Cognitive Radio Cooperation Spectrum Perception Algorithm Research

With the growth of wireless communication applications, the spectrum becomesmore and more congested. Cognitive radio (CR) is considered as an effective way toslove this problem. Under the condition of not interfering the communition of theprimary user (PU), secondary user (SU) can use the unoccupied licensed spectrumbands to communication, thus to enhance the utilization of spectrum. Secondary userneeds to sense the spectrum to find the vacant bands, so spectrum sensing plays the keyrole in cognitive radio.In this paper, we mainly research the spectrum sensing technology of cognitiveradio. At first, we give the introduction of the local sensing technology of cognitiveradio and compare these technologies, summarize the disadvantages of local sensingtechnology. Then we focus on the cooperative spectrum sensing technology andintroduce some cooperative spectrum sensing algorithms. We propose two cooperativespectrum sensing algorithms form different perspective.To improve the detection performance and the reliability, we porpose a cooperativespectrum sensing algorithm based on improved energy detection and Dempster-Shafer(D-S) theory. In this algorithm, SUs uses the improved energy detection to make localsensing and get the trust functions, this make the local detection results more reliable,and improve the whole algorithm’s reliability. Simulation results show that the detectionperformance of this algorithm is better than other cooperative spectrum sensingalgorithms in the same conditions.To reduce the burden of the CR network and the fusion center (FC), we propose anovel cooperative spectrum sensing algorithm combining double threshold energydetection and D-S theory. By using the double threshold the burden can be reduced andsimulation results show that the proposed algorithm has higher detection probability atthe same condition. To reduce the flase alarm probability, we improve the proposedalgorithm by using the tug-of-war rule and the weighted parameter parameter.Simulation results show that the performace of the proposed algorithm is improved afterthe improvement work. The algorithm becomes more reliable, which means our improvement work is meaningful.