Comprehensive analysis of continuous authentication for mobile devices

Abstract

Continuous mobile authentication is an imperative aspect of ensuring the security and privacy of data in the contemporary mobile computing landscape. With the proliferation of mobile devices and the increasing dependency on mobile applications, traditional authentication methods are steadily becoming inadequate to protect sensitive information from unauthorized access. Continuous authentication techniques aim to provide ongoing and seamless verification of user identities, enhancing security while minimizing user inconvenience. This survey paper presents a comprehensive analysis of traditional authentication methods, highlighting key techniques, their integration in mobile systems, applications, and limitations. Then, we elaborate on the various techniques employed in continuous authentication. By examining the strengths and weaknesses of different approaches, this survey aims to provide researchers, practitioners, and developers with a comprehensive understanding of continuous mobile authentication, enabling them to make informed decisions when implementing authentication mechanisms in mobile applications.

Full text

©
Comprehensive analysis of continuous authentication for mobile
devices
Aishani Dagar *
Modern School
Barakhamba Road
New Delhi
India
Karan Singh †
Tayyab Khan §
School of Computer & Systems Sciences
Jawaharlal Nehru University
New Delhi
India
Abstract
Continuous mobile authentication is an imperative aspect of ensuring the security and
privacy of data in the contemporary mobile computing landscape. With the proliferation
of mobile devices and the increasing dependency on mobile applications, traditional
authentication methods are steadily becoming inadequate to protect sensitive information
from unauthorized access. Continuous authentication techniques aim to provide ongoing
and seamless verification of user identities, enhancing security while minimizing user
inconvenience. This survey paper presents a comprehensive analysis of traditional
authentication methods, highlighting key techniques, their integration in mobile systems,
applications, and limitations. Then, we elaborate on the various techniques employed
in continuous authentication. By examining the strengths and weaknesses of different
approaches, this survey aims to provide researchers, practitioners, and developers with a
* E-mail: aishanidagar26@gmail.com (Corresponding Author)
† E-mail: tayyabkhan.cse2012@gmail.com
§ E-mail: karancs12@gmail.com
Journal of Discrete Mathematical Sciences & Cr yptography
ISSN 0972-0529 (Print), ISSN 2169-0065 (Online)
Vol. 26 (2023), No. 7, pp. 2007–2024
DOI : 10.47974/JDMSC-1840

2008 A. DAGAR, K. SINGH AND T. KHAN
comprehensive understanding of continuous mobile authentication, enabling them to make
informed decisions when implementing authentication mechanisms in mobile applications.
Subject Classification: 68P27.
Keywords: Continuous mobile authentication, Security, Data protection, Mobile devices, Seamless
verification, Authentication mechanism, Survey.
I. Introduction
Mobile devices have become an indispensable part of our daily lives,
serving as a hub for communication, connectivity, and a wide range of
activities. With their ever-growing capabilities, mobile phones have
garnered immense popularity and have transformed into essential tools
for telephony, social networking, multimedia entertainment, photography,
online banking, navigation, health and fitness, productivity, education
and business operations. As of 2022, the global number of smartphone
users reached a staggering 6.4 billion individuals [1], showcasing the
ubiquitous nature of these devices and their impact on society. As the
locus of human lives, mobile devices harbor sensitive information which
makes their data privacy, security and protection of paramount importance.
The rising preponderance of mobile devices calls for the adoption of novel
and secure authentication techniques to disable unauthorized access. 
Authentication is the crucial process of verifying a user’s identity.
There are two major categories of authentications: traditional and
biometric. A summary of these two categories are shown in Figure 1 [2].
Conventional mobile authentication methods have significant limitations
in terms of security. They are susceptible to various attacks such as
shoulder surfing and smudge attacks. Additionally, these methods only
provide authentication at the entry point. Once an unauthorized individual
gains access to the device, there are no fallback measures in place to
safeguard the data. In December 2022, the number of global mobile cyber-
attacks was approximately 2.2 million [3].
Conventional authentication techniques are insufficient to adequately
safeguard the vast amount of sensitive data stored on users’ devices.
Therefore, it is advisable to implement continuous authentication, which
falls under the umbrella of behavioral biometrics, as a more robust security
measure. Continuous authentication for mobile users is adaptive and
continuous authentication by multiple ways, such as unique finger
gestures, facial recognition, touch gestures, behavioral attributes, etc of a
mobile user. With minimal user intervention, no additional hardware and

