Impact of Distributed Resource Allocation in Sensor Networks

Sensor networks have generated much interest in the research community because they have a wide range of potentially impactful applications. In this field, one of the major research challenges is the distributed resource allocation problem. This problem concerns how one should allocate or schedule the limited resources in a sensor network to minimize costs and maximize capability. Although various distributed resource allocation techniques have been proposed, the impact of such techniques on real application scenarios of sensor networks is not well understood. In this paper, we study the impact of distributed resource allocation in two realistic application scenarios; namely, the tasking of look-sectors for cameras to monitor vehicle movement, and the sensor mode management in an acoustic sensor network to track vehicle movement. Our simulation results for these two scenarios indicate that with distributed resource allocation, a sensor network can provide a good tradeoff between performance objectives such as target tracking accuracy, coverage area, and network lifetime.