The authentication requires key generation, Key management, revocation,

The sensor network includes spatially distributed small and low
cost autonomous sensing devices together with wireless radio transceiver for
examining the environment and to collect the sensor readings from one or more
base stations. Unlike other traditional networks, the wireless sensor networks
are infrastructure-less. It can also be operated under various environmental
conditions as compared to the other wired and wireless Adhoc networks. It
involves the collection of data and establishing a communication between base
stations to one or more sink nodes 1. WSNs are widely used in many civilian
applications like wild habitat monitoring, forest fire detection, building
safety monitoring, military surveillance and so on 2.

Due to their relatively open nature and unattended operations,
sensor nodes are at high risk of being physically captured by different attacks
and having their security compromised 3. When WSN’s are deployed in hostile
regions where sensor nodes are physically captured and manipulated,
communication security and energy efficiency of the sensor nodes are the major
issues 4.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

The open and broadcasting nature of WSN, the adversary can then
inject false reports via the compromised nodes to overhear transmitted packets
or inject the packets in the sensor network. Thus, WSN’s are easily attacked by
different kinds of internal and external factors.

Traditionally, they have proposed the method of statistical
enrouting filtering which is used to detect and drop injected false reports
during the forwarding process 5. Major type of wireless sensor network attack
is a spoofing attack. In WSN 15,16&17, sensor nodes takes some values and
sends it to the sink or the base station .While routing this information an
attacker may alter or spoof the routing information to disrupt traffic in the
network. 

The traditional approach to address spoofing attacks is to apply
cryptographic key management 6. However, authentication requires key
generation, Key management, revocation, energy and additional infrastructural
overhead. Due to the limited resources of sensor nodes, it is not always
possible to deploy authentication. Key management often has an impact on
significant human management costs in the network.

Beacon based algorithms or signature based algorithms are
implemented in Wireless sensor networks to perform localization. In these
localization algorithms, received signal strength indicator (RSSI) is used as
the parameter 7.Though RSSI plays a significant role it never adds any
algorithmic complexities or overhead to the wireless sensor nodes.

In practical, even under ideal conditions it cannot be used to
determine inter nodal distances in wireless sensor networks. Furthermore, in
RSS the results for detection of identity-based attacks are derived by K-means
algorithm 8. In this, the centroids of the clusters in signal space acts as
the input to the localization system. Hereby, the positions of the attackers
can be localized with the same relative estimation errors under normal
conditions. The probability of the existing algorithm in detecting a spoofing
activity in worst case is lesser.  Our
paper describes about new light weight mechanism to detect and localize the
spoofing attackers effectively with low energy consumption.

Instead of relying on cryptographic based approaches, the spatial
information based on structured SVM is used to assist in attack localization
and detection. In SEECAD mechanism 20, boundaries are created by which the
entry of intruders can easily detected.

The false positive rate can be reduced by increasing the detection
rate using SEECAD mechanism.

This mechanism can also accurately localize multiple adversaries.
This localization happens even when the attackers vary their transmission power
levels and thereby tricking the system for their true locations.

This shows that various localization algorithms also achieve the
same performance while running on SEECAD mechanism rather than the traditional
localization attempts.

The rest of the
paper is organized as follows. The context of the system model and the
performance metrics is measured as the function of spoofing or jamming power is
described in Section 2. The theoretical analysis of RSS and the generalized
attack detection model using SEECAD is described in Section 3. Section 4
contains system simulation results and Section 5 presents the conclusions.

Comments are closed.