Federated Learning Optimization for Privacy-Preserving AI in Cloud Environments

DOI

10.2991/978-94-6463-872-1_51How to use a DOI?

Keywords

Federated Learning; Privacy-Preserving AI; Cloud Computing; Intrusion Detection; Gradient Compression; Differential Privacy; NSL-KDD Dataset; Communication Overhead; Security Optimization

Abstract

The growing dependence on cloud-based AI applications has raised concerns regarding data privacy, security, and computing efficiency. Traditional AI models are susceptible to privacy and security issues due to their reliance on centralized data aggregation. Federated Learning (FL) has emerged as a promising solution that enables decentralized model training without sharing raw data. However, FL faces critical challenges, including communication overhead, slow convergence rates, and privacy leakage risks. This study introduces an optimized FL framework that integrates gradient compression techniques to reduce communication overhead and employs differential privacy mechanisms to enhance data security. We evaluate our approach using the NSL-KDD dataset, which consists of 41 network traffic features. Our experimental results show that the proposed method achieves 97.4% accuracy, surpassing baseline FL techniques such as FedAvg (92.5%), FedAvg with Gradient Compression (93.8%), and FedAvg with Differential Privacy (91.2%). Additionally, our model achieves faster convergence with only 120 communication rounds and significantly reduces privacy leakage to 2.1%. This work proposes a privacy-preserving and scalable FL framework that improves security and efficiency in cloud-based AI environments, providing a workable solution for practical cybersecurity and other applications.

Copyright

© 2025 The Author(s)

Open Access

Open Access This chapter is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International License (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license and indicate if changes were made.

Download article (PDF)

TY  - CONF
AU  - Vineet Kumar Srivastava
AU  - Vishnu Ravi
AU  - Maninder Pal Singh
AU  - Nuzhat Noor Islam Prova
PY  - 2025
DA  - 2025/11/04
TI  - Federated Learning Optimization for Privacy-Preserving AI in Cloud Environments
BT  - Proceedings of the 2nd International Conference on Sustainable Business Practices and Innovative Models (ICSBPIM-2025)
PB  - Atlantis Press
SP  - 825
EP  - 840
SN  - 2352-5428
UR  - https://doi.org/10.2991/978-94-6463-872-1_51
DO  - 10.2991/978-94-6463-872-1_51
ID  - Srivastava2025
ER  -