Research On Resource Optimization In Precoding-based Multi-antenna Networks

With the rapid development of manufacturing technology and mobile communication technology,the number of mobile terminals and the demand of data traffic show explosive growth,which provide many new challenges for the future wireless communication networks,such as the scarcity of spectrum resources and the densification of network deployment.In this context,multi-antenna network has been emerged as one of the primary achievements in wireless communications during the past decades,which can significantly improve users' signal-to-noise ratio and suppress the interference by the precoding technology from the spatial dimension,rather than adding additional spectrum resources.In addition,the frequency resource is generally reused in each cell of the future networks,which results in the serious inter-cell interference.By adopting cooperative precoding strategy among the base stations,the inter-cell interference can be effectively inhibited or even completely eliminated.Due to this fantastic feature,interference coordination technology has also been a hot topic in the academic and industrial community.This dissertation aims to study the precoding-based interference coordination technology and resource optimization scheme in the typical two emerging types of multi-antenna networks,which are developed to deal with the challenges caused by the scarce communications resources and the complex interference management.The main contributions of this dissertation are summarized as follows:Firstly,due to the congested situation for the future networks,we study the joint problem of user grouping,admission control and precoder design in multicast network,and propose the optimal and suboptimal algorithms based on branch and bound method.The main idea of these two developed algorithms is to transform the nontrivial precoder design problem into the semidefinite programming problem by the semidefinite relaxation approach in the case of given the variables about user grouping and admission control,and then the branch and bound method is used to search the solutions of user grouping and admission control.Moreover,we consider the influence of unreasonable admitted order,and also develop a greedy algorithm according to the defined user selection metric,which further reduces the complexity of the algorithm and realizes the effective compromise between the algorithm performance and computational cost.Secondly,since the one-shot scheduling strategy may lead to the unbalance issue,we investigate the multiuser precoding and partitioning with the round-robin service mechanism.Under the condition of given the number of groups,we derive the relationship between channel distance and the total throughput,and propose a user partitioning algorithm based on the hierarchical clustering method.Then,a sequential parametric convex approximation method is applied to relax the non-convex precoder design optimization problem,and an iterative algorithm is designed to solve this approximate problem.By incorporating the solutions of these two subproblems,two algorithms are proposed with different complexities,which can obtain the optimal and suboptimal grouping number,respectively.Thirdly,we consider the situation of base station cooperative transmission.Due to the challenges of the deployment cost,flexibility and scalability for traditional wired backhaul networks,we propose a novel network architecture by jointly designing wireless backhaul and access networks.In order to solve the problem of inter-cell interference in the access network,the cooperative precoding strategy is used to suppress the interference.At the same time,from the user's perspective,we develop a dynamic user association scheme between the users and the base stations,which makes better result of load balance.According to the networking strategy of the access network,we further apply the multiuser precoding scheme in the backhaul links,which can fully exploit the spatial diversity to improve the transmission efficiency.For this joint optimization problem,the discrete association subproblem is first relaxed to the continuous form,and then the DC scheme and the convex approximation method are used to transform the problem,respectively.Finally,we design the corresponding algorithms for these two cases.Accordingly,the balance between the access network and the backhaul network is achieved.Finally,since the ability of interference suppression is always limited for the cooperative precoding strategy,we study the user-centric clustering scheme in access network,which can serve a user by multiple base stations simultaneously to fully exploit the spatial diversity,and then the transmission reliability and rate are enhanced.Furthermore,the TDMA-based multicast transmission scheme is used to deliver user's data to the corresponding the cluster of the serving base station.Based on the characterization of the coupling relationship between backhaul and access networks,the convex approximation method is applied to transform the original problem,and a computation-effective algorithm is proposed.In addition,we further develop a binary search based user clustering search algorithm and Gaussian randomization method based precoding design algorithm,whose objective are to find the feasible solution of the original problem.