In recent years, the use of multiple antennas at both transmitter and receiver ends of a communication link has been identified and widely studied as the most practical method of increasing channel capacity. For the next generation cellular and wireless local area networks, multiple-input multiple-output (MIMO) technology that employs multiple antennas is envisioned to be the core technology to achieve higher data rates. While
MIMO technology promises significant information-theoretic capacity gain for wireless links, there are still many unknowns as to how to efficiently realize such gains in practical communication systems and networks. There are a number of key issues from the physical layer to the network layer that need to be addressed. This proposal aims to take a cross-layer approach to address these issues to facilitate efficient utilization of multiple antennas in wireless networks. MIMO detection is the most fundamental issue in MIMO networking, and it is the complexity bottleneck that limits the employment of MIMO technology. A main objective of this proposal is to develop low-complexity MIMO detectors that scale well with antenna number and modulation size so that it is applicable for practical network setting. We propose a novel MIMO detector based on the Monte-Carlo Markov chain (MCMC) approach which shows performance superior to other existing MIMO detectors at a complexity that is more than one order of magnitude less. Performance analysis of the MCMC detector and its impact on code design will be investigated. One primary focus is on joint optimization of channel codes with MIMO detection for large antenna systems. We plan to develop joint coding design and detection strategies to find capacity-approaching channel codes at high spectral efficiencies. Code design criterions will be derived for short channel codes that are suitable for delay-sensitive applications.
From the network layer this proposal addresses the issues of multiple access and resource allocation for MIMO wireless networks. A central issue in these designs lies in the amount of channel state information (CSI) available at the transmitter. We propose to study practical power control and scheduling algorithms that are robust to channel variations and have the capability of supporting limited CSI. We will address fairness and quality of service for users with heterogeneous channel conditions. Optimal signaling design for practical MIMO detectors will be investigated in order to maximize network throughput and minimize multi-user interference. These are closely tied with our study on the physical layer issues of MIMO detection and coding.
Broader Impact: The educational plan of this project offers diverse opportunities to students at all levels. The proposed research will generate a cluster of undergraduate research projects, which in particular will focus on developing a multiple-antenna test bed for wireless local area networks. The PI plans to encourage students from under-represented groups to participate in such projects. The proposed research may generate industrial interest that can result in undergraduate industry sponsored projects. The proposed research will also attract graduate students to explore modern communication theory and encourage their future careers in this exciting field. The PI plans to develop a new course on software-defined radio (SDR) for wireless communications, and a more advanced course on cross-layer design for wireless networks at the graduate level.
Intellectual merit: The intellectual merit of this proposal lies in the development of new techniques and theories in MIMO networking. The broader impact is in the interdisciplinary dimensions of this research, as well as in the educational program and the exposure of students in all levels to the proposed areas. This plan offers strong integration of the research with education and industry. The impact of this research is expected to be on a variety of fields, including coding and information theory, signal detection and estimation, algorithm design and complexity, network protocol design and more.
