Communication-Aware Multi-Agent Reinforcement Learning for Decentralized Cooperative UAV Deployment

2026-03-17

AI Summary

This work addresses the challenge of cooperative deployment for drone swarms under partial observability and intermittent communication by proposing a graph neural network–based multi-agent reinforcement learning approach grounded in the centralized training with decentralized execution (CTDE) paradigm. The method employs a distance-constrained communication graph and introduces agent-entity attention alongside neighbor self-attention mechanisms, enabling efficient coordination using only local observations and messages from nearby agents. This architecture supports zero-shot generalization to formations of varying scales. Experimental results demonstrate that, in the DroneConnect task, a team of five drones achieves 74% area coverage—approaching the offline upper bound provided by mixed-integer linear programming—and significantly outperforms non-communicating baselines in the DroneCombat task, thereby validating the approach’s effectiveness and scalability.

Technology Category

Multiagent Systems: Agent CommunicationMachine Learning: Graph-based Machine LearningConstraint Satisfaction and Optimization: Distributed CSP/Optimization

Application Category

Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphsSemantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMsSearch and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Autonomous Unmanned Aerial Vehicle (UAV) swarms are increasingly used as rapidly deployable aerial relays and sensing platforms, yet practical deployments must operate under partial observability and intermittent peer-to-peer links. We present a graph-based multi-agent reinforcement learning framework trained under centralized training with decentralized execution (CTDE): a centralized critic and global state are available only during training, while each UAV executes a shared policy using local observations and messages from nearby neighbors. Our architecture encodes local agent state and nearby entities with an agent-entity attention module, and aggregates inter-UAV messages with neighbor self-attention over a distance-limited communication graph. We evaluate primarily on a cooperative relay deployment task (DroneConnect) and secondarily on an adversarial engagement task (DroneCombat). In DroneConnect, the proposed method achieves high coverage under restricted communication and partial observation (e.g. 74% coverage with M = 5 UAVs and N = 10 nodes) while remaining competitive with a mixed-integer linear programming (MILP) optimization-based offline upper bound, and it generalizes to unseen team sizes without fine-tuning. In the adversarial setting, the same framework transfers without architectural changes and improves win rate over non-communicating baselines.

Problem

• multi-agent reinforcement learning
• decentralized cooperation
• UAV deployment
• partial observability
• intermittent communication

Innovation

• communication-aware MARL
• graph-based attention
• decentralized UAV coordination
• CTDE
• partial observability