Research On Non-orthogonal Multiple Access Based On Mixed Multi-dimension Constellations

High spectral efficiency is one of the main requirements in the future 5G system.To meet this requirement,this paper proposes a mixed sparse code multiple access(MSCMA),which is a non-orthogonal multiple access based on mixed multi-dimension constellations.First,this paper designs the MSCMA codebook.To improve the quality of initial information of message passing algorithm(MPA)receiver,the proposed MSCMA codebook design increases the sum of Euclidean distances between interfering nonzero elements of transmitted codewords multiplexed on each resource node.In addition,to improve the convergence reliability of the detected user in each decision process,the proposed MSCMA codebook design increases the diversity over the set of the sums of Euclidean distances between interfering nonzero elements of transmitted codewords multiplexed on all resource nodes.To further improve the spectral efficiency,this paper proposes a downlink multiple input multiple output mixed sparse code multiple access(MIMO-MSCMA)system.Finally,this paper designs two low complexity receivers of downlink MIMO-MSCMA system.In summary,the research achievements of this paper are as below.1.Increasing the minimum intra-partition distance of basic complex multi-dimension constellation can increase the sum of Euclidean distances between interfering nonzero elements of transmitted codewords multiplexed on each resource node.To increase the minimum intra-partition distance,this paper proposes a basic complex multi-dimension constellation design based on lattice code.But the increase of the minimum intra-partition distance is limited.To further increase the minimum intra-partition distance,this paper proposes a threshold basic complex multi-dimension constellation design based on lattice code.The proposed design can change the minimum intra-partition distance based on the performance of system.2.To help the criterion of permutation increase the sum of Euclidean distances between interfering nonzero elements of transmitted codewords multiplexed on each resource node and the diversity over the set of the sums of Euclidean distances between interfering nonzero elements of transmitted codewords multiplexed on all resource nodes,this paper proposes a phase optimization algorithm based on distances between different dimensions.The proposed algorithm can enlarge the range of the sum of Euclidean distances between interfering nonzero elements of transmitted codewords multiplexed on each resource node.However,the proposed algorithm would decreases the distances between some dimensions of any two constellation points.Therefore,this paper proposes an advanced phase optimization algorithm based on distances between different dimensions.3.At the MPA receiver,increasing the differences between the reliability of detection on transmitted users can improve the convergence reliability of the detected user in each decision process.This paper designs the mapping matrix of MSCMA.Benefiting from the proposed design of mapping matrix,the number of nonzero elements in each transmitted codeword is not the same,which can increase the differences between the reliability of detection on transmitted users.4.To improve the quality of initial information and the convergence reliability of the detected user in each decision process of MPA receiver,this paper proposes a criterion of permutation based on the sum of Euclidean distances between dimensions.The propsed criterion of permutation increases the sum of Euclidean distances between interfering nonzero elements of transmitted codewords multiplexed on each resource node and the diversity over the set of the sums of Euclidean distances between interfering nonzero elements of transmitted codewords multiplexed on all resource nodes.To further improve the convergencereliability of some detected users in their corresponding decision processes,this paper proposes an advanced criterion of permutation based on the sum of Euclidean distances between dimensions.5.At the MPA receiver,for each user,to increase the differences between the reliability of detection on the transmitted codewords atdifferent antennas,this paper proposes a downlink MIMO-MSCMA system.For each user of the proposed downlink MIMO-MSCMA system,the transmitted codewords at different antennas come from different codebooks of MSCMA respectively.6.To decrease the complexity of min-sum MPA receiver,with the non-uniform characteristic of MSCMA,this paper proposes a low complexity min-sum MPA receiver for downlink MIMO-MSCMA.The proposed receiver decreases the number of undetected users in the decision process.To improve the performance of Turbo iterative decoder and min-sum MPA receiver,this paper proposes the combined receiver of Turbo iterative algorithm and min-sum algorithm.Finally,the performance of the proposed MSCMA codebook design,the proposed system model of downlink MIMO-MSCMA,and the proposed two low complexity receivers of downlink MIMO-MSCMA will be verified by simulations.