ADRL: A reconfigurable energy-efficient transmission policy for mobile LoRa devices based on reinforcement learning

Project publication · Internet of Things

Summary

The article presents ADRL, a reconfigurable and energy-efficient transmission policy for mobile LoRa devices based on reinforcement learning.

Mecanismo ADRL para seleccionar parámetros de transmisión LoRa — Figure 1. ADRL parameter-selection mechanism with RSSI prediction and reinforcement learning. Source: Internet of Things / Elsevier; reproduced with attribution under CC BY 4.0.

Authors: Laura Acosta-Garcia, Juan Aznar-Poveda, Antonio Javier Garcia-Sanchez, Jakob Hollenstein, Joan Garcia-Haro, Thomas Fahringer

Journal: Internet of Things, volumen 31, artículo 101577

DOI: 10.1016/j.iot.2025.101577

Funding: PID2023-148214OB-C21.

Main contribution

The contribution makes it possible to adapt transmission to mobility and link conditions, aiming to improve performance and energy consumption compared with static policies in LoRa networks.

Resultados y alcance

Las simulaciones muestran que ADRL adapta la política de transmisión de dispositivos móviles y mejora el compromiso entre entrega, energía y ocupación del canal. Su efectividad depende de la movilidad, la recompensa y la capacidad de reconfiguración, aspectos que condicionan el traslado a hardware y tráfico reales.

Link with BoND1

For BoND1, ADRL strengthens the adaptive-control line in non-cellular IoT networks. The combination of mobility, energy efficiency and machine learning fits the objective of deploying scalable and sustainable B6G connectivity.

DOI

Project publication · Internet of Things

Summary

The paper presents ADRL, an energy-efficient and reconfigurable transmission policy for mobile LoRa devices based on reinforcement learning.

ADRL mechanism for selecting LoRa transmission parameters — Figure 1. ADRL parameter selection mechanism with RSSI prediction and reinforcement learning. Source: Internet of Things / Elsevier; reproduced with attribution under CC BY 4.0.

Authors: Laura Acosta-Garcia, Juan Aznar-Poveda, Antonio Javier Garcia-Sanchez, Jakob Hollenstein, Joan Garcia-Haro, Thomas Fahringer

Journal: Internet of Things, volume 31, article 101577

DOI: 10.1016/j.iot.2025.101577

Funding: PID2023-148214OB-C21.

Bibliometric indicators

JCR quartileQ1

Impact factor7.1

Google Scholar citations10

JCR area: Computer Science, Information Systems
JCR year: 2025
Scholar query: 26 October 2026

Main contribution

The contribution allows adapting the transmission to the mobility and link conditions, seeking to improve performance and consumption compared to static policies in LoRa networks.

Results and scope

Simulations show that ADRL adapts the transmission policy of mobile devices and improves the trade-off between delivery, energy and channel occupancy. Its effectiveness depends on mobility, reward and reconfigurability, aspects that determine the transfer to real hardware and traffic.

Link with BoND1

For BoND1, ADRL reinforces the line of adaptive control in non-cellular IoT networks. The combination of mobility, energy efficiency and machine learning fits with the objective of deploying scalable and sustainable B6G connectivity.

DOI

UPCT Repository: https://hdl.handle.net/10317/24193