Date of Submission
7-2025
Date of Award
7-23-2025
Institute Name (Publisher)
Indian Statistical Institute
Document Type
Master's Dissertation
Degree Name
Master of Technology
Subject Name
Cryptology
Department
Cryptology and Security Research Unit (CSRU-Kolkata)
Supervisor
Sasmal, Pradip
Co-Supervisor (if any)
Molla, Anisur Rahman
Abstract (Summary of the Work)
Secure aggregation is a critical component of privacy-preserving federated learning. However, existing fixed-sparsity approaches often incur unnecessary communication overhead. We present DynamicSecAgg, a novel framework that introduces dynamic sparsity while preserving coordinate-level privacy. Our method achieves significant improvements in communication efficiency while maintaining — and in some cases improving — model accuracy across both IID and non-IID user distributions. The framework maintains information-theoretic privacy guarantees via adaptive gradient thresholding and polynomial-based aggregation, proving particularly effective under heterogeneous data settings. These results establish dynamic sparsity as a key optimization for efficient and privacy-preserving federated learning.
Control Number
2023-25
DOI
https://dspace.isical.ac.in/items/747171a8-c428-47ae-9f87-dab53d2e4f3a/full
DSpace Identifier
http://hdl.handle.net/10263/7653
Recommended Citation
Samanta, Bikash, "Dynamic Sparsification in Secure Gradient Aggregation for Federated Learning" (2025). Master’s Dissertations. 414.
https://digitalcommons.isical.ac.in/masters-dissertations/414