Channel reservation and spectrum adaptation strategies in a multi-level prioritized cognitive radio network

Abstract

Wireless technologies have dominated the communication's market by offering reasonable speeds and convenience at low deployment costs. However, due to the significant growth of mobile computing devices and their bandwidth demands, together with the paradigm shift brought by the Internet of things, future wireless networks should become highly dense and heterogeneous, which will hardly cope with the traditional fixed spectrum allocation policy. Some standards such as the Long Term Evolution – Advanced (LTE-A), have already set the precedent for carrier aggregation (CA), aiming at scaling up bitrates, which partially helps solving the problem. However, cognitive radio (CR) has been put forward as the most promising solution to handle this complex ecosystem since it may provide better spectrum utilization and user coordination through non-traditional mechanisms. Among other features, it allows non-licensed users, known as secondary users (SUs) to opportunistically use temporarily idle licensed bands that are used by licensed clients called primary users (PUs). Once PUs and SUs are expected to share the same spectrum bands, a critical issue is to concomitantly avoid primary interference while supporting QoS for the secondary services. This dissertation studies the synergistic integration of cognitive radio networks (CRNs), Dynamic Spectrum Access DSA techniques and resource allocation strategies (e.g., CA) that combined, should improve the overall system’s performance. We have proposed a layered M/M/N/N queue-based model that addresses three user priorities, flexible bandwidth choices, multi-level channel reservation and two channel aggregation strategies. Different network load conditions for each feature were evaluated in terms of four performance metrics: blocking probability, forced termination probability, spectrum utilization and throughput. Such study is particularly useful for understanding the effects of each of these approaches in the secondary network. To the best of our knowledge, our model fulfills almost completely the user bandwidth’s possibilities, improves the existing channel reservation formulation and demonstrates that our proposed dynamic channel aggregation strategy performs similarly to a more complex simultaneous channel aggregation and fragmentation approach, but can be technically more feasible.