適用於毫米波多輸入多輸出通訊系統下以壓縮感知輔助之低複雜度波束域混和預編碼器演算法設計

Compressive Sensing (CS)-Assisted Low-Complexity Beamspace Hybrid Precoding for Millimeter-Wave MIMO Communication Systems

陳敬恒(Chiang-Hen Chen)

指導教授 ： 吳安宇

國立臺灣大學/電機資訊學院/電子工程學研究所/碩士(2016年)

https://doi.org/10.6342/NTU201600782

摘要

下世代通訊系統期待能提供超越4G通訊系統1000倍的傳輸效益。由於下世代通訊期望能支援如車對車 (vehicle-to-vehicle)、等應用提升容量，而透過對預編碼來做波束成形，降低干擾以提升傳輸效益是一種廣泛被使用以提升容量之技巧。在毫米波通道下雖然採用大量數量的天線陣列做波束成形能克服毫米波環境下的高路徑損耗，但採用傳統全數位與編碼技術來做波束成形的同時也增加了射頻鏈 (RF chain) 的成本。因此，有研究提出以混合射頻與基頻之預編碼技術來降低毫米波多輸入多輸出系統下的射頻硬體成本。然而，在設計混合式預編碼器前，大規模的天線數目使得欲拆解之全數位預編碼器變得更難獲得。此外，對於傳統的混合式預編碼器設計方法而言，會需要逆矩陣運算來計算基頻預編碼器。因此本篇論文將著重於適用於毫米波多輸入多輸出通訊系統下以壓縮感知 (Compressive Sensing) 輔助之低複雜度波束域混和預編碼器演算法設計。
本論文中，我們首先利用毫米波通道稀疏的特性，在基於壓縮感知通道估測所獲得的低維度波束域通道資訊下提出了一個低複雜度的全數位預編碼器獲得方法。此方法被稱之為波束域奇異值分解，可以降低目前相關文獻中使用全維度奇異值分解獲得之全數位預編碼器得複雜度達99.4%，且具有相同效能。同時，為了避免計算基頻預編碼器時的反矩陣運算，我們提出透過所提出正交的離散傅立葉轉換矩陣來當作波束成形的基底。然而，採用傳統的混合式預編碼矩陣設計方法在採用此基底挑選射頻預編碼器時會產生大量的運算量，因此我們提出了一個低複雜度的壓縮感知輔助之波束域混合預編碼器設計方法。透過結合所提出之低複雜度全數位預編碼器獲得方法與正交的離散傅立葉轉換矩陣，本方法避免傳統方法中挑選射頻預編碼器挑選時產生巨大運算量，同時又可避免反矩陣預算，所提出的壓縮感知輔助之波束域混合預編碼器設計方法可降低目前現有文獻中方法的複雜度達98.5%，同時達到低於5%全數位預編碼器效能損耗的效果。

關鍵字

多輸入多輸出 ； 毫米波通訊 ； 混合預編碼器 ； 波束成形 ； 低複雜度

並列摘要

The next-generation communication systems expect a 1000x times capacity leap compared with nowadays 4G communication systems to support plenty of applications such as vehicle-to-vehicle communication. Beamforming by precoding to mitigate the interference is a widely used technique to increase capacity. Although adopting large antenna arrays can overcome huge path loss of millimeter-wave (mmWave) multiple-input multiple-output (MIMO) channel, it also increases the RF chain cost over conventional full-digital precoding approach. Hence, hybrid analog/digital precoding techniques are proposed to reduce the hardware cost of RF chains in mmWave MIMO systems. However, before hybrid precoder design, large antenna dimension of makes it difficult to acquire the optimal full-digital precoder. Moreover, it also requires matrix inverse, which leads to high complexity for designing the hybrid precoder.
In this thesis, by exploiting the sparse characteristics of mmWave channel, we propose a low-complexity optimal full-digital precoder acquisition algorithm, named beamspace-SVD, which reduced the complexity of the one that is acquired by full-dimension SVD by 99.4% while retains same performance. This algorithm is proposed based on the reduced-dimension beamspace channel state information (CSI) given by Compressive Sensing (CS)-based channel estimators. Then, to avoid matrix inversion for calculating the baseband precoder, we first propose using orthogonal DFT matrix as beamforming basis. Moreover, we propose a CS-assisted beamspace hybrid precoding (CS-BHP) algorithm to avoid tremendous computation overhead when selecting the RF beamforming vector while also avoid the matrix inversion. By collaborating proposed beamspace-SVD with orthogonal DFT matrix, the proposed CS-BHP reduces the complexity of the state-of-the-art design by 98.5% with less than 5% performance loss of optimal full-digital precoder.

並列關鍵字

MIMO ； Millimeter-Wave Communication ； Hybrid Precoding ； Beamforming ； Low-Complexity

參考文獻

[1] Q. Li, H. Niu, A. Papathanassiou, and G. Wu, "5G Network Capacity: Key Elements and Technologies," IEEE Vehicular Technology Magazine, vol.9, no.1, pp.71-78, Mar. 2014.

[2] D.J. Love, R.W. Heath, Jr., "Limited feedback unitary precoding for spatial multiplexing systems," IEEE Trans. Inf. Theory, vol.51, no.8, pp.2967-2976, Aug. 2005.

[3] S. Yong and C. Chong, “An overview of multigigabit wireless through millimeter wave technology: potentials and technical challenges,” EURASIP J. Wireless Comm. Netw., vol. 2007, no. 1, pp.50-50, 2007.

[4] T.S. Rappaport, S. Sun. R. Mayzus, H. Zhao, Y. Azar, K. Wang, G. Wong, J. Schulz, M. Samimi, F. Gutierrez, "Millimeter wave mobile communications for 5G cellular: It will work!," IEEE Access, vol.1, pp.335-349, May 2013.

[5] F. Rusek, D. Persson, B. K. Lau, E.G. Larsson, T.L. Marzetta, O. Edfors, and F. Tufvesson, "Scaling up MIMO: opportunities and challenges with very large arrays," IEEE Signal Process. Mag., vol.30, no.1, pp.40-60, Jan. 2013.