Hierarchical Spectrum Sharing in Cognitive Radio: A Microeconomic Approach

We consider the problem of hierarchical spectrum sharing in cognitive radio environment. In the system model under consideration, licensed service (i.e., primary service) can share/sell available spectrum to an unlicensed service (i.e., secondary service), and again, this unlicensed service can share/sell allocated spectrum to other service (i.e., tertiary service). We formulate the problem of hierarchical spectrum sharing as an interrelated market model in which a multiple-level market is established among the primary, secondary, and tertiary services. We use the concept of demand and supply functions in economics to obtain the partial equilibrium for which all services are satisfied with the shared spectrum size and the charging price. These functions are derived based on the utility of the connections using the different services. In addition, we consider a system for which the global information is not available. Therefore, each service needs to learn and adapt the strategies to reach an equilibrium. Two iterative algorithms (i.e., excess demand-based and successive overrelaxation (SOR)) are proposed. The stability condition for the learning rate is analyzed for these algorithms.