Design and study of novel anomaly detection for wireless sensor network

Abstract

Wireless Sensor Network (WSN) has been an infrastructure-less, self-organizing and self-configured communication infrastructure that comprehends the environmental parameters of the physical world. WSN consists of a base station and a number of sensor nodes called motes. The sensor nodes sense the environmental parameters such as temperature, light, humidity, fire, etc., and send them to the base station. The main components of a sensor node are micro-controller unit, transceiver unit, memory unit, power unit and sensor unit. WSNs have been used in the field of military, commerce, emergency services, health, communication, critical care and monitoring systems. Since the communication is wireless, WSNs are more vulnerable to various security threats. The most important challenge is the secure communication in developing a WSN routing protocol. In the midst of wide applications of WSN, there are several limitations like bandwidth, limited power and limited resources. WSNs are deployed mostly in unattended areas which are more prone to attacks. Hence, the intrusion can be be detected earlier and isolated from the network. The attack may cause performance degradation of the network in respect of throughput, packet delivery ratio, end-to-end delay and overhead. newline