Third International Conference on Advanced Science and Engineering (ICOASE2020)

University of Zakho, Duhok Polytechnic University, Kurdistan Region, Iraq

Biometric Signature based Public Key Security

System
Huda M Therar EMAD A MOHAMMED Ahmed J Ali
Computer Engineering Technology Computer Engineering Technology Power Engineering Technology Northern
Northern Technical University Northern Technical University Technical University
Mosul, Iraq Mosul, Iraq Mosul, Iraq
hudamm2014@[Link] [Link]@[Link] Ahmedem71@[Link]
2020 International Conference on Advanced Science and Engineering (ICOASE) | 978-1-6654-1579-8/20/$31.00 ©2020 IEEE | DOI: 10.1109/ICOASE51841.2020.9436615

Abstract— Digital signature is a technique that is utilized to the encryption key dissimilar from the decryption key that is
check the authenticity of a message transmitted electronically. preserved secret. The whole asymmetric algorithm is
The digital signature technique is based on a public key
dependent upon the generated two large prime numbers.
method. The meant transmitter signs his / her message of his /
her private key and the meant receiver checks that with the These are prime numbers that can be preserved secret.
public key of the transmitter. Digital signatures also ensure Everyone can always utilize the public key to encrypt
authentication of messages, integrity verification, and non- information but only with knowledge of both the prime
repudiation systems. This paper presents the concept of numbers, the information can be decrypted.
Biometric Signature: a recent technique to digitally sign a
Biometric methods that are focused on every single step
message utilizing biometrics- associated with digital signature
key creation, so, merging the benefits of Public Key of verification and identification gain extra interest in the
Infrastructure (PKI), through the utilize of biometric-based modern age, due to rapid technological progress and a high-
digital signature creation that is secure, reliable, quickly- security demand. Iris recognition system is more reliable and
comfortable, non-invasive, and clearly describes the accurate among various biometrics such as face recognition,
transaction creator. It also proposes biometric signature fingerprints, and ear [2]. Essentially a system is to be
algorithms utilizing the commonly utilized RSA digital performed to offer security to data utilizing iris images saved
signature technique and shows the problems related to them. in the IITD database. There is an important need for personal
This paper analyses the security of the modulus and the size of traits that can offer the highest possible protection toward
the keys used and computes the strength of RSA based on the
impersonation. The RSA algorithm is implemented with
number of bits in the key (iris template). The time of the
scheme is calculated and it is extremely small (in sec). utilize of the iris template as the key. The operations of
encryption and decryption operate independently including a
Keywords— Biometric Signatures, Digital Signatures, ICA key to encrypt and decrypt the data see Fig. 1. The security
algorithm, RSA algorithm.
of the encrypted data completely depends on the key's
confidentiality.
I. INTRODUCTION In biometric encryption, several corporations have
improved biometric with their user recognition and key-
Nowadays, security is a main concept in the digital generation principles [3]. They have designed a system
world. The Internet provides minimal cost but unreliable defined as biometric encryption, where the biometric photo is
medium of connecting people. An imposter can simply merged with a digital key (utilized as an encryption key) to
disguise as an authorized user and deceive the system. establish a more secure block of data named bioscrypt, in a
Robust security techniques and indisputable authentication real way that neither the key nor the biometric can be
methods are needed to supply end-to-end protection and separately obtained from it. During authentication, the
increase consumer trust in multiple internet-related business encryption key is obtained by merging the biometric photo
applications. This is required to encourage consumers to and the bioscrypt. This method solves the issue of key
move from the conventional way of doing business to web- management. Because the key is biometric independent,
based applications, and create e-commerce an entire utilize of the biometric is not lost, and though the key is
success. hacked [17].
The main challenge with PKI implementations is the
Cryptography techniques are utilized to avoid sensitive management of private keys. It is a good idea to save private
data from the attacker. Encryption transforms the keys on smart cards or floppy disks but they could also be
original messages into coded messages by creating a model misplaced or stolen. This can be solved by using biometrics
resistant to multiple attacks. At current, the best excellently- for private key access. Scientists have developed a new
method to reduce the risk of access to private key distributed
known and usually utilized public-key cryptosystem for
generation of RSA keys on more than one server, splitting the
secure data communication is RSA [1]. In this cryptosystem,
key into shares for every server. It will raise the effort needed

