Privacy-utility Trade-offs in IoT Networks: A Comparative Analysis of Differential Privacy Mechanisms for Sensor Data Aggregation
Oleksandr Kuznetsov, Oleksii A. Smirnov, Tatyana Kuznetsova, Aigul Shaikhanova, Igor Svatowsky · River Publishers eBooks · 2025
The rapid growth of Internet of Things (IoT) networks has heightened concerns about data privacy in sensor-based systems. This chapter offers a thorough empirical evaluation of differential privacy mechanisms for IoT data aggregation, emphasizing the essential balance between preserving privacy and maintaining data utility. We analyze four privacy mechanisms – Laplace, Gaussian, Exponential, and Stochastic – across various sensor types and aggregation functions in a large-scale network consisting of 1000 nodes. Our experimental framework takes into account real-world sensor data characteristics, including temperature (18–28 °C), humidity (30%–70%), and 590 energy consumption (0–1000 W) measurements. Through rigorous statistical analysis, we demonstrate that the Stochastic mechanism delivers superior performance, achieving a mean relative error of 0.001304 for mean aggregation, and significantly outperforms traditional approaches. Our findings indicate that the effectiveness of these mechanisms is strongly linked to both data characteristics and aggregation functions, with wide-ranging measurements posing greater challenges for privacy preservation. We provide empirical evidence that effective privacy preservation in IoT networks necessitates a context-aware approach, taking into consideration the relationship between mechanism selection, data characteristics, and aggregation methods. These results lay the groundwork for implementing differential privacy in IoT systems while ensuring optimal trade-offs between utility and privacy.