Deep Belief Network Model for Mobile Terminal Identity Authentication

doi:10.3969/j.issn.1671-1122.2019.03.005

Abstract

Abstract:

Aiming at the information security problem faced by mobile terminal, a deep belief network model for mobile terminal identity authentication is established. The touch screen sensor is used to collect the user’s original touch gesture data sequence. After the data is pre-processed, the gesture features are extracted and passed into the deep belief network model. The layer-by-layer greedy algorithm is used for unsupervised pre-training, and then the back-propagation algorithm is used to supervise and fine-tune the fixed model parameters. The gesture feature data is used as the model input layer, and the output layer data is obtained after the model is calculated, and the output data is classified and authenticated by the Softmax classifier. Compared with the continuous hidden Markov model and back propagation algorithm, the simulation results show that the deep belief network method can achieve a lower error rate and significantly improve the accuracy of authentication.

Key words: deep belief network model, layer-by-layer greedy algorithm, back propagation algorithm, Softmax classifier

CLC Number:

TP309

Ziwen SUN, Yichao ZHANG. Deep Belief Network Model for Mobile Terminal Identity Authentication[J]. Netinfo Security, 2019, 19(3): 34-42.

Figures/Tables 9

References 20

[1]	MENG Yuxin, WONG D S, SCHLEGEL R, et al.Touch Gestures Based Biometric Authentication Scheme for Touchscreen Mobile Phones[C]//Springer. 2012 International Conference on Information Security and Cryptology, November 28-30, 2012, Beijing, China. Heidelberg: Springer, 2012: 331-350.
[2]	LIU Qiang, WANG Mimi, ZHAO Peihai, et al.A Behavioral Authentication Method for Mobile Gesture Against Resilient User Posture[C]//IEEE. 3rd International Conference on Systems and Informatics, November 19-21, 2016, Shanghai, China. New Jersey: 2017: 324-331.
[3]	LIN Xiaofeng, FANG Mu, LI Qiang, et al.A Design and Implementation of Windows Login Based on Fingerprint Identification[J]. Netinfo Security, 2016, 16(9): 130-133.
	林晓锋,房牧,李强,等.一种基于指纹识别的Windows系统登录方法设计与实现[J].信息网络安全,2016,16(9):130-133.
[4]	CHEN Shaokang, MAU S, HARANDI M T, et al.Face Recognition from Still Images to Video Sequences: A Local-Feature-Based Framework[J]. EURASIP Journal on Image & Video Processing, 2011(1): 1-14.
[5]	HUANG Jiaju, HOU Daqing, SCHUCKERS S, et al. Benchmarking Keystroke Authentication Algorithms[EB/OL]., 2017-12-1.
[6]	WANG Kaijie, YANG Tianqi.Gait Recognition Method Based on Column Mass Vector and Support Vector Machine[J]. Computer Engineering & Applications. 2015, 51(7): 169-173.
	王开杰,杨天奇.基于列质量向量和SVM的步态识[J].计算机工程与应用,2015,51(7):169-173.
[7]	LI Daofeng, ZHANG Xiaoping, ZHONG Cheng, et al.Hierarchical Identity Based Signature Scheme from Hardness of Lattice LWE Problem[J]. Journal of Chinese Computer Systems, 2016, 37(1): 96-99.
	李道丰,张小萍,钟诚,等.格LWE 难题下分层的基于身份的签名方案[J].小型微型计算机系统,2016,37(1):96-99.
