You are on page 1of 19

Implementation of Efficient

Fingerprint Authentication Algorithm


using SHA on FPGA
Capstone Project Phase 1
2 Credits
Project ID : W-05
TEAM COMPOSITION

Chetan Ankush Aryan M Nehal M B


Kumar N Gupta
PES1UG20EC050 PES1UG20EC031 PES1UG20EC035 PES1UG20EC122

Guide : Dr. Purushotham U (Department of ECE)


2
Contents
• Introduction
• Problem Statement
• Brief Background
• Literature Survey
• Deliverables
• Project timeline (Gantt Chart)

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 3


Problem Statement

To design and implement an efficient


fingerprint authentication algorithm using
SHA-3 on FPGA.

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 4


INTRODUCTION

● Biometrics is used for identification and authorization of an individual


based on their unique physical characteristics like fingerprint, iris pattern etc.

● Using SHA algorithms in biometric systems provides a high level of security by


protecting the biometric data from unauthorized activities.It enables efficient
storage and transmission of biometric data and also facilitates fast and
accurate matching of biometric data, reducing the computational complexity
and improving the system's performance.

● FPGA can be used for biometric systems for its high-speed processing
capabilities. FPGA devices can be optimized to perform parallel processing,
which can significantly improve the performance of biometric systems.

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 5


Background
● Biometric fingerprint input is a key feature of our
system.
● The run time of the algorithm can be improved through
various enhancements using pipelining and parallelism.
● The Fingerprint template can be stolen and the attacker can
reconstruct it to a fingerprint image.
● FPGA based solutions have emerged as a promising
alternative to traditional software approaches ,due to their
superior performance and reconfigurability.

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 6


Applications of fingerprint with SHA in real
time

1. Banking & Financial services.


2. Law Enforcement.
3. Healthcare.
4. Airport Security.
5. Election Voting System.
Literature Survey
Alfiansyah and R. W. Wardhani, "Implementation of Secure Hash Algorithm – 3 for
Biometric Fingerprint Access Control Based on Arduino Mega 2560," 2018 International
Conference on Applied Information Technology and Innovation (ICAITI), Padang,
Indonesia, 2018, pp. 31-35, doi: 10.1109/ICAITI.2018.8686734

1. This paper highlights the importance of using SHA for biometric fingerprint authentication to protect them
from attacks on fingerprint template from reconstruction.
2. The authors explain the advantages of SHA-3 over other hash functions, including its resistance to various
attacks such as length extension attacks and collision attacks, pre-image attack.
3. This paper demonstrates the feasibility of implementing SHA-3 on a microcontroller, for example Arduino
Mega 2560. Based on the results of performance testing,fingerprint authentication can be 80% successful for
the model proposed in this paper.
4. Research gap :-One potential research gap for the above paper is the lack of discussion on the scalability of
the proposed system.

Learning from the review :- One key learning from the review of the above paper is that the implementation of
secure hash algorithms, on microcontrollers can be a viable solution for biometric authentication.

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 8


Literature Survey
A. M. Tillah, D. Ogi, M. Febriyanto and D. A. Farhatin, "Access Control
System based on Secret Sharing Scheme with Secure Web Database
and SHA-3 Password Authentication," 2021 6th International Workshop
on Big Data and Information Security (IWBIS), Depok, Indonesia, 2021,
pp. 145-152, doi: 10.1109/IWBIS53353.2021.9631847.

1. The applications can pose threats such as unauthorized entities stealing data stored in
databases.We can overcome that by adding encryption to the shared value using AES-256 with
a password as a key, which is implemented using SHA on the access control system, to
strengthen the security.
2. They have made use of the CBC mode,which is more complex than ECB mode,which is simple
and more vulnerable to modification attacks.To overcome that,CBC uses a random initialization
vector which adds uniqueness and randomness.
3. During registering in the database,the system performs a fingerprint scan to generate user
shared values. The fingerprint scan is 512 bytes long.
4. In the login process, the user decrypts the shared value stored in the database by entering the
registered password key and reconstructing the shared value stored .

Research Gaps-Although there are several access control models proposed in the literature, there is
limited research on the practical implementation of access control systems that integrate secret sharing
schemes and SHA-3 password authentication.

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 9


Literature Survey
K. Quist-Aphetsi and I. Baffour Senkyire, "Validating of Digital
Forensic Images Using SHA-256," 2019 International Conference
on Cyber Security and Internet of Things (ICSIoT), Accra, Ghana,
2019, pp. 118-121, doi: 10.1109/ICSIoT47925.2019.00028.

1. Forensic analysis of digital images has helped law enforcement, security and intelligent agencies acquire
supportive information. Intrusive methods such as watermark, extrinsic fingerprint and nonintrusive methods
such as detecting image authenticity or forgery at the image level and detecting on the pixel have been
proposed.
2. They have proposed a novel reversible invisible watermarking and they embedded the watermark into the
satellite image for copyright protection and authenticated the satellite images by inserting a SHA3 generated
hash key watermark.
3. They have worked on file type identification for digital forensics and theyusedcomputational intelligence
techniques in order to identify the JPG,PNG,PDFand GIFfiletypes.
4. They encrypted the picture,by breaking down the image I into 3 channels,namely,R,G and B channels.Then
each channel is hashed using SHA and combined.

Research Gap-Many digital forensic images are compressed to save storage space, but the impact of image
compression on SHA-256 validation has not been extensively studied.
As the volume of digital evidence continues to grow, there is a need for research on the scalability of SHA-256
validation in large-scale investigations.

