AADS: DEVELOPING AGENT ASSISTED DEFENSE SYSTEM AGAINST HARMFUL ATTACKS IN COGNITIVE RADIO NETWORK

Abstract

Cognitive radio network (CRN) is an emerging technology that provides ability of adapting operating parameters to wireless devices in order to overcome spectrum scarcity problems. The paper mainly focused on three major attacks held on CRN such as, primary user emulation (PUE), spectrum sensing data falsification (SSDF), and eavesdropper attacker. To overcome attacks, this paper designs an effective agent assisted defence system in CRN. Initially network is partitioned into four different groups based on communication range of PBS. To mitigate SSDF attack the malicious node is isolated from the network. This decision is made by tri-state Markov chain model (MCM) which predicts future activity of the node. Communication in the network is protected by hybrid advance encryption with Diffie-Hellman encryption (HAES-DHE) algorithm to alleviate eavesdropper attack. Experimental result shows promising improvements in probability of detection, probability of miss detection, estimation of attack strength, average estimate of honest nodes and throughput.