Research On Wideband Compressive Spectrum Sensing Algorithm Based On Sparse Reconstruction And Dynamic Analysis

Wideband spectrum sensing technology aims to detect the main user signal through the direct observation of wideband spectrum,then to achieve the efficient reuse of spectrum resources.Compressive sensing theory provides a theoretical basis for low rate sampling of wideband spectrum sensing.However,in the actual cognitive radio system,there is often no information interaction between cognitive users and primary users,the traditional compressive sensing algorithm can't obtain the prior knowledge of signal sparsity,which affects the signal reconstruction performance of wideband compressive spectrum sensing algorithm.On the other hand,the occupation of authorized frequency band by primary users often changes with time,which makes the sparse structure of wideband spectrum have the characteristics of dynamic change,which also hinders wideband compressive spectrum sensing.Aiming at the above problems existing in the field of wideband compressive spectrum sensing,based on the current research status of wideband compressive spectrum sensing technology.Firstly,this paper analyzes the performance of several typical compressive sensing algorithms and measurement matrices.Secondly,according to the analysis results,the traditional sparsity adaptive matching pursuit algorithm is regularized and improved.Finally,according to the time-varying characteristics of dynamic wideband spectrum sensing,the measurement and linear prediction model of dynamic wideband spectrum sensing system is established by combining Kalman filter and compressive sensing,to realize the dynamic wideband compressive spectrum sensing based on signal correlation.The specific research contents of this paper are as follows:(1)This paper analyzes the influence of signal sparsity and algorithm iteration step on the signal reconstruction performance of traditional compressive sensing algorithm.Aiming at the problems existing in the above wideband spectrum sensing field,and the performance analysis of typical measurement matrix and compressive sensing algorithm,an adaptive matching pursuit(RBVS-SAMP)algorithm with variable step size and sparsity based on weighted regularization is proposed,the weighted regularization idea is introduced into the SAMP algorithm,the iteration step of the algorithm is adjusted by the set residual threshold,so as to improve the accuracy and speed of signal reconstruction.Simulation results show that under the same parameter conditions,The RBVS-SAMP algorithm proposed in this paper has better accuracy and real-time performance,it is superior to other similar algorithms in terms of signal reconstruction success probability,signal reconstruction error and algorithm running time.(2)According to the characteristics of time correlation and dynamic change of wideband spectrum,a dynamic wideband spectrum measurement system model based on sliding window is established.Considering the correlation of time signals before and after different sliding windows and the sparsity of dynamic wideband spectrum signals,a linear prediction model of dynamic wideband spectrum signals based on ARMA model is established.So as to complete the state simulation and prediction of dynamic wideband spectrum.Through simulation and comparative experiments,it is verified that the proposed model has better prediction performance than other similar models under the same experimental conditions.(3)Aiming at the problem that the traditional compressive sensing algorithm has poor noise resistance in the process of processing wideband time-varying sparse signals,a linear prediction wideband compressive spectrum sensing algorithm based on Adaptive Kalman filter is proposed.In the signal reconstruction stage,the support set is continuously improved by greedy algorithm and Kalman filter iteration until the exact solution is obtained.In the simulation experiment,compared with other similar algorithms,under the same signal-to-noise ratio conditions,the linear prediction wideband compressive spectrum sensing algorithm based on Adaptive Kalman filter proposed in this paper significantly reduces the average relative error,it improves the signal detection probability and signal error ratio,and has better advantages in anti noise and dynamic performance of the algorithm.