Research On Spectrum Sensing And Sharing Algorithms In Cognitive Radio Networks

Spectrum resources are non-renewable resource, however, due to the utilization ofspectrum resources by authorization, many authorized spectrum hasn’t been fullyutilized. The technology of Cognitive Radio is an effective approach to solve theproblem of underutilization of valuable spectrum resources. Cognitive Radio is a kindof new technology based on software-defined radio. Cognitive Radio can perceivetemporally unutilized spectrum in the surrounding environment, namely white spaces,and effectively use them on the condition that the transmission of cognitive user doesnot interfere heavily with the primary user outside the band. Multiple cognitive radioswithin a given area form Cognitive Radio Network, which needs to ensure the qualityof services of communication between the secondary users, and to ensure the fairnessof the usage of white spaces within the network.This thesis mainly studies spectrum sensing and sharing technology in theCognitive Radio Networks, Enabling Cognitive Radio Network to untilize whitespaces effectively.In the study of spectrum sensing technology, this thesis makes the most of thecharacteristics of scatterring of secondary users. For the secondary users closer to themain users, they get larger sampled signal power, the use of energy detection canquickly detect the primary users; For weak signal, the thesis makes full use of theinformation contained in the sampled weak signal, and uses Max-Min eigenvaluebased sensing for secondary detection. That is the hybrid spectrum sensing technologybased on Signal Noise Ratio (SNR) proposed in this thesis. The thesis improved thealgorithm of the Max-Min eigenvalue based sensing to reduce its computationalcomplexity before its practical application. The simulation results have shown thereliability of the proposed detection algorithm.In the study of spectrum sharing technology, this thesis analyzes spectrum sharingby auction model. However, the traditional scenario of spectrum sharing pursuesmaximizing network throughput, and hardly concerns about the fairness when thewhite spaces are utilized among the cognitive spectrum users, and users may lie aboutthe valuation of the cognitive channel. In this thesis, the traditional method is improvedby preserving the effectiveness of the original method and making the spectrumsharing process obtain the feature of fairness and truthfulness.