基于GRU的智能手机多场景步态认证

doi:10.11959/j.issn.2096-109x.2022060

网络与信息安全学报 ›› 2022, Vol. 8 ›› Issue (5): 26-39.doi: 10.11959/j.issn.2096-109x.2022060

• 专题：大数据与人工智能安全 • 上一篇下一篇

基于GRU的智能手机多场景步态认证

姜奇¹^,², 冯茹¹, 张瑞杰³, 王金花⁴, 陈婷⁴, 魏福山⁵

¹ 西安电子科技大学网络与信息安全学院, 陕西西安 710126
² 河南省网络密码技术重点实验室，河南郑州 450001
³ 信息工程大学密码工程学院，河南郑州 450001
⁴ 保密通信重点实验室，四川成都 610041
⁵ 信息工程大学网络空间安全学院，河南郑州 450001

修回日期:2022-07-22 出版日期:2022-10-15 发布日期:2022-10-01
作者简介:姜奇（1983- ），男，安徽全椒人，西安电子科技大学教授、博士生导师，主要研究方向为密码协议，物联网安全
冯茹（1997- ），女，山西晋城人，西安电子科技大学硕士生，主要研究方向为生物认证
张瑞杰（1984- ），女，河南郑州人，信息工程大学讲师，主要研究方向为人工智能、网络信息防御等
王金花（1995- ），女，山西太原人，保密通信重点实验室助理工程师，主要研究方向为密码协议设计与分析
陈婷（1995- ），女，江西赣州人，保密通信重点实验室助理工程师，主要研究方向为网络安全、密码协议设计与分析
魏福山（1983- ），男，甘肃武威人，信息工程大学副教授、博士生导师，主要研究方向为安全协议设计与分析
基金资助:
国家自然科学基金重大研究计划(92167203);国家自然科学基金(62072352);国家自然科学基金(62125205);陕西省教育厅科研计划项目(20JY016);陕西省重点产业链项目(2020ZDLGY09-06)

GRU-based multi-scenario gait authentication for smartphones

Qi JIANG¹^,², Ru FENG¹, Ruijie ZHANG³, Jinhua WANG⁴, Ting CHEN⁴, Fushan WEI⁵

¹ School of Cyber Engineering, Xidian University, Xi’an 710126, China
² Henan Key Laboratory of Network Cryptography Technology, Zhengzhou 450001, China
³ College of Cryptographic Engineering, Information Technology University, Zhengzhou 450001, China
⁴ Science and Technology on Communication Security Laboratory, Chengdu 610041, China
⁵ School of Cyber Science and Engineering, Information Technology University, Zhengzhou 450001, China

Revised:2022-07-22 Online:2022-10-15 Published:2022-10-01
Supported by:
The Major Research Plan of the National Natural Science Foundation of China(92167203);The National Natural Science Foundation of China(62072352);The National Natural Science Foundation of China(62125205);Scientific Research Program Funded by Education Department of Shaanxi Provincial Government(20JY016);Key Industrial Chain Projects in Shaanxi Province(2020ZDLGY09-06)

摘要/Abstract

摘要：

基于步态的智能手机认证研究大多针对单一受控场景，而未考虑多场景变化对认证准确性的影响。智能手机以及用户的运动方向在不同场景下会发生变化，当使用对方向变化敏感的传感器采集用户步态数据时，可能会因场景的改变出现一定的偏差。因此，为智能手机提供一种多场景下的高精度步态认证方法已成为亟待解决的问题。此外，模型训练算法的选取是决定步态认证准确率和效率的关键。目前流行的基于长短期记忆（LSTM，long short-term memory）网络的认证模型能够实现较高的认证准确率，但其训练参数较多，内存占用较大，训练效率有待提升。针对现有步态认证方案未满足多场景认证需求、模型认证和训练且难以兼顾高效率与高准确率的问题，提出了基于门控循环单元（GRU，gate recurrent unit）的智能手机多场景步态认证方案。通过小波变换对步态信号进行初步降噪处理，并采用自适应的步态周期分割算法对循环的步态信号进行切分。为满足多步态场景的认证需求，采用坐标系转换方法对步态信号进行方向无关性处理，以消除智能手机方向以及用户运动方向对认证结果的影响。为实现高准确率认证以及高效率训练模型，利用不同体系结构的 GRU 以及多种优化方式训练用户步态模型。在公开数据集 PSR 和ZJU-GaitAcc上对所提方案进行实验分析。与所列方案对比，所提方案提高了认证准确率，较之基于LSTM的步态认证模型，所提模型的训练效率提升了约20%。

