Cross-Layer Joint Optimization Design For 5G In NOMA Systems

The service provision capability of the Fifth Generation mobile communication technology is much higher than that of the Fourth Generation.Instead of emphasizing the peak rate,5G mobile communication technology proposes different communication indexes according to different scenarios to enhance the user experience.In order to adapt to the rapid development of 5G multi-application multi-scene,cross-layer resource allocation is designed to provide different QoS services for mobile multimedia networks.Considering the requirements of user fairness,time delay sensitivity,system throughput,a cross-layer joint resource allocation scheme is studied to provide an effective theoretical support and technical scheme for non-orthogonal multiple access access of the 5G system.The main work of this thesis is as follows:Firstly,the resource allocation scheme based on user fairness in NOMA system is studied,including the optimal user combination on subband and the optimal power allocation scheme for users.For a particular subband,the necessary and sufficient conditions for any two consecutive users in the optimal combination of users to be satisfied are derived by mathematical derivation,and the users in the user set who are not satisfied with the requirements are removed iteratively according to this condition.On this basis,for each group of candidate user combinations,an optimization is established for maximizing the weighted sum rate of user groups.Then,Lagrange multiplier method is adopted to obtain the optimal power allocation of users.The simulation results show that the proposed scheme is superior to the traditional user scheduling and power allocation scheme,and similar to the optimal uplink and downlink duality method but with much lower complexity.Secondly,a cross-layer resource allocation scheme combining time-delay sensitivity and system throughput is studied,and a cross-layer power optimization algorithm combining the parameters of the data link layer and the physical layer is proposed.In the single input and single output scenario,an efficient method for user selection by maximizing the product of average effective capacity is proposed.Then,the delay QoS index is introduced and the maximum-minimum effective capacity model is established.The closed solution of the optimization model can not be obtained directly.In this thesis,the equivalence of the optimization model is transformed into a set of standard convex programming problems by mathematical proof.On this basis,the maximum effective capacity of the system and the corresponding optimal power allocation are obtained iteratively by the binary search method.Finally,the SISO solutions are extended to a multiple-input multiple-output(MIMO)scenario.Simulation results show that compared with the traditional power allocation scheme,the proposed algorithm can not only guarantee the delay QoS index of users,but also maximize the effective capacity of the system.