NeTS: Medium: Collaborative Research: MIMO-Pipe Modeling Scheduling and Delay Analysis in Multi-hop MIMO Networks

Project: Research project

Project Details

Description

The multiple-input multiple-output (MIMO) technology is a major breakthrough in wireless communications, and it is widely believed that the MIMO technology has great potential to significantly improve the performance of multi-hop wireless networks. However, the existing approaches for multi-hop MIMO networks use simplistic abstraction by treating the number of antennas as degrees of freedom while ignoring the transmission reliability, and it remains open to obtain a rigorous understanding of reaping MIMO gains towards quality-of-serice (QoS) provisioning in multi-hop networks. One grand challenge is that multi-hop wireless networks are interference-limited and that the interference introduces coupling across various layers of the protocol stack, including the PHY (physical), MAC, network and transport layers. Clearly, the coupling calls for a synthesis of realistic MIMO link model abstraction and QoS provisioning algorithms. Intellectual merit: The fundamental differences between multi-hop networks and point-to-point settings indicate that leveraging MIMO gains in multi-hop networks requires a paradigm shift from high SNR regimes to interference-limited regimes. This project undertakes a broad research agenda centered around developing fundamental theory towards achieving optimal throughput and delay performance in multi-hop wireless networks. Under a common thread of MIMO-pipe scheduling, the proposed work pursues thorough studies of the following important problems: 1) Developing rate/reliability models for MIMO-pipes in multi-hop networks; 2) MIMO-pipe scheduling for throughput maximization; 3) Real-time scheduling of MIMO-pipes with delay constraints (for time-critical traffic); 4) Delay performance of MIMO-pipe scheduling; 5) Joint congestion control and MIMO-pipe scheduling. The first key step in this project is to take a bottom-up approach for solid model abstraction of MIMO links while taking into account interference, and to extract a set of feasible rate/reliability requirements, corresponding to meaningful MIMO stream configurations. Built on this novel MIMO-pipe model, a thorough investigation of the above challenging problems will be carried out. The study of these open issues clearly makes the proposed research transformative. Broad impacts: The PIs expect that the proposed work will culminate in the formulation of both new fundamental theories and overriding design principles for multi-hop MIMO networks, and will have direct impacts on many wireless applications, including WIMAX and 802.16-based mesh networks, vehicular ad-hoc networks (VANET) and wireless LANs (802.11n). Particularly, real-time scheduling will shed much light on leveraging MIMO gains in VANET and will significantly enhance VANET to deliver timely information reliably to save lives, reduce traffic congestion and improve quality of life. The PIs will establish an annual short-visit program to foster interactions between faculty and students at UIUC and ASU. Moreover, Graduate students participating in this project will get trained on a variety of subjects, ranging from wireless networks, stochastic optimization, combinatorics, distributed algorithms, to queuing analysis. For curriculum development of advanced courses, the PIs will make conscientious efforts to introduce new pedagogical methods, including guest lectures by visitors, replaying plenary talks, reading assignments of short-course lectures. The PIs will also continue and strengthen the efforts in mentoring Hispanic students and female students. Key Words: MIMO networks; scheduling; throughput maximization; real-time scheduling; delay analysis.
StatusFinished
Effective start/end date8/1/097/31/13

Funding

  • National Science Foundation (NSF): $400,000.00

Fingerprint Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.