A privacy-preserving data storage and service framework based on deep learning and blockchain for construction workers' wearable IoT sensors
This work addresses privacy and security issues for construction workers' neurophysiological data, offering a domain-specific solution that is incremental by combining existing deep learning and blockchain methods.
The paper tackles the problem of secure storage and privacy protection for personal neurophysiological data from wearable IoT sensors, proposing a framework that uses generative adversarial networks to generate synthetic signals and blockchain for secure data handling, with results showing superior classification performance when augmenting real EEG data with artificial samples.
Classifying brain signals collected by wearable Internet of Things (IoT) sensors, especially brain-computer interfaces (BCIs), is one of the fastest-growing areas of research. However, research has mostly ignored the secure storage and privacy protection issues of collected personal neurophysiological data. Therefore, in this article, we try to bridge this gap and propose a secure privacy-preserving protocol for implementing BCI applications. We first transformed brain signals into images and used generative adversarial network to generate synthetic signals to protect data privacy. Subsequently, we applied the paradigm of transfer learning for signal classification. The proposed method was evaluated by a case study and results indicate that real electroencephalogram data augmented with artificially generated samples provide superior classification performance. In addition, we proposed a blockchain-based scheme and developed a prototype on Ethereum, which aims to make storing, querying and sharing personal neurophysiological data and analysis reports secure and privacy-aware. The rights of three main transaction bodies - construction workers, BCI service providers and project managers - are described and the advantages of the proposed system are discussed. We believe this paper provides a well-rounded solution to safeguard private data against cyber-attacks, level the playing field for BCI application developers, and to the end improve professional well-being in the industry.