Research On Cooperative Spectrum Sensing Algorithm Based On Data Fusion

Now the wireless communication technology is widely used in mobile, TVbroadcast, public security and other fields. With the increasing demand forwireless services, the wireless spectrum as a kind of important resource is moreand more difficult to supply. However, the investigation and study shows that theaverage utilization rate of current licensed spectrum resources is very low andextremely imbalance. The main reason is the fixed spectrum allocation model. Inorder to solve the problem, cognitive radio system is one of the key technologiesof wireless communication in future. Through the detection of the surroundingenvironment, cognitive radio can find the spectrum holes and allocate them in adynamic way. This method can effectively improve spectrum efficiency. The keyof realizing cognitive radio is spectrum sensing technology.This paper first introduces the related concepts of cognitive radio andspectrum sensing technology, also their current development at home and abroad.At the same time, spectrum sensing technology is the research focus of this paper.Secondly, the single node spectrum sensing under different ways is analyzed andstudied in this paper as the research foundation for cooperative spectrum sensing.Since the selection of single node algorithm directly affects the implementationcomplexity and detection performance of cooperative spectrum sensing, thispaper analyzes single node algorithm by contrast. Energy detection is thesimplest method to realize and the optimal detection algorithm when the testsingle is unknown. Then it mainly studies the principle of energy detectionalgorithm and gives simulation and analysis of detection performance in AWGNchannel and Rayleigh fading channel. The study of weighted criteria found thatits weighted coefficient distribution is not reasonable so that the paper optimizesthe weighted coefficient. In the actual environment, each cognitive user’s locationand the received signal-to-noise ratio is different and they will affect the final results. This paper classifies all cognitive users based on signal-to-noise ratio andallocates different date fusion criterions for them. The simulation shows that thismethod can effectively improve the detection performance of spectrum sensing.The analysis of MATLAB simulation shows that it’s not necessary to haveall cognitive users’ participation to get the best spectrum detection performance.This paper chooses cognitive users who have a better signal to noise ratio toparticipate the cooperation and allocates different data fusion norm base ondifferent situation. This method can reduce the bandwidth requirement in acertain extent. The ROC curves show that the detection performance of theoptimized algorithm is increased by2%-5%and20%-30%compared to EGC andtraditional AND fusion criterion respectively.