|Publication number||USRE42517 E1|
|Application number||US 12/393,959|
|Publication date||Jul 5, 2011|
|Filing date||Jan 26, 2000|
|Priority date||Jan 27, 1999|
|Also published as||CA2360953A1, CA2360953C, DE60000649D1, DE60000649T2, EP1145483A1, EP1145483B1, US7184547, WO2000045549A1|
|Publication number||12393959, 393959, PCT/2000/174, PCT/FR/0/000174, PCT/FR/0/00174, PCT/FR/2000/000174, PCT/FR/2000/00174, PCT/FR0/000174, PCT/FR0/00174, PCT/FR0000174, PCT/FR000174, PCT/FR2000/000174, PCT/FR2000/00174, PCT/FR2000000174, PCT/FR200000174, US RE42517 E1, US RE42517E1, US-E1-RE42517, USRE42517 E1, USRE42517E1|
|Inventors||Marc Girault, Jean-Claude Pailles|
|Original Assignee||Phentam Dire Nv, Llc|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (12), Non-Patent Citations (35), Classifications (11), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
The present invention relates to an authentication or signature process with a reduced calculations set.
More precisely, the invention relates to the public key cryptography domain. Following this process, the entity to be authenticated—the prover—possesses a secret key and an associated public key. The authenticating entity—the verifier—only needs this public key to achieve the authentication.
Even more precisely, the process relates to the set of processes called “zero-knowledge Protocols”, i.e. without any communication of knowledge. According to this kind of process, the authentication is carried out following a protocol that, as it is recognized, and under assumptions considers as perfectly reasonable by the scientific community, discloses nothing about the secret key of the prover.
To be even more precise, the invention relates to zero-knowledge processes based on factoring problems (i.e. on the difficulty to factor large integers into a product of prime numbers).
The invention is applicable in every system where it is necessary to authenticate parties or messages, or to sign messages, in particular in systems where the amount of calculations to be carried out by the prover is critical. This is especially the case for cards that use a standard microprocessor or low cost cards, with no arithmetic coprocessor (which are often called cryptoprocessor) where cryptographic calculations must be accelerated.
A typical application of the invention is the electronic purse that requires a very high security level while discarding the use of a cryptoprocessor, either because of the cost or for technical reasons (for example the use of a contactless interface), or both.
Another possible application is the next generation tele-card, whose cost constraints are by far stricter than those of the electronic purse.
A number of zero-knowledge identification processes have been published. For example:
Generally speaking, most zero-knowledge identification (or message authentication) protocols involve three steps. For the sake of simplicity, we shall assume that the verifier B already knows all the public parameters related to the prover A, i.e. its identity, its public key and so on.
As a first transaction, A supplies B with a value “c” called “opening”, image through a pseudo-random function h of a parameter x (itself derived from a number r selected by A at random), as well as with the message to be authenticated or signed: c=h(x,[M]), where the symbol [M] means that M is optional. This is the first step. Some protocols may involve several openings.
During a second transaction, B sends to A a parameter e selected at random (the “question”). It is the second step.
During a third transaction, A sends to B an “answer” y that is in coherence with the question e, the opening c and the secret key of A (third step).
Then B checks the received answer. More precisely, B recalculates x from the elements y, e and v using the relation x=φ(y,e,v) and verifies that c=h(φ(v,e,y),[M]), which is the fourth step.
When there is no message to authenticate, the use of the pseudo-random function h is optional. In this case, c=x is convenient. The verification consists of checking that x=φ(y,e,v).
In some protocols, there are one or two more transaction(s) between the verifier and the prover.
For a message signature, the two first steps are discarded, as the parameter e is made equal to c; A then successively and only calculates c, e(=c) and y.
The number u of questions to be answered depends directly on the desired protocol security level. This level is defined as the probability p of detecting an impostor. (i.e. an entity C that fraudulently mimics A). It is measured by a parameter k whose value is related to p by the relation p=1−2−k. In other words, the impostor only has 1 chance in 2 k of succeeding. It can be demonstrated in the present case that if a protocol relies on a difficult mathematical calculation, and if the openings are of adequate length, the length of u must simply equal k bits. A typical value of k is 32, which gives the impostor one chance in 4 billion to be successful. In applications where the failure of an identification may have very harmful consequences (e.g. legal proceedings), this length may be reduced to a few bits.