978-1-6654-1579-8/20/$31.00 ©2020 IEEE 133

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on January 17,2023 at [Link] UTC from IEEE Xplore. Restrictions apply.
 Third International Conference on Advanced Science and Engineering (ICOASE2020)
University of Zakho, Duhok Polytechnic University, Kurdistan Region, Iraq

by hackers to access the private key and discourage them by combining fingerprint features with the RSA algorithm
from trying to do, and can also be utilized to transfer the [6]. They suggested saving an encrypted biometric feature on
authority to sign documents to more than one person in the a smart card and utilizing it for authentication and key
organization to reduce errors or misuse by people inside the production purposes. They give a detailed approach to extract
organization. cores and deltas from the fingerprint picture and then utilizing
The advantages of the implemented system are as follows: them to create keys utilizing the Log-Gabor method. While
1. This approach will identify the person correctly
their experimental results show a very encouraging FAR of 0
and not the belonging of a person or what he
percent and FRR of 13.27 percent, they are unable to produce
knows.
robust results based only on core and delta traits. The main
2. No transmission of templates over the internet.
goal of the study is to present a detailed literature search of
3. No storing of the biometric template needed for
the most current biometric signature verification scientific
recovering each private key (because that can be
articles [7]. It discusses the most significant approaches and
reproduced on request).
describes differences in approaches and features that are
4. Offers extra flexibility when signing documents.
being utilized in the most current research in this field, along
You can also sign documents anywhere utilizing
with potential extensions. The researchers in [8] have
cams embedded in your device like PDAs, tablets,
implemented two stages of combination in finger and iris
mobile phones, etc. based on biometrics like Iris.
pictures for biometric authentication. Iris and fingerprint
features are calculated by a feature extraction module that has
updated the LDP and Gabor dependent features. Two
categories of combination are utilized, like SLF and FLF.
Strategies are utilized to combine the characteristics of iris
and fingerprints. In 2017, the researchers suggested the
approach of the digital signing of the document [9]. When the
user sends a document to any users who submit the document
must have their fingerprints on them. Sensor for Smartphone
Fingerprint and the document is expected to be encrypted.
Fig.1. Encryption and Decryption by RSA Then, the person who receives the encrypted file may decrypt
a document utilizing the public key of the person who is a
The paper is organized as follows: Section II contains a
sender. Biometrics is a channel to find the value of a private
review of the suggested methods. The achievement of the
user key from the dataset. This method's key pair technique
biometric digital signature system is defined in section III.
utilizing the RSA algorithm. The authors in [10] proposed a
Section IV discusses the results of the strategies
lightweight shortened complicated digital signature
implemented. The conclusion and future work are explained
mechanism to provide a safe connection with smart devices
in Section V.
in person-customized IoT. We have utilized less
II. RELATED WORK comprehensive procedures to perform signature as well as
verification operations, such as person beings signing legal
This study discusses security systems and digital papers and later checking as per claimant. Improve the
signature systems that depend on the biometric features security intensity to protect against an analysis of traffic. The
produced by various authors based on their experiments. researchers in [11] has developed a new definition of digital
The most significant studies related to this research. signature which the designers name fuzzy signature, which is
The authors in [4] have provided biometrics (iris and a signature system that utilizes a fuzzy string like biometric
fingerprint) to produce an encryption key. Next, the position data mostly as a private key, and therefore does not need user-
and texture features of the minutiae are defined from the specific accessory information (also named an assistance
position of the iris and the finger. So a 256-bit secure string in the sense of fuzzy applicator) to produce a signature.
encryption key was generated by merging the extracted The author in [12] paper gives a critical review of how
features and the key utilized in the BIS. BIS safety is strong biometrics can be effectively implemented to remove one
as well as the best reliability. Low-level encryption methods problem of cloud protection, identity theft. A broad range of
are often utilized in the method of biometrics. The biometric authentication method protocols and cloud-based
researchers in [5] suggested a system to combine the applications, particularly to combat identity theft have been
recognition of iris with a high-security access developed previously proposed [20-22].
smart card. The simulation circuit was constructed with its
software. In security terms, the suggested methodology has
superior efficiency, compared with other technologies. The
template signature was created utilizing the MD5 hash
function and stored on the smart card. In 2012, Conti et al.
suggested a technique for creating a private key on-the-fly

134

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on January 17,2023 at [Link] UTC from IEEE Xplore. Restrictions apply.
 Third International Conference on Advanced Science and Engineering (ICOASE2020)
University of Zakho, Duhok Polytechnic University, Kurdistan Region, Iraq

