NeTS: Small: Dynamic Spectrum Access under Uncertainty: Theory, Algorithm Development, and Evaluation

Ohio State University, The, Columbus OH

Investigators

Abstract

The demand for wireless spectrum is projected to continue growing well into the future, and will only worsen the currently felt spectrum crunch. Rigid licensing policies that give exclusive and permanent right of use of wireless spectrum to lessees further exacerbate this scarcity. This shortcoming has been identified in the famous 2002 FCC study, which estimates the utilization of licensed spectrum between 15-85% depending on time and location, thus, underscoring the critical need for new methods of spectrum sharing. Development of these new methods, coined as Dynamic Spectrum Access (DSA) techniques, is very challenging due to the inherent uncertainty in user traffic demand, spectrum availability, wireless channel conditions, and user locations. Thus, the overarching goal of this project is to efficiently manage dynamic spectrum access in the presence of these uncertainties. The algorithms developed in this project ultimately encourage both federal and commercial spectrum holders to participate in DSA systems. Additional wireless bandwidth is being freed up for essential services to be migrated to the wireless domain, significantly lowering the cost of access to wireless networks for a significant fraction of the society currently shut out of this market. These emerging systems also create new communication-based business models, develop community resources, and improve public safety. Managing dynamic spectrum access faces three major challenges induced by: (1) the dynamics and the possibly correlated nature of spectrum resource from a secondary provider/user's perspective; (2) uncertainty about spectrum availability in terms of long-term channel statistics and real-time channel states; (3) uncertainty of secondary traffic and heterogeneous performance/pricing requirements of secondary users. In this project, efficient information sharing, spectrum sensing, and scheduling policies are designed that take all these three aspects into account. Since jointly optimizing across the three dimensions outlined above is a daunting challenge, the project is organized across two inter-related thrusts. In the first thrust, a given level of spectrum uncertainty is assumed, and efficient scheduling policies are designed for a secondary provider to meet various QoS requirements. In the second thrust, the case where a secondary provider can control information inaccuracy by coordinating SUs to sense channels is considered and the joint sensing and scheduling problem investigated. The developed algorithms are validated through simulations and via testbed implementations. The effect of uncertainties is investigated by developing analytical techniques that combine stochastic optimization, approximation algorithms and game theory. The resulting joint sensing and resource allocation policies are low-complexity and provably efficient. This project will engage underrepresented students and K-12 students.