請用此 Handle URI 來引用此文件:
http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/19608
標題: | 正交分頻多工系統之低複雜度盲式資料估測器 A Low Complexity Blind Data Detector for OFDM Systems |
作者: | Yi-Syun Yang 楊易洵 |
指導教授: | 李學智 |
關鍵字: | 正交分頻多工系統,盲式資料估測器,廣泛概似比估測,模糊性,實數分解,多輸入多輸出,發射分集, Orthogonal frequency division multiplexing (OFDM),blind data detection,generalized likelihood ratio test (GLRT),ambiguity,real value decomposition (RVD),Multiple-Input Multiple-Output (MIMO),transmit diversity, |
出版年 : | 2016 |
學位: | 博士 |
摘要: | 本篇論文在正交分頻多工(Orthogonal Frequency Division Multiplexing, OFDM)系統架構下,以廣泛概似比估測(Generalized Likelihood Ratio Test, GLRT)為準則,提出一低複雜度的盲式資料估測器。傳統的廣泛概似比估測的資料估測器(GLRT Data Detector, GDD)可以表示為一組合最佳化問題,會有複雜度過高以及資料模糊性(Signal Ambiguity)的缺點。在本篇研究中,我們將介紹如何將傳統廣泛概似比估測的資料估測器,經過分組的方式切割成數個子集合廣泛概似比估測的資料估測器(Sub-group GDD, SGDD),以達到降低複雜度的需求。
一般而言,分組切割的方式可以為為兩大類:交錯存取(Interleaving)與部份波段(Subband)。在以交錯存取為切割方式的子集合廣泛概似比估測的資料估測器下,我們提出一兩階段估測(Two-Stage SGDD)的演算法,其在第一階段利用資料模糊性降低搜尋複雜度,第二階段則利用領航訊號(Pilot)解決資料模糊性,完成估測程序。由數學的分析結果發現,使用兩階段估測,在M位元相移調變(M-ary Phase Shift Keying, PSK)下可節省M-1/M的複雜度,而在正交幅度調制(Quadrature Amplitude Modulation, QAM)下可節省約1/4的複雜度,且重要的是,使用該降低複雜度的方法並沒有任何效能上的損失。而在以部份波段為切割方式的子集合廣泛概似比估測的資料估測器下,兩階段估測演算法並不適用,但若部份波段所佔的頻寬較通道同調頻寬(Channel Coherence Bandwidth)來的小,部份波段則可以較小的子集合呈現,以指數遞減的速度降低運算複雜度。除此之外,我們也針對部份波段的特性,提出一星座圖混搭的機制,在無損頻寬的前提下解決資料模糊性。 不論是何種分組切割方式,SGDD皆可透過窮舉搜索法(Exhaustive Search)來求解。然而窮舉搜索並非為一有效率的搜索演算法,為此,我們介紹一常見且高效率的捜索演算法:球體解碼(Sphere Decoding, SD)以及相對應的實數分解(Real Value Decomposition, RVD)。由模擬結果發現,當該SGDD存在有模糊性的的情況下,透過改變實數分解的順序,其位元錯誤率(Bit Error Rate, BER)可達2.5dB左右的差距。此外,我們也將我們提出來的架構延伸到在多輸入多輸出(Multiple-Input Multiple-Output, MIMO)系統下的空頻編碼(Space Frequency Block Code, SFBC)。在多輸入多輸出系統下,除了既有的資料模糊性之外,其特殊的編碼方式會額外造成碼的模糊性(Code Ambiguity)。經由數學推導,證實我們所提出的星座圖混搭機制可以同時解決資料模糊性以及碼的模糊性。最後,我們也透過位元錯誤率的電腦模擬結果,驗證我們的系統效能。 A low-complexity blind data detector is proposed in this research for orthogonal frequency division multiplexing (OFDM) systems, where the generalized likelihood ratio test (GLRT) approach is adopted. Traditional GLRT data detector (GDD) can be viewed as a combinatorial optimization problem, which suffers from prohibitively high computational complexity and signal ambiguity. In the current research, we demonstrate that the computational complexity can be reduced by properly decoupling the GDD into several sub-group GDDs (SGDD). Generally, the method of decoupling can be categorized into two major types, namely, interleaving-based and subband-based. In interleaving-based SGDD, we proposed a novel two-stage SGDD, in which the computational complexity is further reduced by exploiting the signal ambiguity, and then the signal ambiguity is solved by appending a single pilot tone. Mathematical analysis indicates that a reduction of M-1/M (or 3/4) computational complexity is achieved in the case of M-PSK (or M-QAM) without any performance loss. Subband-based SGDD, on the other hand, is not compatible with two-stage SGDD. However, it is shown that the computational complexity of sub-band-based SGDD can be exponentially decreased if the bandwidth of subband is smaller than channel coherence bandwidth. In addition, a mixed-constellation mechanism is proposed to mitigate the problem of signal ambiguity. Both types of SGDD can be solved by exhaustive search, which is a straightforward but inefficient solution. The sphere decoding (SD) algorithm with real value decomposition (RVD) process, which is one of the most popular solution to solve combinatorial optimization problem, is introduced and discussed in the following to suit with the proposed structure. It is shown that the detection performance can be improved up to 2.5dB by switching the detection order in RVD process when ambiguity exists. Furthermore, we extend our proposed scheme to a SFBC-OFDM system. Most of the SFBC are rotatable codes, which result in code ambiguity. It is shown that the proposed mixed-constellation mechanism is capable of mitigating signal ambiguity and code ambiguity simultaneously. Finally, simulation experiments are conducted to evaluate the bit error rate performance. |
URI: | http://tdr.lib.ntu.edu.tw/jspui/handle/123456789/19608 |
全文授權: | 未授權 |
顯示於系所單位: | 電信工程學研究所 |
文件中的檔案:
檔案 | 大小 | 格式 | |
---|---|---|---|
ntu-105-1.pdf 目前未授權公開取用 | 5.12 MB | Adobe PDF |
系統中的文件,除了特別指名其著作權條款之外,均受到著作權保護,並且保留所有的權利。