Research Of Cooperative Spectrum Sensing Algorithm Based On D-S Evidence Theory In Cognitive Radio

As one of the key technologies, spectrum sensing technology is the first step for the realization of cognitive radio.At present, research on spectrum sensing technology is mainly focused on the single-user local spectrum sensing and multi-user cooperative spectrum sensing.Multi-user cooperative spectrum sensing can effectively enhance accuracy of spectrum sensing and solve the hidden terminal problem in single-user sensing by fusing several single-user local spectrum sensing information. So multi-user cooperative spectrum sensing is the hot spot in the research of spectrum sensing. In order to further improve the spectrum sensing performance, the adaptive threshold method on energy detection and fusion algorithm on cooperative spectrum sensing has deeply studied in this thesis. Main work of this thesis is as follows:The concept, research background and research status of spectrum sensing technology are first introduced in thesis.Then single-user spectrum sensing algorithm and multi-user cooperative spectrum sensing algorithm are described in detail.In single-user spectrum sensing algorithm,thesis introduces the energy detection algorithm in particular.Meanwhile the simulation of energy detection in AWGN channel and Rayleigh channel are realized in thesis.In multi-user cooperative spectrum sensing algorithm,the classical fusion algorithm is deeply introduced and analyzed by simulation from both soft decision and hard decision in thesis.For the hard decision fusion algorithm, a cooperative sensing algorithm based on adaptive dual-threshold is proposed in thesis.This algorithm againsts the problem that local energy detection performance will decreased obviously in low signal-to-noise ratio.The advantage of this algorithm is the dual-threshold of energy detection can be adaptively adjusted with the changing signal-to-noise ratio in the wireless environment,making the system can obtain better performance in any signal-to-noise ratio. Compared to the algorithm based on the fixed threshold by the simulation in AWGN channel, the result proves the effectiveness of the proposed algorithm.For the soft decision fusion algorithm, in order to ease the problem that soft decision algorithm will consume a lot of bandwidth, cluster-based cooperative spectrum sensing model is first introduced in thesis. Then a weighted collaboration detection algorithm based on D_S evidence theory is proposed in thesis.On the basis of it, the proposed algorithm is applied in cluster-based cooperative spectrum sensing.So a new algorithm is proposed:the D_S evidence theory weighted cooperative spectrum based on clusters. In this algorithm each cluster can be assigned a different weighting factor by D_S evidence theory, in order to increase the contribution of clusters with better performance and decrease the contribution of clusters with worse performance in cooperative sensing. By simulating and comparing several cluster-based cooperative spectrum sensing algorithms, results show the algorithm that proposed in this thesis can achieve a better detection performance.