III. PROPOSED SYSTEM

Here S is the signature.
A. Biometric Signature-based RSA Algorithm Step 2: Encrypt message and signature (m + S) through each
RSA algorithm is the first public-key cryptography previously accepted private key scheme and transmit that to
method which is commonly utilized for secure information the recipient.
transfer [13], which ensures both confidentiality and
protection. It points to the names of its explorers Ron Rivest, 2) Recipient (signature verification)
Adi Shamir, and Len Adelman in 1977. RSA algorithm is Step 1: Decrypt the obtained encrypted message utilizing a
utilized with a 2048-bit iris template to produce a prívate previously agreed private key mechanism to decrypt the
key by determining the nearest number to the biometric original message and the signature (m + S).
template that is relatively prime with both the Euler totient Step 2: Calculate H'(m) from the message, m utilizing the
equation ø(n) by utilizing it as a decryption exponent, d same hash function as the transmitter. Decrypt the
(maintained secret). We select the size of p and q to be 1024 signature to extract H (m) utilizing:
bit therefore the size of the modulus is 2048 bit. A biometric
signature can is being implemented in the basic steps below 𝑯(𝒎) = 𝑺𝒆 𝒎𝒐𝒅𝒏 (3)
and described in Fig. 2. Step 3: Compare H (m) to H'(m) and check the biometric
signature.
1) Transmitter (key and signature generation)
a) Key generation: If a template itself (without incrementing) is relatively prime
Step1: Generate two prime numbers p, q each1024bit. to ø (n) then utilize the template here as a private key.
Step 2: Find modulus n = p×q then ∅(n) = (p-1)(q-1). Because the private key is never transferred, there is also no
Step 3: Create decryption key, d from the 2048-bit iris possibility of the recipient end being misused for the iris
template by increasing it to the nearest relatively primary templates.
number to ∅ (n).
[Private key = (d, n)].
Step 5: Calculate the exponent of encryption (e) as its
multiplicative inverse of the (d) modulus ∅ (n), termed:
𝒆 = 𝒅−𝟏 𝒎𝒐𝒅(∅(𝒏)) (1)
, [Public key = (e, n)].

b). Signature generation:

Step 1: Calculate messages hash H (m) utilizing SHA1 in

which (m) is the message to be signed digitally. Encrypt H
(m) with d and n utilizing the “(2)”:
Fig. 2 Biometric digital signature by RSA algorithm
𝒅
𝑺 =(𝑯(𝒎)) 𝒎𝒐𝒅 𝒏 (2)

B. Conversion of iris image to a unique code (Iris Code) 3) Image Normalization: Includes conversion from polar to
Every iris image has several features that make uniqueness cartesian coordinates and normalization of iris image. This is
to an individual iris. After executing several steps (like Image achieved utilizing the Daugman rubber sheet technique,
Segmentation and Normalization, Enhancement, where every point is in the iris region is repositioned into
Binarization, and Feature Extraction using the Flexible-ICA polar coordinates (r,θ) [16] as seen in Fig.3.
algorithm), our method creates a template binary code called 𝐼(𝑥(𝑟, θ), y(r, θ)) → I(r, θ) (4)
(Iris Code) from the extracted features. 𝑥(𝑟, θ) = (1 − r)𝑥𝑝 (θ)𝑟𝑥𝐼 (θ) (5)
𝑦(𝑟, θ) = (1 − 𝑟)𝑦𝑝 (θ)r𝑦𝐼 (θ) (6)
C. The basic stages of generating Iris Code
1) Image Acquisition: The first stage of the iris recognition In which I(x,y) in the region of the iris, (x,y) is the main
method is the acquisition of an image. This stage is really Cartesian, coordinates, (r,θ) is the conformable polar
difficult since the size and color of each person's iris is coordinates 𝑥𝑝 , 𝑦𝑝 & 𝑥𝐼 , 𝑦𝐼 are the pupil & iris region
different. It is also very hard to obtain clear images utilizing coordinates along θ the direction[19]
a typical CCD camera under various environmental
conditions. 800 images from the IITD iris dataset have been
used in this study
2) Image Segmentation: Involving the localization of the
internal and external borders of the iris. [15]. A robust
Circular Hough Transformation (CHT), is implemented to
determine both center coordinates & the radius of each pupil
& iris circles.
Fig. 3 Daugman’s rubber sheet algorithm

