運用於上行多點協作通訊下之干擾對齊技術與可彈性規劃之特徵值運算架構設計;An Interference Alignment (IA) Technique for Uplink Coordinated Multi-Point (CoMP) and Eigen-solver for Rank Deficiency Matrix

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Electrical Engineering > Electronic Thesis & Dissertation > Item 987654321/66901

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/66901

Title:	運用於上行多點協作通訊下之干擾對齊技術與可彈性規劃之特徵值運算架構設計;An Interference Alignment (IA) Technique for Uplink Coordinated Multi-Point (CoMP) and Eigen-solver for Rank Deficiency Matrix
Authors:	艾美莉;Rakhmania,Amalia Eka
Contributors:	電機工程學系
Keywords:	干擾對齊;特徵值分解;多點協調;interference alignment;eigenvalue decomposition;Coordinated Multi-Point (CoMP)
Date:	2015-01-15
Issue Date:	2015-03-16 15:58:56 (UTC+8)
Publisher:	國立中央大學
Abstract:	干擾對齊技術可消除無線通訊系統中的干擾源，若結合LTE-A系統中多點協作技術則可以提高蜂胞邊緣使用者的傳輸率與通訊效能。在本論文中，我們提出適用於上行鏈結多點協作系統中之新的干擾對齊技術，在發送端的預編碼器採用最大化每一用戶之信號與洩漏加雜訊能量比(signal-to-leakage-and-noise ratio, SLNR)為設計準則，另一方面，解碼器端則採用以用戶考量為出發點的最大化訊號與干擾加雜訊能量比(signal-to-interference-and-noise ratio, SINR)為設計準則。透過兩者相結合的運算方式使得通道互易性(channel reciprocity)的限制可被去除，因此特別適合操作於上行使用者具有不同發送功率的情形。透過檢視遞迴疊代的過程，以每一用戶為基礎的預編碼與解碼器設計準則可有效抑制同一用戶之天線間干擾，因而比起傳統的干擾對齊演算法，像是最小加權洩漏、以空間串流為基準之最大化訊號與干擾加雜訊能量比或是混和型演算法，具有較佳之性能。在各式干擾對齊演算法中，特徵值分解(eigenvalue decomposition, EVD)扮演著關鍵的角色。在我們提出的方法中，需要計算信號的特徵值子空間，相反的，最小加權洩漏演算法需要計算信號的特徵值零空間，而所需處理的皆為矩陣秩數不足的情況。因此我們基於QR分解(QRD)提出一一致化且具有彈性之特徵值分解運算架構，首先，將矩陣轉換為Hessenberg型式以降低其後的運算複雜度。其次，根據欲求取零空間或是非零特徵值而利用不同的QR分解方式進行運算。我們所提出的方法架構可適用於不同的矩陣大小，相較於滿秩矩陣，所提出的方法可有效解省運算複雜度，並可得到正確的特徵值運算結果。;Interference alignment (IA) is a technique to eliminate the interference in wireless communication system. Combined with coordinated multi-point (CoMP), this method could improve the system sum rate performance for cell edge user in LTE-A system. In this thesis, a new IA algorithm for uplink coordinated multi-point (CoMP) is proposed. The unselfish per-user signal-to-leakage-and-noise ratio (SLNR) criterion is used to design the precoder. On the other hand, the design of decoder adopts the selfish algorithm, per-user signal-to-interference-and-noise-ratio (SINR). The combination of both method does not need the channel reciprocity assumption and thus is suitable to operate in the case of different user transmission power. Through iterative procedure, we show that the per-user-based criterion which keeps user data streams orthogonal can suppress interference effectively and achieve higher sum rate than the conventional IA algorithms, such as minimum weighted leakage interference (min leakage), maximum per-stream SINR algorithms, and the hybrid IA in the multi-user CoMP joint reception scenarios. Eigenvalue decomposition (EVD) plays a key role for our proposed algorithm as well as the conventional algorithm, min leagake method. In our proposed method, EVD is needed to compute the signal subspaces. On the contrary, min leakage method needs the calculation of interference subspace. The decomposed matrix of both IA methods is always rank-deficient. A new eigen-solver based on QR decomposition (QRD) with shift is presented. Hessenberg reduction is implemented in the first stage to reduce the computation complexity. The proposed method could find the correct eigenpairs for both full rank and rank deficient matrix. The architecture of proposed method is shown to be more flexible for any matrix size and has less complexity than the existing method. The proposed eigen-solver could save up to 92% hardware complexity than the conventional EVD to find the nullspacefor hermitian symmetric matrix when the rank of the matrix is quite small . This method is reliable to be implemented due to its equal performance compared to MATLAB “eig” function.
Appears in Collections:	[Graduate Institute of Electrical Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	588	View/Open

社群 sharing

Loading...