Research On Key Techniques Of Broadband Digital Receiver Based On Compressed Sensing

Broadband digital reconnaissance receivers are widely used in the field of communication and radar.With the continuous development of digital signal processing technology,the requirements for receivers' performance are also increasing.At present,a broadband digital reconnaissance receiver needs to meet the requirements of large instantaneous bandwidth,high sensitivity,high resolution and larger instantaneous dynamic range in order to perform high-speed sampling,data storage and real-time processing of signals in wide frequency band.Therefore,the acquisition of large-bandwidth signals and high-speed AD conversion rate are technical bottlenecks that the current broadband digital reconnaissance receivers need to break through.In this dissertation,the outstanding advantages of compressed sensing in signal acquisition and reconstruction are integrated into a broadband digital reconnaissance receiver.In order to solve the receivers' two technical bottlenecks and achieve the compression,acquisition and sample-rate conversion of large bandwidth analog signals,two signal extraction algorithms based on compressed sensing are proposed.The main research contents are as follows:Firstly,this paper studies the basic structure of analog-to-information converter(AIC)and its typical reconstruction algorithm in compressed sensing theory.The atomic structure and the location of reconstructed signal in frequency band have been analyzed in-depth.Based on this,a Partial Orthogonal Matching Pursuit(POMP)algorithm with single frequency signal extraction as its main purpose is proposed,and the applicable conditions of the algorithm are given.The experimental results show that the algorithm can collect and sample-rate conversion of all single-frequency signals in broadband under the condition that the number of signals is unknown.Then,the idea of POMP algorithm based on AIC structure is applied to the reconstruction of multi-band analog signals-Modulation Wideband Converter(MWC)structure.A new multi-band signal extraction algorithm based on wideband digital reconnaissance receiver is proposed,and the applicable conditions of the algorithm are given.The experimental results show that the algorithm can collect,sample-rate convert and down-convert large bandwidth analog signals under the condition that the number of signals is unknown.Based on this algorithm,a “full-band compression+partial reconstruction” signal extraction scheme based on MWC structure is proposed.Last but not least,the "full band compression + partial reconstruction" scheme for both algorithms are compared and analyzed with the "full band sampling + channelization" structure and the "full band compression + full band reconstruction + channelization" structure of the conventional receiver.Experimental results show that the reconstruction scheme proposed by this paper can not only solve the two major technical bottlenecks of traditional receivers,but also be more effective than the other two structures.At the same time,reliability is better than the "full band compression + full band reconstruction + channelization" structure.Therefore,the partial reconstruction signal extraction scheme based on two new algorithms have a good engineering application prospect in the field of wideband digital receivers.