Near-Optimal Opportunistic Spectrum Access via Imperfect Channel Sensing in Cognitive Radio Enabled Vehicular Network

The future of vehicular communications will incorporate cognitive radio technology within vehicles. Cognitive radio-enabled vehicular networks facilitate the opportunistic access of licensed spectrum by vehicles equipped with cognitive radio technology on highways. This study analyzes a cognitive radio-enabled vehicular network including M primary users (transmitters) and K secondary receivers, with the condition that K < M. A channel is designated for each primary user to facilitate data transmission as needed. The secondary user has a backlog of data. It lacks knowledge of the states (either good or bad) of M channels, as well as the statistics of the channel evolution processes. It chooses K from M channels for transmission in each time slot. Upon detecting that a selected channel is in an optimal state, the secondary user employs that channel for data transmission in the designated time slot. Otherwise, the secondary user will choose an alternative channel in following time slot. This problem has been examined under average reward criteria for both finite and infinite time horizons. This study presents the Uniforming Random Ordered Policy with Imperfect Sensing (UROP-IS) and shows that UROP-IS achieves near-optimal throughput, assuming block fading in generic channel evolution processes. We numerically demonstrate that the proposed approach achieve near-optimal throughput, surpassing 80% of optimal throughput in scenarios with a restricted number of available channels, while MP can attain below 75% of the optimal throughput in comparable conditions.