Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. Dilip Menariya, D. B. Ojha / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 5, September- October 2012, pp.328-332 A vital application of security with biometric templates Dilip Menariya1 and D. B. Ojha2 1Department of Computer Science Mewar University Chittorgarh, Raj.India 2Department of Mathematics Mewar University Chittorgarh, Raj.IndiaAbstract In this Paper, We will discuss high level behavioral or physiological feature to identify oneof biometric templates security is critical issue in person form other. Many application for biometricpersonal base authentication system due to addition system are available and all of them fall into twoof error in transmission channel. Therefore we main functionality verification and identificationpropose the framework of security of biometric [5][6][7].template, based on the concept of encryption withsecret key, based on braid group for unauthorized A large number of errors in biometric templates areuser access, Fuzzy error correction code determine encountered in day to day communication due toand encounters error if any error is introduce during environments change. Many approaches used to solvethe transmission between users to database server We the noise problem which is one approach Fuzzystore the biometric templates in encrypted form both commitment scheme is one the scheme for solvingwithout the fusion of score level and decision level in the noise problem in biometric templates of adatabase server Stenography is related to hide biometric recognition system, Jules and Wattenberg’sbiometric templates with help of secret key based on Fuzzy Commitment Scheme[3]has been published tobraid groups. handling difference occurringSection 2 of this paper describe the state of Biometric between two captured of biometric data, using errorSystem, and section 3 as our proposed method and correcting code.finally Section 4 the Conclusion 1.1 Related workKeywords: A handful of papers have been Cryptography, Fuzzy Commitment Scheme, published so far in the area of key generation forBiometric System Templates, Registration Phase, biometric secure templates [3,4,10,12,13] FurtherVerification Phase, Braid Groups more amount of work suggest biometric data gathering one or more actual biometric analysis and1.Introduction: combine their results which increase the reliability of In Ancient times the people recognized the biometric system[1][2]each other by face, voice and their tongs movement.Today’s scenario is different due to large increase in 2.Preliminariesour population. As a byproduct of Biometric systemis activity in our day to day communication ,we need 2.1 Biometric Systemmore secure and authenticated communication. This Biometric System measure and analyzes biologicalrequirement challenged us of think on biometric data and provide capability of identifying personfacilitation[5].Large number of government based on Their intrinsic characteristics that can beorganizations, industry and non government physical such hand shape, a fingerprint, facialorganizations to must attain security with accuracy characteristics, voice, or DNA and proving that theand error free Communication .The various aim is same that is registered in the biometricproperties exhibited by biometric system are security system as a biometric template Basicallyuniqueness collectability, performance, acceptability Biometric–Based Personal authentication systemsand circumvention [3] have five major components 1.Sensor, 2.Feature- Extractor, 3.Template Database Server, 4.Matcher,Various cryptography approaches are used for 5.Decesion Module. Sensor is the interface betweenprotection of biometric templates some are based on users and Biometric System and its function is tothe hardware, some are software base, but common scan biometric traits of the user. Feature extractorencryption technique AES and RAS cannot be used, module process the scan biometric traits to extractbecause of large number of interclass variation in the the silent feature set, in some case Feature. Extractorbiometric templates [5][6][16].Personal base is followed by quality assessment module forauthentication system use checking the sufficient quality of scan biometric traits In Biometric–Based Personal authentication systems works in two steps 328 | P a g e
