Energy Management and Wake-up for IoT Networks Powered by Energy Harvesting

TL;DR

This paper introduces a machine learning-based energy management framework for energy harvesting IoT networks, optimizing duty cycling and wake-up strategies to enhance energy efficiency and event detection accuracy.

Contribution

It proposes a novel KNN-based duty cycling method and evaluates reinforcement learning and decision transformers for energy-efficient IoT operation with energy harvesting.

Findings

All three methods significantly improve energy savings.

RL approach approaches genie-aided benchmark performance.

Methods outperform existing state-of-the-art approaches.

Abstract

The rapid growth of the Internet of Things (IoT) presents sustainability challenges such as increased maintenance requirements and overall higher energy consumption. This motivates self-sustainable IoT ecosystems based on Energy Harvesting (EH). This paper treats IoT deployments in which IoT devices (IoTDs) rely solely on EH to sense and transmit information about events/alarms to a base station (BS). The objective is to effectively manage the duty cycling of the IoTDs to prolong battery life and maximize the relevant data sent to the BS. The BS can also wake up specific IoTDs if extra information about an event is needed upon initial detection. We propose a K-nearest neighbors (KNN)-based duty cycling management to optimize energy efficiency and detection accuracy by considering spatial correlations among IoTDs' activity and their EH process. We evaluate machine learning approaches, including reinforcement learning (RL) and decision transformers (DT), to maximize information captured from events while managing energy consumption. Significant improvements over the state-ofthe-art approaches are obtained in terms of energy saving by all three proposals, KNN, RL, and DT. Moreover, the RL-based solution approaches the performance of a genie-aided benchmark as the number of IoTDs increases.