Research On Blind Detection Algorithm For Primary User Signal

In 21 st century,more and more applications need to use spectrum to transmit information in the fields of technology,life,military and others.On the other side,large portions of certain licensed frequency bands are found unused most of the time.Spectrum sharing mechanism,such as Cognitive Radio,can solve the conflict between limited spectrum resources and rising demands.In Cognitive Radio system,unlicensed users,called secondary users,may utilize spectrum while the licensed users,called primary users,is not occupying it.Therefore accurate detection of primary user signal is the key technology of Cognitive Radio.The spectrum sensing approaches can be classified as blind and knowledge aided approaches.It becomes important that the blind detection without prior knowledge of signals and channel.The paper has four chapters.The first chapter expounds the background and significance of Cognitive Radio research work,the definition of Cognitive Radio,the connotation of Dynamic Spectrum Access and three transmission models of DSA.The summary of this chapter is given about blind signal detection algorithms' research status,this paper's researching significance and author's researching works.In the second chapter,several classical signal detection algorithms are introduced,such as Matching Filter Detection,Cyclic Stationary Detection,Energy Detection,and some higher-level blind detection algorithms.Moreover,it is analyzed that these detection algorithms' advantages and limitations.In the third chapter,a new blind detection algorithm based on the eigenvalues of sampling covariance matrix is proposed.The ratio of the difference and sum of the maximum eigenvalue to the minimum eigenvalue of the sampling covariance matrix is used as the decision statistic.Based on the latest results of the large-dimensional random matrix theory,the distribution of maximum eigenvalues and minimum eigenvalues of the sampled covariance matrices,an effective method for calculating decision threshold is proposed.Compared with the classical eigenvalue detection algorithm,the new algorithm has the advantage of accurate calculation of the decision threshold,and effectively improves the detection performance and the reliability of decision results.The feasibility and superiority of the new algorithm are verified by Monte Carlo simulation experiments.In the last chapter,another new blind detection algorithm is proposed that based on the sample covariance matrix trace of the received signal.The algorithm utilizes the difference of the statistical covariance matrix whether or not primary user signal.The decision statistic is the sample covariance matrix trace of the received signal.Furthermore,a blind sensing algorithm based on noise variance estimation is proposed for white noise sensing scenarios.The new blind spectrum sensing scheme has the advantages of no prior information of the channel and the primary user signal,and is easy to calculate.The feasibility and superiority of the new algorithm are also verified by Monte Carlo simulation experiments.