Research On Change-point Detection In Spectrum Sensing

The short supply of spectrum resources and low utilization of spectrum which has beenauthorized has become a major contradiction that restricting the development in wirelesscommunication. A rapid demand for wireless communication services causing the frequencyband which is suitable for wireless communication has become very crowded. On the otherhand, the wireless spectrum resource which has authorized has a different degree of idle in thespace-time under the management of existing static spectrum. Cognitive radio can ease thecurrent situation of scarce spectrum resources by opportunistic access spectrum holes. So itbecomes one of the hottest wireless communication technologies. Cognitive radio technologyhas the ability to perceive the environment, adjust the communication mechanisms within thesystem adaptively by itself, so it can improve the spectrum efficiency effectively. However,the cognitive radio technology still faces challenges. Spectrum sensing is a key technologywhich is basic and important in cognitive radio technology. The cognitive radio systems cancreate a new communication connection by finding the available spectrum holes. So it is thefocus point and it has difficulty of the study.This paper proposed the spectrum sensing on the basis of change-point detection againstthe idea of detecting the sequence of fixed length by traditional energy detection algorithmand inadequate detection performance affected by the decision threshold. The main work andinnovations are as follows:1. Introduce the background, the concept features key technologies, system architecture,application prospects of cognitive radio technology.2.To sort the basic theory and performance evaluation parameters of signal detection, toanalyze and compare the spectrum sensing algorithm which is based on the transmitter, thereceiver detects and co-detection, to summarize the advantages and disadvantages of variousspectrum sensing technologies and application scenarios.3. Research on the spectrum sensing of the basis of change-point detection against theidea of detecting the sequence of fixed length by traditional energy detection algorithm andinadequate detection performance affected by the decision threshold. Simulation results showthat the algorithm is effective to detect the status of signal activity for authorized usercompared with the traditional energy detection. It needs less signal knowledge. The detectioncan t be affected by the limit of fixed points and the mistake of settings threshold.4. Study the change point detection in the multi-antenna spectrum sensing algorithm, wecan further get the process of change point detection under the generalized likelihood ratio test. Simulation results show that, the multi-antenna mutation detection algorithm has higherprobability of detection and lower probability of false alarm under the same test conditions. Itcan overcome the interference fading channel for single antenna effectively and improve theperformance of spectrum sensing for cognitive radio systems.