Channel Assignment Using Mutual Learning Automata in Wireless Mesh Networks

Wireless Mesh Networks (WMN) provide consumers with reliable internet connectivity. By utilizing the Learning Automata for Channel Assignment (LACA), a single router serves multiple users, resulting in significant data traffic within the mesh network. The primary challenge in a multi-channel environment is managing this substantial data flow. This research presents the Mutual Learning Automata-based Channel Assignment (MLACA) Scheme to address this issue. In this framework, Learning Automata (LA) are deployed at adjacent mesh routers, enabling them to collaborate on data transmission and information sharing while learning from their surroundings. The proposed system's collaborative learning automata dynamically adjust channel assignments based on network conditions, enhancing throughput, packet delivery, and spectral efficiency. Performance evaluations using the NS-2 simulator indicate that MLACA significantly outperforms the existing LACA scheme. Specifically, MLACA achieves a 40% increase in throughput, maintains a higher packet delivery ratio even under increased traffic loads, and reduces switching delay by up to 30% compared to LACA. Additionally, MLACA improves effective channel utilization by 25%, increases remaining bandwidth availability by 20%, and enhances the effective transmission rate under dynamic conditions. These results demonstrate that MLACA not only minimizes interference and switching overhead but also offers a scalable and efficient solution for real-time channel allocation in dynamic WMN environments.