COMPREHENSIVE ANALYSIS OF CONTINUOUS AUTHENTICATION
2009
high accuracy rates, continuous authentication methods promise data
protection on mobile devices. Methods such as gait, power consumption,
voice recognition, keystroke dynamics, behavioural profile etc. comes
under continuous authentication.
A. Background
Mobile phones have become an indispensable component of our
lives. With advent in technology, mobile phones have developed
increasingly secure authentication systems [18-20]. However, as the
preponderance of mobile phones continues to rise, the authentication
systems implemented have struggled to keep pace. In the digital era,
where our phones store vital and sensitive information, it is crucial to
execute robust data protection.
This paper aims to examine the functionality of various traditional
mobile authentication methods and provide an in-depth analysis of
continuous authentication systems. By comparing and contrasting these
different methods based on specific criteria, we will evaluate their
strengths and weaknesses. Additionally, we will explore the future
prospects and potential advancements in the field of continuous
authentication.
B. Motivation
The primary objective of this study is twofold. Firstly, we aim to
provide an overview of traditional authentication techniques, examining
their functionality and identifying their limitations. Secondly, we seek to
Figure 1
Classification of Authentication Techniques

2010 A. DAGAR, K. SINGH AND T. KHAN
classify and analyze Behavioral Biometrics (BB) and Continuous
Authentication (CA) in the context of mobile devices. Moreover, we
conduct a comparative study to differentiate between various CA methods,
outlining the advantages and disadvantages associated with each
approach. Lastly, our discussions delve into the insights gained, the
existing challenges, and the anticipated future trends in the field.
C. Contribution
The contribution of this paper can be summarized as follows:
• We address the inadequacy of traditional authentication methods
in safeguarding sensitive information on mobile devices. The paper
presents an overview of traditional authentication techniques,
analyzes their functioning, and identifies their limitations.
• We explore the emergence of continuous authentication techniques
that provide ongoing and seamless verification of user identities.
• We further classify and analyze the categories of Behavioral
Biometrics (BB) and Continuous Authentication (CA) in mobile
devices, comparing their advantages and disadvantages.
• The discussions cover lessons learned, current challenges, and future
trends in mobile authentication.
II. Traditional Techniques
A. Text-based Passwords
Text-based passwords encompass various forms such as PINs, PUKs,
and alphanumeric passwords. However, this authentication technique
Figure 2
Text-based Passwords

COMPREHENSIVE ANALYSIS OF CONTINUOUS AUTHENTICATION
2011
faces several challenges, including the difficulty of remembering
passwords, vulnerability to shoulder surfing attacks, the tendency to
reuse passwords across various different accounts, the risk of writing
passwords down to keep them in memory, the need for customer service
when wrong PUK code is entered, weak passwords, the absence of
password and no password usage altogether. Despite these issues,
password authentication holds an advantage over other methods as it can
be easily changed at any time (Figure 2).
B. Picture-based Passwords
Graphical passwords were introduced as a solution to the
memorization challenge associated with alphanumeric passwords. The
concept is based on the belief that users find it easier and more natural to
authenticate using image recall. As a result, graphical passwords overcome
the drawbacks of alphanumeric passwords. The benefits of graphical
passwords include the provision of highly secure systems, user-friendly
password options, resistance to dictionary attacks and brute force searches.
However, there are also challenges associated with graphical passwords,
such as the requirement for increased storage space, longer registration
times for passwords, and vulnerability to shoulder-surfing attacks [6].
The Picture Password authentication mechanism comprises two main
components: initial password enrollment and subsequent password
verification. Graphical-based password schemes can be categorized into
four primary types:
• Recognition-Based Systems: Users are challenged to identify images
they had previously selected during the registration process. This
method typically requires users to memorize a set of images and
then identify them during login [7].
• Recall-Based Systems: Pure Recall-Based Systems: Users are
required to reproduce passwords they created or selected during the
registration stage from memory [7].
• Cued Recall-Based Systems: Users are provided with hints or cues to
assist them in recalling their passwords [7].
• Hybrid Systems: These systems combine two or more schemes, such
as a combination of recognition and recall-based methods or a blend
of textual and graphical password schemes [8].

