Agentic Federated Learning: Rule-Based Client and Server Agents for Adaptive Training
Authors: Deepthy K. Bhaskar, VP Binu, B Minimol
Organizations: Department of Computer Engineering, Model Engineering College, APJ Abdul Kalam Technological University, Thiruvananthapuram 695016, Kerala, India. · Department of Biomedical Engineering, Model Engineering College, APJ Abdul Kalam Technological University, Thiruvananthapuram 695016, Kerala, India.
Federated Learning (FL) enables collaborative model training across distributed clients without sharing raw data, making it suitable for privacy-sensitive applications such as healthcare, finance, and edge intelligence. However, conventional FL approaches rely on static client participation and fixed aggregation strategies, which limits their effectiveness under non-IID data distributions, heterogeneous client behavior, and noisy or unreliable updates. To overcome these issuess, this paper proposes an Agentic Federated Learning (AFL) framework that integrates lightweight rule-based autonomous agents at both client and server levels. The proposed framework introduces a Client-Side Agent (CSA) that dynamically adapts local training parameters, controls participa- tion, and evaluates update reliability, while a Server-Side Orchestrator Agent (SSOA) performs quality-aware client selection and adaptive aggregation. Unlike traditional FL methods, AFL enables context-aware decision-making during the training process, improving adaptability and robustness in dynamic distributed environments. Extensive experiments conducted on the CIFAR-10 dataset under IID, non-IID, and noisy-client settings demonstrate that AFL consistently outperforms standard base- lines including FedAvg and FedProx. Experimental results show improvements in classification accuracy, convergence speed, robustness against corrupted updates, and communication efficiency. Ablation studies further confirm the complementary contributions of CSA and SSOA, while statistical analysis validates the significance of the observed gains. The proposed AFL framework demonstrates that incorporating autonomous agentic reasoning into federated learning provides an effective and practical solution for intelligent, adaptive, and robust distributed learning systems.
Figures & tables
Figure 1 : Architecture of the proposed Agentic Federated Learning framework consisting of Client-Side Agents and Server-Side Orchestrator Agent.
Symbol
Description
K
Total number of participating clients
Dk
Local dataset of client k
w
Global model parameters
wk
Local client model parameters
F(w)
Global optimization objective
Fk(w)
Local objective function of client k
Table 1 : Notation and symbol definitions
w(t+1)=w(t)−∑k∈Stαk(t)Δwk(t)
Algorithm 1 Agentic Federated Learning (AFL)
Method
IID
Non-IID
Noisy Clients
FedAvg
78.5
64.2
52.7
FedProx
79.1
66.8
55.3
AFL-CSA
80.3
70.1
58.4
AFL-SSOA
81.0
71.4
60.2
Full AFL
83.2
74.9
63.8
Table 2 : Classification accuracy (%) comparison under different federated learning settings
Figure 2 : Convergence comparison of FedAvg, FedProx, and AFL under different federated learning environments.
Method
Accuracy (%)
FedAvg
52.7
FedProx
55.3
AFL-CSA
58.4
AFL-SSOA
60.2
Full AFL
63.8
Table 3 : Accuracy (%) under noisy-client environments
Figure 3 : Communication cost comparison across federated learning methods.
Method
Communication Reduction
FedAvg
Baseline
FedProx
6%
Full AFL
18%
Table 4 : Communication reduction comparison
Configuration
IID
Non-IID
Noisy Clients
FedAvg
78.5
64.2
52.7
FedProx
79.1
66.8
55.3
AFL-CSA Only
80.3
70.1
58.4
AFL-SSOA Only
81.0
71.4
60.2
Full AFL
83.2
74.9
63.8
Table 5 : Ablation analysis of AFL components
Comparison
p -value
Significance
AFL vs FedAvg (IID)
0.031
Significant
AFL vs FedAvg (Non-IID)
0.008
Significant
AFL vs FedAvg (Noisy)
0.004
Significant
AFL vs FedProx (Non-IID)
0.012
Significant
Table 6 : Statistical significance analysis using paired t-test
Federated learning (FL) is a decentralized approach that enables collaborative model training without exposing raw data. Instead of transferring sensitive data, it allows devices to share only model weights, keeping personal data locally and secure. However, in real world settings, the data held by devices is often not evenly distributed and devices mostly differ in computing power and memory capacity. These differences make FL harder to maintain consistent performance across the system. To address these issues, we propose FedMTFI, a novel architecture that combines multi-teacher knowledge distillation (MTKD) with feature importance to improve the FL process in heterogeneous environments. In FedMTFI, clients are clustered based on similar hardware and model types. Each cluster trains a specific model on not independently and identically distributed (non-IID) data. Within a cluster, every client updates that model using only its own local private data. The server then aggregates the locally trained models in each cluster using FedAvg to form multiple prototype models. Then these prototypes serve as teacher models to train a global generalized student model using MTKD. What makes FedMTFI more unique is the integration of Shapley values (SHAP) to emphasize important features during distillation, which enhances both accuracy and interpretability. Experimental results show that FedMTFI achieves higher accuracy than traditional FL algorithms and performs more effectively under non-IID data conditions.
Federated learning enables collaborative model training across distributed clients, yet vanilla FL exposes client updates to the central server. Secure-aggregation schemes protect privacy against an honest-but-curious server, but existing approaches often suffer from many communication rounds, heavy public-key operations, or difficulty handling client dropouts. Recent methods like One-Shot Private Aggregation (OPA) cut rounds to a single server interaction per FL iteration, yet they impose substantial cryptographic and computational overhead on both server and clients. We propose a new protocol called DisAgg that leverages a small committee of clients called Aggregators to perform the aggregation itself: each client secret-shares its update vector to Aggregators, which locally compute partial sums and return only aggregated shares for server-side reconstruction. This design eliminates local masking and expensive homomorphic encryption, reducing endpoint computation while preserving privacy against a curious server and a limited fraction of colluding clients. By leveraging optimal trade-offs between communication and computation costs, DisAgg processes 100k-dimensional update vectors from 100k 5G clients with a 4.6x speedup compared to OPA, the previous best protocol.
Decentralized intelligence systems with heterogeneous devices and limited coordination increasingly rely on decentralized federated learning (DFL). However, DFL suffers from convergence inefficiency under data heterogeneity due to the use of a uniform learning rate (LR) that ignores layer-specific optimization needs. Foundational layers are responsible for maintaining network consensus, while specialized layers adapt to local data characteristics, leading to conflicting gradients and degraded performance under non-IID conditions. To address this fundamental tension, this work introduces FedA2L, a method that dynamically adjusts layer-wise LRs based on model divergence signals. By leveraging local update intensity and network consensus constraints, FedA2L seamlessly integrates into existing DFL protocols without additional communication or coordination. Extensive evaluations across DFL algorithms, various model architectures, and datasets demonstrate that FedA2L achieves up to 4.94 times faster convergence than vanilla DFL and reduces communication rounds by up to 59% compared to scheduler-based baselines. Furthermore, FedA2L exhibits resilience to severe data heterogeneity, larger network sizes, and sparse topologies, reducing communication overhead and establishing it as a versatile optimization tool for resource-constrained or large-scale distributed learning in edge and IoT deployments. The code is released at https://github.com/nclabteam/FedA2L.
Van Truong Vo, Khoa Nguyen, Taehong Kim
School of Information and Communication Engineering, Chungbuk National University, Cheongju, South Korea