Robust routing and link scheduling in cognitive radio networks under channel uncertainty and demand fluctuations

Abstract

The ability to deal efficiently with uncertainty is one of the fundamental requirements in designing and planning cognitive radio networks (CRNs). However, most of the optimization models proposed so far in the literature assume fixed, often worst-case, problem parameters, which may often result in underutilization of network resources. In this paper, we introduce an optimization approach based on two-stage robust optimization to handle channel uncertainty faced by a CRN operator due to licensed user transmissions. In particular, we formulate two strategies to adapt routing and multi-channel link scheduling in a multihop CRN to current channel occupation with the aim of maximizing the revenue generated from traffic demands. In addition, we study how revenue can be further increased by considering demand fluctuations rather than provisioning peak demand rates. Results show that our approach leads to significant gains compared to nominal models depending on the uncertainty range.