关键词: 持续认证, 步态行为, 多传感器, 门控循环单元

Abstract:

At present, most of the gait-based smartphone authentication researches focus on a single controlled scenario without considering the impact of multi-scenario changes on the authentication accuracy.The movement direction of the smartphone and the user changes in different scenarios, and the user’s gait data collected by the orientation-sensitive sensor will be biased accordingly.Therefore, it has become an urgent problem to provide a multi-scenario high-accuracy gait authentication method for smartphones.In addition, the selection of the model training algorithm determines the accuracy and efficiency of gait authentication.The current popular authentication model based on long short-term memory (LSTM) network can achieve high authentication accuracy, but it has many training parameters, large memory footprint, and the training efficiency needs to be improved.In order to solve the above problems a multi-scenario gait authentication scheme for smartphones based on Gate Recurrent Unit (GRU) was proposed.The gait signals were preliminarily denoised by wavelet transform, and the looped gait signals were segmented by an adaptive gait cycle segmentation algorithm.In order to meet the authentication requirements of multi-scenario, the coordinate system transformation method was used to perform direction-independent processing on the gait signals, so as to eliminate the influence of the orientation of the smartphone and the movement of the user on the authentication result.Besides, in order to achieve high-accuracy authentication and efficient model training, GRUs with different architectures and various optimization methods were used to train the gait model.The proposed scheme was experimentally analyzed on publicly available datasets PSR and ZJU-GaitAcc.Compared with the related schemes, the proposed scheme improves the authentication accuracy.Compared with the LSTM-based gait authentication model, the training efficiency of the proposed model is improved by about 20%.

Key words: continuous authentication, gait behavior, multi-sensor, GRU

中图分类号:

TP393

姜奇, 冯茹, 张瑞杰, 王金花, 陈婷, 魏福山. 基于GRU的智能手机多场景步态认证[J]. 网络与信息安全学报, 2022, 8(5): 26-39.

Qi JIANG, Ru FENG, Ruijie ZHANG, Jinhua WANG, Ting CHEN, Fushan WEI. GRU-based multi-scenario gait authentication for smartphones[J]. Chinese Journal of Network and Information Security, 2022, 8(5): 26-39.

