ABSTRACT
Wireless Powered Communication Networks play a crucial role in Internet of Things scenarios which operate with low energy sensors. In this regard, simultaneous wireless energy transmission (WET) and wireless information transmission (WIT) are proposed for continuous data transmission with low energy sensors. These sensors can be deployed on bridges that are hit by disasters. However, these low-energy networks demand a random data transmission and hence require access points to decode the packets accordingly. Hence it is challenging to design a slot selection algorithm that provides a better throughput. On this subject, we present a Q-Learning-based algorithm that harnesses the qualities of deep reinforcement learning and enhances the throughput. In our algorithm, we enhance the slot allocation of each user with the influence of a few physical layer properties. Results show an improved slot allocation mechanism leads to better throughpu
