Codebook And Low Complexity Receiver Design For SCMA Assisted Visible Light Communication System

With the widespread application and continuous development of wireless technology in various industries,the scarcity of frequency spectrum resources has become the main obstacle to the development of wireless communication technology.Visible light communication(VLC),with its advantages such as non-electromagnetic interference,green environmental protection,and overall lighting,has brought unlimited potential for communication and sensing.Non-orthogonal multiple access(NOMA)is an effective solution to the problems of limited modulation bandwidth and limited transmission rate improvement in VLC.Sparse code multiple access(SCMA)as a code domain NOMA has been widely used.Currently,there have been studies that applied SCMA in visible light communication scenarios,effectively improving the transmission rate and system capacity of VLC systems.However,the SCMA-VLC system has a significant peak-to-average power ratio(PAPR)problem,and the existing small codebooks and high receiver detection complexity limit the practical use of the system when providing services to more users.Therefore,this thesis focuses on the design of low peak-to-average ratio,high overload rate codebooks and low-complexity multi-user detection algorithms for SCMA-VLC systems.The main research work of this thesis is as follows:At the transmission end,this thesis conducts research on codebook design schemes for real and complex fields,respectively,targeting the requirement of smaller PAPR for visible light communication equipment and the problem of multiple user access.Specifically,under the guidance of Star-QAM codebook design and power imbalance codebook design methods in the complex field,this thesis proposes the minimum error rate codebook design scheme based on successive interference cancellation(SIC)and the stacked codebook design scheme based on probability shaping.Simulation compares the differences in performance indicators such as PAPR,minimum euclidean distance,error performance,design complexity,and detection complexity between the codebooks designed by these two schemes and traditional codebooks.The simulation results show that the codebooks designed by these two schemes have lower PAPR and design complexity and better error performance in systems with higher overload rates.In the real field,this thesis derives the expression of the lower bound of system mutual information under shot noise and proposes an real domain codebook design method based on mutual information.Simulation results show that compared with existing real codebooks,this codebook has better error performance.At the receiver end,this thesis addresses the problem of exponential growth in computational complexity of message passing algorithms(MPAs)with the increase of the number of users and codebook size,and designs three low-complexity multi-user detection algorithms to meet different system detection requirements.In the case of limited resources on the mobile end,this thesis proposes a fast decoding detection algorithm and a K-means parallel decoding detection algorithm.In the case of random user access in the system,a compressed sensing-assisted multi-user detection algorithm is proposed.Simulation results show that compared with the MAX-LOG-MPA algorithm,these three algorithms can effectively reduce the complexity of operations while ensuring error rate performance.Among them,the fast decoding detection algorithm has the lowest computational complexity,and the compressed sensingassisted multi-user detection algorithm performs well in error performance when system users randomly access.