图/表 8

图1

图2

图3

图4

表1

表2

图5

表3

参考文献 30

[1]	SUN F , MAO C , FAN X ,et al. Accelerometer-based speed adaptive gait authentication method for wearable IoT devices[J]. IEEE Internet of Things Journal, 2018,6(1): 820-830.
[2]	姜奇, 文悦, 张瑞杰 ,等. 面向智能手机的自适应触屏持续认证方案[J]. 电子学报, 2022,50(5): 1131-1139.
	JIANG Q , WEN Y , ZHANG R J ,et al. Adaptive touch screen co-tinuous authentication scheme for smartphones[J]. Acta Electronica Sinica, 2022,50(5): 1131-1139.
[3]	JIANG Q , HUANG X H , ZHANG N ,et al. Shake to communicate:secure handshake acceleration-based pairing mechanism for wrist worn devices[J]. IEEE Internet of Things Journal, 2019,6(3): 5618-5630.
[4]	ZHU T T , QU Z Y , XU H T ,et al. RiskCog:unobtrusive real-time user authentication on mobile devices in the wild[J]. IEEE Transactions on Mobile Computing, 2020,19(2): 466-483.
[5]	蒋伟, 王瑞锦, 余苏喆 ,等. 基于步态识别的移动设备身份认证模型[J]. 电子科技大学学报, 2019,48(2): 272-277.
	JIANG W , WANG R J , YU S Z ,et al. Research on identity authen-tication model of mobile devices based on gait recognition[J]. Journal of University of Electronic Science and Technology of China, 2019,48(2): 272-277.
[6]	COLA G , AVVENUTI M , MUSSO F ,et al. Gait-based authentication using a wrist-worn device[C]// Proceedings of the 13th International Conference on Mobile and Ubiquitous Systems:Computing,Networking and Services. 2016: 208-217.
[7]	MUAAZ M , MAYRHOFER R . Smartphone-based gait recognition:From authentication to imitation[J]. IEEE Transactions on Mobile Computing, 2017,16(11): 3209-3221.
[8]	杨力, 马卓茹, 张程辉 ,等. 基于步态特征的移动平台持续认证方案[J]. 通信学报, 2019,40(7): 126-134.
	YANG L , MA Z R , ZHANG C H ,et al. Mobile platform continu-ous authentication scheme based on gait characteristics[J]. Journal on Communications, 2019,40(7): 126-134.
[9]	THANG H M , VIET V Q , DINH THUC N ,et al. Gait identification using accelerometer on mobile phone[C]// 2012 International Conference on Control,Automation and Information Sciences (ICCAIS). 2012: 344-348.
[10]	HESTBEK M R , NICKEL C , BUSCH C . Biometric gait recognition for mobile devices using wavelet transform and sup-port vector machines[C]// 2012 19th International Conference on Systems,Signals and Image Processing (IWS-SIP). 2012: 205-210.
[11]	HOANG T , CHOI D , NGUYEN T . On the instability of sensor orientation in gait verification on mobile phone[C]// 2015 12th International Joint Conference on e-Business and Telecommunications (ICETE). 2015: 148-159.
[12]	CHOI S , YOUN I H , LEMAY R ,et al. Biometric gait recognition based on wireless acceleration sensor using k-nearest neighbor classification[C]// 2014 International Conference on Computing,Networking and Communications (ICNC). 2014: 1091-1095.
[13]	FERNANDEZ-LOPEZ P , LIU-JIMENEZ J , KIYOKAWA K ,et al. Recurrent neural network for inertial gait user recognition in smartphones[J]. Sensors, 2019,19(18): 4054.
[14]	GIORGI G , MARTINELLI F , SARACINO A ,et al. Walking through the deep:gait analysis for user authentication through deep learning[C]// IFIP International Conference on ICT Systems Security and Privacy Protection. 2018: 62-76.
[15]	ZHANG M , . Gait activity authentication using LSTM neural networks with smartphone sensors[C]// 2019 15th International Conference on Mobile Ad-Hoc and Sensor Networks (MSN). 2019: 456-461.
[16]	GAFUROV D , HELKALA K , SONDROL T . Biometric gait authentication using accelerometer sensor[J]. Journal of Computers, 2006,1(7): 51-59.
[17]	GAFUROV D , SNEKKENES E , BOURS P ,et al. Gait authentication and identification using wearable accelerometer sensor[C]// 2007 IEEE Workshop on Automatic Identification Advanced Technologies. 2007: 220-225.
[18]	MANTYJARVI J , LINDHOLM M , VILDJIOUNAITE E ,et al. Identifying users of portable devices from gait pattern with accelerometers[C]// Proceedings of IEEE International Conference on Acoustics,Speech,and Signal Processing. 2005: 973-976.
[19]	XU W , SHEN Y , ZHANG Y ,et al. Gait-watch:a context-aware authentication system for smart watch based on gait recognition[C]// Proceedings of the Second International Conference on Internet-of-Things Design and Implementation. 2017: 59-70.
[20]	XU W , REVADIGAR G , LUO C ,et al. Walkie-talkie:Motion-assisted automatic key generation for secure on-body device communication[C]// 2016 15th ACM/IEEE International Conference on Information Processing in Sensor Net-works (IPSN). 2016: 1-12.
[21]	GADALETA M , ROSSI M . IDNet:Smartphone-based gait recognition with convolutional neural networks[J]. Pattern Recognition, 2018,74: 25-37.
[22]	FERREIRA A , SANTOS G , ROCHA A ,et al. User-centric coordinates for applications leveraging 3-axis accelerometer data[J]. IEEE Sensors Journal, 2017,17(16): 5231-5243.
[23]	ABUHAMAD M , ABUHMED T , MOHAISEN D ,et al. AUToSen:deep learning-based implicit continuous authentication using smartphone sensors[J]. IEEE Internet of Things Journal, 2020,7(6): 5008-5020.
[24]	ZOU Q , NI L , WANG Q ,et al. Robust gait recognition by integrating inertial and RGBD sensors[J]. IEEE Transactions on Cybernetics, 2018,48(4): 1136-1150.
[25]	FONG W T , ONG S K , NEE A Y C . Methods for in-field user calibration of an inertial measurement unit without external equipment[J]. Measurement Science and Technology, 2008,19(8): 085202.
[26]	戴邵武, 王克红, 庆曾宏 . 基于混合 PSO 算法的加速度计快速标定方法[J]. 压电与声光, 2015,37(2): 232-236.
	DAI S W , WANG K H , QING Z H . Rapid calibration of accel-erometer based on hybrid particle swarm optimization algorithm[J]. Piezoelectrics ＆ Acoustooptics, 2015,37(2): 232-236.
[27]	KINGMA D P , BA J . Adam:A method for stochastic optimization[C]// Proceedings of the 2015 International Conference on Learning Representations. 2015: 1-15.
[28]	SRIVASTAVA N , HINTON G , KRIZHEVSKY A ,et al. Dropout:A simple way to prevent neural networks from overfitting[J]. Journal of Machine Learning Research, 2014,15(1): 1929-1958.
[29]	SHOAIB M , BOSCH S , INCEL O ,et al. Fusion of smartphone motion sensors for physical activity recognition[J]. Sensors, 2014,14(6): 10146-10176.
[30]	ZHANG Y , PAN G , JIA K ,et al. Accelerometer-based gait recognition by sparse representation of signature points with clusters[J]. IEEE Transactions on Cybernetics, 2014,45(9): 1864-1875.

