当前位置: X-MOL 学术J. Inf. Secur. Appl. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Open code biometric tap pad for smartphones
Journal of Information Security and Applications ( IF 3.8 ) Pub Date : 2021-01-05 , DOI: 10.1016/j.jisa.2020.102688
Carl Haberfeld , Md Shafaeat Hossain , Lisa Lancor

Poor security practices among smartphone users, such as the use of simple, easily guessed passcodes for logins, are a result of the effort required to memorize stronger ones. In this paper, we devise a concept of “open code” biometric tap pad to authenticate smartphone users, which eliminates the need of memorizing secret codes. A biometric tap pad consists of a grid of buttons each labeled with a unique digit. The user attempting to log into the phone will tap these buttons in a given sequence. He/she will not memorize this tap sequence. Instead, the sequence will be displayed on the screen. The focus here is how the user types the sequence. This typing behavior is used for authentication. An open code biometric tap pad has several advantages, such as (1) users do not need to memorize passcodes, (2) manufacturers do not need to include extra sensors, and (3) onlookers have no chance to practice shoulder-surfing. We designed three tap pads and incorporated them into an Android app. We evaluated the performance of these tap pads by experimenting with three sequence styles and five different fingers: two thumbs, two index fingers, and the “usual” finger. We collected data from 33 participants over two weeks. We tested three machine learning algorithms: Support Vector Machine, Artificial Neural Network, and Random Forest. Experimental results show significant promise of open code biometric tap pads as a solution to the problem of weak smartphone security practices used by a large segment of the population.



中文翻译:

适用于智能手机的开放式代码生物识别点击板

智能手机用户中不良的安全习惯,例如使用简单,容易猜到的登录密码,是记忆更强密码的努力的结果。在本文中,我们设计了一种“开放代码”生物特征点击垫的概念来对智能手机用户进行身份验证,从而无需记住密码。生物识别点击垫由按钮网格组成,每个按钮均标有唯一数字。尝试登录手机的用户将按给定的顺序点击这些按钮。他/她将不会记住此点击顺序。而是将序列显示在屏幕上。这里的重点是用户如何键入序列。此键入行为用于身份验证。开放式代码生物识别分录器具有几个优点,例如(1)用户不需要记住密码,(2)制造商不需要包括额外的传感器,并且(3)围观者没有机会进行肩膀冲浪。我们设计了三个触摸板,并将它们合并到Android应用中。我们通过试验三种序列样式和五种不同的手指(两个拇指,两个食指和“通常”手指)来评估这些敲击垫的性能。我们在两周内从33名参与者那里收集了数据。我们测试了三种机器学习算法:支持向量机,人工神经网络和随机森林。实验结果表明,开放代码生物识别分接板有望解决很大一部分人使用的智能手机安全操作不力的问题,具有很大的前景。我们通过试验三种序列样式和五种不同的手指(两个拇指,两个食指和“通常”手指)来评估这些敲击垫的性能。我们在两周内从33名参与者那里收集了数据。我们测试了三种机器学习算法:支持向量机,人工神经网络和随机森林。实验结果表明,开放代码生物识别分接板有望解决很大一部分人使用的智能手机安全操作不力的问题,具有很大的前景。我们通过试验三种序列样式和五种不同的手指(两个拇指,两个食指和“通常”手指)来评估这些敲击垫的性能。我们在两周内从33名参与者那里收集了数据。我们测试了三种机器学习算法:支持向量机,人工神经网络和随机森林。实验结果表明,开放代码生物识别分接板有望解决很大一部分人使用的智能手机安全操作不力的问题,具有很大的前景。和随机森林。实验结果表明,开放代码生物识别分接板有望解决很大一部分人使用的智能手机安全操作不力的问题,具有很大的前景。和随机森林。实验结果表明,开放代码生物识别分接板有望解决很大一部分人使用的智能手机安全操作不力的问题,具有很大的前景。

更新日期:2021-01-06
down
wechat
bug