135

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on January 17,2023 at [Link] UTC from IEEE Xplore. Restrictions apply.
 Third International Conference on Advanced Science and Engineering (ICOASE2020)
University of Zakho, Duhok Polytechnic University, Kurdistan Region, Iraq

of modulus n bits used in the RSA. The equation NIST

4) Feature extraction: The Flexible-ICA algorithm has recommends to compute the approximate length for the key
been used to obtain powerful vector features. Then, every iris is found in FIPS 140-2 Implementation Guidance [18]. In
vector is encoded into an Iris Code, illustrated in Fig. 4. The 2003, RSA Security mentioned that 1024-bit keys were
distinctive features of the human iris are present in some expected to be crackable sometime during 2006 and 2010,
scales like the Independent Component Analysis (ICA) is while 2048-bit keys were sufficient until 2030.
utilized to take distinctive information in an iris and to 3 3
1.923 × √𝐿× ln(2) × √[ln( 𝐿×ln(2))]2
produce a collection of features for an efficient recognition 𝑋= ln(2)
process [14]. Typically, the most common noise-free linear (8)
ICA model is described below.
𝑋 is strength of key, L the lengths of the public and private
𝑋 = 𝐴𝑆 (7) keys.

In which X is a vector parameter of dimension N, and S is a The NIST suggested the highest strength value is 112 and
vector parameter of dimension M, in which every parameter above, so that the encryption is considered strong, while the
it's the source signal. We're assuming that N > M. Mixing it values less than 112 are considered easily hackable (this is the
Matrix A determines the linear transformation of S. Rang).
In another very powerful algorithm, AES (only brute force
breaks it), this algorithm is used as a reference for encryption
strength. The meaning of this is the strength of the RSA of n=
2048 bit in length in front of the method of breaking GNFS is
equivalent to the strength of the AES with a key length of
112bit with respect of the brute force breaking method.

Add to that computes the strength of RSA based on the

number of bits in the key (iris template). At the first the iris
template is calculated using one iris image, which is equal to
256 bits, then the iris template is calculated using two images
of the iris of the eye (of the same person) Which equals 512
Finally, the iris template is calculated using five iris images
(of the same person). So they have three different templates
for the same person, note that the length of modulus n= (p ×
Fig. 4 The basic stages of generating Iris Code q) is equal to the iris template as shown in table II.

D. Generation of public and private keys from iris image

The features derived from the iris of the user have been IV. RESULTS AND DISCUSSION
converted into a binary template called Iris Code (The iris This study focused on the following validation:
image after processing & encoding). The feature extracted by 1. There is a one-to-one correlation between the iris
the ICA algorithm is mapped to the feature vector then apply template and the created Iris Code; i.e. no two iris
the threshold value when the value in the feature vector templates should have the same Iris Code created.
greater than the threshold value is set, when the value smaller 2. There is a one-to-one correlation between the public and
than the threshold value is reset then the feature vector is private key pairs; i.e. Data encrypted utilizing a private
flattened to 1D. In the RSA algorithm, a private key is key can only be decrypted by utilizing its paired public
represented as {d,n}, where d is the private part, So d here is key and conversely.
the private key = iris template. Create decryption exponent 3. No two iris templates can produce the same set of public-
(d) acquired from SHA1 of the 2048-bit iris template by private key pairs.
increasing it to get the nearest relatively primary number with
∅ (n), by Select an integer d such that, GCD [Φ(N), d] = 1. The result of the implementation of the (Biometric
[Private key = (d, n)]. Calculate encryption exponent (e) as Signature utilizing RSA algorithm) is calculated and we are
its multiplicative inverse of the (d) modulus ∅ (n), termed providing a comparison of key generation speeds. TABLE I
𝒆 = 𝒅−𝟏 𝒎𝒐𝒅(∅(𝒏)) , [Public key = (e, n)]. The public and shows the time required for biometric signature utilizing RSA
private keys are created from it utilizing the RSA algorithm. for various modulus lengths and iris recognition (template
size = 2048 bits). SHA1 was utilized as the one-way function
E. Security strength of RSA based the modulus size to convert the template to (120 & 240) bit of hash by
Strength is a bit of security that has been adopted by the concatenating the results acquired by utilizing (6 & 12)
American organization NIST (National Institute of Standards various keys for modulus sizes of (1024 & 2048) bits
and Technology) to determine the strength or weakness of the respectively. The key generation (calculating n, d, and e)
algorithm. The simplest way to do this is by using the “(7)” speed has been raised for the modulus of 2048 bits because of
to determine the strength of the RSA algorithm for this a larger key length included. The time of signature and
approach provides a bit of security depending on the number verification is extremely small.

