Wireless Sensor Network, 2010, 2, 264-266
doi:10.4236/wsn.2010.23036 Published Online March 2010 (http://www.scirp.org/journal/wsn)
Self Umpiring System for Security in Wireless Mobile Ad
Hoc Network
Ayyaswamy Kathirvel, Rengaramanujam Srinivasan
Assistant Professor, B.S.A. Crescent Engineering College, Chennai, India
Professor, B.S.A. Crescent Engineering College, Chennai, India
Received December 22, 2009; revised January 6, 2010; accepted January 8, 2010
Abstract
A wireless mobile ad hoc network (MANET) is a self-created self-organized and self-administering set of
nodes connected via wireless links without the aid of any fixed infrastructure or centralized administrator. In
this paper we propose a solution of self-umpiring system that provides security for routing and data forward-
ing operations. In our system each node in the path from source to destination has dual roles to perform:
packet forwarding and umpiring. In the umpiring role, each node in the path closely monitors the behavior of
its succeeding node and if any misbehavior is noticed immediately flags off the guilty node. The umpiring
system proposed is sufficiently general and can be applied to any networking protocol. For demonstration,
we have implemented the self-umpiring system by modifying the popular AODV protocol. Simulation stud-
ies show that the proposed system increases throughput by 166.9% with an increase in communication over-
head of 13.3% as compared to plain AODV, when 40% of the nodes are malicious and are roaming with a
mobility of 20 m/s.
Keywords: MANET, Security, AODV, Self-Umpiring System
1. Introduction
A wireless mobile ad hoc network (MANET) is a
self-created self-organized and self-administering set of
nodes connected via wireless links without the aid of any
fixed infrastructure or centralized administrator. Each
node moves and operates in a distributed peer-to-peer
mode, generating independent data and acting as a router
to provide multi-hop communication. MANET is ideally
suited for potential applications in civil and military en-
vironments, such as responses to hurricane, earthquake,
tsunami, terrorism and battlefield conditions. Security is
an important aspect in such mission critical applications.
In this paper we tackle the problem of securing the
network layer operations from malicious nodes. Mali-
cious nodes may disrupt routing algorithms by transmit-
ting a false hop count; they may drop packets, route the
packets through unintended routes and so on. Our work
rests on the foundations of two excellent systems already
proposed: the twin systems of watchdog and pathrater [1]
and SCAN [2].
Our self-umpiring system has been strongly influenced
by the above two schemes. In our system all the active
nodes have dual roles just as in watchdog; we also ex-
ploit promiscuous hearing functionality as done by both
SCAN and watchdog. We have adopted the token con-
cept from SCAN. However we have dropped partially
the pathrater functionality. We believe link reliability
assessment of pathrater may not be correct; a proper re-
liability metric for path assessment should consider the
direction and velocity of movement of active nodes.
Having dropped the link reliability factor from the path-
rater, the only other functionality that remains is avoid-
ance of malicious nodes. We achieve the avoidance of
malicious nodes by a system of tokens, which is similar
to the ones used in SCAN. Token is a pass or validity
certificate enabling a node to participate in the network.
It contains two fields: nodeID and status bit; nodeID is
considered to be immutable. Initially the status bit of all
participating nodes is set as 0 indicating “green flag”
with freedom to participate in all network operations. It
is assumed that a node cannot change its own status bit.
When an umpiring node finds its succeeding node mis-
behaving it sends a M-Error message to the source and
malicious node’s status bit is changed using M-Flag
message (set to 1 indicating “red flag”). With “red flag”
on the culprit node is prevented from participating in the
network.

A. Kathirvel ET AL.265
The rest of the paper is organized as follows: Section 2
provides an overview of Self_USS models. Section 3
presents simulation results; Section 4 gives the related
work and Section 5 gives the conclusions
2. Self-Umpiring System Security Model:
Self_Uss
In the self-umpiring system each node is issued with a
token at the inception. The token consists of two fields:
NodeID and status. NodeID is assumed to be unique and
deemed to be beyond manipulation; status is a single bit
flag. Initially the status bit is preset to zero indicating a
green flag. The token with green flag is a permit issued
to each node, which confers it the freedom to participate
in all network activities. Each node in order to participate
in any network activity, say Route Request RREQ, has to
announce its token. If status bit is “1” indicating “red
flag” protocol does not allow the node to participate in
any network activity. The working of the self-umpiring
system is explained with reference to Figure 1.
