| User-centric non-cellular networks,also referred to in this thesis as non-cellular UC networks,refer to the mobile communication network coverage architecture that some distributed access points(APs)are controlled by several central processors.Because this architecture does not have a traditional cellular structure and activates several APs to provide user-centric service,it is known as the user-centric non-cellular network.The existing studies show that since non-cellular UC networks can shorten the average distance between mobile users and APs and increase the degree of spectrum reuse,it can greatly improve the spectrum efficiency and energy efficiency of the network.Therefore,non-cellular UC networks have been widely studied in the academic fields in recent years.In this dissertation,the multi-user downlink scenarios of non-cellular UC networks are considered to maximize or minimize system performance functions such as total energy efficiency,power consumption and spectrum efficiency,by jointly optimizing the parameters of wireless resource such as AP clustering,channel allocation and beamforming.The main contributions are summarized in the following four aspects.Firstly,in the multi-user downlink transmission scenario,AP clustering and beamforming are jointly optimized under the constraints of user rate and transmitting power to maximize the total energy efficiency of the network.The main work includes: 1)The mathematical optimization problem corresponding to the problem is formulated.Due to the characteristics of discrete AP clustering variables and continuous beamforming variables,the joint optimization problem is decomposed into two sub-problems: ”given beamforming variables,AP clustering problem” and ”given AP clustering,beamforming problem”.2)For the former subproblem,an AP clustering scheme based on channel gain threshold is proposed,and a method to determine channel gain threshold based on multidimensional grid search is proposed.3)For the latter one,the subproblem is solved through the following three steps: x The maximum and minimum rate problem is formulated to determine whether the beamforming problem is feasible,and the Lagrangian dual and WMMSE methods are used to solve the problem.y When the problem is feasible,an auxiliary variable and Lagrangian dual method are introduced to solve the energy efficiency maximization problem after relaxing the rate constraints.z When the problem is infeasible,the rate constraint is relaxed to make the problem feasible,and the feasible problem is transformed into a power minimization problem.Then,the approximate solution of the original problem is obtained by using Lagrange dual and WMMSE method.Simulation results show that the proposed AP clustering and beamforming optimization algorithm can converge to a better solution than the existing classical algorithms,thus improving the energy efficiency of the network to a certain extent.Secondly,in the multi-user downlink transmission scenario,AP clustering,transmission time and beamforming are jointly optimized,under the constraints of fronthaul link capacity,computing resources,user rate and transmitting power,to minimize the total power consumption of the network.The main work includes: 1)The mathematical optimization problem corresponding to the problem is formulated.Due to the characteristics of discrete AP clustering variables,continuous transmission time and beamforming variables,the total power minimization problem is a mixed integer non-convex optimization problem.In order to solve the problem,the original problem is converted into an equivalent one by using semi-positive definite relaxation,and then the Taylor expansion method is used to approximate the transformed problem.2)Based on the successive convex approximation(SCA)method,an iterative joint optimization algorithm is proposed.In addition,a distance-based AP clustering scheme with low computational complexity is designed to generate good initial solution,so as to reduce the number of iterations of the joint optimization algorithm.The simulation results show that compared with the existing classical algorithms,the proposed joint optimization algorithm can converge to a better solution,thus improving the power consumption performance of the network to a certain extent.Thirdly,in the multi-user multi-slot downlink transmission scenario,AP clustering and beamforming are jointly optimized,under the constraints of transmitting power and AP clustering between adjacent time slots,to maximize the total spectrum efficiency of the network.The main work includes: 1)The user mobility is modeled by the random walk model,and then the dynamic AP clustering and beamforming design problem is formulated.In order to solve this complex problem,it is decomposed into two sub-problems: ”given beamforming,dynamic AP clustering problem” and ”given AP clustering,beamforming problem”.2)For the former subproblem,a particle swarm optimization method based on the constraint of AP clustering is proposed,so as to realize the dynamic optimization of AP clustering scheme.3)For the latter one,the non-convex beamforming problem is transformed into matrix weighted mean square error minimization problem by using WMMSE and Lagrange methods,and an iterative beamforming algorithm is proposed.Simulation results show that the convergence performance of the algorithm can be effectively improved by exploiting the relationship between the AP clustering solutions of adjacent time slots.In addition,compared with the existing classical algorithms,the proposed algorithm can improve the spectrum efficiency of the network to a certain extent.Fourthly,the V2 I networks is a typical application scenario of multi-user non-cellular UC networks.In the downlink transmission scenario of such networks,dynamic resource(vehicle user-road side unit pairing and subcarrier)allocation and beamforming are jointly optimized,under the constraints of transmitting power and resource allocation between adjacent time slots,to minimize the total power consumption of the network.The main work includes: 1)Under the high-mobility time-varying channel,the dynamic resource allocation and beamforming joint optimization problem is formulated.In order to solve this complex problem,a two-stage dynamic resource allocation and beamforming algorithm is proposed.2)In the first stage,given beamforming matrices,the resource(VUE-RSU paring and subcarrier)allocation are dynamically optimized.A particle swarm optimization algorithm is proposed to solve the subproblem of dynamic resource allocation.In the second stage,given the resource allocation variable,the beamforming matrices are optimized.The maximum and minimum rate problem is firstly formulated to determine the feasibility of beamforming problem,and then the beamforming problem is transformed and solved by using the semi-definite relaxation and Lagrangian dual method.3)Based on the above two stages,an effective dynamic resource allocation and beamforming algorithm is proposed.Compared with the existing classical algorithms,the proposed algorithm can improve the power consumption performance of the network to a certain extent. |
PDF Full Text Request