A Novel Dynamic Channel Assembling Strategy in Cognitive Radio Networks With Fine-Grained Flow Classification

Abstract

With the rapid development of various applications in the Internet of Things (IoT), we have witnessed much progress with very wide differences in characteristics and requirements. In this article, we propose a novel dynamic channel assembling (DChA) strategy for channel access of heterogeneous secondary user (SU) flows in IoT-oriented cognitive radio networks (CRNs), making use of the priority queues based on fine-grained flow classification. Specifically, three categories of SU flows are considered, so-called the real-time SU (RSU) flows, the elastic large SU flows, and the elastic small SU flows. On top of this, channel access opportunities are distributed to the SU flows in three specially designed queues performing the channel access algorithm. The highlight of our main idea is that the RSU flows with higher priority are only supposed to assemble as few channels as possible, so long as their minimum requirements are fulfilled, thereby minimizing the impact on elastic SU (ESU) traffic. For the sake of performance evaluation, we utilize the continuous-time Markov chain to model our proposed strategy and conduct theoretical analyses. With the detailed theoretical analyses and extensive simulations, the proposed DChA strategy is demonstrated to be able to fulfill the deadline of SU flows, while significantly reducing the blocking probability as well as the completion time of the ESU flows.