This project studies a paradigm in which nodes cooperate by pooling power and bandwidth resources and where flows interact opportunistically to avoid interference and increase network utilization. The PIs will leverage their existing expertise in cooperative and opportunistic communications to analyze the implications for broader networks of communication nodes. In particular, they will instantiate their design philosophy in three ways:
Node Information Management: While previous network analyses considered only isolated aspects of a node (e.g., channel gain), the project studies a comprehensive network state information, which captures not only physical-layer conditions but also higher-layer information such as queue state, processing power, and availability of forwarding routes.
Novel Network Representations: Instead of regarding the network as a simple connectivity graph, the PIs will introduce and develop a network representation which incorporates both temporal and spatial relationships between nodes. The PIs refer to this as the trellis representation of the network, and it will enable us to describe cooperative and opportunistic communication in a wide area network. The trellis will provide a structure in which to identify opportunities for physical layer cooperation, determine the impact of cooperation on neighboring nodes and flows, and opportunistically schedule and route competing flows at fine grained time scales.
Distributed Cooperative Discovery: Traditional discovery protocols for determining network connectivity are unable to identify cooperative links. New techniques will be developed that leverage existing discovery protocols to efficiently locate potential cooperative topologies, which are a key to opportunistic communication. These discovery protocols will recover network state information and enable the use of the trellis representation to identify the optimal cooperative route through the network.
With these tools, the PIs will develop and analyze protocols for coordinating cooperative and opportunistic communications in heterogeneous networks. The new protocols will expand access in underserved areas while increasing throughput in existing networks.
This research will have a broad impact on education by engaging undergraduate and graduate students in the Rice Center for Multimedia Communication (CMC) laboratory. Cooperative communication will be integrated into several courses at Rice in the wireless communication and networking areas. Software and firmware modules as well as publications will be distributed through the WARP open-access repository (http://warp.rice.edu/trac).
