Sparse Feature Extraction Method Of Wireless Channel Based On Compressed Sensing

Under the rapid development of information technology,wireless communication plays an increasingly important role in data transmission.Wireless channel feature extraction is the premise of wireless channel transmission characteristic estimation and detection,and it is the key to solving the multi-dimensional complexity and time-varying of the channel.With the increasing transmission bandwidth and the shortage of channel resources,the transmission characteristics of wireless channels are greatly suppressed.Therefore,it is of great significance to extract the sparse characteristics of wireless channels.This paper first introduces the transmission characteristics of wireless channels and traditional channel feature extraction methods.The shortcoming of the traditional method is that the sparse characteristics of the channel are not fully utilized,resulting in low spectral efficiency and low algorithm accuracy.Compressed sensing theory provides a new solution to this problem.Combining the essential characteristics of wireless channel and the theoretical basis of compressed sensing,this paper finds a new idea to improve the accuracy of the algorithm by selecting compressed sensing measurement matrix,and determines the information entropy theory as the preferred criterion.The algorithm pre-judging mechanism proposes a wireless channel sparse feature extraction method based on the compressed sensing measurement matrix,which aims to improve the accuracy of the algorithm and alleviate the spectrum pressure.Finally,the performance of the proposed algorithm is comprehensively investigated by using simulation software.Sparse feature extraction experiments show that the algorithm can accurately extract sparse features and meet the requirements of small errors and real-time.The conventional performance and statistical characteristics test show that the information entropy criterion has good reliability under the condition of channel parameter variation and algorithm parameter variation,and can adaptively select the optimal measurement matrix to complete the accurate extraction of sparse features.This proves that the algorithm in this paper is effective in real time.