Cooperative Spectrum Sensing With Adaptive Energy Threshold Control and Efficient Data Fusion

We propose a framework for cooperative spectrum sensing in cognitive radio (CR) networks with adaptive energy threshold control for local spectrum sensing and an efficient approach for data fusion. Each secondary user (SU) carries out energy detection based local spectrum sensing, where the energy threshold is adaptively adjusted in real time according to both the SINR (signal to interference plus noise ratio) observed at the SU receiver and the statistics of the PU transmissions. A distinct linear threshold control function is adopted at each SU, which formulates the dynamic energy threshold as a function of the real-time SINR. The threshold control function is determined by solving a MAR optimization problem, which maximizes the achievable average rate of the SU under a given missed detection probability constraint. We also propose an efficient data fusion mechanism to well balance the impacts of the local sensing beliefs and the number of SUs supporting each hypothesis. As far as we know, this is the first work that adopts adaptive energy threshold control for cooperative spectrum sensing, through which data fusion can directly interplay with not only the diversified time-varying channel states of individual SUs but also the statistics change of PU transmissions. Numerical results show that the proposed framework significantly reduces the false alarm probability while keeping a low missed detection probability as compared with other existing works.