Recovery in Channel-Hopping Cognitive Networks Under Random Primary-User Activity

We investigate the performance of a recovery mechanism in channel-hopping cognitive networks. The recovery algorithm is based on a list of backup channels to be used as alternatives if primary-user activity interrupts ongoing communication on the current channel; if all backup channels are exhausted without success, the nodes revert to a blind-rendezvous mechanism. We evaluate the performance of the recovery algorithm using the tools of probabilistic analysis and renewal theory and validate the results through simulation. We discuss the impact of various network and channel parameters on the performance of the recovery algorithm and show the importance of accurate sensing information for successful recovery.