For protocols using factoring, the calculation of x in terms of r, or the calculation of y in terms of e, or both, involve(s) operations modulo n, where n is a compound number that is hard to factor. This number is said to be of the universal type, generated by a trustworthy third party. It is stored and used by all authorised entities. The “universal” character of n implies that it is a large number (usually 1024 bits), as breaking the factoring of n should compromise the secret keys of all accredited users.
In their basic versions, none of the above mentioned protocols can be implemented in an application that has to comply with severe specifications (low cost, low sophistication), as described in the previous section, as the required calculations could not be performed by a microprocessor card without a cryptoprocessor.
Though the French patent application FR-A-2 752 122 describes an optimization of these protocols, it is restricted to protocols involving the discrete logarithm method following a mode called “with pre-calculations” that has the drawback of implying regularly scheduled reloads.
The document from J. BRANDT et al. entitled “zero-knowledge Authentication scheme with Secret Key Exchange” published in Advances in Cryptology, Crypto 88 Proceedings, XP 000090662, pp. 583-588, describes a zero-knowledge authentication scheme with exchange of secret keys between two users, a scheme wherein the prover calculates its own modulus n=pq and carries out an operation of the type md (mod n).
The present invention aims to reduce the number of calculations to be carried out by the prover when using zero-knowledge identification (or message signature or authentication) protocols involving factoring, the gain being liable to reach a factor 2 or 3 when using a particular operation v=s−t (mod n).
It also makes possible—and in particular when coupled with the GUILLOU-QUISQUATER protocol—the fast completion of an identification (or message authentication or signature) with public key included in a low cost standard microcircuit card, for applications such as the electronic purse or next generation telecard.
The modulus n being an individual parameter (in other words each user owns his own n value), this selection may be exploited in the following two ways (which may be advantageously combined):
1) first by retaining a length of n lower than the currently used values (typically lower than 1000 bits and for example, ranging between 700 and 800 bits); this is possible as breaking the factoring of n only compromises the secret key of the related user and in no way the secret keys of other users; this modification alone reduces the duration of calculations carried out modulo n by 40%;
2) If the user has stored the prime factors of n in the memory of his security device, he may use the Chinese remainders technique to further reduce the duration of modulo n calculations by 40%, when there are two prime factors; this reduction may be increased when using several prime factors (typically 3 or 4).
On the whole, the modulo n calculations can then be reduced by 60%, that is a factor 2, at least.
Precisely, the invention relates to a process of identification involving a first entity called a “prover”, owning a public key v and a secret key s, these keys being related by a modulo n calculation, where n is an integer called modulus, specific to the prover, and a second entity called a “verifier”, which knows the public key v, these entities being provided with means to exchange information in a zero-knowledge context and to carry out cryptographic calculations on this information, some calculations being performed in the modulo n mode, the process being characterised by the fact that the modulus of the modulo n operation expressed as v=s−t (mod n), t being a parameter.
The aforementioned entities may be, for example, micro-circuit cards, electronic purses, telecards, and so on . . .
Following a preferred implementation, the zero-knowledge information exchanges and the cryptographic calculations are as follows:
The size of the number n, expressed in number of bits, is less than 1000. For example, it may be between 700 and 800.
The present invention also relates to a message signature process to be used by an entity called a “signatory”, this entity being provided with a public key v and a secret key s, which are related by a modulo n operation, where n is an integer called modulus and t is a parameter, a process in which the signatory calculates an opening c that is notably a function of the message to be signed and a number y that is a function of the secret key, transmits the numbers y and c that are the signature and the message, the process being characterised in that the modulus n is specific to the signatory.
Following a preferred implementation, the signatory selects an integer r at random between 1 and n−1, calculates a parameter x equal to rt (mod n), calculates a number c that is a function of the parameter x and of the message to be signed, calculates a number y using the secret key s, as a function of numbers r and e, then transmits the numbers c and y as signature.
In the following description, the invention is assumed to be combined with the protocol GUILLOU-QUISQUATER, as an example. It is clear that the invention is not restricted to this protocol.
Note that the universal parameters of the GUILLOU-QUISQUATER protocol are the modulus n, products of prime numbers, comprising at least 1024 bits, and an integer value t.
The public key v and the secret key s verify the relation v=s−t (mod n).
The retained security level is u (lower than or equal to t, commonly equal to t)
The authentication of A by B, which are named Alice and Bob, following the usual terminology, is completed as follows:
1. Alice selects r within the range [1,n−1], calculates x=rt (mod n) then c=h(x,[M]) and sends c to Bob.
2. Bob selects e within the range [1,u−1] and sends e to Alice.
3. Alice calculates y=rse (mod n) and sends y to Bob.
4. Bob calculates x=ytve (mod n) and verifies that c=h(x, [M])
When no message is to be authenticated, it is optional to involve the pseudo random function h: c=x can be used. The verification then consists of checking that x=ytve (mod n).
