A Resource Sharing Platform for Resource-Constrained Software Defined Cognitive Radio Networks

In this paper we introduce a novel collaborative spectrum sensing/resource sharing scheme that is employed in resource-constrained cognitive radio networks (CRNs) where there is a limited resource budget, and in the meanwhile each secondary user (SU) has a certain transmission cost. The problem is to decide the SUs that are allowed to transmit without exceeding the resource budget constraint. The proposed scheme fits well in the Software Defined Networking (SDN) paradigm as it can be reconfigured/reprogrammed to support a wide variety of (limited) resources, such as limited data rate, limited tolerable interference, etc. The proposed platform is based on modeling the CRN as a normal graph over which the belief propagation algorithm is run. Simulation results demonstrate many advantageous features such as high accuracy and minimal amount of message exchange.