Energy-Efficient UAV-assisted LoRa Gateways: A Multi-Agent Optimization Approach
Abdullahi Isa Ahmed, Jamal Bentahar, El Mehdi Amhoud
TL;DR
This work tackles energy-efficient data collection in UAV-assisted LoRa networks with multiple gateways under dynamic channels and mobility. It models the problem as a partially observable stochastic game (POSG) and proposes a two-stage solution: a channel-aware ED-UAV association algorithm and a MAPPO-based resource allocation framework under centralized training with decentralized execution (CTDE). The approach achieves faster convergence and superior system energy efficiency compared to state-of-the-art MARL baselines across varying numbers of active EDs and UAVs, with gains of up to substantial percentages over benchmarks. This demonstrates a scalable, practical method for optimizing LoRa networks in dynamic NG-IoT deployments, with potential extensions to UAV trajectory optimization and hybrid-band operation. The key result is the effective combination of channel-aware association and cooperative MARL to maximize $EE_{sys}$ in a complex, multi-agent setting: $EE_{sys} = \sum_{t=1}^T \sum_{u=1}^U \left[ \dfrac{ \sum_{v=1}^V a_{u,v}[t] \Re_{uv}[t] }{ ( \sum_{v=1}^V a_{u,v}[t] P_v[t] ) + P^{hover}_u } \right]$.
Abstract
As next-generation Internet of Things (NG-IoT) networks continue to grow, the number of connected devices is rapidly increasing, along with their energy demands, creating challenges for resource management and sustainability. Energy-efficient communication, particularly for power-limited IoT devices, is therefore a key research focus. In this paper, we study Long Range (LoRa) networks supported by multiple unmanned aerial vehicles (UAVs) in an uplink data collection scenario. Our objective is to maximize system energy efficiency by jointly optimizing transmission power, spreading factor, bandwidth, and user association. To address this challenging problem, we first model it as a partially observable stochastic game (POSG) to account for dynamic channel conditions, end device mobility, and partial observability at each UAV. We then propose a two-stage solution: a channel-aware matching algorithm for ED-UAV association and a cooperative multi-agent reinforcement learning (MARL) based multi-agent proximal policy optimization (MAPPO) framework for resource allocation under centralized training with decentralized execution (CTDE). Simulation results show that our proposed approach significantly outperforms conventional off-policy and on-policy MARL algorithms.