136

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on January 17,2023 at [Link] UTC from IEEE Xplore. Restrictions apply.
 Third International Conference on Advanced Science and Engineering (ICOASE2020)
University of Zakho, Duhok Polytechnic University, Kurdistan Region, Iraq

TABLE I. BIOMETRIC SIGNATURE UTILIZING RSA SPEEDS FOR VARIOUS

MODULUS LENGTHS AND TEMPLATE SIZE OF 2048 BITS.

1024 2048
bits(in sec) bits(in sec)
Key generation (n,d,e) 11.23 50.05
Signature S 0.32 1.29
Verification 0.13 1.196
Note: SHA1 is utilized as the hash function to produce message-
digest
Fig. 6 Comparison of execution time of different template size

TABLE II shows the cryptographic algorithm provides The experiments of the system have been performed on
different security “strengths” based on a modulus and the key a Laptop PC (64-bit operating system, x64-based CPU, (4)
size utilized. GB RAM, Core i3-7900 processor, 2.40GHz) with the help
of python (hmac, hashlib, sympy, skimage, and rsa) libraries
and Jupyter notebook environment.
TABLE II .SECURITY STRENGTH IN CONNECTION WITH MODULUS
Template No of iris modulus Security
size image Size levels(in bits)
Time(s) V. CONCLUSION
256 1 n=256 40 8.99 A new method for implementing digital signatures
512 2 n=512 60 10.2 utilizing biometrics has been developed as well as named as
1024 4 n=1024 80 12.50 biometric signatures. It's doesn't completely change digital
2048 5 n=2048 112 50.50
signatures but is only utilized to produce a private key. So,
the biometric signature identifies a unique person and solves
Figure 5 Shows the RSA algorithm strength(bits) in relation key management problems by preventing the storage of
with (512,1024,2048,4096,7680) modulus size. private keys or biometric templates anyplace. The uniqueness
of every person's iris, reliability, and high accuracy enable iris
recognition system a completely matchless approach for the
biometric security of e-commerce implementations. The iris
is then utilized as a biometric to be combined with the PKI
for digitally signing of documents. The biometric signature
system utilizes the current and commonly utilized RSA
digital signature that has been introduced. Speed of Biometric
Signatures utilizing iris image and comparison key generation
speeds for different biometrics are provided. Biometric
signatures can ensure long-term stability and high accuracy
without limiting the size of the template. Private keys may be
Fig. 5 RSA strength (bits) with different modulus size easily updated periodically or on request.
The implemented system confirms that it is possible to create
a public / private key pair in less than a minute. Such that, it
Figure 6 shows the comparison speed of signature confirms that is capable of supporting real-time
key generation for the different iris templates. As well as the communication.
increase in the size of the template, the time taken for the key This paper explains on the cryptographic algorithm
generation is also increasing as predicted. Biometric provides different security “strengths” based on the modulus
Signature utilizing RSA can therefore be applied with any and the key size utilized. Also, the time performance of the
biometric without limiting on template size. And, because the keys and the result experiment data is discussed and
verification time is very smaller, the validity of the private examined in detail.
key produced could be tested locally in a very short time (in In the future, the use of biometric could be modified to
a millisecond). utilize DNA as the best biometric to be combined with a
digital signature that will aid to improve the system
performance.

137

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on January 17,2023 at [Link] UTC from IEEE Xplore. Restrictions apply.
 Third International Conference on Advanced Science and Engineering (ICOASE2020)
University of Zakho, Duhok Polytechnic University, Kurdistan Region, Iraq

