Impulsive interference avoidance in dense wireless sensor networks

Abstract

Wireless sensor networks (WSNs) are subject to interference from other users of the radio-frequency (RF) medium. If the WSN nodes can recognize the interference pattern, they can benefit from steering their transmissions around it. This possibility has stirred some interest among researchers involved in cognitive radios, where special hardware has been postulated to circumvent non-random interference. Our goal is to explore ways of enhancing medium access control (MAC) schemes operating within the framework of traditional off-the-shelf RF modules applicable in low-cost WSN motes, such that they can detect interference patterns in the neighbourhood and creatively respond to them, mitigating their negative impact on the packet reception rate. In this paper, and based on previous work on the post-deployment characterization of a channel aimed at identifying "spiky" interference patterns, we describe (a) a way to incorporate interference models into an existing WSN emulator and (b) the subsequent evaluation of a proof-of-concept MAC technique for circumventing the interference. We found that an interference-aware MAC can improve the packet delivery rates in these environments at the cost of increased, but acceptable, latency. © 2012 Springer-Verlag.