2012 A. DAGAR, K. SINGH AND T. KHAN
C. One-time Passwords
OTP systems can be considered as a bridge between a static password
authentication and a better authentication method. As OTP generates a
password, the verification requires synchronisation between the token
and the monitor (Figure 3). There are several categories of OTP, depending
on counter synchronised, time synchronised, involving a secure channel,
or with a shared list of passwords.
There are two primary OTP algorithms: HMAC-based One-time
Password (HOTP) and Time-based One-time Password (TOTP). HOTP
utilizes a Hash-based Message Authentication Code (HMAC) and operates
as an event-based OTP, where each code is derived from a counter. In
simpler terms, pressing a button generates a new password for logging in,
and with each validation, the moving factor is incremented based on a
counter.
Time-based One-time Password (TOTP) is an OTP that operates
based on time. Similar to HOTP, TOTP uses a static seed, but the significant
difference is that the moving factor in TOTP is based on time instead of a
counter. The TOTP token contains an internal clock that synchronizes
with the server’s clock. New passwords are generated by utilizing the
value of the current timestamp [9][10].
One-time passwords (OTPs) offer several advantages, including their
resistance to cracking during replay attacks, avoidance of risks associated
with traditional passwords, and enhanced security. However, there are
also drawbacks to consider. These include the need for additional
Figure 3
OTP-based Passwords

COMPREHENSIVE ANALYSIS OF CONTINUOUS AUTHENTICATION
2013
technology, the potential for security tokens to fail or break, and the
potential complexity of the OTP generation process [11].
D. Face Recognition
Face recognition is a method of biometric identification that enables
the identification, verification, and authentication of individuals by
analysis of their unique facial biometric patterns and data (Figure 4).
Unlike other authentication methods that require memorization, face
recognition offers the advantage of utilizing built-in sensors in
smartphones, eliminating the need for additional equipment or
requirements.
The development of a robust face recognition system involves three
fundamental steps: face detection, feature extraction, and face recognition.
Face detection is responsible for locating and identifying human faces
within the captured images. Feature extraction extracts the relevant
characteristics of the detected faces, generating feature vectors. Finally, in
the face recognition step, the extracted features are compared with
template face databases to determine the identity of the person [12].
However, this authentication model faces several challenges that
affect its accuracy. These challenges include factors such as aging, occlusion
(obstacles in the image, such as objects or parts of the face being covered),
Figure 4
Face Recognition

2014 A. DAGAR, K. SINGH AND T. KHAN
variations in pose (different angles or orientations of the face), and
variations in illumination (changes in lighting conditions) .
E. Fingerprint
Fingerprint authentication involves scanning a user’s fingerprint
using a mobile device’s fingerprint sensor to verify their identity (Figure
5). This process analyzes various characteristics of the fingerprint, such as
pressure, the three-dimensional shape of the finger, ridges, valleys, and
other features through different sensors like optical, capacitive and
ultrasonic sensors [13]. The information is then processed by the device’s
pattern analysis/matching software, which compares it to the list of
registered fingerprints on file. A successful match means that an identity
has been verified, thereby granting access. Attacking applications that use
biometric-based authentication poses a lower risk compared to those
relying on password-based authentication. However, this method faces
several challenges in accurately extracting fingerprint information. Genetic
factors, injuries to the fingers, and the natural aging process can affect the
human fingerprint. Additionally, fingerprint readers are unable to
distinguish between a live finger and a severed one [2].
Figure 5
Fingerprint
F. Iris recognition
Iris recognition is a method of mobile authentication that relies on
scanning a person’s iris. This process involves using a specialized digital
camera that captures a clear and high-contrast image of the iris using
visible and near-infrared light. The camera focuses on the eye, identifying

