Research On Cooperative Spectrum Detection Technology

In the era of Internet of things,the access of massive terminals has made the shortage of spectrum resources increasingly serious.Cognitive radio(CR),as an intelligent technology with the opportunity to access the idle spectrum,provides an effective way to solve the shortage of spectrum resources,and spectrum detection is an important basis for CR implementation.In spectrum detection,due to the complex electromagnetic environment,multipath fading and shadow effect are very easy to exist.Energy detection has the advantages of not requiring signal prior information and simple implementation.Therefore,this paper mainly studies the cooperative spectrum detection algorithm based on energy detection.Aiming at the problems of poor detection performance and high transmission overhead of cooperative spectrum detection algorithm in low SNR environment and uncertain noise,this paper introduces the idea of stochastic resonance into energy detection algorithm.Then,by using the characteristic that the channel state has a certain correlation in time and the advantage of fuzzy logic in solving the uncertainty problem,to meet the requirements of spectrum detection for detection performance and transmission overhead.The main work is as follows:Firstly,this paper studies the principle of energy detection algorithm,and then discusses the problems of energy detection algorithm.Aiming at the problem of poor detection performance of energy detection algorithm in low signal-to-noise ratio environment,this paper introduces the idea of stochastic resonance,which can improve the signal-to-noise ratio of the system output signal to improve the detection performance in low signal-to-noise ratio environment.The simulation results also verify the conclusion that stochastic resonance can improve the detection performance of energy detection algorithm.Secondly,in order to solve the problem of energy detection algorithm which is difficult to judge between the double thresholds when the noise has uncertainty,a cooperative spectrum detection algorithm based on historical perception information prediction is proposed.The algorithm uses the correlation of channel state in time between the dual thresholds to make decision by exponential smoothing prediction,which solves the problem of local detection failure.The fusion center adopts the "or" criterion to judge the local hard judgment results.Simulation results show that under the same transmission overhead condition,compared with the other two simultaneous interpreting algorithms using the or fusion criterion,the algorithm can effectively reduce the impact of noise uncertainty on detection performance without increasing the computational complexity significantly.Thirdly,in order to balance the performance of detection and transmission cost,a collaborative sensing algorithm based on fuzzy logic is proposed.The algorithm uses discrete fuzzy quantization to the region between the local double thresholds,and then uploads local quantization information through 2-bit data to reduce the transmission cost of the system.In order to reduce the computation of fusion algorithm as much as possible,a three-step collaborative fusion decision method combining fuzzy logic decision and historical perception information prediction is adopted to judge the local detection results.The simulation results show that the proposed fusion algorithm can achieve the tradeoff between detection performance and transmission cost without greatly increasing the computational complexity.