Joint Resource Management for Energy-efficient UAV-assisted SWIPT-MEC: A Deep Reinforcement Learning Approach

TL;DR

This work tackles energy-efficient UAV-assisted SWIPT-MEC in infrastructure-free settings by formulating a bi-objective problem that minimizes total energy and maximizes a Jain-fairness-based terminal battery metric. The authors reformulate the problem as an MDP and introduce SAC-SK, a DRL framework that handles a hybrid action space through action simplification, SRU-based temporal encoding, and a modified KAN for improved function approximation. Empirical results show SAC-SK achieves substantial gains in terminal energy retention and charging fairness while reducing overall energy consumption, with strong generalization across diverse terminal distributions. The approach offers a practical pathway to deploy UAV-based MEC with simultaneous charging and communication capabilities, enabling resilient IoT operations in disaster or remote scenarios.

Abstract

The integration of simultaneous wireless information and power transfer (SWIPT) technology in 6G Internet of Things (IoT) networks faces significant challenges in remote areas and disaster scenarios where ground infrastructure is unavailable. This paper proposes a novel unmanned aerial vehicle (UAV)-assisted mobile edge computing (MEC) system enhanced by directional antennas to provide both computational resources and energy support for ground IoT terminals. However, such systems require multiple trade-off policies to balance UAV energy consumption, terminal battery levels, and computational resource allocation under various constraints, including limited UAV battery capacity, non-linear energy harvesting characteristics, and dynamic task arrivals. To address these challenges comprehensively, we formulate a bi-objective optimization problem that simultaneously considers system energy efficiency and terminal battery sustainability. We then reformulate this non-convex problem with a hybrid solution space as a Markov decision process (MDP) and propose an improved soft actor-critic (SAC) algorithm with an action simplification mechanism to enhance its convergence and generalization capabilities. Simulation results have demonstrated that our proposed approach outperforms various baselines in different scenarios, achieving efficient energy management while maintaining high computational performance. Furthermore, our method shows strong generalization ability across different scenarios, particularly in complex environments, validating the effectiveness of our designed boundary penalty and charging reward mechanisms.

Figures (8)

• Figure 1: The directional antenna-enhanced UAV-assisted multi-terminal SWIPT-MEC system.
• Figure 2: The time slot division model in the offloading scenario.
• Figure 3: The architecture of SAC-SK, where the SRU layer and the modified KAN layer are integrated into the neural networks.
• Figure 4: The schematic of SRU calculative architecture.
• Figure 5: The neural network architecture of SAC-SK.