Privacy-Preserving Intrusion Detection for IoT Networks Using Federated Learning with FedAvgM, FedProx, and Best-K Aggregation

Aakash Chouhan; Mukul Shukla; Puja Gupta

Privacy-Preserving Intrusion Detection for IoT Networks Using Federated Learning with FedAvgM, FedProx, and Best-K Aggregation

Author(s)	Mr. Aakash Chouhan, Mr. Mukul Shukla, Ms. Puja Gupta
Country	India
Abstract	In light of the fact that Internet of Things sensor networks are now at the center of essential infrastructure, it is becoming more crucial to address the susceptibility of these networks to assaults at the routing layer. In bandwidth-constrained installations, traditional machine learning techniques establish a single point of failure and add privacy problems by requiring raw network information to transit to a central server. An analysis of the differences between centralized and federated training setups for RPL-based Internet of Things intrusion detection, with the intention of maintaining the localization of data on the device. A deep Multi-Layer Perceptron [512→256→128→64] locally trained on each of 20 client nodes using the RPL-IDS-Beh dataset (158,254 samples, 26 behavioural features). Three server-side techniques were combined: Federated Averaging with Momentum (FedAvgM, β = 0.9), Proximal-term Regularisation (FedProx, μ = 0.01), and Quality-Weighted Best-K Client Selection (16 of 20 clients per round). The federated configuration reached 90.28% accuracy and a macro F1-score of 61.80%, compared to 88.85% accuracy and 52.68% macro F1 for the centralised baseline. Rank-Attack F1 climbed from 0.06 to 0.19, and DIS-Flood detection rose from 0.44 to 0.61.Under the evaluated experimental conditions, the federated configuration achieved higher overall accuracy and macro F1-score relative to the centralised baseline while maintaining on-device data locality.
Keywords	Federated Learning; IoT Intrusion Detection; RPL Protocol; FedAvgM; FedProx; Privacy-Preserving Machine Learning; Neural Networks; Network Security
Published In	Volume 8, Issue 3, May-June 2026
Published On	2026-05-10

View / Download PDF File

E-ISSN 2582-2160

doi

CrossRef DOI is assigned to each research paper published in our journal.

IJFMR DOI prefix is
10.36948/ijfmr

Downloads

Research Paper Format Copyright Permission Form and Undertaking Form Cover Page Vol 8 Isu 2 Cover Page Vol 8 Isu 1 Cover Page Vol 7 Isu 6

All research papers published on this website are licensed under Creative Commons Attribution-ShareAlike 4.0 International License, and all rights belong to their respective authors/researchers.

CC-BY-SA

About IJFMR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+91-9687-828-838	editor@ijfmr.com

International Journal For Multidisciplinary Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Privacy-Preserving Intrusion Detection for IoT Networks Using Federated Learning with FedAvgM, FedProx, and Best-K Aggregation

Share this