Energy Efficient Task Offloading in UAV-Enabled MEC Using a Fully Decentralized Deep Reinforcement Learning Approach

2025-08-09

AI Summary

Joint optimization of task offloading, user association, and UAV trajectory in UAV-assisted multi-access edge computing (MEC) is challenged by non-convexity, unpredictable channel conditions and user positions due to mobility, and poor scalability and robustness arising from semi-centralized architectures with high communication overhead. Method: We propose a fully decentralized deep reinforcement learning framework that integrates graph attention networks (GATs) to model local topological relationships and employs experience- and parameter-sharing proximal policy optimization (EPS-PPO) for distributed cooperative decision-making—relying solely on local observations and neighbor-to-neighbor communication. Contribution/Results: Experiments demonstrate significant improvements over baseline methods (e.g., MADDPG) in energy efficiency, task completion rate, and latency. The framework further achieves superior scalability and system robustness under dynamic network conditions and heterogeneous user mobility.

Technology Category

• Multiagent Systems: Mechanism Design
• Machine Learning: Distributed Machine Learning & Federated Learning
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Unmanned aerial vehicles (UAVs) have been recently utilized in multi-access edge computing (MEC) as edge servers. It is desirable to design UAVs' trajectories and user to UAV assignments to ensure satisfactory service to the users and energy efficient operation simultaneously. The posed optimization problem is challenging to solve because: (i) The formulated problem is non-convex, (ii) Due to the mobility of ground users, their future positions and channel gains are not known in advance, (iii) Local UAVs' observations should be communicated to a central entity that solves the optimization problem. The (semi-) centralized processing leads to communication overhead, communication/processing bottlenecks, lack of flexibility and scalability, and loss of robustness to system failures. To simultaneously address all these limitations, we advocate a fully decentralized setup with no centralized entity. Each UAV obtains its local observation and then communicates with its immediate neighbors only. After sharing information with neighbors, each UAV determines its next position via a locally run deep reinforcement learning (DRL) algorithm. None of the UAVs need to know the global communication graph. Two main components of our proposed solution are (i) Graph attention layers (GAT), and (ii) Experience and parameter sharing proximal policy optimization (EPS-PPO). Our proposed approach eliminates all the limitations of semi-centralized MADRL methods such as MAPPO and MA deep deterministic policy gradient (MADDPG), while guaranteeing a better performance than independent local DRLs such as in IPPO. Numerical results reveal notable performance gains in several different criteria compared to the existing MADDPG algorithm, demonstrating the potential for offering a better performance, while utilizing local communications only.

Problem

• Optimizing UAV trajectories for energy-efficient MEC task offloading
• Decentralizing UAV control to reduce communication overhead and bottlenecks
• Enhancing performance using local DRL without global network knowledge

Innovation

• Fully decentralized deep reinforcement learning approach
• Graph attention layers for local observations
• Experience sharing proximal policy optimization