Distributed Energy-Efficient Detection in Sensor Networks with an Unknown Number of Sensors

This work considers the problem of collaborative detection in sensor networks with an unknown number of operating nodes. In wireless sensor networks, both the energy resource and the bandwidth of communication channel are limited. This work employs the sensor censoring scheme to achieve energy-efficiency and low communication rate on the design of distributed decision fusion when the number of operating sensors is unknown to the fusion center. Very surprisingly, in this work, we showed that the energy conservation does not necessary result in the degradation of fusion performance in both theoretical analysis and numerical simulations. Indeed, in many cases, utilizing more energy or bandwidth actually degrades the fusion performance, and the design of energy-efficient local detection rule should start from a nonzero censoring rate, which gives the optimal fusion performance