A novel coupled queueing model to control traffic via QoS-aware collision pricing in cognitive radio networks

We consider a cognitive radio network, where primary users have priority over the spectrum resources, and secondary users can exploit the unused resources through channel sensing. Due to sensing inaccuracy, the secondary traffic may obstruct the primary traffic. A penalty for collision has been used to protect the primary traffic, which is often designed to provide a fixed per-collision compensation or to restrict the collision rate at an acceptable level. In this work, we develop a framework that can protect the primary traffic taking into account the Quality of Service of the primary traffic. In particular, we pay attention to the delay performance, which is determined not only by the collision rate but also by the amount of traffic in both networks. We design a novel model with coupled queues, and successfully incorporate dynamic interactions between the two systems through the standard optimization problem. We also consider the practical requirement of no direct sharing of the system information between the two networks, and develop a close-to-optimal solution of per-collision price and channel sensing under mild assumptions. We evaluate its performance through simulations.