CRYPTOGRAPHIC PROTOCOL
DEPENDING ON BIOMETRIC
AUTHENTICATION
SANJUKTA PAL1, PROF (DR.) PRANAM PAUL2Sudent,
M
.Tech. (CST) Dr. B C Roy Engineering College, Durgapur, Fuljhore, Durgapur-713206,W.B., India1 Computer Application, Narula Institute of Technology, Agarpara, W.B., India2[email protected], [email protected]
ABSTRACT:
In modern age, security is the most challenging issue for using the secure data used by computer. This cryptographic protocol on biometric authentication is nothing but the combination of cryptography and biometric authentication. Here the same idea of cryptography is working (i.e. using key, conversion of plain text into cipher text called encryption and the reverse, means cipher text to plain text called decryption). Here the most promising method fingerprint geometry of biometric authentication is used as the key for encryption and decryption.
Here this cryptographic protocol is just an algorithm for matching the key means matching of stored fingerprint images say DB Images with further given fingerprint image say Final Image. For matching purpose we used binary conversion of images.
This algorithm is suitable for any type of data (means text data, multimedia data etc.)
Keywords: Encryption, Decryption, Plain Text, Cipher text, Biometric Authentication, DB Image, Final Image.
1. INTRODUCTION
In modern age in each and every field we use computer. The main reason is the efficiency in each and every case, because we can access any kind of data (confidential or non-confidential) in a fraction of time interval. Regarding the confidential issue, security is required. There are so many approaches which are used for security purpose. Two of them are cryptography and biometric authentication.
This protocol is nothing but a combination of cryptography and biometric authentication. The main idea behind cryptography is to convert a plain text into cipher text using key. Now a day one of the most promising approach in biometric authentication is fingerprint geometry. Here we take the concept of fingerprint geometry to construct the key.
So, the concept in a whole is
1st, for constructing a key, one person will give (five or six) images of a particular finger from different angles. The images must be stored in the database (say DB Images). These images are said to be key.
2nd, to convert a plain text into a cipher text the key must be attached with the plain text. After attaching the key with the plain text, it would be converted into cipher.
3rd, to derive the proper information that means for getting the plain text ,if that person gives that fingers image (say, Final Image) ,it will check db images one by one.
4th, ifthe final image would match with any one of the db images above the given threshold value (may be 80% or 85%, which depends on code), that means if the key matches, then the plain text would be read.
But, if the final image does not match with any one of the db images, then the key may not be decrypted.
2. ALGORITHM
1st,Read five or six fingerprint images from different angle (say DB Images). 1a: convert images into gray scale image.
1b: convert gray scale image into binary image.
Finally for matching purpose one fingerprint image would be taken (say, Final Image) 2a: convert Final Image into gray scale image
2b: convert this gray scale image into binary image.
Step1: Derive maximum no of pixels in length and in width (say m and n) from Final Image. Step2: Make g= gcd (m, n)
.Step3: Derive all the factors of gcd (g).
Step-4: Select the factor f, where f= max(1 to 9) and <10.
Step-5: Select centre position c of the final image by (m/2, n/2) or ((m+1)/2,n/2) or (m/2,(n+1)/2) or ((M+1)/2,(n+1)/2).
Step-6: Select a pixel chunk of area (f*f) in the centre position of final image.
Step-7: Using the pixel chunk , traverse one db image, pixel by pixel in row wise and column wise(not by pixel chunk wise).
7a: for each traverse store left most pixel position (p) of DB Image and no of pixel match (q) of two chunks at that position.
7b: store left most pixel position (p) and number of pixel match (q) of maximum matching of any two traverses.
Step-8: If the maximum matching of pixel is greater than the threshold value, then go to step 9, Else go to step12.
Step-9: Using the left most pixel position fix up the pixel chunk in the DB Image. Step-10: Match the neighboring pixels row wise and column wise.
Step-11: If the total pixel matching is greater than the threshold value, then continue and go to step 10. Else go to step 12.
Step-12: Go to the next db image, and again check it from step1.
3. EXAMPLE
The DB Image and the Final Image may not be of same size because they may be taken from different device. 1st the DB Image must be converted gradually into gray scale image and finally into binary image, which will give the 0’s and 1’s sequence of that image.
2nd the Final Image also be taken from device and also be converted gradually into gray scale image and binary image.
Step-1: Maximum no. of pixel in length (m) = 6 Maximum no. of pixel in width (n)= 9. Step-2: g= gcd(m,n)= gcd(6,9)= 3.
Step-3: Factors of gcd= 1,3.
Step-4: Factor (f) =3= max(1,3) and <10
Step-6:Select pixel chunk of area (f*f)= (3*3) at the centre position.
Step-7:
Step-7a: For 1st traversing Store(p,q) Where p= Left most pixel position. q=Percentage of pixel match
In the example percentage of pixel match=(6*100)/9.
