A Threat-Aware and Efficient Wireless Charging Scheme for IoT Networks

The Internet of Things (IoT) is a breakthrough that enables many applications and improves our daily life. Since IoT networks deploy many devices which are generally energy or computationally-constrained, it is necessary to efficiently manage their energy to maximize network lifetime. Moreover, IoT networks may face multiple security threats that must be dealt with defense mechanisms. However, these defense mechanisms reduce network lifetime. Hence, it is necessary to design solutions that reduce the impacts of defense solutions on network lifetime. In this paper, we propose a solution based on wireless Mobile Chargers (MCs) which proactively and preventively charge devices that may need energy for the execution of security services that protect the IoT network. By using Deep Reinforcement Learning (DRL), especially Deep-Q learning, our solution determines, from the current threat level, the remaining energy of the devices, and the distance from the charger, the next device to charge. Compared to approaches that are not threat-aware, our solution improves the lifetime of a rechargeable IoT network of 16 devices by 21.59%.