COMPREHENSIVE ANALYSIS OF CONTINUOUS AUTHENTICATION
2015
the center of the pupil, the pupil’s edge, the iris’s edge, as well as the
eyelids and eyelashes. This information is then processed by Iris
Recognition software, which analyzes the unique iris pattern and converts
it into an iris template[14] (Figure 6).
Figure 6
Iris Recognition
III. Continuous Authentication
A. Gait dynamics-based continuous authentication systems
Continuous authentication systems that are based on gait dynamics
use individuals’ walking patterns to identify them. The required gait data
is captured by embedded accelerometer and gyroscope in mobile devices
(Figure 7). Classifiers distinguish between different users using unique
Figure 7
Gait-based Authentication

2016 A. DAGAR, K. SINGH AND T. KHAN
characteristics extracted from raw data. Over the past few years, diverse
methodologies have come to light for gait-based authentication on mobile
devices, depending on the types of features extracted from raw data and
the authentication methods employed.
By granting a non-intrusive authentication method for mobile devices
with accelerometers, gait recognition offers a significant advantage. As a
result, this technology facilitates seamless and continuous verification of
user identity without the need for user intervention [15].
The overall process of gait recognition typically involves five steps
that are highlighted in the figure: video capture silhouette segmentation,
contour detection, feature extraction and classification [2].
On the other hand, gait-based approaches do not showcase a high
accuracy level in user authentication. Gait based authentication systems
encounter several challenges namely requirement for numerous data
sources, optimal sensors placement, lack of authentication in stagnant
user, and concerns regarding reliability. First and foremost, acquiring
sensory data related to gait necessitates the collection of both visual
information and motion data using multiple sensors. Furthermore, device
placement can greatly alter sensory readings. When the user is not moving,
alternative authentication tasks are requires for CA. Lastly, the state if the
user during enrolment might not align later on due to the dependency of
gait-based traits on the user’s physical condition [16].
B. Voice Recognition
Voice-related characteristics encompass both physiological aspects
and behavioral traits, enabling the analysis of speech across a wide range
of features (Figure 8). A distinctive combination of the anatomical feature
of one’s vocal tract, mouth, larynx and nose and influence of life experiences
Figure 8
Voice Recognition

COMPREHENSIVE ANALYSIS OF CONTINUOUS AUTHENTICATION
2017
that mould one’s manners, rhythm, accent, vibration of their speech form
an individual person’s voice.
A speaker can be authenticated through two ways: text-dependent
approach and text independent approach. In the former approach, users
voice a pre-defined phrase for verification and identification. Although
this method is straightforward and highly accurate, it disables transparent,
continuous authentication and does not rely on a secret-based mechanism.
The latter approach grants higher flexibility, particularly for transparent
authentication, but low accuracy attributable to the dynamic alterations in
the feature space of voice input, which encompass factors such as user
condition and environmental conditions. Under this approach, users are
identified by voice regardless of spoken words.
User identification through voice features adheres to the standard
process, commencing with data collection and preprocessing, progressing
to feature extraction and selection, and concluding with modeling and
pattern recognition. The accuracy of this approach is majorly affected by
feature quality, evaluated on feature distinctiveness and ability to
withstand potential noice sources introduced by factors such as user
condition and environmental variables.
Initial voice recognition applications were numerous, namely access
control, personalization, and forensic and criminal investigations.
However, the scope of applications has expanded to include online
banking, where customers can conduct transactions through voice
communication as the voice recognition system ensures transparent and
continuous authentication of the customer.
The advantages of employing voice recognition for biometric
authentication systems is the prevalence of microphones in mobile devices
and the inherent naturalness associated with both speaking and being
recognized by one’s voice.
The shortcomings of this authentication model are background noise,
voice variations due to user physical and emotional state, adverbial attacks,
system overhead in terms of both power and computation and low
usability due to variable accuracy rates [16].
Furthermore, speaking continuously is often unnatural in most
applications. Therefore, for the purpose of continuous recognition, the
optimal approach is to combine voice with other traits. It is also advisable
to utilize it in applications where frequent speech is anticipated to avoid
compelling users to speak unnecessarily [2].