放置部位	步态场景		GRU			LSTM
放置部位	步态场景	准确率	EER	AUC	准确率	EER	AUC
	步行	94.41%	6.45%	97.33%	92.47%	8.62%	95.26%
左口袋	慢跑	95.68%	5.32%	98.73%	95.53%	3.91%	98.82%
	上楼	89.05%	13.21%	88.34%	90.17%	14.22%	90.41%
	下楼	87.42%	12.65%	82.12%	87.50%	13.86%	83.77%
	步行	94.41%	12.72%	93.65%	92.47%	13.31%	92.38%
右口袋	慢跑	95.90%	4.91%	99.52%	95.64%	2.49%	98.72%
	上楼	89.66%	16.73%	92.33%	87.27%	15.62%	87.54%
	下楼	82.74%	14.24%	83.98%	82.97%	14.73%	84.13%
	步行	95.26%	4.48%	98.79%	94.90%	6.42%	97.53%
腕部	慢跑	93.22%	3.91%	98.89%	97.99%	3.83%	98.92%
	上楼	88.74%	14.11%	87.25%	88.68%	15.12%	88.33%
	下楼	84.73%	15.98%	82.54%	85.92%	16.23%	81.77%
	步行	93.82%	12.34%	94.21%	89.86%	12.17%	94.78%
上臂	慢跑	97.03%	3.47%	98.97%	96.73%	6.38%	97.35%
	上楼	89.90%	13.92%	87.88%	90.21%	14.17%	86.31%
	下楼	85.95%	15.42%	83.47%	87.20%	15.66%	82.79%
	步行	96.34%	6.71%	96.89%	95.47%	6.39%	97.22%
腰部	慢跑	93.43%	4.23%	98.92%	93.60%	7.21%	96.54%
	上楼	96.37%	2.76%	98.33%	94.90%	5.91%	96.97%
	下楼	95.85%	4.11%	99.32%	94.09%	6.12%	98.23%

放置部位	会话状态		GRU			LSTM
放置部位	会话状态	准确率	EER	AUC	准确率	EER	AUC
上臂	会话内	85.21%	13.28%	85.74%	86.18%	12.88%	86.08%
	会话间	78.59%	19.86%	79.11%	77.36%	20.13%	78.25%
手腕	会话内	88.47%	10.74%	88.56%	87.25%	11.21%	87.42%
	会话间	80.53%	17.49%	81.22%	81.89%	16.78%	81.68%
骨盆	会话内	93.22%	9.28%	92.88%	93.76%	8.94%	93.41%
	会话间	84.68%	13.77%	85.14%	85.26%	13.16%	85.49%
大腿	会话内	94.58%	8.75%	94.51%	94.31%	9.11%	93.79%
	会话间	88.57%	12.64%	88.79%	88.72%	12.47%	88.90%
脚踝	会话内	91.16%	9.87%	91.49%	91.82%	9.76%	91.73%
	会话间	82.62%	14.50%	81.88%	81.98%	14.69%	81.65%

