User Scheduling And Resource Management Of Uplink Multi-user MIMO

Radio resource management(RRM) is in charge of reasonably allocating and effectively managing radio resource in mobile communication system. This enables an optimum combination of user performance and system capacity.Under the TD-LTE system in 3GPP, RRM includes resource management and bearer management, this paper mainly studies scheduling methods in uplink resource management, including user scheduling in time-domain and resource allocating in frequency-domain.In chapter one first expounds research background, raising the research plot. Later Basic theory of resource management and system model of uplink CoMP are summarized.At last, this chapter explains the research content and paper structure.In chapter two studies classic MU-MIMO pairing methods as follows: orthogonal pairing, determinant pairing, proportional fair pairing, max carrier to interference pairing and rand pairing. In uplink cellular wireless system, virtual MIMO makes use of user CSI to schedule at least two mobile set(MS) sharing all antennas and using the same time/frequency resource, thereby improving the uplink spectral efficiency. The key to the question of realizing V-MIMO is how to design efficient user pairing methods to obtain multiple user diversity gain, guaranteeing each user’s quality of service and at the same time improving system throughout. This chapter first overviews uplink MU-MIMO system model and sets up simulation platform using Matlab. Then different pairing methods are discussed under SCM channel、path loss、shadow fading to get the available system capacity, analyzing the difference of normalized system capacity curve can induce the features of each algorithm. At last, this chapter counts the complexity of floating-point calculation for different algorithms, comprehensively exploring the advantages and dis advantages of each algorithm, pointing out the superiority of DP.In chapter three studies the joint optimization problem of uplink user pairing and resource allocating. Because all users in MU-MIMO pairing group must use the same time/frequency resource, it is unreasonable to allocate resource for single user. Regarding pairing users in V-MIMO as one to allocating resource is necessary. Allocating resource of different user groups is a non-deterministic polynomial(NP) problem. Obtaining the optimal solution needs to use exhaustive search method, which means complicated arithmetic, so the key point to solve the problem is how to design low complicated suboptimal algorithm. In LTE system, the essence of the problem is jointly optimizing user pairing and resource allocating.Three suboptimal solutions are researched in this chapter and mathematical modeling is constructed for jointly optimizing problem. From the point of the sequence order of solving sub-problem, they are : user pairing followed resource allocating、resource allocating followed by user pairing、combined resource allocating and user pairing. The core concept is to split the optimization problem into two sub-problems, distinguished in the different sequence order of solving the sub-problem. It first analyzes the modeling of optimization problem and sketches the optimal solution based on exhaustive search method. Then specific steps of three algorithms are given, including the description of uplink resource allocating methods. At last, the normalized user throughout under three algorithms are simulated, comparing the throughout curves we prove that the “user pairing followed resource allocating” algorithm can ensure better user performance.In chapter four does some study on the RRM problem in uplink CoMP system, whose nature is how to realize the optimal joint scheduling between users. At the same time, it is necessary to user IRC and link adaptive technology in TD-LTE system to jointly optimize its performance. The forth chapter first proposes the coordinated system modeling of three cell, constructing simulation platform based on Matlab. After that, it introduced two common joint scheduling methods, based on orthogonality in single cell and global orthogonality in CoMP, using PF and DP. Based aforementioned two methods, a new algorithm employing switching technique is proposed which combines their advantages. The performance of the new method is proved through simulation of system capacity and user throughout with three methods. At the end of this chapter, IRC and MCS are introduced and employed in uplink CoMP scène, by means of simulating system capacity we confirm that IRC and MCS can further optimize system performance and improve system throughout.In chapter five summaries the article and point out the further research direction.