Research On Spectrum Sensing Of Cognitive Radios

With the rapid development of wireless applications, the demand for spectrum is on the increase. However, it is found out that the fixed spectrum allocation makes most of the spectrum underutilized. As a revolutionary spectrum sharing technology, cognitive radio can significantly improve the spectrum utilization so as to solve the problem of current spectrum resource scarcity. Spectrum sensing, one of the key techniques in cognitive radios, can reliably detect the spectrum holes and avoid interference to the licensed users. The thesis concentrates on spectrum sensing technologies in cognitive radios.This thesis firstly introduces the basic concepts and key technologies of cognitive radio, and then studies the detection principles, usage situations, advantages and disadvantages of some typical spectrum sensing algorithms. Subsequently the limitations of the local spectrum sensing algorithms are analyzed.Then, cooperative spectrum sensing mechanism is studied in detail and the fusion rules are summed up. Quantitative analyses of the reliability and efficiency of spectrum sensing are made particularly. On this basis, how to reduce out-band sensing time while providing the primary user with its desired level of interference protection is proposed in this thesis for the case of cooperative spectrum sensing. By investigating the intrinsic relationship between local processing and cooperation, our research makes a tradeoff between the two and formulates a nonlinear constrained optimization problem to improve the uncooperative scheme. Simulation shows that the"majority"rule performs best.In order to allow secondary users to access the unused spectrum based on spectrum sensing results, a preliminary exploration on cross-layer optimization is conducted by proposing a novel joint design approach of spectrum sensing and access in this thesis. Consider the interaction between the physical layer and MAC layer, and coordinate the characteristic parameters of the two layers may optimize the overall system performance. This approach selects the optimal sensing duration, packet length and probability of random access to maximize the secondary network throughput under constrains of collision probability. Compared with the layered approach, Simulation results indicate that the joint PHY-MAC design obtains high network throughput gain.