An intelligent Hybrid‐Q Learning clustering approach and resource management within heterogeneous cluster networks based on reinforcement learning

• Published in: Transactions on emerging telecommunications technologies Vol. 35; no. 4
• Main Authors: Mughal, Fahad Razaque, He, Jingsha, Zhu, Nafei, Almutiq, Mutiq, Dharejo, Fayaz Ali, Jain, Deepak Kumar, Hussain, Saqib, Zardari, Zulfiqar Ali
• Format: Journal Article
• Language: English
• Published: 01.04.2024
• ISSN: 2161-3915; 2161-3915
• DOI: 10.1002/ett.4852

Recently, heterogeneous cluster networks (HCNs) have been the subject of significant research. The nature of the next‐generation HCN environment is decentralized and highly dynamic; optimization techniques cannot quite express the dynamic characteristics of node resource utilization and communication of HCN networks. In this article, we present an intelligent Hybrid‐Q Learning approach (Hybrid QL)‐based clustering approach for IoT and WSN. Using the self‐learning abilities of (HCNs), we propose a model for dynamic accessing systems on nodes and agents that identify the best possible paths and communication over heterogeneous cluster networks using reinforcement learning. In addition to reducing energy consumption, it creates efficient and effective resource utilization and node communication performance. Through increased throughput and link management, the HCN aims to reduce energy consumption. The proposed model is compared to existing approaches based on various scenarios. Finally, the results of the evaluation tasks demonstrate high accuracy, low‐level complexity, fast dynamic response times, and scalability for heterogeneous cluster networks. Our model showed exceptional node allocation efficiency for dynamic IOT environments and WSNs. We propose a Hybrid‐Q Learning (Hybrid QL)‐based clustering for IoT and WSN. Self‐learning solution to solve the problem of decentralized and dynamic self‐access for heterogeneous nodes. Our proposed model dynamic accessing system on node/agents identifies the best possible paths and communication over heterogeneous cluster networks using self‐learning abilities (HCNs).