Performance Evaluation of Reinforcement Learning Based Distributed Channel Selection Algorithm in Massive IoT Networks

In recent years, the demand for new applications using various Internet of Things (IoT) devices has led to an increase in number connected wireless networks. However, owing limitation available frequency resources IoT devices, degradation communication quality caused by channel congestion is a practical problem developing technology. Many have hardware and software limitations that prevent centralized allocation, even more severe massive networks without central controller. Therefore, distributed sophisticated selection algorithm necessary. previous studies, each device was modeled as multi-armed bandit (MAB) problem, method based on MAB algorithm, which simple reinforcement learning, proposed. particular, it been shown tug-of-war (TOW) dynamics can efficiently select channels with much lower computational complexity power compared other learning-based channel-selection methods. This paper proposes TOW fully decentralized We evaluate effectiveness proposed methods success rate experiments simulations. The results show improves than dense dynamic network environment.