User Association And Resource Management For 5G Wireless Heterogeneous Networks

With the rapid development of information technology,the requirement of mobile data traffic and the number of terminal equipments will increase dramatically.Meanwhile,more and more multimedia applications will appear in smart terminal equipments.Thus,the fifth generation of wireless mobile communication system(5G)needs to deal with the explosive growth of mobile data traffic,massive interconnected equipments,continuous emergence of new services and application scenarios.In order to meet the demands of wireless communication,heterogeneous networks(HetNets)can effectively improve the system coverage probability and the system capacity by increasing the number of small base stations and shortening the distance between terminal equipment and base station.In HetNets,the terminal equipment will be covered by multiple types of base stations.Since different types of base stations have different transmit power and deployment density,the traditional user association strategy based on the maximum received signal stfength will not be directly applied to 5G wireless HetNets.This will cause inefficient small base s-tation deployment and increase the energy consumption of the terminal equipment.Due to the shortage of spectrum resources in 5G wireless HetNets,spectrum is often shared in HetNets,which will cause severe co-channel interference.Facing the co-channel in-terference and emergence of new application for terminal equipment,the contradiction between the energy demand of terminal equipment and the limited battery capacity be-comes increasingly prominent.In such resource-limited circumstances,how to extend the standby time and improve the energy efficiency of terminal equipment has become an urgent problem.In addition,HetNets are gradually applied to high-rise buildings.Due to the coexistence of a large number of base stations and terminal equipments,the dis-tance between the terminal equipment and the base station is closer,and the interference is more serious,then the terminal equipments need to determine which base station to be associated with before communication.Different user association schemes lead to differ-ent system performance.Therefore,in this paper,under the framework of 5G wireless HetNets,with the goal of improving the system performance such as energy efficiency,the resource management is studied through key issues such as user association and power control.The main contributions are summarized as follows:1)From the perspective of network energy consumption and terminal equipment energy efficiency,the user association and power control problems in HetNets have been studied.Aiming at the problem of excessive energy consumption in the network,a heuris-tic user association strategy is proposed and the energy consumption of HetNets is reduced through switching off some base stations which are idle.Based on this,a distributed pow-er control strategy is proposed,which effectively reduces co-channel interference and improves the energy efficiency of terminal equipment.Finally,simulation results show that compared with the existing resource management strategies,the proposed algorithms can effectively reduce the network energy consumption and improve the energy efficiency of terminal equipment.2)By considering the fairness for energy efficiency,joint user association and power control problems have been studied in the multiple input multiple output(MIMO)en-abled HetNets.Firstly,the expression of terminal equipment’s rate is derived,then the energy efficiency expression has been given in combination with the energy consump-tion model of terminal equipment.The energy efficiency optimization problems have been established from the proportional fairness and the max-min fairness criteria.Since each optimization problem is a mixed integer non-convex problem,it is difficult to obtain the optimal solution.The formulated optimization problems are solved through dividing the original problem into two subproblems(i.e.user association subproblem and power control subproblem).For the proportional fairness problem,the dual decomposition and Newton methods have been utilized to solve the user association subproblem and power control subproblem,respectively.Through successive convex approximation and dual de-composition methods,the power control subproblem has been solved based on max-min fairness.For the user association subproblem,by the relaxation and equivalent change of the user association variable,the subproblem has been transformed to a convex prob-lem and a dual method has been used to solve it.Finally,the problem modeled by each fairness criterion is solved by an iterative algorithm.Simulation results show that the max-min fairness criterion has higher fairness and the proportional fairness criterion has higher energy efficiency.Compared with other resource management strategies,the pro-posed algorithms can significantly improve the energy efficiency of terminal equipments.3)From the perspective of reducing overhead in wireless communication system,the user association and power control problems based on reinforcement learning have been studied.Different from the traditional methods,such as game theory,fractional programming and convex optimization,which need more and accurate network informa-tion in practice,the multi-agent deep Q-learning network(DQN)method which requires less communication information of the environment is studied in orthogonal frequency division multiple access(OFDMA)based uplink HetNets.According to the application scenario,the action space,state space and reward function of the multi-agent DQN based framework are redefined and formulated.Simulation results demonstrate that the multi-agent DQN method has the best performance on convergence and energy efficiency com-pared with the traditional reinforcement learning(Q-learning)method.4)From the perspective of user association,a maximum instantaneous signal to interference plus noise ratio(MISINR)user association scheme has been studied in the high-rise building 3 dimension(3D)environment.The coverage probability and average user rate expressions have been theoretically derived by using stochastic geometry in a general case and an interference-limited case,respectively.The system performance expressions based on the maximum long-term averaged received power(MLARP)and minimum transmission distance(MTD)user association strategies have also been studied.Under certain special conditions,the expressions of coverage probability and average user rate based on the MLARP and MTD user association strategies are equivalent.Results show that MISINR user association scheme offers the best performance in terms of both the coverage probability and the average user rate through comparisons with the MTD and MLARP user association schemes.In summary,for the 2 dimension(2D)HetNets,several efficient user association and power control strategies have been proposed to improve the energy efficiency of terminal equipments.The performance analysis of different user association strategies in 3D Het-Nets has been studied and this work expanded the theory of 3D HetNets research based on stochastic geometry theory.