Research On Distributed Algorithm For Resource Allocation In Dense Wireless Communications

In order to support the high-speed transmission of massive data,the cell radius of the 5G or B5 G mobile communication system will become smaller and smaller,and the multi-antenna coverage makes the coupling degree of the cell larger and larger.For such highly-coupled cellintensive mobile communication networks,the conventional cell-by-cell radio resource allocation method is no longer applicable in this scenario since the neighbor cell interference can no longer be ignored.In order to obtain a better radio resource allocation scheme,a global radio resource allocation optimization problem needs to be calculated for the entire network of mobile stations and base stations.Due to the huge number of mobile stations and antennas,this optimization problem is not feasible for centralized calculations on a network node.One of the best ways to solve the problem is that the global wireless resource allocation optimization problem is decomposed into various small problems for distributed computing.This paper focuses on distributed algorithms for global wireless resource allocation in dense cellular networks.The main tasks are as follows:1.Distributed algorithms for downlink global time slot and power resource allocation in dense cellular networks based on minimizing total power and maximizing total rate.The main work consists of two parts: 1)The signals of neighboring base stations are modeled as interference.The closed-form expressions of optimal solutions for time slots and power resources in a single cell are deduced based on minimizing the total power and maximizing the total rate,respectively.Since the corresponding optimization problem is non-convex,the optimization problem can be equivalently transformed into a convex problem through variable transformation.By solving the optimal conditions,a closed-form solution is obtained.2)Based on the closed-cell expressions of the optimal solution for time slots and power in a single cell,two low complexity distributed time and power resource allocation algorithms are designed for the allocation of downlink global time slots and power resources in a coupled dense network based on minimizing the total power and maximizing the total rate,respectively.The distributed algorithm for minimizing the total power or maximizing the total rate has the following characteristics:each base station updates the time slot and power resource of the user of the current cell synchronously,and broadcasts the updated time slots and power sum information to other base stations.3)The convergence of the proposed distributed algorithm is proved.For total power minimization,the algorithm is proved to be convergent and globally optimal by using the properties of standard interference function.For total rate maximization,the convergence of the distributed algorithm is proved by using the monotonically increasing and boundedness of the series generated by the iteration.2.Distributed algorithms for the allocation of downlink global carriers and power resources for intensive cellular networks based on minimizing total power,maximizing total rate,and maximizing total energy efficiency.The main work includes the following two points: 1)In consideration of the carrier and power coupling between multiple cells,we derive the unified optimization problem for the allocation of downlink global carriers and power resources in dense cellular networks to minimize total power,maximize total rate,and maximize total energy efficiency.2)For the problem of minimizing the total power,maximizing the total rate,or maximizing the total energy efficiency,the corresponding distributed algorithm is proposed.Besides,the complexity of the algorithm is analyzed,and the implementation scheme of the algorithm is given.3.Distributed algorithms for the allocation of power resources for intensive cellular networks with NOMA based on minimizing total power and maximizing total rate.The main work includes the following two points: 1)A distributed algorithm is proposed for downlink power resource allocation in dense cellular networks using NOMA based on minimizing total power.Taking into account the minimization of total power,a closed-form solution of the optimal power allocation strategy is first obtained.Then the original problem is successfully converted into a linear problem with fewer variables,which can be optimally solved by using the standard interference function.2)A distributed algorithm is proposed for the allocation of downlink power resources for dense cellular networks using NOMA based on the maximum total rate.In order to solve the problem of maximizing the non-convex total rate,it is first proved that the power allocation problem of a single cell is a convex problem.By analyzing the optimal conditions,the optimal power allocation for users in a single cell can be derived as a closed expression.Based on the optimal solution for each cell,a distributed algorithm is accordingly proposed to obtain an effective solution to solve the problem of maximizing the total rate of multiple cells.Numerical results validate the theoretical findings and show the superiority of the proposed solution compared to orthogonal frequency division multiple access and broadcast channels.4.A distributed algorithm for joint allocation of antennas,time slots,and power resources in dense and heterogeneous networks based on rate and fairness.The main work includes the following three points: 1)For the non-convex optimization problem of joint allocation of threedimensional resources,the closed-form expression of the optimal time slot allocation strategy is deduced under the given conditions of antenna and power resource allocation.It is proved that the optimal base station time slot allocation pattern is obtained.2)Under the given conditions of antenna resource allocation,the problem of non-convex time slot and power resource allocation can be transformed into a convex problem with fewer constraint conditions by using exponential transformation.It can be proved that the original non-convex problem is equivalent to the variable substituted convex problem.3)A low-complexity distributed iterative algorithm is proposed to obtain a suboptimal solution to the joint optimization problem.The feature is that all users update the antenna selection information according to the broadcast information of the base station and report it to the selected base station.The base station updates the broadcasted parameter information based on user-uploaded information,and the parameters of the user and the base station are iterated until convergence.