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 10


Literature Survey
M. Ibrahim, K. Ravindran, H. Lee, O. Farooqui and Q. H. Mahmoud, "ElectionBlock: An
Electronic Voting System using Blockchain and Fingerprint Authentication," 2021 IEEE 18th
International Conference on Software Architecture Companion (ICSA-C), Stuttgart, Germany,
2021, pp. 123-129, doi: 10.1109/ICSA-C52384.2021.00033.
1. Registered voters will have their biometrics registered in a database accessible by the fingerprint reader.
Upon receiving a fingerprint reading, the system will perform two checks; first, it will authenticate the user
against the database to verify the user is a registered voter, and secondly, it will check the blockchain to
verify the user has not already cast a vote.
2. The application server communicates with the central server in order to pass voting information to the
centralized queue(holds 16 per block).
3. This information is hashed and processed as a Merkle tree(data sequence).Once this new block has been
appended to the existing chain, it is pushed to the subscribing voting stations using the pub-sub broker.
4. The voting stations subscribe to blockchain updates published by the central server. The received
updates are checked and validated by recalculating hash values based on blockchain data.

Research gaps-ability to withstand cyberattacks, legal and regulatory frameworks has framed challenges
and barriers to adoption.Scalability: how the system would scale to handle large-scale elections (throughput).

Learning from the paper-The proposed ElectionBlock system combines blockchain technology and
fingerprint authentication to create a secure and transparent electronic voting system that addresses the
challenges of traditional voting systems
28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 11
Literature Survey
A. D. Dinesh, C. Durga Prasad Reddy, G. V. Gopi, R. Jain and T. N. Shankar, "A Durable
Biometric Authentication Scheme via Blockchain," 2021 International Conference on
Advances in Electrical, Computing, Communication and Sustainable Technologies
(ICAECT), Bhilai, India, 2021, pp. 1-5, doi: 10.1109/ICAECT49130.2021.9392415.
1. The grayscale image of the fingerprint is represented with pixel values 0 for black and 1 for white. A
fingerprint is not saved directly, only a series of numbers (a binary code).
2. It is used for identification Arches, loops, or whorls are spotted as black, and the rest of the portion
is white.
3. Each block has NONCE or PREVIOUS HASH,HASH VALUE,DATA.(1Block=1Voter)
4. Every block in the blockchain contains transactions. When a new transaction is observed in the
blockchain, a record or copy of that transaction is added to every user’s ledger.

Research Gaps - Cost-effectiveness and Privacy concerns surrounding the


use of biometric data for authentication is a potential research gap that could be
addressed in future work
Learnings from the paper - The review of the paper highlights the potential of
using blockchain technology for secure biometric authentication schemes

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 12


Literature Survey
M. M. Sravani and S. A. Durai, "On Efficiency Enhancement of SHA-3 for FPGA-
Based Multimodal Biometric Authentication," in IEEE Transactions on Very
Large Scale Integration (VLSI) Systems, vol. 30, no. 4, pp. 488-501, April 2022,
doi: 10.1109/TVLSI.2022.3148275.

1. This paper proposes an efficient implementation of SHA-3 on a FPGA by using


multimodal biometric model
2. The enhancement was done by reducing the number of XOR operation and replacing the
operation by bit wise operation like AND,OR.
3. The enhanced SHA was implemented on FPGA and compared to standard SHA-3
algorithm and the graphs and plots were obtained.
4. The enhanced SHA showed 20% reduction in power consumption and 15% increase in
throughput.
5. This paper also describes the importance biometrics and an efficient algorithm in real
time application.

RESEARCH GAP-
Though the number of XOR operations were reduced the number of other bitwise operations
like AND,OR were increased during the implementation of SHA and will make use more FPGA
resources

This paper was implemented on a single FPGA and on a specific input data set so we don’t
know how the system will work on other FPGA devices and on a broader input data set.
Literature Survey
M. M. Sravani and S. A. Durai, "On Efficiency Enhancement of SHA-3 for FPGA-
Based Multimodal Biometric Authentication," in IEEE Transactions on Very
Large Scale Integration (VLSI) Systems, vol. 30, no. 4, pp. 488-501, April 2022,
doi: 10.1109/TVLSI.2022.3148275.

Learning from the paper-the hardware acceleration techniques can be optimized by


using unrolling and pipeline that makes use of less FPGA resources

Alternative parallel techniques like SIMD(single instruction,multiple data) can be used along
with traditional parallelism to increase performance
Literature Survey
K. Han and X. Li, "Application of Partial Differential Equation Method in
Fingerprint Image Enhancement," 2021 IEEE/ACIS 19th International
Conference on Software Engineering Research, Management and
Applications (SERA), Kanazawa, Japan, 2021, pp. 33-38, doi:
10.1109/SERA51205.2021.9509273.
1. PDE is a powerful mathematical tool that is used in image enhancement including
fingerprint image enhancement
2. In the context of image enhancement,PDE method is used to to smooth out the
noise and unwanted artifacts while preserving the edges and the ridge valley
3. One common approach is antisotropic diffusion which involves applying PDE to
image selectively and removing noise while retaining edges and valleys

Research Gap-The above method is computational intense and requires significant


processing power

The efficiency of PDE methods are limited by the quality of input images poor quality or
degraded images will not be able to improve it

28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 15


Deliverables
1.Design details:

Pre-Processing SHA-512 FPGA

2.Performance improvement graphs / statistics


3.Validation Report (with benchmark application specified)
28/03/23 Capstone Project B.Tech 2018-22 - Feasibility 16
PROJECT
TIMELINE

17
Q&A
Thank You

You might also like