Energy-Efficient Distributed Spectrum Sensing for Cognitive Sensor Networks

TLDR

Conventional distributed sensing improves detection with more radios but also increases network energy consumption. The study seeks to minimize energy consumption in distributed spectrum sensing while preserving detection performance by optimally selecting sleeping and censoring parameters in cognitive sensor networks. The authors enforce a target detection probability and a false‑alarm limit, and assess the scheme under blind and knowledge‑aided scenarios depending on prior knowledge of primary‑user presence. In an IEEE 802.15.4/ZigBee network, the proposed sleeping‑censoring scheme delivers significant energy savings.

Abstract

Reliability and energy consumption in detection are key objectives for distributed spectrum sensing in cognitive sensor networks. In conventional distributed sensing approaches, although the detection performance improves with the number of radios, so does the network energy consumption. We consider a combined sleeping and censoring scheme as an energy efficient spectrum sensing technique for cognitive sensor networks. Our objective is to minimize the energy consumed in distributed sensing subject to constraints on the detection performance, by optimally choosing the sleeping and censoring design parameters. The constraint on the detection performance is given by a minimum target probability of detection and a maximum permissible probability of false alarm. Depending on the availability of prior knowledge about the probability of primary user presence, two cases are considered. The case where a priori knowledge is not available defines the blind setup; otherwise the setup is called knowledge-aided. By considering a sensor network based on IEEE 802.15.4/ZigBee radios, we show that significant energy savings can be achieved by the proposed scheme.