Research On The Spectrum Sensing In Cognitive Radio

Static spectrum allocation strategy is the main reason that leads to the shortage of spectrum resources.Cognitive radio technology can effectively relieve the stress of spectrum resources.Spectrum sensing is one of the key technologies of cognitive radio.Because of the characteristics of primary users' frequency offset and random phase in real communication system,the performance of the existed narrowband spectrum sensing algorithms using covariance and power spectrum density deteriorates;The characteristics of each channel through which signals pass vary greatly in wide spectrum bands,the performance of wideband spectrum sensing algorithm is bad;In order to deal with the above problems,this thesis will deeply study the narrowband and wideband spectrum sensing algorithms.Firstly,the research background and significance of cognitive radio are studied.Then the existed narrowband and wideband spectrum sensing technologies are analyzed and summarized.Secondly,spectrum sensing algorithm using covariance matrix is proposed.Since there is a correlation between the same primary user signals received by non-identical antennas at the receiving end,the local variance spectrum sensing algorithm has made full use of the second-order correlation of the received signal.However,when there are not sufficient receiving antennas,the detection performance of the algorithm is below the expectation.Resulting from the increase of the noise uncertainty,as well as,impacts of frequency offset?random phase,the algorithm's detection performance is decreased.To solve this problem,a spectrum sensing algorithm is proposed,whose detection statistic is the ratio of the sum of the squares of the diagonal elements' absolute values of the covariance matrix to their mean variance.The expressions of the detection threshold are deducted.The simulation results in Additive White Gaussian Noise(AWGN)channel and Rayleigh fading channel show that the performance of the proposed algorithm is better than that of the local variance method under the condition of frequency offset,limited number of antennas,limited number of sampling points,low signal-to-noise ratio and noise uncertainty.Then,a spectrum sensing algorithm using the ratio of average value of maximum and minimum power spectral is proposed.Because there are the unknown carrier frequency offset and the noise uncertainty,the Power Spectral Ratio Averaging(PSRA)algorithm's maximum and minimum values of signal's power spectral are unable to be evaluated accurately,and its performance of spectrum sensing is deteriorated.To solve this problem,a spectrum sensing algorithm using the ratio of average value of maximum and minimum power spectral is proposed.The expressions of the detection threshold are deducted.The simulation results in AWGN channel and Rayleigh fading channel show that the performance of the proposed algorithm is better than that of the Power spectral density Segment Cancellation(PSC)?PSRA algorithm,and the impacts of the unknown carrier frequency offset and the noise uncertainty on the algorithm are reduced.Finally,the pheromone-driven distributed cooperative broadband spectrum sensing algorithm is proposed.Because of the impacts of various factors such as channel fading on a single node,the spectrum sensing performance needs to be further improved.The filter bank is used to divide each node's broadband signal into multiple channels' narrowband signals.The energy detection method is used in each channel.The detection statistics and the detection results of every channel in all nodes are obtained.The concept of pheromone is introduced into the distributed collaborative spectrum sensing process.The methods of calculation and updating pheromone,as well as,the pheromone-based information fusion methods are proposed.The multiple-cycle fusion method is used to enhance each node's sensing performance.The simulation results show that the distributed cooperative broadband spectrum sensing algorithm is better than the sensing algorithm based on a single node.