Article Details

Abstract

Smartwatches are equipped with motion sensors such as accelerometers and gyroscopes, and they move with hand movements, which makes it possible to steal a user's external keyboard input through a smartwatch. This study focuses on the side-channel privacy leakage of motion sensors in smartwatches. It combines theoretical analysis with experimental verification, reviews the relevant leakage scenarios, propagation paths and risk levels, and constructs Machine-Learning-based side-channel attack models for testing. The results show that motion sensors expose sensitive privacy information of users under specific conditions, and the Convolutional Neural Network (CNN) attack model achieves the highest recognition success rate of 91.2%, which indicates that this type of privacy threat is highly realistic. Based on these findings, this study further proposes a comprehensive protection scheme at the hardware, software and user levels, which can suppress the risk of privacy leakage to a certain extent. The conclusions provide a theoretical basis for the privacy and security protection of motion sensors in smartwatches, and also provide a reference for subsequent practical applications.

Keywords

smartwatch; motion sensor; side-channel attack; privacy leakage; privacy protection

References

1. [1] Song Chengguang, Liu Jianwei, Wu Qianhong, etc A new type of attack and prevention based on smart watch motion sensors [J]. Journal of Communications, 2015, 36 (S1): 235-242
2. [2] Liu Qiang, Li Tong, Yu Yang, etc Overview of Data Security and Privacy Protection Technologies for Wearable Devices [J]. Computer Research and Development, 2018, 55 (1): 14-29
3. [3] Ni Xueli, Wang Qun, Liang Guangjun Research on Security and Privacy Threats of Intelligent Wearable Devices [J]. Information Network Security, 2022, 22 (10): 98-107
4. [4] Feng Dengguo, Zhang Min, Li Hao Big Data Security and Privacy Protection [J]. Journal of Computer Science, 2014, 37 (1): 246-258
5. [5] Zhao Ge, Zheng Yang, Tao Zelin Analysis of System Risk Assessment for Intelligent Wearable Devices [J]. Information Network Security, 2024, 24 (10): 1595-1603
6. [6] Wang Yongjuan, Fan Haopeng, Dai Zhengyi, etc Research progress on side channel attack and defense technology [J]. Journal of Computer Science, 2023, 46 (1): 2022-228
7. [7] Dai Xiaolu, Wang Tinghua, Hu Zhenwei Research progress on fuzzy multi-core support vector machine [J]. Computer Application Research, 2021, 38 (10): 2896-2903
8. [8] Lv Hongyan, Feng Qian A Review of Random Forest Algorithm Research [J]. Journal of Hebei Academy of Sciences, 2019, 36 (3): 37-41
9. [9] Zhou Feiyan, Jin Linpeng, Dong Jun A Review of Research on Convolutional Neural Networks [J]. Journal of Computer Science, 2017, 40 (6): 1229-1251
10. [10] Song Lei, Ma Chunguang, Duan Guanghan Research progress on machine learning security and privacy protection [J]. Journal of Network and Information Security, 2018, 4 (8): 1-11
11. [11] Tang Benxiao, Wang Lina, Wang Run, etc Android Physical Sensor Side Channel Defense Method Based on Differential Privacy [J]. Computer Research and Development, 2018, 55 (7): 1371-1392
12. [12] Tang Benxiao, Wang Lina, Wang Run, etc A universal motion sensor side channel defense scheme based on Laplace mechanism [J]. Journal of Software, 2019, 30 (8): 2392-2414
13. [13] Zhang Yan, Xie Hong Research progress on sensing technology of intelligent wearable flexible data gloves [J]. Journal of Command and Control, 2024, 10 (3): 264-275

About the authors

Dai Guangzhao (1990-), male, Han ethnicity, from Xuzhou, Jiangsu, engineer, master's student, research interests include information security and computer technology; Kong Juan (1987-), female, Han ethnicity, from Yangzhou, Jiangsu, confidentiality officer, bachelor's degree, research interests include confidentiality management.