GAIA-DRL: A Geoenvironmental Agent for Energy Optimization in Batteryless IoT Networks with Ambient Backscatter

Edwardes Amaro Galhardo, Wesley dos Reis Bezerra, Carlos Becker Westphall, Antonio Oliveira-Jr

DOI: 10.14209/sbrt.2025.1571156219

Evento: XLIII Simpósio Brasileiro de Telecomunicações e Processamento de Sinais (SBrT2025)

Keywords: GAIA-DRL Batteryless IoT Ambient Backscatter Remote Sensing

Abstract

This work addresses sustainability-driven network design for future wireless systems by integrating batteryless dense IoT networks-based on ambient backscatter communication-with geospatial intelligence and deep reinforcement learning. We propose GAIA-DRL (Geoenvironmentally-Aware Intelligent Agent with Deep Reinforcement Learning), an approach that jointly optimizes throughput, latency, energy efficiency, and interference, while incorporating land-use information such as NDVI and pasture coverage. A geospatially-informed agent controls communication policies using real territorial data. Geographic Information Systems (GIS) and remote sensing techniques enable context-aware decisions, supporting low-cost, low-carbon applications in 6G-ready IoT networks for greenhouse gas monitoring, smart agriculture, and sustainable environmental management.

Download