Research On Key Technologies Of Sparse Code Multiple Access

With the development of wireless communication technology,it is difficult to meet the needs of high rate,low delay and massive connection because of the spectrum resource is limited.Driving by the impetus of people's growing communication demand,there is an urgent need to propose a new technology to improve spectrum efficiency.By transmitting user data overlaying on non-orthogonal resources with the help of sparse performance of codebook,Sparse Code Multiple Access(SCMA)technology can greatly improve the spectral efficiency.SCMA,as a potential non-orthogonal multiple access technology,has attracted extensive attention in academia and industry,in this thesis,we mainly focus on the study of SCMA codebook and the construction of large-scale irregular SCMA factor graph matrix.Firstly,with the rules of minimizing the upper bound of error probability,we jointly optimize the parameters of mother codebook and its rotation angles,in this scheme,bare bones particle swarm optimization(BBPSO)algorithm was used to search the optimal codebook parameters and their corresponding rotation angles quickly.Then,in the condition of different channel models and overload rate,we take bit error rate(BER)analysis with our proposed SCMA scheme and the other SCMA scheme,the results show that the proposed SCMA code has high validity and robustness,and compared with previous SCMA codebook can obtain obvious gain in terms of BER performance.Secondly,large-scale irregular SCMA factor graph matrix is constructed,and the degree distribution is optimized.Inspired by the fact that SCMA factor graph matrix and LDPC basis matrix have the same sparse property.we utilize progressive edge-growth(PEG)algorithm to optimize the degree distribution of large-scale irregular factor graph matrix.In addition,the extrinsic information transfer diagram demonstrates that the proposed structure has advantages in multi-user iterative detection,and the simulation results also indicate that the construction factor diagram matrix in this thesis has a good performance in reducing block error rate.