Cognitive radio networks with elastic traffic

Abstract

We model and evaluate an infrastructure based cognitive radio network where impatient unlicensed or secondary users (SUs) offer elastic traffic. The interference created by SUs to the licensed users is analyzed when the secondary network uses two different spectrum access schemes, the conventional random access scheme and a new scheme we refer to as access with preference. To further control the interference, a limit is set to the number of channels SUs have access to. With this new constraint, the abandonment probability, the mean delay and throughput of the SUs degrades significantly. To improve the QoS perceived by the active SUs, we define and evaluate an admission control scheme for the SUs.