Short-Term and Long-Term Throughput Maximization in Mobile Wireless-Powered Internet of Things
Kechen Zheng, Rongwei Luo, Zuxin Wang, Xiaoying Liu, Yuan Yao
Abstract
With the evolution of Internet of Things (IoT), some IoT nodes possess a certain degree of mobility, and the gains of the corresponding channels vary dramatically, incurring the energy supply problem for IoT nodes. To tackle this problem, we study a wireless-powered IoT (WPIoT), where a static <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$U$ </tex-math></inline-formula> -antenna hybrid access point (HAP) coordinates the wireless energy transfer to mobile single-antenna IoT nodes and receives data from these IoT nodes. When IoT nodes have sufficient energy for transmitting generated data packets, we propose a generated data packets-based throughput maximization (GDPTM) algorithm for the short-term throughput maximization, and the GDPTM algorithm is designed to save nodes’ energy while transmitting all the generated data packets. Through monotonicity analysis, we prove the existence of the optimal transmit power that maximizes the throughput. When IoT nodes do not have sufficient energy for transmitting generated data packets, we propose a deep deterministic policy gradient (DDPG)-based multinode resource allocation (DMRA) algorithm. Through designing the action space, we find that the HAP under the DMRA algorithm manages the time, transmit power, and channel allocation of IoT nodes to improve the throughput. Numerical results validate that, when IoT nodes have sufficient energy, the GDPTM algorithm saves nodes’ energy and improves the throughput. When IoT nodes do not have sufficient energy, the DMRA algorithm also improves the throughput.