Evaluation of Security Problems and Intrusion Detection Systems for Routing Attacks in Wireless Self-Organised Networks

Abstract

Security in Wireless Self-organised Networks (WSoN) could be looked at from various view points with respect to different communication layers through which their functionalities are attained. Multihop communication between mobile devices may require specially designed routing algorithms for self-organised networks; those routing protocols can become primary targets for attacks. In this article we run a simulation study of an attack causing an 86% decrease in packet delivery ratio and an increase in packet drops by the same proportion for a selected scenario. We simulated a wireless sensor network with the presence of phenomenon, to investigate and extended it to generate denial of service attacks. We have used the phenomenon contribution to generate a realistic traffic pattern for accurate evaluation of protocols, and compared it to the traditional method of using only Constant Bit Rate (CBR) traffic, usually employed by most researchers. We then adapted an Intrusion Detection System (IDS) into our simulated sensor network, and achieved detection rate of greater than 90% with a very low false positive rate of less than 1%. Each node was configured to independently monitor, detect and report intrusions, providing us with more accurate results and avoiding the additional delays and existence of a central point of failure that a centralised IDS would introduce. We also investigated the impact of a Black-hole attack on MANETS and we observed that under the on-demand routing protocol, the closer a malicious node is to the source of traffic, the greater the extent of damage inflicted on the networks. This study on security attacks and intrusion detection systems has led us to illustrate some of the particular security issues in WSoN.