Fig. 7 Screenshot of encryption and generated signatures

REFERENCES Shamir-Adleman Algorithm and End of File Method for Data

Security,” 2018.
[1] D. Boneh and M. Franklin, “Efficient generation of shared RSA [14] M. E. Fouda, S. Member, E. Neftci, A. Eltawil, and S. Member,
keys,” J. ACM, vol. 48, no. 4, pp. 702–722, 2001. “Independent Component Analysis using RRAMs,” IEEE Trans.
[2] J. Daugman, “New methods in iris recognition,” IEEE Trans. Syst. Nanotechnol., vol. PP, no. c, p. 1, 2018.
Man, Cybern. Part B Cybern., vol. 37, no. 5, pp. 1167–1175, 2007. [15] A. S. Al-Waisy, R. Qahwaji, S. Ipson, and S. Al-Fahdawi, “A fast
[3] R. Jayapal and P. Govindan, “Biometric Encryption System for and accurate iris localization technique for healthcare security
Increased Security,” vol. 16, no. 1, pp. 75–80, 2018. system,” Proc. - 15th IEEE Int. Conf. Comput. Inf. Technol. CIT
[4] A. Jagadeesan, T. Thillaikkarasi, and D. K. Duraiswamy, 2015, 14th IEEE Int. Conf. Ubiquitous Comput. Commun. IUCC
“Cryptographic Key Generation from Multiple Biometric 2015, 13th IEEE Int. Conf. Dependable, Auton. Se, pp. 1028–
Modalities: Fusing Minutiae with Iris Feature,” Int. J. Comput. 1034, 2015.
Appl., vol. 2, no. 6, pp. 16–26, 2010. [16] R. S. Choraś, “Iris recognition,” Adv. Intell. Soft Comput., vol. 57,
[5] M.A.M. Abdullah, et al. "Smart card with iris recognition for high pp. 593–600, 2009.
security access environment." 2011 1st Middle East Conference on [17] S. Nisha and M. Farik, “RSA Public Key Cryptography
Biomedical Engineering. IEEE, 2011. Algorithm – A Review,” no. July, 2017.
[6] V. Conti, S. Vitabile, and F. Sorbello, “Fingerprint traits and RSA [18] National Institute of Standards and Technology, “Implementation
algorithm fusion technique,” Proc. - 2012 6th Int. Conf. Complex, Guidance for FIPS PUB 140-2 and Cryptographic Module
Intelligent, Softw. Intensive Syst. CISIS 2012, pp. 351–356, 2012. Validation Program National Institute of Standards and
[7] S. Jabin and F. J. Zareen, “Biometric signature verification,” Int. Technology,” 2019.
J. Biom., vol. 7, no. 2, 2015. [19] Al-Ataby, A., Al-Nuaimy, W., & M. Abdullah, M. A. (2012).
[8] V. Sireesha and S. R. K. Reddy, “Two levels fusion based Wavelet Transform-Multidisciplinary Applications. Wavelet
multimodal biometric authentication using iris and fingerprint Transforms and Their Recent Applications in Biology and
modalities,” Int. J. Intell. Eng. Syst., vol. 9, no. 3, pp. 21–35, 2016. Geoscience. [Link]
[9] E. Rahmawati et al., “Digital signature on file using biometric [20] S. B. Sadkhan, B. K Al-Shukur, A. K. Mattar, “Human Voice
fingerprint with fingerprint sensor on smartphone,” Proc. IES-ETA Extracted Biometric Features: What can be used for”, 2017
2017 - Int. Electron. Symp. Eng. Technol. Appl., vol. 2017- International Conference on Current Research in Computer
December, pp. 234–238, 2017. Science and Information Technology (ICCIT).
[10] A. T. A. Ullah, S. Ullah, and Z. Mahmood, “A Lightweight [21] S. .B Sadkhan, B. K Al-Shukur, A. K Mattar, “Biometric voice
Digital Signature Based Security Scheme for Human-Centered authentication auto-evaluation system”, 2017 Annual Conference
Internet of Things,” IEEE Access, vol. 6, pp. 31630–31643, 2018. on New Trends in Information & Communications Technology
[11] K. Takahashi, T. Matsuda, T. Murakami, G. Hanaoka, and M. Applications (NTICT).
Nishigaki, Signature schemes with a fuzzy private key, vol. 18, no. [22] S. B. Sadkhan, B. S. Yaseen,” DNA-based Cryptanalysis:
5. Springer Berlin Heidelberg, 2019. Challenges, and Future Trends”, 2019 2nd Scientific Conference
[12] R. Article, “Review Article A SURVEY ON BIOMETRIC of Computer Sciences (SCCS), 24-27.
AUTHENTICATION SYSTEMS IN CLOUD TO COMBAT
IDENTITY THEFT,” vol. 7, no. 3, pp. 540–547, 2020.
[13] D. Rachmawati, A. Amalia, D. I. Komputer, F. Ilmu, and T.
Informasi, “Combination of Rivest-Shamir-Adleman Algorithm
and End of File Method for Data Security Combination of Rivest-

138

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on January 17,2023 at [Link] UTC from IEEE Xplore. Restrictions apply.