Improved Attack Detection Using Decision Tree Based Algorithm

Abstract

Security is an essential requirement in order to provide protected communication in wireless network and wired networks. Due to the dynamic topology and shared medium of MANETs, attackers can easily disrupt these networks. This paper focuses on detection of the multiple attacks. This paper introduces a new technique for the detection of wormhole and gray hole and other attacks in DSR protocol in order to make a wireless ad-hoc network secure. This technique detects the malicious nodes and avoids these nodes for participating in the network. The parameters used for evaluating the network performance are average end to end delay and throughput.