Research Of Spectrum Sensing On Cognitive Radio

With increasing demand for high-speed data transmission in wireless communications, new chanlleges, such as effectively use of the spectrum resource, generated. At present, the spectrum resource has become ever more scarce, because of the allocation policy is not reasonable and the communications users increase. Cognitive Radio (CR) technique is considered as one of the effective solutions of this problem. To improve utilization of the spectrum, CR employ spectrum sensing to detect the spectrum holes that it is the unused spectrum of a licensed primary user then opportunistic use it. Spectrum sensing is the one of the key technologies in CR networks, so it is the premise and basis of the CR technology.This paper research focuses on CR spectrun sensing technology. Paper first introduces the research background, signigicance and current status, and the characteristics of CR technology and key technology. Then introducing classification of the spectrum sensing algorithm, which the algorithms principle, application congitions,advantages and disadvantages. Finally focuses on research on the cooperative spectrum sensing algorithm based on interference temperature mechanism.Based on the interference temperature mechanism, this paper research on the local multitaper singular value decomposition (Local MTM_SVD) algorithm, then popose a improve spectrum sensing algorithm. The improve algorithm employ the multiple signal classification (MUSIC) algorithm to estimate the directions of arrival (DOA) of the interference source, and can be computed the coordinate position and distance, after can be decided the coefficient in the spatiotemporal complex matrix constructed by MTM_SVD. In the original algorithm, the power spectrum density was estimated by the nearly optimal multitaper method (MTM). But use the singular value decomposition (SVD) with the large amount of calculation at each interest the frequency bin. So the computational complexity of the original algorithm is very large, it is cannot meet the requirement for the real-time. The improve algorithm proposed the matrix F norm with small computational complexity instead the singular value decomposition. Finally comparative and analysis the performance of the improve algorithm. Theoretical analysis and simulation results show that the improve algorithm not only can accurately locate to licensed primary user, greatly improve the algorithm of the real-time, but also can provide the good detection performance.