  2. 2. Dilip Menariya, D. B. Ojha / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 5, September- October 2012, pp.328-332Step1.Registered: In the first step the scan biometrictraits of user R are processed and then stored in the Otherwise rejected strings remain uncrossed. Thedatabase Server in Encrypted or secure form defining relations areStep2. Verification: In this step again new scan i j = ji |i j |2,biometric traits of user R ’is captured and processed i j i = j i j For |i j |its then process for Comparing it with alreadyRegistered scan biometric traits as template and if An n-braid has the following geometricboth R and R’ is matches then it is valid for interpretation: It is a set of disjoint n-strands all ofValidation in biometric system otherwise which are attached to two horizontal bars at the topVerification is invalid in biometric system Basically and at the bottom such that each strands alwaystwo types of errors are introduce In the Biometric– heads downward as one walks along the strand fromBased Personal authentication systems[14] the top to the bottom. In this geometric interpretation, each generator i i represents the process of swapping1. False Reject: Valid user (Probability of un the ith strand with the next one (with ith strand goingmatching of two traits R and R’ of same user) under the (i+1)th one). Two braids are equivalent if2. False Acceptance: Invalid User (Probability of one can be deformed to the other continuously in thematching of two traits R and R’ of same user) set of braids. Bn is the set of all equivalence classes of geometric n-braids with a natural group structure.2.2 Error Correction Code The multiplication ab of two braids a and b is the braid obtained by positioning a on the top of b. The2.2.1 Definition: A Metric space is a set B with a identity e is the braid consisting of n straight verticaldistance function dist B X B→B+= {0, ∞), which strands and the inverse of a is the reflection of a withobeys the usual properties (symmetric, triangle 1 respect to a horizontal line. So  can be obtainedinequalities, Zero distance between equal points) from by switching the overstrand and under- -[13]. strand. ∆=(I, 2,……..n-1) (I, 2,……..n-2) (I, 2)2.2.2 Definition : Let B(0,1)n be a code set which (I)is called the fundamental braid. We describeconsists of a set code words bi of length n ,The some mathematically hard problems in braid groups.distance metric between any two code words b i and We say that x and y are conjugate if there is anbj in B is defined By element a such that y = 1 . For m <n; Bm can axa 𝑛 dist ( bi , bj)= 𝑟=1 |𝑏 𝑖𝑟 − bjr | bi , bj ЄC be considered as a subgroup of Bn generated by I,This is known as Hamming Distance [13]. 2,……..m-12.2.2 Definition: An Error Correction Function f fora code B Defined as 3. Proposed SchemeF(bi)={bj/dist(bi ,bj) is the minimum, over B-{bi }}. Here we propose biometric basedHere, bj= f(bi) is called the nearest neighbor of Authentication system using the concept of braidbi[16]. group based cryptosystem for security purpose and fuzzy commitment schema for error correction. We2.2.3 Definition: the measurement of nearness describe the complete process step of biometricbetween two code word v and v’ is defined by Authentication system will achieve the high securitynearness (v,v’)=dist(v.v’)/n,.it is obvious that 0 ≤ and accuracynearness(v.v’) ≤ 1[13]. Step 1: Scanning Scan the biometric traits of users2.2.4 Definition: the fuzzy membership function for using sensor at the user side 1. False Reject: Valida code v’ to be equal to given v is defined as[17] user (Probability of un matching of two traits R and FUZZ (v’) =0 if nearness (v,v’) =z ≤ zo ≤1) R’ of same user) =z otherwise2.3 Braid GroupEmil Artin[11] in 1925 defined B ,the braid group of Step2: Quality Assessment Checking n This process is to check the quality is sufficient ofindex n, using following generators and relations: scanned traits of user for further processingConsider the generators , ……. , nwhereirepresents the braid in which the(i1) string st Step3: Process for convert scans biometric traits of th users to Message Bitscrosses over the i string while all other First the scan biometric traits images decomposed into w/8 *h/8 blocks where each one contain a fix number of pixel where w=height and h=height 329 | P a g e
  3. 3. Dilip Menariya, D. B. Ojha / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 5, September- October 2012, pp.328-332Step 4: Feature Extraction: At this step the silent c=commita lg(XOR,g(m1), Fkshared (m),R) wherefeature are extracted from transformed block of Fkshared (m)=m1Kshared+e, m is the K-bit message ,biometric traits and generate the biometric templates m=m1∆ m2, Fkshared (m) is an n-bit unreadable text ,Ras set of blocks is random braid (R can be changed in every Message that sent),e=h(m2),here h is an invertible functionStep 5: Cryptographic key generation based on braid which maps m2 in to an n-bit error vector of weight αGroupOur scheme is made up of Five algorithms: setup, Decryption: in the open Phase Alice open theencryption, decryption, enrollment, verification commitment at time t2 which was sent by Bob at time t1using the inverse procedure of commita lg(a2)andInitially set up phase: the environment is set up Alice make a calculation c’ using with help of secretinitially for generating public and private key,the key c’=commita lg(XOR,g(m1), F kshared (m),R)andsecret key according to the algorithms using braid check its time t3 result is same as the sent by Bobgroup at time to where event time is function Fuzzy decision MakingE={ti,ai} of time and event algorithms and publish If (near(c,c’) ≤Zo) if result is less than Acceptsecret key between two parties requires for otherwise rejectsecures communication startingEncryption: According to the commit phase, Bobcommits to a Message m€ M to Alice, send as: 330 | P a g e
