Paper 2023/141

A Secure Bandwidth-Efficient Treatment for Dropout-Resistant Time-Series Data Aggregation

Abstract

Aggregate statistics derived from time-series data collected by individual users are extremely beneficial in diverse fields, such as e-health applications, IoT-based smart metering networks, and federated learning systems. Since user data are privacy-sensitive in many cases, the untrusted aggregator may only infer the aggregation without breaching individual privacy. To this aim, secure aggregation techniques have been extensively researched over the past years. However, most existing schemes suffer either from high communication overhead when users join and leave, or cannot tolerate node dropouts. In this paper, we propose a dropout-resistant bandwidth-efficient time-series data aggregation. The proposed scheme does not incur any interaction among users, involving a solo round of user→aggregator communication exclusively. Additionally, it does not trigger a re-generation of private keys when users join and leave. Moreover, the aggregator is able to output the aggregate value by employing the re-encrypt capability acquired during a one-time setup phase, notwithstanding the number of nodes in the ecosystem that partake in the data collection of a certain epoch. Dropout-resistancy, trust-less key management, low-bandwidth and non-interactive nature of our construction make it ideal for many rapid-changing distributed real-world networks. Other than bandwidth efficiency, our scheme has also demonstrated efficiency in terms of computation overhead