Malicious Node Detection in Wireless Sensor
Networks using Weighted Trust Evaluation
ABSTRACT
Deployed in a hostile environment, individual nodes of a wireless sensor network
(WSN) could be easily compromised by the adversary due to the constraints such as
limited battery lifetime, memory space and computing capability. It is critical to
detect and isolate the compromised nodes in order to avoid being misled by the
falsified information injected by the adversary through compromised nodes.
However, it is challenging to secure the flat topology networks efficiently because of
the poor scalability and high communication overhead. On top of a hierarchical WSN
architecture, in this paper we proposed a novel scheme based on weighted-trust
evaluation to detect malicious nodes. The hierarchical network can reduce the
communication overhead between sensor nodes by utilizing clustered topology.
Through intensive simulation, we verified the correctness and efficiency of our
detection scheme.
EXISTING SYSTEM
Wireless sensor networks are often deployed in a hostile environment and work
without human supervision, individual node could be easily compromised by the
adversary due to the constraints such as battery lifetime, smaller memory space and
limited computing capability. Security in WSN has been one of the most important
topics in the WSN research community [1], [8], [22]. Here we only briefly review the
reported works closely related to malicious node detection due to the limited space.
It is critical to detect and isolate the compromised nodes in order to avoid being
misled by the falsified information injected by the adversary. Luo et al. [11] have
pointed out that infrastructureless ad hoc networks rarely have a real defense
mechanism against most of the attacks, including both outsider and insider attacks
such as compromised node attacks. They suggested a system design like this – if one
node is named trusted by certain number of its neighboring nodes, that particular
node is trusted both locally and globally. However, since the system uses a minimum
number of trusted nodes it is not so applicable to sensor networks where the nodes
are randomly spread out. In other words, it is possible that under certain conditions
nodes cannot find the minimum number of neighboring nodes in order to be named
trusted.
One solution for locationized anomaly detection in a group of nodes is suggested in
[4]. Every node gets the localization information from the neighboring nodes and
also computes the localization information itself and compares these two values. If
the difference is small enough, that node decides there is no adversary around
causing the localization problem in its location. Researchers also suggested detecting
malicious node using signal strength [7]. The idea here is to depend on neighborhood
monitoring of the nodes. Every sensor node monitors its surrounding and whenever a
transmission signal is detected by a sensor node, it would check if the signal strength
of the transmitting node is compatible with the originator node's geographical
position. Even though this approach is applicable, it is not efficient in many ways.
The large overhead needed for transmitting data is a problem both for sending and
processing. Also it is not energy efficient since all nodes are monitoring and
processing data all the time.
The work reported in [3] is the most close to our approach. They proposed to detect
malicious node by comparing its output with an aggregation value. Inspired by the
Byzantine problem, our approach is more straightforward and incurs much less
overhead since there is no expensive calculation involved. Karlof and Wagner [9]
suggested to construct efficient random sampling mechanisms and interactive proofs,
then a user can verify that the answer given by the aggregator is a good
approximation of the true value even when a fraction of the sensor nodes are
compromised. Furthermore, in other fields Byzantine program is considered as an
important issue. For example, in cognitive radio network, Byzantine problem in
spectrum sensing is also investigated [2].
Disadvantages
An existing system didn’t explore Implementation of Malicious Nodes
Detection using WEIGHTED TRUST EVALUATION TECHNOLOGY.
An existing system, it is critical to detect and isolate the compromised nodes in
order to avoid being misled by the falsified information injected by the
adversary.
Proposed System
In this work, we addressed an even trickier scenario. When an adversary has gained
control over certain sensor node(s), he/she does not launch direct attacks against the
network. Since once the misbehavior is detected, the operator may forsake these
compromised nodes and turn to other data sources. Instead, the attacker let those
compromised nodes behave normally but report false data to the data collector. The
purpose of the adversary is to mislead the operator with falsified data. This may lead
to more serious consequences; for instance, in the battlefield a false report regarding
the operations of the enemy may lead to extra casualties.
In this paper, we proposed a weighted-trust evaluation (WTE) based scheme to
detect the compromised nodes by monitoring its reported data. It is a light-weighted
algorithm that would incur little overhead. Considering the scalability and flexibility,
hierarchical network architecture is adopted. Through intensive simulation, we
verified that our WTE scheme detects misbehaved nodes accurately with very short
delay.
Advantages
The proposed system’s approach is based on the assumption that base stations are
trusted. In fact, if the adversary can gain control over the base stations, he/she can do
any possible attack against the WSN. This is an interesting open problem; however it
is beyond the scope of this paper. Another critical assumption is that the majority of
the sensor nodes are working properly. If the number of compromised nodes is more
than the number of normal nodes, the legal nodes will be reported as malicious one
and being isolated.
SYSTEM REQUIREMENTS
➢ H/W System Configuration:-
➢ Processor - Pentium –IV
➢ RAM - 4 GB (min)
➢ Hard Disk - 20 GB
➢ Key Board - Standard Windows Keyboard
➢ Mouse - Two or Three Button Mouse
➢ Monitor - SVGA
Software Requirements:
Operating System - Windows XP
Coding Language - Java/J2EE(JSP,Servlet)
Front End - J2EE
Back End - MySQL
