LiPI: Lightweight Privacy-Preserving Data Aggregation in IoT

TL;DR

LiPI introduces a lightweight, decentralized privacy-preserving data aggregation method for IoT devices that reduces communication and energy consumption while maintaining privacy, suitable for resource-constrained environments.

Contribution

The paper presents LiPI, a novel lightweight, decentralized data obfuscation strategy that leverages synchronous transmission protocols for efficient privacy-preserving aggregation in IoT.

Findings

LiPI achieves at least 51.7% faster performance.

LiPI reduces energy consumption by 50.5%.

Effective privacy preservation in resource-constrained IoT systems.

Abstract

In the modern digital world, a user of a smart system remains surrounded with as well as observed by a number of tiny IoT devices round the clock almost everywhere. Unfortunately, the ability of these devices to sense and share various physical parameters, although play a key role in these smart systems but also causes the threat of breach of the privacy of the users. Existing solutions for privacy-preserving computation for decentralized systems either use too complex cryptographic techniques or exploit an extremely high degree of message passing and hence, are not suitable for the resource-constrained IoT devices that constitute a significant fraction of a smart system. In this work, we propose a novel lightweight strategy LiPI for Privacy-Preserving Data Aggregation in low-power IoT systems. The design of the strategy is based on decentralized and collaborative data obfuscation and does not exploit any dependency on any trusted third party. In addition, besides minimizing the communication requirements, we make appropriate use of the recent advances in Synchronous-Transmission (ST)-based protocols in our design to accomplish the goal efficiently. Extensive evaluation based on comprehensive experiments in both simulation platforms and publicly available WSN/IoT testbeds demonstrates that our strategy works up to at least 51.7% faster and consumes 50.5% lesser energy compared to the existing state-of-the-art strategies.