REDCATLE: 节点前端IoT流数据实际适用的$k_s$-匿名 (RedCASTLE: Practically Applicable $k_s$-Anonymity for IoT Streaming Data at the Edge in Node-RED)

from arxiv, Accepted for publication as regular research paper for the "8th International Workshop on Middleware and Applications for the Internet of Things". This is a preprint manuscript (authors' own version before final copy-editing)

In this paper, we present RedCASTLE, a practically applicable solution for Edge-based $k_s$-anonymization of IoT streaming data in Node-RED. RedCASTLE builds upon a pre-existing, rudimentary implementation of the CASTLE algorithm and significantly extends it with functionalities indispensable for real-world IoT scenarios. In addition, RedCASTLE provides an abstraction layer for smoothly integrating $k_s$-anonymization into Node-RED, a visually programmable middleware for streaming dataflows widely used in Edge-based IoT scenarios. Last but not least, RedCASTLE also provides further capabilities for basic information reduction that complement $k_s$-anonymization in the privacy-friendly implementation of usecases involving IoT streaming data. A preliminary performance assessment finds that RedCASTLE comes with reasonable overheads and demonstrates its practical viability.

翻译：在本文中,我们介绍RedCASTLE,这是在节点-RED中将IoT流数据以美元为基值的以美元为单位的以美元为单位的匿名化的实用解决方案。RedCASTLE以CASTLE算法的原有初步实施为基础,并大大扩展了该算法,为现实世界IoT情景提供了不可或缺的功能。此外,RedCASTLE为将美元以美元为单位的匿名化顺利融入Node-RED提供了一个抽象层,这是在Edge基于IoT的情况中广泛使用的数据流流流的可视可编程中间软件。最后但并非最不重要的是,RedCASTLE还提供了进一步的基本信息减少能力,补充了在涉及IoT流数据的对隐私友好使用案例实施中以美元为单位的匿名化。一项初步的绩效评估发现,RCASTLE有合理的间接费用,并展示了其实际可行性。