Optimal threshold and weighted cooperative data combining rule in Cognitive Radio network

This work was supported by following projects: Shanghai Leading Academic Discipline Project (S30108), STCSM (08DZ2231100) and the National Natural Science Foundation of China (60872021). Abstract-For spectrum sensing in Cognitive Radio (CR), previous researches mostly considered that distributed CR nodes are under identical channel conditions, so uniform and fixed detection threshold is set normally with Energy Detector. However, the distribution of nodes is not quite the same due to different channel states. In this paper, the optimal threshold to minimize the total error rate over AWGN channel is derived firstly. Then the effect of variable SNR and noise power on the optimal threshold is investigated. Based upon the optimal threshold theory, we propose Dynamic Threshold Scheme to reduce average total error rate for local CR node. Finally, for cooperative sensing, to distinguish the role of different nodes at Base Station, weighted rule for data combining, based on the difference between energy decision statistic and the optimal threshold, is proposed. Simulations have shown the performance is conditionally improved a lot compared with other rules.