標題: 多使用者多輸入多輸出線性前置編碼系統之傳送天線及使用者選取的設計與效能分析
Designs and Performance Analysis of Transmit Antenna and User Selections for Linear Precoded Multiuser MIMO Systems
作者: 林普暄
Lin, Pu-Hsuan
蔡尚澕
Tsai, Shang-Ho
電控工程研究所
關鍵字: 多樣性選擇;線性前性編碼;多使用者多輸入多輸出;碼書;有限回傳;selection diversity;linear precoding;MU-MIMO;codebook;limit feedback
公開日期: 2013
摘要: 在本篇論文中,我們探討在多使用者多輸入多輸出系統下分在別傳送端及接收端的選擇多樣性的研究。在傳送端,我們先考慮基於線性前置編碼傳輸天線選擇機制。在有些前置編的碼單使用者多輸入多輸出系統中,由於有些不良傳輸通道的影響,使用所有的傳輸天線來傳送並無法獲得最好的傳輸效能。這個情形使我們對多使用者多輸入多輸出系統是否有類似單一使用者的這些效應產生興趣。透過本篇論文推導分析的結果,我們發現對於一給定的傳輸天線個數,減少傳輸天線使用的個數對於線性前置編碼多使用者多輸入多輸出會有單一減低系統效能的現象。因此,我們更深入去分析由於天線選擇所造成的系統效能損失。這些分析結果提供在現實系統實作上的選擇傳輸天線一些良好的設計參考。然後,基於這些分析推導的結果,我們提出數個簡單低複雜度的對於多使用者多輸入多輸出之傳輸天線選擇的演算法。此外,演算法複雜度的分析以及模擬結果都顯示出我們提出的演算法能大幅地減低系統運算複雜度而且保有與最佳演算法接近的系統效能。結果而論,這些分析的結果可以讓我們能更有效的了解傳輸天線選擇是如何影響到多使用者多輸入多輸出系統;然後我們提出的演算法使得傳輸天線選擇在實際系統使用上能更佳的有彈性。在接收端使用者上,我們考慮對於迫零前置編碼多使用者多輸入多輸出系統在完整通道狀態資訊及有限通道狀態資訊環境下的多使用者選擇機制。我們使用基於貪婪架構下考量總系統流量準則的使用者選擇演算法用來避免窮舉法產生的龐大複雜度。然後藉由分析此演算法,我們推導出一個等效的選擇準則還有無損效能的快速演算法來大幅減低運算複雜度。此外,在有限通道狀態資訊的環境下,我們發現最大化每個接收端使用者的接收能量在最大化總流量準則下扮演者相當重要的角色。然後我們證明出接收端訊號能量在碼書,也就是前置編碼使用公正矩陣群組下能被最大化。根據這些結果,我們建議在現實系統下如何建構實作可用的碼書。另一方面,我們也提出一個公平的使用者分配機制來達到良好的系統效能及公平權責。透過模擬結果我們可以發現提出的使用者選擇演算法、碼書設計以及使用者分配在完整通道狀態資訊跟有限通道狀態資訊環境下都優於傳統的方法。
In this dissertation, selection diversity in both transmitter and users for multiuser MIMO (MU-MIMO) systems is considered. In the transmitter, transmit antenna selection for linearly precoded MU-MIMO systems is investigated. In some precoded single-user MIMO systems, using all transmit antennas does not always lead to the best performance due to ill-conditioned channel matrix. This motivates us to investigate whether similar result can be obtained in MU-MIMO systems. From the derived analytical results, it was found that for a given number of transmit antennas, decreasing the number of active transmit antennas (number of RF units) always degrades system performance in the linearly precoded MU-MIMO systems. The performance loss due to transmit antenna selection is further analyzed. These analytical results provide good design references for using transmit antenna selection in practical systems. Moreover, based on the analytical results, several simple transmit antenna selection algorithms for linearly precoded MU-MIMO systems are proposed. Complexity analysis and simulation results show that the computational complexity of the proposed algorithms can be significantly reduced while the performance is still comparable to the optimal selection scheme. As a result, the analyzed results enable us to better understand how transmit antenna selection affects the MU-MIMO systems; also, the proposed algorithms make transmit antenna selection more feasible to be used in practical systems. In the user terminals, user selection for zero-forcing precoded MU-MIMO systems is considered in both full and partial CSI (channel state information) environments. A greedy-based user selection algorithm using a sum throughput criterion is proposed to avoid exhaustive search. By analyzing the performance of the proposed algorithm, an equivalent design criterion is derived and a fast algorithm without degrading the performance is proposed for further reducing the computational complexity. Moreover, in the partial CSI environments, it is found that maximizing the received signal power for each user plays an important role in maximizing the sum throughput. Then, it is shown that the received signal power can be maximized if the codebooks as well as the precoders are designed using groups of unitary matrices. From the results, it is suggested that how to construct the codebooks for practical systems. Furthermore, a fair user scheduling algorithm is proposed based on the proposed user selection algorithms and codebook designs for achieving good trade-off between fairness and system performance. Simulation results show that the proposed user selection algorithms, codebook designs and user scheduling outperform conventional schemes in both full and partial CSI environments.
URI: http://140.113.39.130/cdrfb3/record/nctu/#GT079612547
http://hdl.handle.net/11536/74245
Appears in Collections:Thesis