Study On The Technology Of Signal Superposition And Detection For Non-orthogonal Multiple Access Systems

With the worldwide spread of 4G wireless technology,mobile internet has been developed rapidly.As the next generation of mobile communication system,5G has become the focus of all sectors of the community.Among all the current researches for 5G,the non-orthogonal multiple access is regarded as one of the most promsing technology due to its advantages in significantly increasing system spectrum efficiency.The sparse coding multiple access(SCMA)and pattern division multiple access(PDMA)based on the code domain are the industry-accepted non-orthogonal multiple access technologies.At the transmitting end of those systems,user information are mapped into corresponding codewords and sparsely multiplexed on limited time-frequency resources.At the receiving end,the use of message transmission algorithm or successive interference cancellation detection combining decoder recovers the information bits of each user.Conventional SCMA and PDMA systems based on factor diagram and successive interference cancellation algorithm are affected by the short cycle and error propagation phenomenon.In order to overcome the shortcomings of SCMA and PDMA systems,this paper proposed a signal superposition design scheme based on factor diagram rearrangement during resource groups and multiple decision aided successive interference cancellation algorithm respectively at the sender and the receiver.The key of the scheme of rearrangement during resource groups is to re-permutate the graph diagram during different resource groups in order to reduce the numbers of short cycles.The speed of the correction of error messages in the process of iteration is then accelerated finally.Multiple decision aided successive interference cancellation put the codewords with relative high probabilities into a candidate collection for further comparison when detecting each user.In addition,the detection algorithm can also adaptively allocate computing resources and reduce the computational complexity as much as possible through a predetermined threshold.This article focuses on the overall framework of the rearrangement schemes for graph diagrams during different resource groups and the multiple decision aided successive interference cancellation detection scheme.The simulation results show that the scheme of factor graph rearrangement among resource groups can effectively mitigate the impact of cycle loop,accelerated the convergence speed of the systems.The multiple decision aided successive interference cancellation detection can overcome the error propagation in conventional successive interference cancellation algorithm,and finally improve the performance of the whole system with a relatively low computational complexity.