Research On Spectrum Detection Algorithm Of OFDM In Dynamic Spectrum Access

With the rapid growth of wireless communications, there are larger demandsfor spectrum, and at the same time we have clearly recognized that the scarcity ofspectrum is caused by its low utilization. As an efficient and flexible technologyto improve spectrum utilization, dynamic spectrum access has been widelyappreciated, and pointed a new direction for the future development of wirelesscommunications. Dynamic spectrum access technology detects frequency bandswhether used by primary users or not, and then dynamically accesses them.However, spectrum detection is the first step towards dynamic spectrum access,and has a vital role significance to research it.This thesis conducted in-depth study on single-user spectrum sensing andmultiple-users spectrum sensing on Orthogonal Frequency Division Multiplexing(OFDM). Firstly, the cyclostationary feature detection algorithm on OFDMsignal was discussed. By deriving the spectral correlation function of OFDM,this thesis detected the known OFDM signal using the spectral peaks on thespecific cycle frequency section of its cyclic spectrum. Based on comprehensiveconsideration of efficient and reliable detection, cyclic prefix (CP) correlationdetection algorithm on OFDM signal is also discussed. The CP correlationdetection algorithm detects OFDM signal in time domain using its time-domainfeatures. Secondly, the collaborative detection algorithm was discussed for thepoor detection performance of single-user detection at low signal-noise-radio orunder problems of shadow fading, hidden terminals, etc. The local detectionalgorithm in discussed collaborative detections is the CP correlation detectionalgorithm. Against the shortages of traditional hard decision collaborationalgorithm when low signal-noise-ratio local users participated in, selectiondiversity collaboration algorithm and set diversity collaboration algorithm werederived by improving the “or” algorithm. Finally, on the base of clusteringnetwork structure, combining the selection diversity collaboration algorithm, acluster-based selection diversity collaboration algorithm was established, toreduce the number of result transmissions to the base station and improve thedetection performance of the traditional collaboration algorithm.