An improved user authentication scheme on smartphone using dominating attribute of touch data. (17th November 2019)

• Record Type: Journal Article
• Authors: Shankar, Vishnu; Singh, Karan
• Is Part Of: Journal of discrete mathematical sciences & cryptography. Volume 22:Number 8(2019)
• Page Start: 1549
• Page End: 1561
• Publication Date: 2019-11-17
• Subjects: Authentication; Continuous smartphone authentication -- Smartphone authentication -- Touch biometrics -- machine learning -- Feature selection; Computer science -- Mathematics -- Periodicals; Cryptography -- Periodicals; Computer science -- Mathematics; Cryptography; Periodicals; 004.0151
• DOI: 10.1080/09720529.2019.1695903
• Languages: English
• ISSNs: 0972-0529

Abstract

Abstract: In the era of science and technology, the user of Smartphone is omnipresent, as the data of research shows the number of Smartphone users is rising exponentially. Earlier Mobile phones used for calling and sending messages. Technology has evolved, and presently Smartphone is used as multipurpose device. The smartphone is used for online shopping, reading books, listing music, clicking photo, recording video, internet access, digital banking, and many more sophisticated services. Access of Smartphone data must be by secure and user-friendly system of authentication. Authentication methods of traditional computer systems cannot fit as it is into Smartphone authentication due to small size screen for input. Presently fingerprint authentication, pattern authentications systems and many more are used for Smartphone authentication, but they are just providing the security at entry-level authentication. Once a user does entry-level authentication device is open for use, and unattended phone can be unauthorized access. There is research going on in the area where the unauthorized person must not have access even the device is in unlock state. This dimension of authentication is called continuous Smartphone authentication. Our focus in this paper is on continuous mobile authentication. In this paper dataset of 41 users is used. Feature selection methods for continuous Smartphone authenticationuses dominating features. For the better tradeoff OptAT performance parameter is proposed. The improvement of 16% has been achieved using this parameter. The learning curve also has been drowned for the features.