Gait-Based Privacy Protection for Smart Wearable Devices

TL;DR

This work tackles privacy risks in gait-based identification on smart wearables by introducing AmSoBe, a three-component framework that combines an ABLSTM-based gait-identification network with a CPA-resistant stochastic orthogonal transformation (SOT) for feature privacy and a biometric-based encryption (BBE) scheme for secure message exchange. The ABLSTM achieves $95.28\%$ accuracy, significantly reducing prior error rates, while SOT protects gait features against leakage and CPA attacks with about $30\%$ faster performance. The BBE component leverages biometric distance and identity-based encryption (IBE) to enable secure communications using gait-derived keys, with security grounded in the DBDH assumption. Together, AmSoBe provides high-accuracy gait recognition and end-to-end privacy guarantees for SWDs, addressing both template security and secure inter-device interactions in practical IoT-enabled wearables.

Abstract

Smart wearable devices (SWDs) collect and store sensitive daily information of many people. Its primary method of identification is still the password unlocking method. However, several studies have shown serious security flaws in that method, which makes the privacy and security concerns of SWDs particularly urgent. Gait identification is well suited for SWDs because its built-in sensors can provide data support for identification. However, existing gait identification methods have low accuracy and neglect to protect the privacy of gait features. In addition, the SWD can be used as an internet of things device for users to share data. But few studies have used gait feature-based encryption schemes to protect the privacy of message interactions between SWDs and other devices. In this paper, we propose a gait identification network, a bi-directional long short-term memory network with an attention mechanism (ABLSTM), to improve the identification accuracy and a stochastic orthogonal transformation (SOT) scheme to protect the extracted gait features from leakage. In the experiments, ABLSTM achieves an accuracy of 95.28%, reducing previous error rate by 19.3%. The SOT scheme is proved to be resistant to the chosen plaintext attack (CPA) and is 30% faster than previous methods. A biometric-based encryption scheme is proposed to enable secure message interactions using gait features as keys after the gait identification stage is passed, and offers better protection of the gait features compared to previous schemes.

Figures (12)

• Figure 1: The process of the SWD gait registration stage and identification stage.
• Figure 2: Gait cycles distribution of the OU-ISIR and the whuGait databases.
• Figure 3: An example segmentation of a 6-channel gait signal. The upper (lower) part corresponds to three-axis acceleration (gyroscope) signals.
• Figure 4: The structure of the ABLSTM network.
• Figure 5: The structure of the BLSTM.