A New Strategy for Dynamic Channel Allocation in CR-WMN Based on RCA

Abstract

Channel assignment has emerged as an essential study subject in Cognitive Radio-based Wireless Mesh Networks (CR-WMN). In an era of alarming increase in Multi-Radio Multi-Channel (MRMC) network expansion interference is decreased and network throughput is significantly increased when non-overlapping or partially overlapping channels are correctly integrated. Because of its ad hoc behavior, dynamic channel assignment outperforms static channel assignment. Interference reduces network throughput in the CR-WMN. As a result, there is an extensive research gap for an algorithm that dynamically distributes channels while accounting for all types of interference. This work presents a method for dynamic channel allocations using unsupervised Machine Learning (ML) that considers both coordinated and uncoordinated interference. Unsupervised machine learning uses coordinated and non-coordinated interference for dynamic channel allocation. To determine the applicability of the proposed strategy in reducing channel interference while increasing WMN throughput, a comparison analysis was performed. When the simulation results of our proposed algorithm are compared to those of the Routing Channel Assignment (RCA) algorithm, the throughput of our proposed algorithm has increased by 34% compared to both coordinated and non-coordinated interferences.