Performance Improvement Based On Adaptive Weighted Graph Matrix For Uplink SCMA With Randomly Distributed Users

The limited accessibility to the spectrum due to the fifth generation(5G)networks requirements has led to severe challenges,which can be overcome through sparse code multiple access(SCMA)as one of the robust schemes of non-orthogonal multiple access(NOMA)techniques.Stochastic spatial modelling of SCMA networks is essential for their accurate visualization,design methods,and performance analysis,which,in some special situations,can lead to closed-form relationships.These equations enable the understanding of SCMA behaviour and provide an insightful design mechanism.It is challenging for SCMA,to deal with the multidimensional codebooks involved in the multiuser random scenarios.This thesis provides a systematic treatment for performance metrics analysis of uplink SCMA networks where users are randomly distributed in a disk-shaped cell.First,in this thesis,new closed-form expressions are derived for the average sum rate and users’ individual rate based on an adaptive weighted graph matrix(AWGM).A novel joint resource allocation method as a multi-objective optimization(MO)problem is developed to maximize the average sum rate and fairness among users.Thanks to AWGM,which eliminates the need to separate the assignment and power allocation problems,the MO problem is proposed by a heuristic search approach and a four-step algorithm.It is confirmed that our proposed method,compared to other algorithms,can compromise and improve the multiple objectives’ performance and guarantees a stable range of network performance at different times.Second,this thesis investigates average sum rate maximization by separating assignment and power allocation optimization problems.A new assignment algorithm based on the individual rate with the interference updating capability is proposed.After calculating the factor graph matrix,the power allocation problem is solved by the waterfilling method.The simulation result shows that our proposed method can guarantee optimality compared to the other sub-optimal algorithms.Third and last,the symbol error rate(SER)performance is characterized by assuming joint maximum likelihood(ML)receivers.The new analytical expression for SER is derived.Essential insights are concluded from the simulation and analytical results.