Detecting Denial of Service attack in Wireless Sensor Networks

Abstract

Wireless sensor networks, thanks to recent technological advances, has become prevalent and offer a variety of applications ranging from environmental monitoring to support and automate chores fields. However, this very promising technology faces many inherent constraints (sensor node architecture, runtime, etc...). All these, because the network face many challenges such as energy efficiency, routing, self-organization and self-maintenance, data aggregation, security, mobility, etc. A wireless sensor network is a special case of ad hoc networks and therefore inherits certain characteristics of ad hoc networks. Due to the nature of the wireless environment, the sensor nodes face many security challenges. Intruders may enter the network and cause disruption of its normal operation. Nodes usually perform energy-saving mechanisms that allow them to switch to standby (sleep) mode from time to time. However, an evil intentioned node can join the network and thus prohibit nodes wishing enter standby mode from turning off their radio. This can be termed as sleep deprivation torture also known as Denial of sleep attacks. It is achieved by making them believe that there is data to be transmitted or just has to stay awake for monitoring. Much overhead is introduced in most of the existing works on sleep deprivation attacks detection, leading to poor performance. The need of the day is to therefore develop energy efficient methods by which the attack can be mitigated. In this work, a strong link-layer authentication and Anti-replay protection is proposed for TMAC protocol to mitigate Denial of sleep attacks. Simulation results show that our proposed mechanism is able to reduce the effects of Denial of sleep attacks in Wireless Sensor Networks.