2018 A. DAGAR, K. SINGH AND T. KHAN
C. Keystroke Dynamics
Keystroke dynamics involves recognizing an individual based on
their typing rhythm (Figure 9). The process of keystroke dynamics includes
extracting various features such as duration, latency, pressure, and
location. Duration refers to the time span between pressing and releasing
a key. Latency represents the time between releasing one key and pressing
the next key. Pressure indicates the force applied to a key, while location
refers to the specific finger position or area on the screen. Additional
metrics, such as error rate (frequency of using backspaces or deletion
options), can also be employed. It is possible to utilize other metrics as well
in keystroke dynamics analysis.
Keystroke dynamics-based methods have several advantages, namely
high authentication accuracy, high power efficiency, hardware
independence, implicit processing and resistance to shoulder surfing
attacks.
However, implementing a keystroke dynamics-based approach can
be challenging for several reasons, namely variations in user behaviour,
vulnerability of extracted metrics to noise and behavioural changes.
Additional challenges may arise in relation to typing with different
languages and potential variations in a user’s typing behavior across
languages [16].
D. Signature Recognition
In this authentication technique, the system analyses how a user
signs on the touch-screen of a mobile by extracting some features, namely
time, speed, acceleration, pressure, and direction [2] (Figure 10).
Figure 9
Keystroke Dynamics

COMPREHENSIVE ANALYSIS OF CONTINUOUS AUTHENTICATION
2019
Various devices are utilized for capturing and verifying mobile-
biometric signatures, with the gyroscope and accelerometer being the two
crucial components in mobile devices [17].
The accelerometer in mobile devices is responsible for determining
the phone’s orientation. On the other hand, the gyroscope is utilized to
estimate the angular rotational velocity. These devices enable
measurements of pen-tilt angles, hand movements, tablet direction, and
other related aspects. Electronic pens used for signing purposes are also
employed to extract various signature features, including position
coordinates, velocity, pen pressure, and pen angles. This process involves
the use of laser diodes, resonance frequency, magnetoelastic sensors,
strain gauges, and similar technologies [17].
Since handwritten signature is an established means of individual
authentication and is socially accepted; therefore, combining these two
into one system would produce a foolproof and efficient biometric
authentication system on the go [17]. There are five modules to a standard
signature verification system: pre-processing, feature extraction,
enrollment, similarity computation and score normalization. Firstly, signal
obtained from the touch signature or pen movement on screen are
digitised. In pre-processing, any missing parts of the signature are
reconstructed. During feature extraction, a feature vector is generated
from each acquired signal, consisting of signature duration or average
speed. These features are either enrolled as templates or utilized by a
statistical model that represents the generated signatures. The similarity
computation module compares the claimed identity with the enrolled
templates by calculating a similarity score. To grant access to the claimed
user, score normalization is applied to standardize the similarity scores
within a predefined range and compare the resulting score with a
Figure 10
Signature Recognition

2020 A. DAGAR, K. SINGH AND T. KHAN
predetermined threshold value. Score normalization is valuable when
multiple algorithms are employed in a system.
An advantage of this technique is that it has a high user’s acceptance.
As smartphones do not require any additional hardware.
However, signatures seem different on smart phones, signature pads
or pen-based tablets. Also an individual’s signature can vary due to
different style over time or in case of injury, or many other causes [2].
E. Fusion
Multimodal authentication systems have gained immense popularity
because they leverage numerous biometrics and thus, offer accurate
results in comparision to unimodal systems that depend on a single
biometric modality. These systems provide enhanced security and offer
flexibility in authentication. Implementing multimodal authentication
may involve combining data from multiple sources, extracting features
from different modalities, and employing various algorithms and models.
Utilizing multimodal authentication on smartphones is feasible as modern
devices are equipped with diverse sensors capable of capturing multiple
biometrics. However, there are drawbacks to consider including the
quality of input data from different sources, as poor data quality can lead
to decreased performance. Additionally, utilizing multiple data sources
requires reading from various sensors, which can be computationally
demanding and energy-intensive. Multimodal-based approaches can
result in longer training times, larger model sizes, increased memory
usage, and longer inference times [16].
IV. Comparision
As discussed earlier, each technique has its own advantages and
drawbacks. While no technique is considered optimal, it can still meet the
user’s needs to a certain extent. There are notable differences between
traditional and biometric authentication methods. Traditional techniques
are active, whereas biometric authentication is passive. Moreover,
biometric data is inherently linked to its owner, unlike traditional
credentials which can be forgotten, shared, lent, or stolen. Biometrics
provide a reliable and natural means of identification as the user must be
physically present during authentication and cannot deny access to the
system. In contrast, with traditional techniques, users can deny login by
sharing passwords. Lastly, biometric data is generally unique to each

