Trade-offs in Decentralized Agentic AI Discovery Across the Compute Continuum

2026-05-12

AI Summary

This study addresses the lack of efficient and adaptive decentralized discovery mechanisms for multi-agent systems operating in intermittently connected cloud-edge-end environments. It presents the first systematic quantitative evaluation of mainstream distributed hash table (DHT) protocols—Chord, Pastry, and Kademlia—in the context of agent discovery, analyzing their trade-offs in reliability, bootstrap behavior, and control overhead. Using a unified simulation framework at a scale of 4,096 nodes under dynamic perturbations, the work reveals significant differences among these protocols in terms of discovery success rate, convergence speed, and communication cost. The findings delineate the operational boundaries of each DHT within the computing continuum, thereby providing principled guidance for protocol selection in cross-domain multi-agent systems.

Abstract

Agentic systems deployed across the compute continuum need discovery mechanisms that remain effective across cloud, edge, and intermittently connected domains. In some emerging agentic architectures, decentralized discovery is already an active design direction, placing DHT-based lookup on the path toward agent directories. This paper studies the trade-offs among major structured-overlay families for agent discovery, comparing Chord, Pastry, and Kademlia as candidate indexing substrates within a shared control-plane framework. Using a benchmark subset centered on a 4096-node stationary comparison and a representative 4096-node churn benchmark, the paper characterizes how discovery reliability, startup behavior, and control-plane overhead vary across these overlays. The goal is to clarify the operating points they expose for agent discovery across edge-to-cloud environments.

Problem

• Decentralized Discovery
• Agentic AI
• Compute Continuum
• Structured Overlay
• Agent Directory

Innovation

• Decentralized Discovery
• Agentic AI
• Structured Overlay
• Compute Continuum
• DHT-based Lookup