Frameworks to enhance performance in precoded multi-user MIMO systems
Abstract
Multiple-input multiple-output (MIMO) wireless communication systems, where both the transmitter and receiver have multiple antennas, have been a major factor in facilitating the proliferation of high-data rate applications in wireless systems. Compared to single-antenna systems, these MIMO systems are able to exploit the extra spatial dimensions in order to provide multiplexing gains for higher data rates, and diversity gains for better reliability. MIMO system technology can presently be found in the IEEE 802.11n standard for wireless local area networks, and is also a building component for future-generation (3.5G, 4G and beyond) wireless broadband communication standards such as LTE and WiMAX. In single cell multi-user MIMO (MU-MIMO) communication, the base station (BS) communicates with multiple mobile station (MS) users. In downlink transmission, the BS transmits independent data streams to each of the MSs. However, because this transmission utilizes the same frequency resource for all users, each user experiences some distortion in the form of intra-cell interference (ICI), as well as additive white Gaussian noise (AWGN). In multicell MU-MIMO, the interference effect is further complicated by other-cell interference (OCI). In this work, we develop techniques and algorithms to enhance the performance of precoded multiuser MIMO systems. First, we consider the single-cell precoded multiuser MIMO system, which is severely limited due to ICI. Moreover, it is very difficult for users in MU-MIMO to obtain the interference statistics without user cooperation. For this system, we present a technique through which each user is able to obtain the statistics of the interference that it experiences. By doing so, optimal detection of the data streams through maximum-likelihood detection can be achieved. Without this statistic, the users would perform sub-optimal minimum-distance decoding. In addition, we develop a low-complexity perturbation-codebook decoder that attempts to mitigate the effects of both interference and noise on the received data streams, and this decoder is able to achieve near-optimal performance. Secondly, we extend consideration to the multi-cell multiuser MIMO system, which, apart from IUI, also includes additional interference in the form of other-cell interference (OCI). We consider the case where the cells are grouped into cooperative clusters, and the BSs within the cells are allowed to share limited-rate information. Usually, the cell-edge users within the multi-cell multiuser MIMO system experience the most interference. Thus, the sum-rate that is achievable by the cell-edge users is much less than that which is achievable by inner-cell users. Hence, the cluster sum-rate is severely reduced. We develop an algorithm to improve the cell-throughput and cluster-throughput by enhancing the cell-edge user performance. This algorithm provides a much suitable sum-rate for the inner-cell users and improves the cluster sum-rate over other conventional algorithms. In addition, this algorithm requires significantly less feedback than is required for the standard Precoding Matrix Index restriction (PMI-Restriction) algorithm. Thirdly, we consider the problem of sharing channel state information (CSI) in multiuser MIMO systems. More importantly, we consider the problem of attaining global CSI knowledge, where each user in the system knows the channel between every pair of nodes in the system. We leverage this global CSI knowledge to improve performance in multiuser MIMO systems by considering its applications in cooperative transmission and decoding. We utilize an algorithm that is able to significantly reduce the training time in which all nodes need to learn global CSI knowledge. We also develop decoding algorithms that use this global CSI knowledge to improve decoder performance for limited-cooperation multiuser MIMO systems. We are able to show the performance of the algorithms described above through simulations, and detail how these algorithms enhance the performance of precoded multi-user MIMO systems for both the single-cell and multi-cell.
Degree
Ph.D.
Advisors
Krogmeier, Purdue University.
Subject Area
Electrical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.