Dynamic Privacy Preservation Architecture for Trustworthy Autonomous Data Exchange Environments
TL;DR
The results show that the proposed architecture is suitable to optimise the data utility-privacy trade-off in absolute terms while preserving strict privacy, thereby providing a reliable, secure and highly scalable ecosystem for autonomous data stakeholders.
Abstract
In autonomous data exchange environments, where data is flowing in real-time, it is crucial to have comprehensive security solutions to ensure user privacy and system performance. This paper presents a Dynamic Privacy Preservation Architecture specifically developed for untrusted, decentralised networks of autonomous nodes that frequently interact. The framework uses adaptive anonymisation algorithms and decentralised trust verification mechanisms to dynamically protect private data tokens based on context sensitivity and the recipient's risk profile. The evaluation of this architecture was conducted through an experimental study using a synthetic dataset from the operational Internet of Things network, comprising 159 distinct communications. Using the Python programming language and Python-specific data science packages such as Pandas for data manipulation, Scikit-learn for metric evaluation, and Matplotlib for visual plotting, system performance, computational overhead and privacy metrics were simulated and analysed. The results show that the proposed architecture is suitable to optimise the data utility-privacy trade-off in absolute terms while preserving strict privacy. The system dynamically scales its defence mechanisms to keep processing latencies low and effectively prevent unauthorised reconstruction attacks, thereby providing a reliable, secure and highly scalable ecosystem for autonomous data stakeholders.