As cognitive radio (CR) technology becomes more mature in the near future, it is not hard to foresee that CR will eventually play a pivotal role in wireless networking, such as mesh and ad hoc networks. However, due to its unique characteristics in spectrum usage, a CR network differs significantly from the existing multi-channel multi-radio (MCMR) network. As a result, networking problems for CR networks are much more challenging and interesting.
A major objective of this project is to develop network theoretical bounds and performance limits for future cognitive radio networks. Such efforts are not only of theoretical interest, but also offer fundamental understanding of the potential and limits of such networks, as well as provide performance benchmarks for the design and evaluation of distributed algorithms and protocols. New performance metrics such as Bandwidth-Footprint Product (BFP), which is unique to CR, will be employed, along with traditional performance metrics (e.g., network capacity). Special consideration will be given to interference modeling, power control, scheduling, and routing. An analytical framework will be developed that allows one to unify the multi-dimensional cross-layer design space into a tractable, understandable research problem. We will develop efficient computational procedures by employing powerful optimization techniques for given network instances. In addition to theoretical study of performance bounds, we also plan to develop distributed algorithms for optimal radio resource utilization.
The educational objective of this project is to develop teaching materials for future CR-based wireless networks. The investigator aims to develop a new course that intersects networking, communications, and software curricula in both the ECE and CS departments. Special efforts will be made to involve students from underrepresented groups in the form of wireless summer school to encourage participation in this project.