In the self-umpiring system all the nodes have dual
roles—packet forwarding and umpiring. In the forward
path during data forwarding, each node monitors the per-
formance of immediate next node. That way, node A can
tell correctly whether B is forwarding the packet sent by
it, by promiscuously hearing B’s transmissions. Similarly
during reply process RREP, C can verify whether B is
unicasting the route reply RREP and whether the hop
count given by B is correct. Thus during forward path A
is the umpire for B and C is the umpire for B during re-
verse path operations.
When a node is found to be misbehaving—say drop-
ping packets, the corresponding umpire immediately
sends a M-ERROR message to the source and the status
bit of guilty node is set to “1”—red flag using M-Flag
message. In order to correctly correlate the overheard
messages an additional field next_hop has been intro-
duced in all routing messages as done in SCAN [2].
Though there are several kinds of misbehavior that could
be captured by promiscuous hearing we are focusing
only on two types of malicious actions: dropping packets
and transmitting false hop count.
Figure 1. Self umpiring system model.
3. Simulation and Results
We use a simulation model based on QualNet 4.5 in our
evaluation [3]. Our performance evaluations are based on
the simulations of 100 wireless mobile nodes that form a
wireless ad hoc network over a rectangular (1500 × 600
m) flat space. The MAC layer protocol used in the simu-
lations was the Distributed Coordination Function (DCF)
of IEEE 802.11 [4]. The performance setting parameters
are given in Table 1.
Before the simulation we randomly selected a certain
fraction, ranging from 0% to 40% of the network popula-
tion as malicious nodes. We considered only two attacks
—modifying the hop count and dropping packets. Each
flow did not change its source and destination for the
lifetime of a simulation run.
3.1. Throughput
In the world of MANET, packet delivery ratio has been
accepted as a standard measure of throughput. Packet
delivery ratio is nothing but a ratio between the numbers
of packets received by the destinations to the number of
packets sent by the sources.
From packet delivery ratio the following conclusions
can be drawn:
1) In general packet delivery ratio decreases as mobil-
ity and percentage of malicious nodes increase.
2) We observe that the same results are obtained with
Self_USS also. With zero percentage malicious nodes,
self-umpiring system and plain AODV have almost iden-
tical performances.
3) We find similar increase in throughput at all other
combinations of malicious node percentages and mobil-
ity values, with self-umpiring system.
From the above results we conclude that self-umpiring
system leads to a substantial improvement over plain
AODV, from the point of view of throughput.
3.2. Failure to Deduct (False Negatives) Probability
False Negatives Probability can be defined as:
False Negatives Probability = number of malicious
nodes left undetected/total number of malicious nodes.
Table 1. Parameter settings.
Simulation Time 1500 seconds
Propagation model Two-ray Ground Reflection
Transmission range 250 m
Bandwidth 2 Mbps
Movement model Random way point
Maximum speed 0-20 m/s
Pause time 0 seconds
Traffic type CBR(UDP)
Payload size 512 bytes
Number of flows 10/20

A. Kathirvel ET AL. 266
The above definition requires some elaboration. We
can think of two groups of malicious nodes that are left
undetected. In the first group are those nodes, which
never played a part in the network operation; they were
probably traveling along the boundaries and never had a
chance to participate in the network activity.
The second groups of malicious nodes are those that
played a role as a forwarding node, but went undetected.
Clearly our umpiring system is responsible only for the
second group. The first group of nodes is similar to re-
serve players in the sidelines and clearly any umpire
cannot show red flag and march off players in the side-
lines. Appropriately we have done the failure to detect
probability calculation taking into consideration only
those nodes, which took part in the network activity.
