Privacy-Preserving Federated Heavy Hitter Analytics for Non-IID Data

TL;DR

This paper presents a new privacy-preserving federated algorithm for heavy hitter analysis in non-IID data, effectively balancing privacy and utility through hierarchical techniques and demonstrating its practical application with FedCampus.

Contribution

The paper introduces a novel hierarchical, privacy-preserving algorithm for federated heavy hitter analytics on non-IID data, addressing privacy and utility challenges.

Findings

Effective identification of local and global heavy hitters in non-IID data

Maintains privacy while achieving high utility in federated settings

Demonstrated success on synthetic and real datasets

Abstract

Federated heavy-hitter analytics involves the identification of the most frequent items within distributed data. Existing methods for this task often encounter challenges such as compromising privacy or sacrificing utility. To address these issues, we introduce a novel privacy-preserving algorithm that exploits the hierarchical structure to discover local and global heavy hitters in non-IID data by utilizing perturbation and similarity techniques. We conduct extensive evaluations on both synthetic and real datasets to validate the effectiveness of our approach. We also present FedCampus, a demonstration application to showcase the capabilities of our algorithm in analyzing population statistics.