aiqbench analysis
This research introduces "Fulcrum," a novel mechanism for topology-aware differential privacy in hierarchical federated learning that optimizes noise allocation based on regional exposure rather than applying uniform multipliers. By deriving an explicit solution to the min-max optimal allocation problem, the method significantly improves model accuracy—up to 14.84 percentage points on image classification—while maintaining strict privacy guarantees against observers at higher aggregation tiers. The approach provides a closed-form metric called "exposure dispersion," allowing practitioners to evaluate privacy risks and efficiency before training begins without relying on empirical calibration for the worst-case scenario.
Topology-Aware Differential Privacy in Hierarchical Federated Learning
arXiv:2506.19260v3 Announce Type: replace-cross Abstract: Hierarchical federated learning places regional aggregators between clients and the cloud, so a participant's update is observed only alongside its neighbours'. The concealment this arrangement provides depends on the size of the aggregation region, and regions in operational deployments vary widely. Prevailing practice applies a single noise multiplier to every participant, calibrated for the most exposed region, so every other participant carries more noise than its own exposure requires. We show that this allocation problem admits an explicit solution. We first give a silo-level differential privacy guarantee for the mechanism, then bound the mutual information between a participant's local class distribution and any estimate an observer positioned above the regional tier could form of it, using an adjacency notion matched to the quantity being protected. Minimising the worst-case bound under a fixed utility budget yields a min-max optimal allocation, which we call Fulcrum. The budget it recovers has a closed form we term the exposure dispersion, a measure of how unevenly aggregation weight is concentrated within regions relative to the most exposed one. Because this quantity follows from the region structure and the aggregation weights alone, a practitioner can evaluate it before training begins, and it vanishes precisely when all regions are equally exposed. On image and text classification at $\varepsilon = 0.99$, accuracy at a matched worst-case per-client guarantee improves by up to $14.84$ and $12.16$ percentage points where the dispersion is large, and is exactly zero on a balanced control for which the theory predicts parity.
Abstract:Hierarchical federated learning places regional aggregators between clients and the cloud, so a participant's update is observed only alongside its neighbours'. The concealment this arrangement provides depends on the size of the aggregation region, and regions in operational deployments vary widely. Prevailing practice applies a single noise multiplier to every participant, calibrated for the most exposed region, so every other participant carries more noise than its own exposure requires. We show that this allocation problem admits an explicit solution. We first give a silo-level differential privacy guarantee for the mechanism, then bound the mutual information between a participant's local class distribution and any estimate an observer positioned above the regional tier could form of it, using an adjacency notion matched to the quantity being protected. Minimising the worst-case bound under a fixed utility budget yields a min-max optimal allocation, which we call Fulcrum. The budget it recovers has a closed form we term the exposure dispersion, a measure of how unevenly aggregation weight is concentrated within regions relative to the most exposed one. Because this quantity follows from the region structure and the aggregation weights alone, a practitioner can evaluate it before training begins, and it vanishes precisely when all regions are equally exposed. On image and text classification at $\varepsilon = 0.99$, accuracy at a matched worst-case per-client guarantee improves by up to $14.84$ and $12.16$ percentage points where the dispersion is large, and is exactly zero on a balanced control for which the theory predicts parity.
27 pages, 5 figures, 8 tables. Data from the experiments and source code can be found here: this https URL
Cryptography and Security (cs.CR); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
I.2.6; C.2.4; K.6.5
arXiv:2506.19260 [cs.CR]
(or arXiv:2506.19260v3 [cs.CR] for this version)
https://doi.org/10.48550/arXiv.2506.19260 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 373 words
Related coverage
News
Quantum Computing Report · Jul 25, 2026
A Constraint-Aware Generative Framework for Synthetic Origin-Destination Demand in Logistics Networks
arXiv · Sep 7, 2026
Structural Entropy-Driven Graph Diffusion Generation for One-Shot Federated Graph Learning
arXiv · Sep 9, 2026
Graph-Based Inference and Topology-Aware Multi-Agent Reinforcement Learning for Large-Scale Railway Network Management
arXiv · Sep 26, 2026