Research On Algorithm Of Cooperative Spectrum Sensing In Cognitive Radio

The shortage of spectrum resource has been a serious problem nowadays because of the fast development of smart electronic products and techniques of wireless communication. As the first step of cognitive radio, spectrum sensing has to discern unoccupied frequency band quickly and accurately, which means a hard mission for it. By analyzing the single node spectrum sensing, several problems about it has been discovered like hidden terminal so that centralized cooperative spectrum sensing is studied in this paper. According to the sensing time and accuracy of cooperative sensing, the performance of cooperative sensing is improved by double threshold energy detection, clustering and improved two-stage cooperative sensing algorithm. Finally, a union optimization algorithm based on cluster and improved two-step cooperative sensing is proposed to be helpful for application in cognitive radio system.In this paper, single node spectrum sensing is analyzed in different ways and relevant algorithms are researched by introducing procedure and theory of them. Then, energy sensing is also studied under different background of noise for determining the threshold of double threshold energy detecti on. On top of that, influence of thresholds and noise background is studied by computer experiments. Compared with energy detection, cyclostationary feature detection is robust to noise and processes a high detective accuracy despite of complicated algorit hm.Then, according to the limitation of two threshold energy cooperative detection and advantages of cyclostationary feature detection, an improved two-stage cooperative sensing algorithm is proposed. In this algorithm, definitions of coarse sensing and fine sensing are offered. In coarse sensing, double threshold energy sensing is used and in fine sensing, cyclostationary feature detection is used. By some experiments, it can be known that improved sensing algorithm could reduce cost of sensing and raise the efficiency.Finally, for hard decision cooperative spectrum sensing, by introducing ‘and’ criterion, ‘or’ criterion, ‘K-rank’ criterion and cluster, the performance of detection, accuracy and detective time are all improved totally. These algorithms are combined together and made to be a new union cooperative spectrum sensing algorithm. The experiments showed that the results of improved algorithm on detection probability and detection time are very effective.