Adaptive Medium Access Control for Distributed Processing in Wireless Sensor Networks

Abstract

Signal and information processing tasks over Wireless Sensor Networks can be successfully accomplished by means of a distributed implementation among the nodes. Existing distributed schemes are commonly based on iterative strategies that imply a huge demand of one-hop transmissions, which must be efficiently processed by the lower layers of the nodes. At the link layer, general purpose medium access (MAC) policies for wireless communications usually focus on avoiding collisions. These existing approaches result in a reduction of the number of simultaneous transmissions, and an underutilization of the channel as a consequence. This leads to a decrease in the performance of the distributed tasks, since an efficient channel occupation is not generally accomplished. In this paper, we propose a new MAC protocol that, besides focusing on the reliability of the communications, provides an efficient channel occupation. While the former has a direct impact on the energy consumption of usually battery powered devices, the latter affects the performance of the distributed task executed. We include both aspects in a global utility function that the nodes, relying just on local available information, aim to increase at every communication step. Furthermore, our proposal combines in a unique framework both unicast and broadcast scenarios. Through several simulation results, we show that our adaptive protocol outperforms the related literature.