Improved IDR Response System for Sensor Network

Abstract

Wireless sensor network (WSN) is highly sophisticated than ad hoc wireless network. Ad hoc wireless network is mostly affected by different resources such as high processing energy, storage capabilities and battery backup and etc. Due to the open nature, poor infrastructure, quick deployment practices, and the conflict environments, make them susceptible to a wide range of attacks. Recently, the network attack affects the performance of networks such as network lifetime, throughput, delay, energy consumption, and packet loss. The conventional security mechanisms like intrusion detection system (IDS) of network security are not enough for these networks. In this thesis, we introduce an enhanced intrusion detection and response (EIDR) system using two tire processes. The first contribution of proposed EIDR system is optimal cluster formation and performed by the chaotic ant optimization (CAO) algorithm. The second contribution is to calculate the trust value of each sensor node using the multi objective differential evolution (MODE) algorithm. The computed trust value is used to design the intrusion response action (IRA) system, which offers additional functions and exhibit multiple characteristics of response to mitigate intrusion impacts. The simulation results display that the proposed EIDR system has a better detection rate and false positive rate without affecting network performance.