Cognitive Radio (CR) has emerged a potential solution which addresses the challenges of radio frequency (RF) spectrum scarcity, underutilization and inefficient utilization. Due to recent advancements in multimedia, social networking, wireless applications and mobile application technologies, the demand of RF spectrum has increased exponentially. The CR technology enables the RF spectrum to be shared amongst the licensed and unlicensed users in an intelligent and opportunistic manner. However, the conception of a shared spectrum evolution has introduced a number of unique challenges in wireless networking, requiring efficient novel techniques to transmit data in wireless networks. The focus of this paper is on the routing layer of the multi-hop cognitive radio based wireless mesh networks. Routing in multi-hop networks plays a vital role in the establishment and maintenance of communication routes. We analyze and evaluate the performance of the three candidate routing protocols, namely: Ad-Hoc On-Demand Distance Vector (AODV), Weight Cumulative Expected Transmission Time (WCETT) and the proposed Extended Weight Cumulative Expected Transmission Time (xWCETT). The three routing protocols are evaluated using the end-to-end delay, jitter, throughput, packet delivery ratio and normalized routing overhead metrics. We simulated the performance of the these protocols using NS2 with Cognitive Radio Cognitive Network (CRCN) patch. The numerical results obtained show that the proposed xWCETT protocol offered better performance in term of the end-to-end average latency, throughput and packet delivery ratio. It also attained the lowest routing overhead as compared to the AODV and CRCN WCETT, reserving more bandwidth for data transmission.
