Research On Wideband Spectrum Detection Algorithms Based On Compressed Sensing

Spectrum resources become increasingly scarce with the development of wireless communication services and the growing number of users.However,there are many spectrum holes in these frequency bands which are allocated.Finding and using these idle frequency bands to achieve communication,it will further improve the utilization ratio of spectrum resources.Compressed sensing makes a use of the sparsity of the signal,and it can be applied to the spectrum detection,such as maritime communication,where abundant idle frequency bands exist.This paper makes a research on wideband spectrum detection algorithms based on compressed sensing,mainly in the wideband spectrum detection scenarios where the carrier frequency and bandwidth of the occupied frequency bands are unknown.First of all,a wideband spectrum detection algorithm based on modulated wideband converter(MWC)is designed in this paper.Under the MWC compressed sensing framework,it proposes to utilize Compressive Sampling Matching Pursuit(CoSaMP)algorithm to improve the compressed sensing spectrum detection algorithm which uses Orthogonal Matching Pursuit(OMP).The simulation results demonstrate that such this algorithm can achieve higher percentage of correct support recovery than the algorithm based on MWC with OMP,especially in low Signal Noise Ratio(SNR)regime.In order to utilize maritime spectrum holes more effectively,a multi-vessels collaborative wideband spectrum detection algorithm based on compressed sensing is also designed in this paper.It uses Block Sparse Simultaneous Compressive Sampling Matching Pursuit(BSCoSaMP)algorithm as the improved joint signal reconstruction algorithm for collaborative spectrum detection,the simulation results demonstrate that such this algorithm can achieve higher percentage of correct support recovery than the collaborative spectrum detection algorithm with Block Sparse Simultaneous Orthogonal Matching Pursuit(BSOMP)algorithm,especially in low SNR regime.The algorithms designed in this paper can improve the accuracy of wideband spectrum detection in low SNR.It contributes to the detection of idle frequency bands and improve the utilization rate of spectrum resource.