方案	步态活动	手机放置部位	用户数	准确率
文献[5]	步行、慢跑、上楼梯、下楼梯	固定部位	未公开	85.79%（步行）
				83.70%（慢跑）
				85.00%（上楼梯）
				80.00%（下楼梯）
文献[8]	步行、上楼梯、下楼梯	手持	20	94.86%（步行）
				87.77%（上楼梯）
				89.42%（下楼梯）
文献[1]	步行	左上臂、右手腕、右骨盆、左大腿、右脚踝	153	91.75%
文献[30]	步行	左上臂、右手腕、右骨盆、左大腿、右脚踝	153	95.80%
本文方案	步行、慢跑、上楼梯、下楼梯	左口袋、右口袋、腕部、上臂、腰部	10	96.83%（步行）
				95.66%（慢跑）
				94.96%（上楼梯）
				94.74%（下楼梯）
	步行	左上臂、右手腕、右骨盆、左大腿、右脚踝	153	96.83%

基于GRU的智能手机多场景步态认证

GRU-based multi-scenario gait authentication for smartphones

在线阅读

pdf下载

可视化

摘要/Abstract

引用本文

使用本文

图/表 8

参考文献 30

相关文章 15

Metrics

推荐阅读 0

[1]	陈先意, 顾军, 颜凯, 江栋, 许林峰, 付章杰. 针对车牌识别系统的双重对抗攻击[J]. 网络与信息安全学报, 2023, 9(3): 16-27.
[2]	叶天鹏, 林祥, 李建华, 张轩凯, 许力文. 面向雾计算的个性化轻量级分布式网络入侵检测系统[J]. 网络与信息安全学报, 2023, 9(3): 28-37.
[3]	祖立军, 曹雅琳, 门小骅, 吕智慧, 叶家炜, 李泓一, 张亮. 基于隐私风险评估的脱敏算法自适应方法[J]. 网络与信息安全学报, 2023, 9(3): 49-59.
[4]	夏锐琪, 李曼曼, 陈少真. 基于机器学习的分组密码结构识别[J]. 网络与信息安全学报, 2023, 9(3): 79-89.
[5]	袁静怡, 李子川, 彭国军. EN-Bypass：针对邮件代发提醒机制的安全评估方法[J]. 网络与信息安全学报, 2023, 9(3): 90-101.
[6]	余锋, 林庆新, 林晖, 汪晓丁. 基于生成对抗网络的隐私增强联邦学习方案[J]. 网络与信息安全学报, 2023, 9(3): 113-122.
[7]	朱春陶, 尹承禧, 张博林, 殷琪林, 卢伟. 基于多域时序特征挖掘的伪造人脸检测方法[J]. 网络与信息安全学报, 2023, 9(3): 123-134.
[8]	李晓萌, 郭玳豆, 卓训方, 姚恒, 秦川. 载体独立的抗屏摄信息膜叠加水印算法[J]. 网络与信息安全学报, 2023, 9(3): 135-149.
[9]	蔡召, 荆涛, 任爽. 以太坊钓鱼诈骗检测技术综述[J]. 网络与信息安全学报, 2023, 9(2): 21-32.
[10]	潘雁, 林伟, 祝跃飞. 渐进式的协议状态机主动推断方法[J]. 网络与信息安全学报, 2023, 9(2): 81-93.
[11]	杨盼, 康绯, 舒辉, 黄宇垚, 吕小少. 基于函数摘要的二进制程序污点分析优化方法[J]. 网络与信息安全学报, 2023, 9(2): 115-131.
[12]	肖天, 江智昊, 唐鹏, 黄征, 郭捷, 邱卫东. 基于深度强化学习的高性能导向性模糊测试方案[J]. 网络与信息安全学报, 2023, 9(2): 132-142.
[13]	袁承昊, 李勇, 任爽. 多关键词动态可搜索加密方案[J]. 网络与信息安全学报, 2023, 9(2): 143-153.
[14]	侯泽洲, 任炯炯, 陈少真. 基于神经网络区分器的SIMON-like算法参数安全性评估[J]. 网络与信息安全学报, 2023, 9(2): 154-163.
[15]	郭学镜, 方毅翔, 赵怡, 张天助, 曾文超, 王俊祥. 基于传统引导机制的深度鲁棒水印算法[J]. 网络与信息安全学报, 2023, 9(2): 175-183.