Learning to Charge RF-Energy Harvesting Devices in WiFi Networks

TL;DR

This paper introduces two novel methods for managing energy in RF-energy harvesting WiFi networks, improving energy efficiency and user satisfaction without requiring perfect channel information.

Contribution

It proposes DQN and MPC-based solutions for energy management in RF-energy harvesting WiFi networks, suitable for current systems and not reliant on perfect channel data.

Findings

DQN improves energy efficiency by up to 35%.

MPC enhances user satisfaction by up to 42%.

Both methods outperform existing algorithms.

Abstract

In this paper, we consider a solar-powered Access Point (AP) that is tasked with supporting both non-energy harvesting or legacy data users such as laptops, and devices with Radio Frequency (RF)-energy harvesting and sensing capabilities. We propose two solutions that enable the AP to manage its harvested energy via transmit power control and also ensure devices perform sensing tasks frequently. Advantageously, our solutions are suitable for current wireless networks and do not require perfect channel gain information or non-causal energy arrival at devices. The first solution uses a deep Q-network (DQN) whilst the second solution uses Model Predictive Control (MPC) to control the AP's transmit power. Our results show that our DQN and MPC solutions improve energy efficiency and user satisfaction by respectively 16% to 35%, and 10% to 42% as compared to competing algorithms.