Analysis of spectrum adaptation in cognitive radio networks with elastic traffic and buffering

In this paper, the performance of spectrum adaptation in cognitive radio networks (CRNs) with elastic traffic is analytically evaluated. Contrary to most of the previously published related works on spectrum adaptation in CRNs, in this paper both call buffering and unencumbered service time with non-exponential probability distribution are considered. To reduce the negative impact of the interruption of ongoing calls of secondary users (SUs) due to the arrival of primary users (PUs) on the quality of service and to exploit the delay tolerance of elastic traffic, call buffering is employed. Although in cognitive radio networks with elastic traffic and spectrum adaptation there is a tradeoff among blocking/interruption probability and transmission delay, transmission delay has been not previously evaluated as performance metric in these systems. The developed mathematical model additionally considers both impatient time and non-homogeneous bandwidth requirement (i.e., different number of channels utilized) for PUs and SUs. System performance is evaluated in terms of new call and forced call termination probabilities, throughput, and transmission delay. Different probability distributions of the unencumbered service time are considered.