Study On Spectrum Sensing Technology Based On Cognitive Radio

As the rapid growth of the wireless communication, the conflict between the increasing demands and the limited spectrum resources has become one of the primary bottlenecks that constrains the development of wireless communication. As a smart spectrum sharing technology, cognitive radio can sense the wireless communication environment with the support of artificial intelligence; changing the system parameters automatically according to certain learning and decision making algorithms; testing idle spectrums dynamically and making best use of them. Theoretically speaking frequency division multiplexing and sharing are allowed in time, space and frequency domain. Spectrum sensing is one of the basis of constructing a cognitive radio system. The statistical information about wireless environment can be collected by spectrum sensing. It can provide both assistance for higher level spectrum management and necessary data to the power control of the system. Research about cognitive radio spectrum sensing technology is carried out in this dissertation with the main contents:(1) Principles, characteristics and performance in cognitive radio spectrum sensing technology are summarized fully in this dissertation. And the shortcomings of energy detecting method in sensing authorized users in both ideal and fading conditions are analyzed according to the type of unknown signal. Then an improved algorithm combined multiple users collaborative sensing and diversity reception technology into energy detection is proposed. The advantages of easy to implement and needless of prior information are retained while the reliability of detecting is improved effectively. In this dissertation performance simulation and analysis are done in both additive white Gaussian noise channel and typical multi-path fading channel.(2) With respect to the linear frequency modulated signal and phase modulated signal model, the performance of using matched filter to sense the signal of authorized users is analyzed. As the Gauss noise's higher order cumulant permanent is zero, a new method combined higher order cumulant and matched filter to eliminate Gauss noise and improve sensing ability is proposed. Performance simulation and analysis are done according to the proposed algorithm.(3) After recognizing and understanding cognitive radio-based wireless area network interface standard --IEEE802.22, which is applicable for the TV band, a method using the matched filter to sense spectrum by detecting local oscillator leakage of receivers is described.(4) The transmission scene analysis in cognitive radio is constituted by spectrum sensing, spectrum analysis and spectrum decision. In order to analyze and estimate the spectrum, performance simulations aiming at the Welch's method and Multitaper spectral estimation method are carried out. Finally the impact of important parameters in the two methods on estimation ability was analyzed.