Fuzzy Based Secure Intrusion Detection System for Authentication in Wireless Sensor Networks

Authors:

Abstract
Wireless Sensor Network is a indivisible part of network where it has no infrastructure. In the past, Intrusion detection systems were used to detect intrusions in network effectively. Most of the systems are able to detect intrusions with high false alarm rate. In this paper, we propose a Fuzzy based Secure Intrusion Detection System (FSIDS) for detecting malicious activities and providing authentication as well as data integrity. To achieve this, Cluster based routing is established based on trust vector of neighbor nodes in random topology. Certificate based trust recommendation is estimated based on packet identification, certification revocation record and packet receiving capability etc. Including this, clock based intrusion detection is also proposed for identifying and isolating the malicious nodes in network. Security is also enhanced with the proposed IDS using fuzzy encryption and decryption to provide data integrity and authentication. Simulation results shows that the FSIDS provides better detection efficiency, packet delivery ratio, low end to end delay, successful certification rate and low overhead than existing schemes.