Performance of a collaborative target detection scheme for wireless sensor networks

A sensor network is an ad hoc network of a large number of sensor nodes, each equipped with embedded processor, memory, and a wireless transceiver. The collective resources of such a system of networked sensors can be used to perform powerful distributed target detection and monitoring operations over a wide area. Key concerns for designing such networks are the energy efficiency and scalability of its communication and signal processing tasks. In addition, spatial variations of noise characteristics and differences in signal sensitivities amongst the sensors lead to challenging problems for designing a reliable distributed detection scheme. This paper addresses these issues and presents a collaborative signal processing technique that reduces the amount of traffic in the network and also maintains a constant false alarm rate under varying noise powers and sensor sensitivities.