[8]	LIU Weibo, WANG Zidong, LIU Xiaohui, et al. A Survey of Deep Neural Network Architectures and Their Applications[EB/OL].2018-7-11.
[9]	LI Hongling, ZOU Jianxin.Research of SQL Injection Detection Based on SVM and Text Feature Extraction[J]. Netinfo Security, 2017, 17(12): 40-46.
	李红灵,邹建鑫.基于SVM和文本特征向量提取的SQL注入检测研究[J].信息网络安全,2017,17(12):40-46.
[10]	LI Dongjie, LI Yangyang, LI Junxiang, et al.Gesture Recognition Based on BP Neural Network Improved by Chaotic Genetic Algorithm[J]. International Journal of Automation and Computing. 2018, 15(3): 1-10.
[11]	DAI Yukun, ZHOU Zhiheng, CHEN Xi, et al. A Novel Method for Simultaneous Gesture Segmentation and Recognition Based on HMM[EB/OL]., 2018-7-12.
[12]	MIAO Yunqi, WANG Linna, XIE Chunyu, et al.Gesture Recognition Based on Deep Belief Networks[C]//Springer. 12th Chinese Conference on Biometric Recognition, October 28-29, 2017, Beijing, China. Heidelberg: Springer, 2017: 439-446.
[13]	DU Yongping, YAO Changqing, HUO Shuhua, et al.A New Item-based Deep Network Structure Using a Restricted Boltzmann Machine for Collaborative Filtering[J]. Frontiers of Information Technology & Electronic Engineering, 2017, 18(5): 658-666.
[14]	LI Fu, SUN Ziwen.Continuous Hidden Markov Model for Hand Gesture Recognition[J]. Journal of Chinese Computer Systems, 2018, 39(3): 474-477.
	李富,孙子文.手势识别身份认证的连续隐马尔可夫模型[J].小型微型计算机系统,2018,39(3):474-477.
[15]	CHEN Luefeng, ZHOU Mengtian, SU Wanjuan, et al. Softmax Regression Based Deep Sparse Autoencoder Network for Facial Emotion Recognition in Human-robot Interaction[EB/OL]. , 2018-7-12.
[16]	RZECKI K, PŁAWIAK P, NIEDŹWIECKI M, et al. Person Recognition Based on Touch Screen Gestures Using Computational Intelligence Methods[EB/OL]. , 2018-7-12.
[17]	JAIN A, KANHANGAD V.Exploring Orientation and Accelerometer Sensor Data for Personal Authentication in Smartphones Using Touchscreen Gestures[J]. Pattern Recognition Letters, 2015, 68(P2): 351-360.
[18]	DIAZ M M, J FIERREZ, GALBALLY J. Graphical Password-based User Authentication With Free-form Doodles[J]. IEEE Transactions on Human-Machine Systems, 2016, 46(4): 607-614.
[19]	SHAHZAD M, LIU A X, SAMUEL A.Secure Unlocking of Mobile Touch Screen Devices by Simple Gestures: You Can See It But You Can Not Do It[C]//ACM. 19th International Conference on Mobile Computing & Networking, September 30-October 4, 2013, Miami, Florida, USA. New York: ACM, 2013: 39-50.
[20]	YAO Luying, HUO Zhiyong.Stereo Matching Algorithm Based on Steady-state Matching Probability and Edge-aware Disparity Propagation[J]. Journal of Nanjing University of Posts and Telecommunications(Natural Science Edition), 2016, 36(2): 124-130.
	姚璐莹,霍智勇.基于稳态匹配概率和边缘感知视差传播的立体匹配算法研究[J].南京邮电大学学报(自然科学版),2016,36(2):124-130.

	Number	Epoch	C	FAR/%	FRR/%
A	(10,10)	(225,75)	0.001	3.68	4.65
B	(20,20)	(225,75)	0.001	2.34	2.17
C	(25,25)	(225,75)	0.001	2.63	3.03
D	(20,20)	(150,50)	0.001	2.51	2.56
E	(20,20)	(300,100)	0.001	2.33	2.14
F	(20,20)	(225,75)	0.01	4.82	3.79
G	(20,20)	(225,75)	0.0001	2.34	2.17

认证方法	FAR/%	FRR/%
BP算法	4.53	4.01
HMM算法	3.44	3.46
DBN算法	2.37	2.17