Other researchers adopt the same approach also. The
results are similar that of SCAN [2].
3.3. False Accusation (False Positives) Probability
This is the probability of wrongly booking innocent
nodes. We find false positive probability increases with
increasing percentage of malicious nodes and increased
mobility. The values vary between 0 to 10% and are
similar to the patterns obtained for SCAN [2].
3.4. Communication Overhead
Communication overhead can be evaluated based on the
number of transmissions of control messages like RREQ,
RREP, RERR in the case of plain AODV and in addition
M_ERROR, M-Flag messages in the self umpiring sys-
tem. RREQ are to be decimated to the entire network,
where as RREP messages are unicasts.
From communication overhead following inferences
can be drawn:
1) The communication overhead increases with in-
creasing percentage of malicious nodes and mobility for
both plain AODV and Self_USS.
2) Further we find that when there is no malicious
nodes (0% malicious nodes) the nodes in their umpiring
role have very few message packets to send and the
communication overheads for plain AODV and Self_USS
are nearly same.
4. Related Works
The key distribution center (KDC) architecture is the
main stream in wired network because KDC has so many
merits: efficient key management, including key genera-
tion, storage, distribution and updating. The lack of
trusted third party (TTPs) key management scheme is a
big problem in mobile ad hoc network [5–7].
All the above schemes only try to protect the system
from the attacker, but not bother about quarantining at-
ackers. The twin systems of watchdog and pathrater [1]
not only detect the mischievous nodes but also prevent
their further participation in the network. SCAN [2] also
has similar action, but is more comprehensive, in the
sense not only packet dropping but also other misbehav-
iors like giving wrong hop count are covered. Our self-
USS is an extension of the above two works.
t
Routeguard [7] is similar to pathrater and is run by
each node. Routeguard introduces more detailed and
natural classification system that rates each node into one
of the five classes: fresh, member, unstable, suspect or
malicious. Accordingly each node is treated differently.
5. Conclusions
A self-umpiring system for security for wireless mobile
ad hoc network has been proposed. Simulation studies
show that the proposed system increases throughput by
166.9% with an increase in communication overhead of
13.3% as compared to plain AODV, when 40% of the
nodes are malicious and are roaming with a mobility of
20 m/s. Research work is in progress.
6. Acknowledgements
We express our thanks to Prof. V. M. Periasamy, the Reg-
ister and Prof. K. M. Mehata, the Head, Department of CSE,
B. S. A. Crescent Engineering College Chennai, Tamilnadu,
India for the encouraging environment provided.
7. References
[1] S. Marti, T. J. Giuli, K. Lai, and M. Baker, “Mitigating
routing misbehavior in mobile ad hoc networks,” Pro-
ceedings of the Sixth Annual International Conference on
Mobile Computing and Networking, Boston, USA, pp.
255–265, 6-11 August, 2000.
[2] H. Yang, J. Shu, X. Meng and S. Lu, “SCAN: Self-or-
ganized network-layer security in Mobile ad hoc net-
works,” IEEE Journals on Selected Areas in Communica-
tions, Vol. 24, No. 2, February 2006.
[3] 3. Scalable Networks Technologies: QualNet simulator
4.5, http://www.scalable-networks.com/
[4] IEEE 802.11g. Part 11: Wireless LAN Medium Access Control
(MAC) and Physical Layer (PHY) specifications, August, 1999.
[5] M. A. Azer, S. M. El-Kassas, and M. S. El-Soudani, “Certi-
fication and revocation schemes in ad hoc networks sur-
vey and challenges,” Proceeding of IEEE International Con-
ference on Systems and Networks Communications, 2007.
[6] J. Kong, P. Zerfos, H. Luo, S. Lu, and L. Zhang, “Pro-
viding robust and ubiquitous security support for
MANET,” Proceeding of International Conference on
Network Protocols, pp. 251–260, 2001.
[7] N. Nasser and Y. Chen, “Enhanced intrusion detection
system for discovering malicious nodes in mobile ad hoc
networks,” Proceeding of International Conference on
Communications, pp. 1154–1159, 2007.