Detection of selective forwarding attack using BDRM in wireless sensor network

Abstract

With rapid development of sensor network technology, reputation model has an important role in the critical process of routing, aggregation, localization, synchronization and security on wireless sensor network. The selective forwarding attack is a kind of packet drop attack which drops sensitive packets received from sensors and degrades the performance of the wireless sensor network. It is difficult to detect in an open wireless environment. In this paper, we propose a Beta Distribution Reputation Model (BDRM) to detect the selective forwarding attack quickly. BDRM utilizes modified beta distribution to compute the reputation of the sensors. BDRM monitors the neighbour behaviours and effectively filters the fluctuations among sensor nodes for better regulations. The result of proposed BDRM shows that the reputation value computed is used to detect selective forwarders in different attack probability ratio. The BDRM increases the packet delivery ratio and throughput of wireless sensor network by isolating selective forwarders.