COMPREHENSIVE ANALYSIS OF CONTINUOUS AUTHENTICATION
2021
individual. However, it is important to note that biometric data can be
noisy, requiring accurate measurements, which presents significant
challenges for biometric authentication as a developing field [2]. Figure 11
compares and contrasts each of the discussed techniques.
V. Conclusion and Future Scope
In conclusion, this survey paper has examined various aspects of
continuous mobile authentication. We have explored different techniques,
such as gait recognition, voice recognition, keystroke dynamics, and
multimodal authentication, discussing their advantages, challenges, and
applications. It is evident that continuous mobile authentication offers
benefits in terms of unobtrusive and seamless user verification, enhancing
security and user experience.
However, there are still areas that require further exploration and
research. One such area is the integration of multiple biometric modalities
for robust and accurate authentication. Additionally, the impact of
changing user conditions, environmental factors, and different languages
on the performance of continuous authentication systems warrants further
investigation.
Furthermore, the cost-effectiveness of continuous authentication
methods, as well as the management of privacy concerns and data security,
are important considerations for future developments in this field.
Figure 11
Comparison Chart

2022 A. DAGAR, K. SINGH AND T. KHAN
However, there are still areas that require further exploration and
research. One such area is the integration of multiple biometric modalities
for robust and accurate authentication. Additionally, the impact of
changing user.
In terms of future scope, advancements in machine learning, artificial
intelligence, and sensor technologies hold promise for improving the
accuracy and reliability of continuous mobile authentication systems.
Exploring novel approaches and algorithms, as well as addressing
usability and acceptance issues among users, will contribute to the
widespread adoption of these systems.
Overall, continuous mobile authentication presents a promising
direction for enhancing security and user convenience in mobile devices.
Continued research and innovation in this field will pave the way for more
sophisticated and effective authentication solutions in the future.
References
[1] Laricchia, F., Number of smartphone users by leading countries in
2022. Statista (2023). https://www.statista.com/statistics/330695/
number-of-smartphone-users-worldwide/.
[2] Amin, R., Gaber, T., ElTaweel, G., Hassanien, A.E., Biometric and
Traditional Mobile Authentication Techniques: Overviews and Open
Issues. In: Hassanien, A., Kim, TH., Kacprzyk, J., Awad, A. (eds) Bio-
inspiring Cyber Security and Cloud Services: Trends and Innova-
tions. Intelligent Systems Reference Library, vol 70. Springer, Berlin,
Heidelberg (2014). https://doi.org/10.1007/978-3-662-43616-5_16.
[3] Ceci, L., Number of mobile cyber attacks against users worldwide
from January 2020 to December 2022. Statista (2023). https://www.
statista.com/statistics/1305965/mobile-users-cyber-attacks/.
[4] Ambimat Electronics. SMS-based OTP Authentication and Its Dis-
advantages. Ambimat Electronics (2021, June 1). https://ambimat.
com/sms-based-otp-authentication-and-its-disadvantages/.
[5] Chen Wang, Yan Wang, Yingying Chen, Hongbo Liu, Jian Liu.,
User authentication on mobile devices: Approaches, threats and
trends.Computer Networks,170 (2020). https://doi.org/10.1016/j.
comnet.2020.107118.
[6] Er. Aman Kumar, E. N. B., A Graphical Password Based Authentica-
tion Based System For Mobile Devices.International Journal of Com-

