Equilibrium Analysis and Social Optimization in Cognitive Radio Networks

This paper investigates the admission control problem in cognitive radio networks (CRNs) from the economic viewpoint. Each secondary user (SU) has the option of accessing the spectrum or balking on account of the trade-off between the service profit and the delay cost. In order to avoid disturbing the transmission of primary users (PUs), the secondary user requires detecting the presence of PUs before starting its own communication. In addition, whenever a PU appears on the working channel, the SU must halt its transmission to guarantee the communication quality of PU. Therefore, regarding the actions of PU and SU, the licensed channel can be seen as a server switching between two working modes ON/OFF and each SU's data packet, call or connection as a customer. We propose a partially observable M/M/1 queueing model to describe the decision process of SUs. Based on the queueing model, we present two admission control strategies to maximize the welfare of individual SU and the whole society respectively.