发表机构
University of Calgary; University of Michigan（卡尔加里大学; 密歇根大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
FL-MAESTRO是面向资源受限联邦学习的多智能体LLM编排框架，通过三个专业LLM智能体联合决策，在非IID CIFAR-10基准上，准确率与最强能量感知基线相当，浪费轮次能量从超三分之一降至近零。
AI 中文摘要
在联邦学习（Federated Learning, FL）中，通信拓扑是运行时变量，而非固定设计选择，因为训练过程中链路和边缘设备会随机加入或退出。每一轮训练中，服务器必须做出三个相互关联的决策：通信拓扑、每客户端的资源分配，以及用于合并本地更新的聚合规则。近期的智能体系统已开始将大语言模型（Large Language Model, LLM）引入联邦学习领域，但现有研究要么在设置阶段运行，要么仅处理单一运行时维度（如客户端选择）。本文提出FL-MAESTRO，一种多智能体编排框架，它通过三个专业LLM智能体直接做出联邦学习的联合运行时决策，每个智能体对应一个决策维度；协调器将各智能体的分析整合为单一决策，并在该轮执行前通过非LLM可行性检查确认其可行性。由于该编排框架使用服务器预测的故障设备列表，它会剔除那些其更新永远不会被聚合的客户端，从而消除了传统联邦学习在不稳定边缘网络中浪费轮次能量的主要来源。由于客户端状态以自然文本形式的配置文件读取，该编排框架可扩展到异构设备类别，无需为每个类别单独建立能量模型。在非独立同分布（non-IID）CIFAR-10基准测试中，FL-MAESTRO的准确率与最强的能量感知基线相当，同时将浪费的轮次能量从超过三分之一降至接近零。代码可在该https URL获取。
英文摘要
In Federated Learning (FL), the communication topology is a runtime variable rather than a fixed design choice, since links and edge devices drop in and out during training. Each round, the server must commit three coupled decisions, namely the communication topology, per-client resource allocation, and the aggregation rule for combining local updates. Recent agentic systems have begun bringing large language models (LLM) into FL, but the existing line of work either operates at setup time or handles a single runtime dimension such as client selection. We propose FL-MAESTRO, a multi-agent orchestrator that makes the joint runtime FL decision directly through three specialist LLM agents, one per decision dimension. A coordinator combines their analyses into a single decision, and a non-LLM feasibility check confirms it before the round executes. Because the orchestrator consumes the server's predicted-failure list, it withholds clients whose updates would never be aggregated, which removes the dominant source of wasted round energy in classical FL on volatile edge networks. Because client state is read as natural-text profiles, the same orchestrator extends to heterogeneous device classes without per-class energy models. On a non-IID CIFAR-10 benchmark, FL-MAESTRO matches the accuracy of the strongest energy-aware baseline while cutting wasted round energy from over a third to near zero. Code is available at https://github.com/denoslab/FL-MAESTRO.
CommentsAccepted at IEEE GLOBECOM 2026