COMPREHENSIVE ANALYSIS OF CONTINUOUS AUTHENTICATION
2023
puter Science and Mobile Computing, 3(4), 744–754 (2014). https://
www.ijcsmc.com/docs/papers/April2014/V3I4201499a50.pdf.
[7] Bandakkanavar, R.,Graphical Password Authentication. KrazyTech
(2017, August 19). https://krazytech.com/technical-papers/graph-
ical-password-authentication.
[8] Towseef Akram Vakeel Ahmad Israrul Haq Monisa Nazir. Graphical
Password Authentication.International Journal of Computer Science and
Mobile Computing,6(6), 394–400 (6, June 2017).
[9] What’s the Difference Between OTP, TOTP and HOTP? (n.d.).
Onelogin Logo. https://www.onelogin.com/learn/otp-totp-
hotp#:~:text=There%20are%20two%20types%20of%20OTP%3A%20
HOTP%20and%20TOTP.
[10] Syed Zulkarnain Syed Idrus, Estelle Cherrier, Christophe Rosenberg-
er, Jean-Jacques Schwartzmann. A Review on Authentication Meth-
ods. Australian Journal of Basic and Applied Sciences, 7 (5), pp. 95-107
(2013). hal-00912435.
[11] One-time password (OTP) – more security online. (2020, January 10).
IONOS. https://www.ionos.com/digitalguide/server/security/
what-is-a-one-time-password-otp/.
[12] HUSEYNOV, Emin, SEIGNEUR, Jean-Marc. WifiOTP: Pervasive
Two-Factor Authentication Using Wi-Fi SSID Broadcasts. In: ITU /
IEEE. Kaleidoscope International Conference. 2015.
[13] Goodner, S., What Are Finger Scanners and How Do They
Work? Lifewire (2021, August 29). https://www.lifewire.com/un-
derstanding-finger-scanners-4150464.
[14] What is Iris Recognition and how does it work? NEC (2022, June 15).
https://www.nec.co.nz/market-leadership/publications-media/
what-is-iris-recognition-and-how-does-it-work/.
[15] M. O. Derawi, C. Nickel, P. Bours and C. Busch, “Unobtrusive Us-
er-Authentication on Mobile Phones Using Biometric Gait Recogni-
tion,”2010 Sixth International Conference on Intelligent Information
Hiding and Multimedia Signal Processing, Darmstadt, Germany, pp.
306-311 (2010), doi: 10.1109/IIHMSP.2010.83.
[16] Abuhamad, M., Abusnaina, A., Nyang, D., & Mohaisen, D., Sensor-
based Continuous Authentication of Smartphones’ Users Using
Behavioral Biometrics: A Contemporary Survey (2020). ArXiv. /
abs/2001.08578.

2024 A. DAGAR, K. SINGH AND T. KHAN
[17] Zareen, F.J. and Jabin, S., Authentic mobile-biometric signature veri-
fication system. IET Biom., 5: 13-19 (2016).https://doi.org/10.1049/
iet-bmt.2015.0017.
[18] Shankar, V., & Singh, K., An improved user authentication scheme on
smartphone using dominating attribute of touch data.Journal of Dis-
crete Mathematical Sciences and Cryptography,22(8), 1549-1561 (2019).
[19] Shekhawat, K., & Bhatt, D. P., A novel approach for user authentica-
tion using keystroke dynamics.Journal of Discrete Mathematical Sci-
ences and Cryptography,25(7), 2015-2027 (2022).
[20] Ara, A., Sharma, A., & Yadav, D., An efficient privacy-preserving
user authentication scheme using image processing and blockchain
technologies.Journal of Discrete Mathematical Sciences and Cryptogra-
phy,25(4), 1137-1155 (2022).
[21] Dinker, A. G., Sharma, V., Mansi, & Singh, N., Multilevel authentica-
tion scheme for security critical networks.Journal of Information and
Optimization Sciences,39(1), 357-367 (2018).
[22] Joshi, A., & Mohapatra, A. K., A novel lightweight authentication
protocol for body area networks based on elliptic-curve cryptogra-
phy.Journal of Information and Optimization Sciences,41(7), 1645-1672
(2020).
[23] Kumar, A., Singh, K., & Khan, T., L-RTAM: Logarithm based reliable
trust assessment model for WBSNs.Journal of Discrete Mathematical
Sciences and Cryptography,24(6), 1701-1716 (2021).
Received September 2023