Distributed Slot Allocation in Capillary Gateways for Internet of Things Networks

The applications and usage of the internet is expanding on a daily basis and the Internet of Things (IoT) is fast becoming the new approach for incorporating the internet into our personal, professional and social lives. IoT enables a wide variety of devices to inter-operate through the existing internet infrastructure. Capillary networks are proposed as a fundamental part of loT development, and will enable local sensor and devices to connect efficiently with other ubiquitous communication networks such as cellular systems. In this paper, we apply the Q-learning algorithm for the scheduling of capillary gateways for (M2M) communication in IoT networks. Q-learning algorithm is used to select conflict-free slot assignment for these gateways in a self-organizing manner. We analyze the performance of the proposed algorithm with respect to learning rates and rewards.