Optimal spectrum utilisation in cognitive network using combined spectrum sharing approach: overlay, underlay and trading

Cognitive radio technology enables unlicensed users (secondary users, SUs) to access the unused licensed spectrum which is allocated to primary users (PUs). In the literature, there are three traditional spectrum sharing paradigms that enable SUs to access the licensed spectrum. These access techniques include underlay, overlay and spectrum trading, and have their own drawbacks. To combat these drawbacks, we propose a new approach for each of them and merge them into one combined complete distributed system for cognitive network that contains all cognitive network functions. Our overlay scheme provides quick access to the unused spectrum because of the advantages of a distributed system where there is no contention for a central resource. We propose a new cooperative sensing protocol to enable the overlay scheme to reduce the likelihood of interfering with PUs and to avoid service interruption. We propose using our underlay scheme when the traffic loads at PUs are high that enables SUs to transmit simultaneously with PUs. The scheme uses reinforcement learning (RL) to manage the power of SUs and to protect PUs against harmful interference. Our trading scheme allows PUs to trade the unused spectrum for the SUs that require better quality of service. RL is used to control the size and the price for the rented spectrum. The new combined scheme increases the size of spectrum in the cognitive network because of using different access techniques based on their availabilities and requirements. Simulation results show the ability of the new scheme to serve extra traffic in the cognitive network.