Adaptive Event-Triggered Policy Gradient for Multi-Agent Reinforcement Learning

2025-09-24

AI Summary

Conventional multi-agent reinforcement learning (MARL) suffers from excessive computational and communication overhead due to fixed-interval triggering mechanisms. Method: This paper proposes AET-MAPG, an event-triggered MARL framework that jointly models action policies and event-triggering policies for the first time. It incorporates a self-attention mechanism to dynamically select communication partners and content, enabling co-optimization of execution timing and collaborative behavior. Built upon policy gradients, AET-MAPG is compatible with mainstream MARL algorithms. Contribution/Results: Evaluated on multiple standard benchmarks, AET-MAPG achieves state-of-the-art (SOTA) performance while significantly reducing computational load and communication frequency. These results empirically validate the effectiveness and practicality of the event-driven paradigm in multi-agent coordination.

Technology Category

• Multiagent Systems: Agent Communication
• Humans and AI: Human-Aware Planning and Behavior Prediction
• Game Theory and Economic Paradigms: Coordination and Collaboration

Application Category

• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Search and Retrieval-Augmented AI: Agentic search

Abstract

Conventional multi-agent reinforcement learning (MARL) methods rely on time-triggered execution, where agents sample and communicate actions at fixed intervals. This approach is often computationally expensive and communication-intensive. To address this limitation, we propose ET-MAPG (Event-Triggered Multi-Agent Policy Gradient reinforcement learning), a framework that jointly learns an agent's control policy and its event-triggering policy. Unlike prior work that decouples these mechanisms, ET-MAPG integrates them into a unified learning process, enabling agents to learn not only what action to take but also when to execute it. For scenarios with inter-agent communication, we introduce AET-MAPG, an attention-based variant that leverages a self-attention mechanism to learn selective communication patterns. AET-MAPG empowers agents to determine not only when to trigger an action but also with whom to communicate and what information to exchange, thereby optimizing coordination. Both methods can be integrated with any policy gradient MARL algorithm. Extensive experiments across diverse MARL benchmarks demonstrate that our approaches achieve performance comparable to state-of-the-art, time-triggered baselines while significantly reducing both computational load and communication overhead.

Problem

• Reducing computational and communication costs in multi-agent reinforcement learning
• Integrating action and event-triggering policies into unified learning process
• Optimizing selective communication patterns using attention mechanisms

Innovation

• Event-triggered policy gradient for multi-agent learning
• Integrated learning of control and triggering policies
• Attention mechanism for selective communication optimization