WiFiUS: Message and CSI Sharing for Cellular Interference Management with Backhaul Constraints

University Of Illinois At Urbana-Champaign, Urbana IL

Investigators

Abstract

The past decade has witnessed significant growth in the use of cellular networks, reflecting increased data demands. This has been the driver of recent research for new ways of managing interference in wireless networks. The focus of this project is to explore the potential benefits of exploiting anticipated infrastructure enhancements, particularly in terms of a more flexible use of the backhaul network connecting the base-stations, through the lens of information-theoretic models of single hop wireless networks. The goal is to show how centralized channel-aware decisions for assigning backhaul resources, by sharing messages and channel state information, can lead to rate gains and simplification of interference management schemes in network models that prove relevant to practial deployments. The project connects three research thrusts. The first is to study the fundamental limits of the rate of communication in large interference networks with a rate-limited backhaul that allows for sharing messages between base stations in both downlink and uplink scenarios. This study uses high level channel state information to identify the topology of the network and carry out a degrees of freedom based analysis and design of message sharing schemes. The second thrust builds on the first thrust to design distributed multi-antenna interference control and resource management schemes using more detailed channel state information and exchange. The third thrust addresses interference management in dynamic networks, accounting for dynamics in topology due to fading as well as due to users entering and leaving the cellular network. An important component of this thrust is the joint optimization of the rate-limited backhaul for message sharing and channel state information exchange. Research on interference management, coordination and related resource allocation will advance our understanding of how to significantly improve the energy and spectral efficiency of evolving wireless networks. While this project will advance the science and engineering of wireless networking, it should also impact the national and international wireless industry by developing new design principles and methodologies for the design and implementation of future wireless networks. By enabling more efficient use of wireless spectrum, such networks have the potential to improve broadband access across the world and provide the communication infrastructure for numerous new wireless services. Furthermore, these advances are necessary to enable the evolving internet of connected objects or the Internet of Things (IoT). Education initiatives include the involvement of undergraduate students in the research project; integration of the research results developed under this project into graduate and undergraduate curricula; presentation of tutorials on topics related to the project at international conferences, seasonal schools and WiFiUS meetings; and facilitating student exchanges between the two collaborating institutions.