N=80
90.000%
80.000%
70.000%
60.000%
50.000%
40.000%
30.000%
20.000%
10.000%
0.000%
A
ver.
C
onnect
i
v
i
ty
Efficiency
0 50 100 150 200 250 300
Radio Range
0 50 100 150 200 250 300
Radio Range
Speed=15
N=20
N=30
N=50
N=80
Aver. Connectivity
Efficiency
90.000%
80.000%
70.000%
60.000%
50.000%
40.000%
30.000%
20.000%
10.000%
0.000%
80.000%
70.000%
60.000%
50.000%
40.000%
30.000%
20.000%
10.000%
0.000%
A
ver.
C
onnect
i
v
i
ty
Efficiency
0 50 100 150 200 250 300
Radio Range
N=20
N=30
N=50
N=80
Speed=20

54 Research on Survivability of Mobile Ad-hoc Network
Figure 5. Average connectivity efficiency vs. different event
number per minute
In this paper, attacks of network layer, data-link layer
and physical layer are the three types of attacks that are
included in the experiment. Energy consumption attacks,
for network layer attack testing, and jamming attacks, for
physical layer attack testing, are selected as examples to
test the influences on MANET survivability under attacks.
Node mobility is not considered either.
We make the simulation in a 1000*1000 bounded
square area under different event frequency, which means
we set different number of event that happen per minute.
For every minute we choose these events randomly.
Parameters used in the experiment of the influences on
MANET survivability of faults or attacks are as follows:
·Simulation time:300
·Average fault/attack amount per minute:3-20
·Node amount:100
·Radio range of each node:100
·Mobile way of each node: quite
·Routing protocol of each node: AODV
·Bandwidth of each link:1Mb
From the results we can find out that when event fre-
quency of faults or attacks increases, more nodes can’t
provide stable services for routing and forwarding, and
the connectivity efficiency decreases.
5. Conclusions and Future Work
In this paper, we systematically analyzed MANET sur-
vivability and the impacts that affect it. Based on this, we
designed and implemented the simulation environment.
The simulation environment represents characteristics of
MANET towards survivability and events that impact
survivability of MANET like fault events and attack
events. We then tested the influences of the above- men-
tioned impacts under the simulation environment. Users
can also configure new topologies, attacks and faults to
test MANET survivability in this simulation environment
using their own scenario. In our current research, the in-
fluences of each factor were tested independently. That
means, future work shall be done to present
cross-analysis of the influences of these impacts.
REFERENCES
[1] P. Papadimitratos and Z. Haas, “Handbook of ad hoc
wireless networks,” chapter Securing mobile ad hoc
networks, CRC Press, 2002.
[2] F. Adelstein, S. K. S. Gupta, and G. G. Richard III,
“Funda-mentals of mobile and pervasive computing,”
McGraw-Hill, 2005.
[3] D. Djenouri, L. Khelladi, and A. N. Badache, “A survey of
security issues in mobile ad hoc and sensor networks,”
IEEE Communications surveys & tutorials, 7(4): pp. 2-28,
2005.
[4] M. G. Zapata, “Secure ad hoc on-demand distance vector
routing,” ACM SIGMOBILE Mobile Computing and
Communications Review, 6(3): pp. 106-107, 2002.
[5] Y. Xue and K. Nahrstedt, “Providing fault-tolerant ad hoc
routing service in adversarial environments,” Wireless
personal communications: an international journal, 29
(3-4): pp. 367-388, 2004.
[6] K. Paul, R. R. Choudhuri, and S. Bandyopadhyay,
“Survivability analysis of ad hoc wireless network architecture,”
Proceedings of the IFIP-TC6/European Commission
International Workshop on Mobile and Wireless
Communication Networks, pp. 31-46, May 16-17, 2000.
[7] J. Short, R. Bagrodia, and L. Kleinrock, “Mobile wireless
network system simulation,” Wireless Network Journal,
Vol. 1, No. 4, 1995.
[8] R. J. Ellison, D. A. Fisher, and R. C. Linger, “An approach
to survivable systems,” In: NATO IST Symposium on
Protecting Information Systems in the 21st Century, 1999.
[9] M. S. Deutsch and R. R.Willis, “Software quality engineering:
A total technical and management approach,” IST Symposium
on Protecting Information Systems in the 21st Century, 1999,
Englewood Cliffs, NJ: Prentice- Hall, 1988.
[10] B. Ellison, D. Fisher, R. Linger, H. Lipson, T. Longstaff,
and N. Mead, “Survivable network systems: An emerging
discipline,” Technical Report CMU/SEI-97-TR-013, Software
Engineering Institute, Carnegie Mellon University, November
1997.
[11] D. Moitrasoumyo and L. K. Suresh, “A simulation model
of managing survivability of network of emergent Systems,”
[J], TechniealRePort CMU/SEI-2000-TR-020, 2000.
[12] K. J. Sanjay, M. W. Jeannette, and C. L. Richard, “Survivability
Analysis of Network Spesifications,” In: Workshop on
Depended ability Despite Malicious Fault, 2000 Intemational
Conefreneeon Dependable Systemsand Networks (DSN2000);
2000, NewYork USA: IEEE ComPuterSoeiety: 2000,
June 25-28, 2000.
[13] R. W. Makr, “The quantitative impact of survivable
network architectural on service availbaility,” [J], IEEE
Communications Magazine, 36(5): pp. 71-77, 1998.
[14] D. Y. Chen, S. Garg, and K. S. Trivedi, “Network
survivability performance evaluation: A quantitative
approach with applications in wireless ad-hoc networks,”
Proceedings of the 5th ACM international workshop on
Modeling analysis and simulation of wireless and mobile
systems, Atlanta, Georgia, USA, pp. 28-28, September 2002.
[15] Dahlberg, T. S. Ramaswamy, and D. Tipper, “Issues in
the survivability of wireless networks,” Proceedings IEEE
Mobile and Wireless Communication Networks Workshop,
May 1997.
[16] B.Wu, J. M. Chen, and J. Wu, “A survey of attack and
countermeasures in mobile ad hoc networks,” Wireless
Network Security, Springer US, pp. 103-135, 2007.
[17] G. F. Riley, “The georgia tech network simulator,” In
Proceedings of the ACM SIGCOMM, New York, pp.
5-12, August 2003.
100.000%
90.000%
80.000%
70.000%
60.000%
50.000%
40.000%
30.000%
20.000%
10.000%
0.000%
0 5 10 15 20 25
Event Num
Attack Event
A
ver.
C
onnect
i
v
i
ty
Effi
c
i
enc
y
Fault Event