  4. 4. Dilip Menariya, D. B. Ojha / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 5, September- October 2012, pp.328-332 Liveness check Templates D(m1)= t1 X Kshared User Generation t1 Є H Biometric (v) and Acquisition Pre-Processing Feature Shared Key Extraction Generate (Kshared) Pre-Processing Feature Extraction User Templates D(m2)= t2 X Kshared Liveness check Biometric (v’) Generation t2 Є H and Acquisition DV(M)=biX Kshared+Δe Database (a) Enrollment Phase Server……………………………………………………………………………………………………………………… ………………………………………… Verified F(DV(M))( Kshared)-1 Accept Fuzzy Decision Reject (b)Identification Phase Figure(1)Enrollment/Identification Phase of Biometric Database server where database server foundSystem encrypted biometric templates with error are introduce in transmission channel so finallyIn the Identification phase Ev(M)=Ma+@e Biometric template’s of user is u’ at receiveat receive side and compared it with one is stored Step 6 error correction codespreviously in database if they are matching then user If Error introduce in transmission ofis validate for the system biometric template we can correct using fuzzy error correcting code if any error are introduce during theIn the Registration Phase transmission of biometric templates User biometric scan traits as biometrictemplate as encrypted using secret key and send to 331 | P a g e
  5. 5. Dilip Menariya, D. B. Ojha / International Journal of Engineering Research and Applications (IJERA) ISSN: 2248-9622 Vol. 2, Issue 5, September- October 2012, pp.328-332Receiver check the dist(t(c),c’)>0 he realize that there match score data”, Proc. Canadian an error occur during transmission .receive applies Electrical Computer Eng., pp.469-472.the error correction function c’:f(c)The receive will compute nearness [9] Sunil V. K. Gaddam, Manohar Lal(t{c},f{c}=dist{t{c},f(c’)}/n (2010), “Efficient Cancellable Biometric AFuzz(c’)=0 if nearness(c,c’)=Z<Z<1 Multi- Biometric Template Security: An=z otherwise Application of Code-Based Cryptosystem4. Conclusion [10] Deo Brat Ojha, Ajay Sharma (2010), “A This paper presents fuzzy commitment fuzzy commitment scheme with McEliece’sMethod for error correcting code and braid group to cipher”, Survey in Mathematics and Itscompute a cryptographic key for biometric data .The Application Vol.5pp73-83.method not only meet the security requirements, butalso produce comparable accuracy to well-known K- [11] F. J. MacWilliams and N. J. A. SloaneNN classification method. Experiments on real (1991), Theory of Error-Correcting Codes.biometric data, particular fingerprints and voice, are North Holland.being conducted and will be reported in the nearfuture. [12] A. A. Al-saggaf, H. S. Acharya (2007), “A Fuzzy Commitment Scheme”, IEEEREFERENCES International Conference on Advances in [1] K. NandaKumar (2008), “Multibio- Computer Vision and Information metric Systems: Fusion Strategies and Technology, 28-30 November– India. Template Security”, PhD Thesis, Department of Computer Science and [13] Andrew Burnett, Adam Duffy, and Tom Engineering, Michigan State University, Dowling (2004) “A Biometric Identity January. Based Signature Scheme”, [2] Anil K. Jain and Arun Ross (2004),“Multibiometric systems,”Communi- [14]. V. Pless (1982), Introduction to theory of cations of the ACM, January, Volume 47, Error Correcting Codes, Wiley, New York. Number 1. [15] T. van der Putte and J. Keuning (2000), [3]. A. Juels and M. Wattenberg (1999), “A “Biometrical Fingerprint Recognition: Don’t fuzzy commitment scheme”, In Proceedings Get Your Fingers Burned”. Proceedings of of the 6th Security, pp.28-36, the Fourth Working Conference on Smart November.ACM Conference on Computer Card Research and Advanced Applications. and Communication [16] J. Bringer and H. Chabanne (2008), [4] M. Blum (1982), “Coin flipping by “An Authentication protocol with encrypted telephone: a protocol for solving impossible biometric data”, Proc. Int. con cryptology. problems”, Proc. IEEE Computer Africacrypt. pp-109-124,. Conference, pp. 133-137. [5] A. K. Jain, S. Pankanti, S. Prabhakar, L. Hong, A. Ross (2004), “Biometrics: A Grand Challenge”, Proc. of the International Conference on Pattern Recognition, Vol. 2, pp. 935–942, August. [6] J. Wayman, A. Jain, D. Maltoni, D. Maio(2005),Biometric Systems:Technology, Design and Performance Evaluation, Springer-Verlag,. [7] D. Maltoni, D. Maio, A. K. Jain, S.Prabhakar (2003), Handbook of Fingerprint Recognition, Springer,. [8] A. Adler (2004), “Images can be regenerated from quantized biometric 332 | P a g e