Information-Theoretic Secure Aggregation With Heterogeneous Client Quantization
Abstract
In federated learning (FL), secure aggregation enables a server to compute the sum of local model updates of N clients, without revealing any information on the individual model updates to the server, beyond their sum. Different clients may be subject to heterogeneous bandwidth constraints, and consequently may choose different numbers of quantization levels to represent the information they send. Most secure aggregation works require the clients to project the values of their update parameters to a common finite group (or field), which does not directly accommodate heterogeneous precision constraints on their transmitted masked updates. Existing approaches of secure aggregation with heterogeneous quantization levels reveal partial aggregates of model-updates, and can reveal segments of individual model updates, under certain dropout and client-server collusion patterns. To prevent this privacy breach, we propose an information-theoretic secure aggregation scheme that tolerates client drop-outs, and up to T semi-honest clients that collude with each other and with the server, under heterogeneous quantization at clients. Our protocol utilizes the (T+1, N) threshold black-box secret sharing construction, where secret shares of the client updates, despite belonging to distinct algebraic groups owing to different quantization, are generated from identical linear integer systems.
Related
- Fundamental Limits of Information-Theoretic Secure Aggregation with User Dropouts in Federated Learning
- SILMARILS-MV: Information-Theoretically Secure Multi-Verifier Signatures with Dealer-Free Key Issuance and Linear Aggregation for Permissioned Blockchains
- Progressive-Resolution Secure Aggregation for Federated Learning
- Measuring Learned Equivariance: An Information Theoretic Approach
- A Sample-Based Approach for Hierarchical Information-Theoretic Compression of Probabilistic Occupancy Grids
BibTeX
@misc{meel2026informationtheoretic,
title = {{Information-Theoretic Secure Aggregation With Heterogeneous Client Quantization}},
author = {Shreya Meel and Sennur Ulukus and Itzhak Tamo},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.35833},
eprint = {2610.35833},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.35833},
}
RIS
TY - UNPB
TI - Information-Theoretic Secure Aggregation With Heterogeneous Client Quantization
AU - Shreya Meel
AU - Sennur Ulukus
AU - Itzhak Tamo
PY - 2026
DA - 2026/10/26
PB - arXiv
DO - 10.48550/arxiv.2610.35833
UR - https://arxiv.org/abs/2610.35833
AB - In federated learning (FL), secure aggregation enables a server to compute the sum of local model updates of N clients, without revealing any information on the individual model updates to the server, beyond their sum. Different clients may be subject to heterogeneous bandwidth constraints, and consequently may choose different numbers of quantization levels to represent the information they send. Most secure aggregation works require the clients to project the values of their update parameters to a common finite group (or field), which does not directly accommodate heterogeneous precision constraints on their transmitted masked updates. Existing approaches of secure aggregation with heterogeneous quantization levels reveal partial aggregates of model-updates, and can reveal segments of individual model updates, under certain dropout and client-server collusion patterns. To prevent this privacy breach, we propose an information-theoretic secure aggregation scheme that tolerates client drop-outs, and up to T semi-honest clients that collude with each other and with the server, under heterogeneous quantization at clients. Our protocol utilizes the (T+1, N) threshold black-box secret sharing construction, where secret shares of the client updates, despite belonging to distinct algebraic groups owing to different quantization, are generated from identical linear integer systems.
ER -