Research On Cooperative Noma Based On Factor Graph Construction

Internet of Things is a network of connected terminals that grows exponentially in recent years.Such a huge growth demands that future wireless communications must support high spectral efficiency,massive connectivity,and low latency.Non-orthogonal Multiple Access(NOMA)has attracted much attention for its ability to support the above requirements better than the traditional orthogonal Multiple Access.In order to further improve the cell coverage and the ability to resist channel fading,the cooperative NOMA(Co-NOMA)has attracted much attention of many researchers.Therefore,we study Co-NOMA system based on factor graph design and optimization in this paper.We apply multi-carrier NOMA to the relay network and study the performance evaluation and optimization of Co-NOMA communication system in this paper.In this model,the source node broadcasts their messages to the base station and the relay node.Then,the rely node decodes and forwards the euqipments’ messages to the base station.The base station performs joint multi-user detection.We study the design and optimization of Co-NOMA communication system over Gaussian channel and fading channel respectively.In the Gaussian channel,with the goal of minimize the pairwise error probability we propose multiple optimization schemes,including the optimization of the superposition matrix,the selection of matrix elements,constellation rearrangement,etc.In the fading channel,different from the selection combining utilized in the literature,we employ maximum ratio combining in the joint detection.We study evaluation indicators and optimization methods of system performance.And we derive the analytical expressions of the outage probability and the sum data rate for the Co-NOMA system.Due to the high complexity of the above optimization scheme,it is difficult to obtain the optimal solution by exhaustive search.So we use genetic algorithm and simulated annealing algorithm to approximate the optimal solution in this paper.The simulation results show that the above optimization schemes can eliminate the overlapped points of the joint constellation,and improve the symbol error rate performance of the system over the Gaussian channel.In the fading channel,the joint detection,the matrix optimization and constellation rearrangement can improve the outage performance of the system.