Research On Spectrum Sensing Using Asymptotic Spectrum Theory Of Random Matrix Theory

With the rapid development of wireless communication technology,radio spectrum resources and its use efficiency has become increasingly important for the development of wireless mobile communication applications and services.Random matrix theory, as an effective tool for large-dimensional data processing, has been used in fields as nuclear physics, digital communications, finance and so on. In the field of wireless communications, random matrix theory has been successfully applied to the capacity analysis、 performance analysis and receiver design.Therefore, the combination of the two aspects is very important for the technologies of Cognitive Radio. This paper mainly studies eigenvalue-based spectrum sensing algorithms for Cognitive Radio, this kind of methods overcome the noise uncertainty problem and can be used for various signal detection applications without knowledge of the signal, the channel and noise power. Firstly, eigenvalue-based spectrum sensing algorithm is analyzed and acquires the new decision threshold under the situation of noise-related, and the simulation results show that the new decision threshold is much better than M-P approach.Secondly, MME algorithm’s energy efficiency is analyzed in the case of noise uncertainty, and optimizes the energy efficiency through the optimization of sensing time.Finally, taking into account of the possibility of collision during spectrum sensing, this paper studies eigenvalue-based spectrum sensing algorithm in this situation by mathematical analysis and simulative verification. At last, the optimal energy efficiency is achieved through the joint optimization of the frame period and sensing time.