Abstract:
The advancement in technology has led to the requirement of wireless communication at longer distances in areas such as wide range environmental monitoring, automated meter reading and wide area sensors networks. The principle of Long-Range WAN is to provide reliable communication shifting the focus of wireless systems from static spectrum to dynamic spectrum access. Variation of channel state is the key to designing the channel selection mechanism in LoRaWAN. The mechanism must be predictive and must consider the multi-step decision-making based on the reward and penalty response for channel selection. High efficiency and reliability in decision making concerning various channel states can be determined by sequential decision processes. For LoRaWAN Learning Automata is the ideal process to integrate the channel selection mechanism. Learning Automata has been used as AI powerful tool in different area to solve different problems including the one for channel selection. Despite its variant version Learning automata improved its power invoking the concept of Pursuit. This research aims to integrate a Learning Automata-based channel selection framework for LongRange Wide Area Network (LoRaWAN) deployments. The framework leverages Hierarchical Discrete Pursuit Learning Automata (HDPA) to select optimal channels among multiple channels, aiming to enhance network performance and mitigate interference. The study investigates LoRaWAN communication characteristics, integrates the HDPA model, evaluates its performance through simulations, and compares it with existing methods. The research methodology involves theoretical modeling, simulation, to evaluate the proposed Hierarchical Discrete Pursuit Learning Automata-based channel selection algorithm's effectiveness. The outcomes are expected to advance the efficiency and reliability of LoRaWAN networks, benefiting applications across smart cities, IoT, and industrial sectors. The simulation shows that our HDPA outperformed the HCPA in terms of speed and accuracy.
