Proper

Abstract

wireless networks. In these networks to identify the attacker node, it will be checked by the cryptographic authentication, everytime authentication is not possible because it needs key management and additional infrastructural overhead. The proposed system focus on 1.) Detecting spoofing attacks, 2.) identifying the no of attackers when conflicts masquerading the same node identity, 3.) localizing multiple conflicts. To detect the spoofing attacks received signal strength (RSS) is used. To determine the no of attackers cluster based mechanisms are introduced. In this paper, K-Nearest-Neighbor classifier (KNN) is proposed to improve the performance of determining the number of attacks.