Statistical delay control of opportunistic links in cognitive radio networks

Cognitive radio (CR) technology has been considered promising to enhance spectrum efficiency via opportunistic transmission at link level. To make CR useful, networking CRs to form a cognitive radio network (CRN) is able to support end-to-end transmission from CR source to CR destination. However, the opportunistic nature of CR link for interference avoidance to primary users degrades the quality-of-service (QoS) of end-to-end CR transmission and challenges the CRN toward a completely successful operation. Through queueing analysis, we propose a statistical control mechanism to deal with such opportunistic links in CRN by cooperative relaying the same packet flows into several opportunistic paths simultaneously. The availability and reliability of redundant transmission over the end-to-end paths in the same group is enhanced. By maximizing the number of groups with bounded statistical delay, the spectrum efficiency is enhanced. This optimized grouping problem is mathematically equivalent to the bin covering problem with NP-hard complexity. By the proposed Round-Robin algorithm, simulation results show that the optimal performance can be achieved in the case that the statistical availabilities of all opportunistic links are the same. This work therefore provides an essential viewpoint via cooperative relay among CRs in CRN, the QoS (i.e., average delay) can be guaranteed and spectrum can be efficiently utilized.