Research On Non-Orthogonal Multiple Access Detection Technology Assisted By Compressed Sensing

As one of the promising techniques in the fifth generation of communication system(5G),non-orthogonal multiple access(NOMA)can meet the demands of massive connectivity and high throyghput for 5G.In order to solve the multi-user detection problem in NOMA,the schemes with compressive sensing(CS)theory has been proposed.The feature of the proposed scheme is to fully explore the potential sparsity of the system models,and then CS approach can be exploited to recover the sparse signal.As one of the code-domain NOMA technique,in sparse code multiple access(SCMA),low-density spreading and multi-dimensional modulation techniques are merged together to directly map incoming data streams to multidimensional complex codewords selected from a predefined SCMA codebook.In SCMA multi-user detection(MUD),message passing algorithm(MPA)is used to recover signals,which can approximate the optimal performance of the maximum likelihood methed.Despite of exploring the sparsity of SCMA codes,the computational complexity of MPA is still prohibitively high.A novel low-complexity multi-user detection has been proposed for SCMA systems,combined MPA and CS technique,referred to as Compressive Sending assisted MPA detector.Specifically,the proposed CS-assisted MPA detector is divided into two-procedure,initial detection and sparse error correction respectively.In the first step,MPA with a few iterations is employed as initial detection.Furthermore,we observe that the symbol error produced by initial detection tends to be sparse.In the second step,based on the observation,threshold-aided adaptive subspace pursuit is proposed to estimate the error vector.Finally,the final estimated vector can be obtained by adding the initial estimated vector produced by MPA to the estimated error vector.Simulation results demonstrate that the proposed CS-assisted MPA detector can reduce the computational complexity substantially with similar SER performance compared to the original one.In massive machine type communications(mMTC)scenario,grant-free NOMA technique can remarkably reduce signaling overhead,data transmission latency and terminal power consumption.For the hybrid sparsity transmission model in grant-free NOMA system,we propose a multi-user signal detection scheme based on Distributed Compressive Sensing.Firstly,the block sparse signal recovery is exploited to estimate the common active user set in hybrid sparsity transmission model.Then,the proposed algorithm utilizes the estimated common set as prior information and detects the single set in each time slot individually based on belief support.Simulation results show that the proposed dectection scheme can gain good SER performance.Furthermore,the proposed dectection scheme that does not require the exact number of active users as prior knowledge is highly practical in real communication scenario.