Research On SCMA Low Complexity Detection Algorithms Based On Message Passing

In the case of increasingly limited frequency resources,traditional multiple access technologies cannot meet the business needs of massive access scenarios.In recent research,non-orthogonal multiple access technology has obvious advantages in improving spectral efficiency,expanding access,and reducing communication delay.The sparse code multiple access technology with good prospects in non-orthogonal multiple access technology can effectively improve the general efficiency and expand the access amount.In this thesis,the receiver of SCMA is studied,and two low-complexity optimization schemes are proposed for the problem of high complexity of SCMA receiver,and the performance analysis of the optimization scheme is given.The receiver scheme proposed in this thesis adds parallel computing to CMPA.The optimization of delete rule and the iterative update rule are optimized to maximize the minimum ring length of the subgraph and parallel calculation between the subgraphs.Compared with CMPA,the Parallel based Message Passing Algorithm(PMPA)has better BER performance and lower complexity.According to the idea of reducing the number of iterations by the Partial Marginalization Based Message Passing Algorithm(PM-MPA),this thesis proposes a receiver that reduces the number of iterations of the user's iterations by verifying that when a user's codeword confidence information is much larger than other codewords.Compared with Message Passing Algorithm(MPA),the Low-Iteration Message Passing Algorithm has close bit error rate performance and lower complexity.In addition,this thesis combines PMPA and Low-Iteration MPA to form a joint receiver that can reduce more complexity.Compared with MPA,the joint receiver has close bit error rate performance and lower complexity.