This research involves one of the key challenges in wireless networks: how to take advantage of new advances at the physical layer to push for new design at the upper layers. The approach used is hitchhiking-based cooperative communication (CC) that takes advantage of the physical layer design that facilitates the combining of partial information. A node can receive several partial signals and combine these signals to retrieve the complete signal. Through effective use of partial signals, a packet can be delivered with fewer nodes and/or less transmission power at each node. The investigator proposes a new weighted graph model that can capture the nature of CC. One key concept proposed is the new notion of ?link? and ?path? on which other graph terminologies can be defined.
Based on this new graph model, the research focuses on two types of power-efficient design under CC: (1) Power saving protocols that put wireless nodes into periodical sleep states while maintaining global ?domination? of active nodes. (2) Power control for transmission energy consumption by adjusting transmission ranges while maintaining global ?connectivity?. With these, this research presents a promising and unique way of applying this graph model to energy-efficient design in wireless networks. This research also involves the design of a general methodology of localized solutions and applies it to address various energy-related optimization problems under CC. The central theme of this research fits well with the objective of the SING program on fundamental theoretical and algorithmic studies involving coordination and communication. The insights and results of this research are expected to provide guidelines for energy-efficiency for a wide range of wireless network applications.
In an early NSF project dated back to 2003, we proposed hitchhiking-based cooperative communication (CC) that takes advantage of the physical layer design that facilitates the combining of partial information. A node can receive several partial signals and combine these signals to retrieve the complete signal. Through the effective use of partial signals, a packet can be delivered with fewer nodes and/or less transmission power at each node. This project focuses on a new weighted graph model that can capture the nature of CC. One key concept we introduce is the new notion of ?link? and ?path? on which other graph terminologies can be defined. Based on this new graph model, we systematically study two types of power-efficient designs under CC: (1) Power saving protocols that put wireless nodes into periodical sleep states while maintaining global ?domination? of active nodes. (2) Power control for transmission energy consumption by adjusting transmission ranges while maintaining global ?connectivity?. The proposed graph theoretic model is the first for CC that is novel in its way of defining ?link? and ?path? under both static and dynamic frameworks. The proposal presents a promising and unique way of applying this general graph model to an energy-efficient design in wireless networks. The proposal also studies efficient solutions to various energy-related optimization problems through a proposed methodology for localized solutions, in which nodes make decisions based only on their local surrounding information. Finally, the proposed graph theoretic model will enrich the general body of network science. We envision that insights and results from this research will provide guidelines for energy-efficiency for a wide range of wireless network applications. This research will also exploit and contribute to fundamental theories on topology control and virtual backbone under the new graph abstraction.