Step-7b: For any two traversing store p, where max (q) is received. If the 1st traverse q1=81% and for 2nd traverse q2=84%,
Store left most position for 2nd traverse.
Step-8: After final traversing, if the percentage of maximum pixel match is greater than the given threshold value, then go to next step.
Step-9:
Step-10:
Then the percentage of pixel match= (Total number of pixel match*100) / Total number of pixel in the chunk with neighbor pixel.
Step-11: If the percentage of pixel match of previous step is greater than the pre defined threshold value then continue step 10.
After full checking if the percentage of pixel match is greater than the given threshold value then the image is matched, i.e. the key is matched. So the plain text must be decrypted.
Step-12: If the percentage of pixel matches is less than the threshold value (at step 8 or at step 11) then check the next DB Image.
4. ANALYSIS ANDCONCLUSION
The above algorithm is only effective on the binary conversion of an image. So our first task before implementation of algorithm is to convert an image into its binary form. Fingerprint geometry is mostly effective in the centre of the image. So at first we choose the pixel chunk at centre position. After matching the pixel chunk with DB Image above threshold value we can further proceed.
Here the algorithm is not implemented by checking pixel by pixel. Because this technique would be complex than it
Using this algorithm, after the key matching the encrypted plain text must be decrypted.
5. REFERENCES
[1] J. K. Mandal, S. Dutta, “A 256-bit recursive pair parity encoder for encryption”, Advances D -2004, Vol. 9 nº1, Association for the Advancement of Modelling and Simulation Techniques in Enterprises (AMSE, France), www. AMSE-Modeling.org, pp. 1-14 [2] Pranam Paul, Saurabh Dutta, “A Private-Key Storage-Efficient Ciphering Protocol for Information Communication Technology”,
National Seminar on Research Issues in Technical Education (RITE), March 08-09, 2006, National Institute of Technical Teachers’ Training and Research, Kolkata, India
[3] Pranam Paul, Saurabh Dutta, “An Enhancement of Information Security Using Substitution of Bits Through Prime Detection in Blocks”, Proceedings of National Conference on Recent Trends in Information Systems (ReTIS-06), July 14-15, 2006, Organized by IEEE Gold Affinity Group, IEEE Calcutta Section, Computer Science & Engineering Department, CMATER & SRUVM Project-Jadavpur University and Computer Jagat
[4] Dutta S. and Mandal J. K., “A Space-Efficient Universal Encoder for Secured Transmission”, International Conference on Modelling and Simulation (MS’ 2000 – Egypt, Cairo, April 11-14, 2000
[5] Mandal J. K., Mal S., Dutta S., A 256 Bit Recursive Pair Parity Encoder for Encryption, accepted for publication in AMSE Journal, France, 2003
[6] Dutta S., Mal S., “A Multiplexing Triangular Encryption Technique – A move towards enhancing security in
[7] ECommerce”, Proceedings of IT Conference (organized by Computer Association of Nepal), 26 and 27 January, 2002, BICC, Kathmandu. Paul Reid, 2003, Biometrics for Network Security, Prentice Hall PTR, chapter-5
[8] A white paper by the University of Southern California and VeriSign 2005 Building a Security Framework for Delivery of Next Generation Network Services United States.
[9] L. Podio and Jeffrey S. Dunn 2002, Biometric Authentication Technology: From the Movies to Your Desktop, National Institute of Standards and Technology (NIST), Information Technology Laboratory 497
[10] Edited by Lori Ayre, Infopeople Project, 2003, Library Computer and Network Security Infopeople Project, http://infopeople.org/howto/security/.
[11] Sarbari Gupta, 2004, Identity Authentication Identity Authentication using the using the PIV Token PIV Token, National Institute of Standards and Technology, India.
[12] Secure Computing Corporation, 2001, Authenticating with one of the safest devices: the biometric Sony Puppy, Secure Computing Corporation, 4810 Harwood Road, San Jose, CA 95124 USA.
[13] Biometric Consortium web site: http://www.biometrics.org 2006 [14] International Biometric Industry Association, http://www.ibia.org 2005
[15] Bioenable Technologies Pvt. Ltd. 2004-2005 http://www.bioenabletech.com/ biometrics_india_pune_contact.htm [16] Securitex Electronic Systems Engineering, 2006, Fingerprint Identification system http://www.securitex.com.sg/ [17] Manvish Embedded Services, 2006, Finger print sensors technology overview
http://www.manvish.com/embedded/miFAUN/techoverview.php
[18] TopAZ Solutions Pte Ltd, 2006, Biometric Fingerprint Security, http://www.topazsol.com/bio_door_access.htm
6. Biography of Authors
Sanjukta Pal is MTech (CST) final year student of Dr. B. C Roy Engineering College, Durgapur,W.B.,India. She had completed MCA under West Bengal University of Technology in 2010.
