Research On High Dimensional SCMA Codebook Design And Low Complexity Detecting Algorithm

Sparse Code Multiple Access(SCMA)is one of the candidates of multiple access technologies for the 5th generation communication system.SCMA is a multi-dimensional codebook based non-orthogonal spreading technique.In SCMA system,the data symbols are spread over multiple channel resources.However,the number of multiplexing layers is normally larger than its spreading factor.Hence,the SCMA system can accommodate more users than other multiple access technologies,in another word,while requiring the same number of channel resources,system overload rate can be greatly increased.On the basis of the Low Density Signature-Multiple Access(LDS-MA)technique,the SCMA combine the modulator and the spreader,which directly maps the information bit stream to a complex codeword.Compared with the QAM modulation used in LDS,the multi-dimensional modulation technique employed in SCMA is capable of obtaining a higher constellation coding and shaping gain.Thus,the design of SCMA multi-dimensional codebook is one of its hot topics.However,current codebook design of SCMA is constrained to a low dimension,There is no research about the performance and detection complexity of a SCMA system which is employed high dimensional codebook.However,in fact,SCMA system is designed for supporting massive-user access.it is important to investigate high dimensional SCMA codebook design and analyze its performance both in AWGN and Rayleigh channel.Another important contribution of our work is we demonstrate the significant benefit of increasing the dimension of SCMA codebook,where the detailed method in support of the high-dimensional codebook design,i.e.the "Latin or Semi-Latin" matrix based construction is also found and verified.The thesis then analyzes the system performance both in Gaussian channel and Rayleigh fading channel.At the receiving part,we inherit the improved MPA algorithm which is invented by us early.The high-dimensional codebook combining with the improved MPA algorithm achieves a lower detection complexity and a better system performance.In a SCMA system,each user has its unique codebook.Benefit from the fact that the codebooks are sparse,the low complexity MPA Detection algorithm can be implemented into the SCMA system.The traditional MAP requires an exhaustive search,and the detection has to perform a large number of calculations.Along with the increase number of users and iterations,the MAP complexity increases rapidly.In this thesis,the SCMA system are holistically discussed.Particularly,we focus on the design of the high-dimensional SCMA codebook and the low-complexity MPA.In more details,we analyze the system performance in AWGN channel when the improved MPA algorithm is employed.