Research On Blind Spectrum Sensing Based On Compressed Sensing

Cognitive radio technology is an emerging technology in recent years.It aims to discover spectrum opportunities through spectrum sensing to achieve efficient reuse of spectrum resources.It is considered to be the most effective solution to solve the current shortage of radio spectrum resources.In spectrum sensing,the traditional sampling method is difficult to meet the requirements of hardware when facing high frequency and large bandwidth spectrum.At this time,compressed sensing theory can effectively solve this problem.However,most of the measurement and restoration methods in compressed sensing need to determine some parameters of the signal to be measured in order to accurately reconstruct the signal.When processing the completely unknown signal,the effect of these methods will become very bad,which is often encountered in the actual spectrum sensing.In this paper,we conduct research on this issue.Firstly,we evaluate and compare some existing compressed sensing methods,and find the technical characteristics of these methods are suitable for the application of blind spectrum estimation.Then we combine and improve these methods,and propose a new method which can complete the compression measurement and signal reconstruction efficiently and accurately in the case of blind spectrum sensing.The specific research content of this paper is as follows:1.Aiming at the problem that many measurement and restoration methods of compressed sensing signal can not be applied to blind spectrum sensing,this paper sorts out several typical measurement matrices and signal restoration algorithms in compressed sensing system,explains the construction characteristics of various matrices and the principle steps of restoration algorithm,and then carries out a series of simulation experiments to evaluate the performance of these matrices and restoration algorithms.2.To solve the problem that it is impossible to determine the appropriate sampling times of the spectrum signal in the case of blind spectrum sensing,based on the study of the characteristics of matrix construction and the principle of the restoration algorithm,this paper proposes an adaptive sampling spectrum sensing algorithm composed of Toeplitz matrix and BCS-RVM restoration algorithm,which adopts the strategy of incremental measurement,so that the algorithm can stop measuring when it obtains enough information to recover the signal successfully,and the effectiveness of the algorithm is verified by simulation experiments.3.In view of the over estimation problem of the fixed step size strategy adopted by the SAMP algorithm,this paper analyzes the principle of its self-adaptive to the sparsity in detail,and proposes a variable step size strategy optimization method based on iterative residual,which can make the algorithm have the advantages of fast iteration and accurate location of signal sparsity.Experimental results show that the improved algorithm can reconstruct the signal quickly and reduce the possibility of over estimation.4.In order to solve the problem of low efficiency of spectrum sensing algorithm based on adaptive sampling,this paper first introduces the improved samp algorithm with faster restore speed to replace bcs-rvm algorithm,and then analyzes the change trend of sparse degree of estimated signal when the amount of information contained in the measured signal is insufficient,and proposes a variable increment measurement strategy based on the sparse degree of estimated signal,so that the algorithm can reduce the number of signal recoveries in the case of blind spectrum sensing.Simulation results show that the optimized algorithm can greatly improve the speed and accuracy of perception.