Research On Technologies Of User Scheduling And Beamforming In Multi-user MIMO System

Due to its advantage of effectively improving the spectrum utilization andcapacity, thereby significantly improving the data transmission rate, MIMO technologyis widely used among various wireless communication systems, such as3GPPLTE/LTE-Advanced, IEEE WiMAX, WLAN and so on. However, on the issues ofrealization of the MIMO technology, the implementation of the uplink MIMO islimited by the terminal size and complexity, so it is difficult to achieve a plurality ofantennas in the terminals’ side, therefore people consider the introduction of a newtechnology named V-MIMO. V-MIMO technology allows two or more terminals totransmit simultaneously in the same time-frequency resources, so that in the opinion ofbase station, a MIMO structure is configurated with multiple single-antenna terminalsand itself composed of multiple antennas, so as to improve the reliability and capacityof the uplink communication, with help of spatial diversity and multiplexing. The keyof V-MIMO technology lies in scheduling schemes, that is, how to choose theco-channel users, which directly affects the performance of the system. With the factthat performance of multi-user MIMO system is limited by inter-user interference,correct choice of users as to minimum the interference between users, can enhancespectral efficiency of the V-MIMO system, on the contrary, incorrect selection mayeven lead to poorer performance contrasting to single input multiple outputtransmission.Beamforming technology is also an effective interference suppression technology,which can be used to improve the performance of capacity and BER of a wirelesscommunication system. The principle of beamforming is that users’ data are multipliedby precoding vectors before transmitted over channels, or received user data stream ismultiplied by a receive beamforming vector before processing. So when base stationtransmits or receives data of a user, the main lobe is made in the direction of certainuser channel, and zero-value point is aligned with other users who interfere. For thedownlink transmission with beamforming, useful signals can be extracted by means ofreceiver combinations after users receive signals, while receive beamforming for uplink transmission is also a kind of technology for multi-user detection, with whichbase station extracts the data of a specific user from mixed data of multiple users. Withreasonable design of beamforming vectors, co-channel interference between users canbe effectively eliminated or weakened, and this technology has been used as part of theIEEE802.16e standard, as well as LTE-Advanced standard.In this paper, studies of joint user scheduling and beamforming are made inmulti-user MIMO systems. The main tasks are as follows:Firstly, for both uplink and downlink, latest progress and research status of userscheduling and beamforming in multi-user systems are summarized, and the existingproblems are analyzed.Secondly, for uplink communications of multi-user V-MIMO, whose performanceis greatly affected by user scheduling, we summarize several traditional algorithms foruser pairing based on its system model. On basis of the analysis that traditionalalgorithms only support to schedule no more than two users, which can not sufficientlymake use of the spatial multiplexing of MIMO channels, we expand these algorithmsto support the user selection of more than two users, and then simulation comparisonsare made with the traditional algorithms.Thirdly, based on the multi-user MIMO uplink, taking the calculation of SINRinto account we propose two algorithms for joint user scheduling and receivebeamforming, and principles and flows of the two algorithms are recounted. Then theyare compared with other algorithms to show their applicable scenarios.Finally, based on the multi-user MIMO downlink, on purpose of improving theexisting algorithm, we propose a successive algorithm for joint precoding and userselection in broadcast channel with limited feedback, and the derivation and flow ofthis algorithm is detailed. At last this new algorithm is contrasted with other algorithms,and simulation results show the advantage of the proposed scheme under certainsystem conditions.The end of this paper concludes this dissertation and discusses the future researchwork.