A novel node selection scheme for energy-efficient cooperative spectrum sensing using D-S theory

Energy-efficient and reliable detection of available spectrum are fundamental objectives for cooperative spectrum sensing (CSS) in cognitive radio sensor networks (CRSNs). In this paper, a novel node selection scheme for energy-efficient CSS based on Dempster-Shafer (D-S) theory is proposed. Firstly, taking into account the historical data of nodes with historical reliability and residual energy of nodes, we propose a filtering strategy to filter out ineligible nodes, in order to reduce the computation loads in later steps and the amount of sensing results to be transmitted. Secondly, taking into account energy consumption balance of the network with the distance from the remaining nodes to fusion center, we propose a representative node selection algorithm for CSS, in order to reduce energy consumption. Thirdly, we consider that some representative nodes (R-Nodes) may not work as expected. Hence, facing this problem of malicious nodes in CRSNs, we propose an evaluation method based on D-S theory which considers simultaneously the current reliability and the mutually supportive degree among different R-Nodes to derive final decision. Simulation results show that the proposed scheme can not only reduce energy consumption, but can guarantee spectrum sensing accuracy, even in the presence of malicious nodes.