Towards Adaptive, Scalable, and Robust Coordination of LLM Agents: A Dynamic Ad-Hoc Networking Perspective

2026-02-08

AI Summary

This work addresses the challenge of enabling adaptive, scalable, and robust communication among large language model (LLM) agents without human intervention. To this end, it introduces the concept of dynamic self-organizing networks into multi-agent LLM systems for the first time, proposing an intent-driven, reputation-aware publish-subscribe framework. The architecture supports dynamic subscription based on agent intentions, integrates a Bayesian reputation mechanism to detect and mitigate malicious nodes, and eliminates reliance on predefined network topologies. Experimental results across five benchmark tasks demonstrate that the proposed approach significantly enhances collaborative efficiency, system scalability, and resilience against adversarial interference.

Technology Category

• Multiagent Systems: Agent Communication
• Machine Learning: Large Multimodal Models (LMMs)
• Cognitive Modeling & Cognitive Systems: Agent Architectures

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Multi-agent architectures built on large language models (LLMs) have demonstrated the potential to realize swarm intelligence through well-crafted collaboration. However, the substantial burden of manual orchestration inherently raises an imperative to automate the design of agentic workflows. We frame such an agent coordination challenge as a classic problem in dynamic ad-hoc networking: How to establish adaptive and reliable communication among a scalable number of agentic hosts? In response to this unresolved dilemma, we introduce RAPS, a reputation-aware publish-subscribe paradigm for adaptive, scalable, and robust coordination of LLM agents. RAPS is grounded in the Distributed Publish-Subscribe Protocol, allowing LLM agents to exchange messages based on their declared intents rather than predefined topologies. Beyond this substrate, RAPS further incorporates two coherent overlays: (i) Reactive Subscription, enabling agents to dynamically refine their intents; and (ii) Bayesian Reputation, empowering each agent with a local watchdog to detect and isolate malicious peers. Extensive experiments over five benchmarks showcase that our design effectively reconciles adaptivity, scalability, and robustness in a unified multi-agent coordination framework.

Problem

• LLM agents
• dynamic ad-hoc networking
• agent coordination
• scalability
• robustness

Innovation

• RAPS
• publish-subscribe
• dynamic ad-hoc networking
• Bayesian reputation
• LLM agents