Wideband Spectrum Sensing Algorithms Based On Compressed Sensing

In order to meet the increasing demands in wireless communication services,more and wider spectrum bandwidth needs to be used to transmit information.In view of the problem between the shortage of available spectrum and the low utilization rate of authorization spectrum,cognitive radio technology can effectively improve the usage of spectrum and achieve the efficient spectrum utilization rate.However,as the key technology of cognitive radio in the wireless wideband network,spectrum sensing based on the nyquist sampling approach not only leads to a severe challenge of high-speed analog-to-digital converter about hardware implementation,but also brings high communication overhead to the sensing process.Therefore,how to effectively realize the wideband spectrum sensing with low sampling rate has become a hotspot in cognitive radio field.As a new sampling method,compressed sensing can realize the low speed sampling of the signal and reconstruct high dimensional signal from low dimensional sampling data.Unlike the nyqusit sampling methods,the sparse or compressible signal can be sampled and compressed simultaneously according to compressed sensing.Compressed sensing can reduce the high sampling rate and avoid producing a large number of redundant data.Spectrum sensing based on compressed sensing in cognitive radio can reduce both sampling pressures and communication costs in sensing process.For this reason,this paper mainly studies the wideband spectrum sensing algorithm based on sub-band matching selection and the wideband distributed cooperative compressed spectrum sensing algorithm based on weighted consensus optimization.The main innovation points are as follows:1.For spectrum reconstruction algorithm design in wideband spectrum sensing,an improved reconstruction algorithm based on sub-band matching selection is proposed.With a single frequency point matching selection transforming into a sub-band matching selection,this algorithm can not only determine the existence of primary users in the sub-band,but also greatly reduce the number of iterations in spectrum reconstruction process,Furthermore,the regularized least square method is used to improve the accuracy of the reconstruction in theresidual signal update process.The proposed method can not only improve the accuracy of spectrum reconstruction,but also effectively shorten spectrum sensing time and achieve better wideband spectrum sensing performance.2.For cooperative spectrum reconstruction algorithm design in wideband spectrum sensing,an improved wideband distributed cooperative spectrum reconstruction algorithm based on weighted consensus optimization is proposed.In this algorithm,the next iterative reconstruction weights are determined according to the current iterative reconstructed spectrum signal,which can encourage the sub-band occupied by primary user to generate signal value and decrease the likelihood of incorrect reconstruction.The proposed algorithm can not only increases the spectrum reconstruction accuracy,but also reduces time and communication costs of the sensing process,and improves the spectrum sensing performance.