! "
##"$ %%#
& ’’’(
© 2010, IJARCS All Rights Reserved 192
ISSN No. 0976-5697
A Modified Grayhole Attack Detection Technique in Mobile Ad-hoc Networks
Mangesh M. Ghonge
*
Faculty
Department of Computer Science & Engineering
Jawaharlal Darda Institute of Engineering &
Technology, Yavatmal, India
Pradeep M. Jawandhiya
Assistant Professor & Head of Department
Department of Computer Science & Engineering
Jawaharlal Darda Institute of Engineering &
Technology, Yavatmal, India
Abstract: Mobile Ad Hoc Networks are vulnerable to various types of Denial of Service (DoS) attacks for the absence of fixed network
infrastructure. The Gray Hole attack is a type of DoS attacks. In this attack, an adversary silently drops some or all of the data packets sent to it
for further forwarding even when no congestion occurs. Our proposed scheme comprises two steps of detecting malicious node in the network:
Detection of malicious activity, Identification of malicious node. The first step is to detect any malicious activity took place in network or not
and second if malicious activity took place in the network then identification of that malicious node.
Keywords: MANET, Security attacks, Grayhole attack
I. INTRODUCTION
An ad hoc network is a collection of nodes that do not
rely on a predefined infrastructure to keep the network
connected. So the functioning of Ad-hoc networks is
dependent on the trust and co-operation between nodes.
Nodes help each other in conveying information about the
topology of the network and share the responsibility of
managing the network. Hence in addition to acting as hosts,
each mobile node does the function of routing and relaying
messages for other mobile nodes. Early research work on
route establishment in MANET has mainly focused on the
probability and the efficiency, and assumes nodes are
trustworthy and cooperative. Recently, more attention has
been given to security problems in MANET
The Gray Hole attack is a kind of Denial of Service
(DoS) attacks. In this attack, an adversary first exhibits the
same behavior as an honest node during the route discovery
process, and then silently drops some or all of the data
packets sent to it for further forwarding even when no
congestion occurs. The malicious nodes could degrade the
network performance; disturb route discovery process, etc. In
this paper, we proposed a simple two step method to detect
the malicious node in the network and isolate the node from
the network.
II. RELATED WORKS
Marti et al [1] proposed to trace malicious nodes by
using watchdog/pathrater. This scheme was consisted of two
related algorithms: 1) the watchdog algorithm. When a node
forwards a packet, the node’s watchdog verifies that the next
node in the path also forwards the packet. The watchdog
does this by promiscuously listening to the next node’s
transmissions. If the watchdog finds the next node does not
forward the packet during a certain period of time, the next
node will be suspected as a malicious node. If the next
node’s tally exceeds a predefined threshold, the watchdog
will accuse the next node as a malicious node to the source
node; 2) the pathrater algorithm. The source node selects the
path that most likely to deliver packets, according to the
reports provided by watchdogs equipped with each node in
the network. The proposal has two shortcomings: 1) to
monitor the behavior of nodes two or more hops away, one
node has to trust the information from other nodes, which
introduces the vulnerability that good nodes may be bypassed
by malicious accusation; 2) bi-directional communication
links are needed. Awerbuch et al [2] proposed to detect
malicious nodes by using acknowledgements sent by
destination node. This scheme was consisted of three related
algorithms:
1) The route discovery with fault avoidance. By using
flooding, cryptography algorithms and weight list, the source
nodes could discover route that will deliver packets;
2) The Byzantine fault detection. Based on binary search
algorithm and the input path, the source node could detect
malicious nodes with Byzantine behavior;
3) The link weight management. This algorithm is used
to update the link weight. The proposal has three
shortcomings: 1) the bandwidth overhead is significant, as
the destination node will send an acknowledgement
whenever it receives a packet; 2) it is a challenging work to
make sure that the source node has a shared key with each
node in the network; 3) the probe packet is easily to be
distinguished from other general packet, as the probe packet
contains a probe list. Just el al [3] have reviewed the related
works on tracing packet dropping nodes, and proposed to
detect malicious nodes by using the probe technique. This
scheme was consisted of three related algorithms:
1) The probing path selection algorithm. This algorithm
is used to select the probing paths;
2) The probing algorithm. This algorithm is used to detect
possible malicious nodes in the probing path; the diagnosis
algorithm. This algorithm is used to test the possible
malicious nodes by using the property of bi-directional
communication link.
The proposal has four shortcomings: 1) the source node
will begin to probe malicious nodes when it finds that Gray
Hole attack has taken place. On considering the dynamic
topology of MANET and the random behavior of malicious
nodes, this method is not satisfying; 2) bi-directional
communication links are needed; 3) the efficiency of this
method is related to the location of the malicious nodes in the
source route; 4) in order to keep malicious nodes from
distinguishing probing packets, the probing packets must be
encrypted. Huang el al [4] proposed to detect malicious