大型天線陣列(Massive-MIMO)為5G行動通訊系統中重要技術之一，藉由基地台擺放大量天線與線性預編碼(例如: 迫零預編碼, Zero-Forcing Precoding)來大幅提升通道傳輸量。但在實際FDD系統中，基地台執行線性預編碼計算前用戶須回報與基地台端天線數正相關之通道資訊(Channel State Information, CSI)量，造成實現上之困難。因此，相關研究提出一整合空間分割及多工(Joint Spatial Division and Multiplexing, JSDM)之預編碼技術來降低高用戶通道回饋量與高預編碼計算複雜度，該技術透過用戶分組，並結合預波束集成(prebeamforming)及傳統迫零預編碼技術來實現。
文獻提出之整合空間分割及多工預編碼技術大多假設群組內用戶存在相同的天線通道相關性，並藉此來設計預波束集成矩陣。但實際通信環境中，用戶分散不同地方，即使同一群組(或同一beam內)其天線通道相關性也無法相同，而用戶分組即為應用JSDM技術的重要議題。有鑑於此，本論文將整合空間分割及多工預編碼應用至實際用戶分佈環境，且討論適應性預波束集成矩陣與固定預波束集成矩陣設計及其對應之用戶分組演算法，並探討與比較不同做法間之複雜度、通道資訊回饋量之降低與系統效能。

Massive MIMO is one of the important technologies in 5G mobile communication system development. System can get much spectrum efficiency improvement by placing a great number of antennas at base station (BS) and using liner precoding scheme (ex. Zero-Forcing Precoding).
But, in practice, base station needs the channel state information (CSI) from all users to carry out the liner precoding process and it will cause dramatic feedback overhead in FDD systems. Previous literature proposed a Joint Spatial Division and Multiplexing (JSDM) precoding scheme, which combined the prebeamforming and conventional Zero-Forcing precoding, to reduce CSI feedback overhead and the complexity for precoding matrices computation.
Most literature related to the JSDM precoding scheme usually assumed that all users in the same group or same beam have identical antenna correlation matrices and design the prebeamforming matrices based on this assumption. But in real communication scenario, users have different antenna correlation matrices even in same the group due to their different geographic locations. Hence, user grouping algorithm based on the similarity between the antenna correlation matrices becomes an important issue in JSDM precoding scheme.
In this thesis, the user grouping algorithms for the JSDM precoding scheme with both the adaptive prebeamforming and fixed prebeamforming schemes are studied and proposed to apply on the practical user distribution scenario. The numerical results show the comparison of the performance and complexity among different grouping algorithms.

[1] F. Rusek, D. Persson, B. Lau, E. Larsson, T. 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, 2013.
[2] A. Adhikary, J. Nam, J. Ahn and G. Caire, "Joint spatial division and multiplexing-the large-scale array regime", IEEE Trans. Inf. Theory, vol. 59, no. 10, pp. 6441-6463, 2013.
[3] D. Shiu, G. Foschini, M. Gans, and J. Kahn, “Fading correlation and its effect on the capacity of multielement antenna systems,” IEEE Trans. on Commun., vol. 48, no. 3, pp. 502–513, 2000.
[4] L. Lu, G. Y. Li, A. L. Swindlehurst, A. Ashikhmin and R. Zhang, "An overview of massive MIMO: Benefits and challenges", IEEE J. Sel. Areas Commun., vol. 8, no. 5, pp. 742-758, 2014.
[5] X. Mestre, “Improved estimation of eigenvalues and eigenvectors of covariance matrices using their sample estimates,” IEEE Trans. On Inform. Theory, vol. 54, no. 11, pp. 5113–5129, 2008.
[6] J. Nam, A. Adhikary, J. Ahn and G. Caire, "Joint spatial division and multiplexing: Opportunistic beamforming, user grouping and simplified downlink scheduling", IEEE J. Sel. Areas Commun., vol. 8, no. 5, pp. 876-890, 2014.
[7] A. Barg and D. Y. Nogin, "Bounds on packings of spheresin the Grassmann manifold", IEEE Trans. Inf.Theory, vol. 48, no. 9, pp. 2450-2454, 2002.
[8] T. Y. Young and T.W. Calvert, Classification, Estimation and PatternRecognition. New York, NY, USA: American Elsevier, 1974.
[9] S. Lloyd, “Least squares quantization in PCM,” IEEE Trans. Inf.Theory, vol. 28, no. 2, pp. 129–137, Mar. 1982.
[10] MacQueen, J. B. (1967). Some Methods for classification and Analysis of Multivariate Observatio
[11] Rui Xu, Donald Wansch II, "Survey of Clustering Algorithms", IEEE Trans. On Neural Networks, Vol. 16, NO. 3, May 2005.