In the protocol modified in accordance with the invention, t is the only universal parameter.
The public key is (n,v), where n has at least 768 bits. The public key v and the secret key of Alice satisfy the relation v=s−t (mod n).
The secret key may include prime factors from n to take advantage of the second aspect of the invention.
The parameter t may be included in the public key (in this case, there is no longer any universal parameter).
The security level retained by Alice and Bob is u (lower than or equal to t; usually u=t).
The authentication of Alice by Bob is performed as described above, but with faster calculations, which results from a smaller modulus n.
As all Alice's calculations are carried out modulo n, the gain factor resulting from only one modular multiplication affects the complete set of calculations completed by Alice when carrying out the protocol. This should be the same, for example, with Fiat-Shamir or Girault protocols (in the latter case, no gain should be expected in step 3, as there is no modular computation, but the execution time of this step is negligible with respect to the modular exponentiation of the first one).
The invention may also be implemented by the Chinese remainders technique, which consists of calculating the values modulo n of the prime factors of n. As these numbers are inevitably smaller, these operations are quickly done. The result modulo n is still to be obtained through a “reconstitution” operation. This technique is described in the article of J. J QUISQUATER and C. COUVREUR entitled (Fast Decipherment algorithm for RSA public-key crypto-system” published in “Electronic Letters”, vol. 18, October 1982, pp. 905-907.
Let's consider the case when n is the product of two prime factors p and q.
From the Bezout theorem, it is known that two integers exists, such as ap+bq=1
To calculate y=xe (mod n), we start by reducing x modulo each prime factor by calculating xp=x (mod p) and xq=x (mod q). We also reduce e modulo (p−1) and (q−1) by calculating ep=e modulo (p−1) and eq=e (modulo q−1) (in the protocol of Quillou-Quisquater, e is always lower than e−1 and q−1, then ep=eq=e.
We then calculate yp=xp e p (mod p) and yq=xq e q (mod q). When p and q are of similar size, each of these calculations is about 8 times faster than the calculation y=xe (mod n) when e and n are of similar size (first case); 4 times faster when the size of e is lower than or equal to the size of p (second case as, for example, in the algorithm). The set of two calculations is then either 4 times faster or 2 times faster.
y is still to be reconstructed from yp and yq, which is carried out using the relation:
On the whole, the method of Chinese remainders leads to an acceleration of calculations by a factor ranging from 3 to 4 in the first case, and from 1.5 to 2 in the second case, when the number of prime factors (assumed to be of similar sizes) is larger than 2 and equal to k; the acceleration factor is nearing k2 in the first case and close to k in the second case.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4964164 *||Aug 7, 1989||Oct 16, 1990||Algorithmic Research, Ltd.||RSA computation method for efficient batch processing|
|US5140634 *||Oct 9, 1991||Aug 18, 1992||U.S Philips Corporation||Method and apparatus for authenticating accreditations and for authenticating and signing messages|
|US5218637||Jun 11, 1991||Jun 8, 1993||L'etat Francais Represente Par Le Ministre Des Postes, Des Telecommunications Et De L'espace||Method of transferring a secret, by the exchange of two certificates between two microcomputers which establish reciprocal authorization|
|US5787178 *||Apr 12, 1996||Jul 28, 1998||Deutsche Telekom Ag||Computerized method for signing a message|
|EP0325238A2||Jan 18, 1989||Jul 26, 1989||Yeda Research And Development Company Limited||Improved variants of the Fiat-Shamir identification and signature scheme|
|FR2716058A1||Title not available|
|FR2752122A1||Title not available|
|JPH118616A||Title not available|
|JPH08149124A||Title not available|
|WO1989011706A1||May 4, 1989||Nov 30, 1989||Ncr Corporation||Method and device for authentication|
|WO1998042173A2||Mar 23, 1998||Oct 1, 1998||Fd Finanssidata Oy||Use of banking services in a digital cellular radio system|
|WO1998051038A1||May 5, 1998||Nov 12, 1998||Gemplus S.C.A.||Pseudo-random generator based on a hash coding function for cryptographic systems requiring random drawing|
|1||A. Fiat, et al., Advances in Cryptology-Crypto '86 Proceedings, pp. 186-194, "How to Prove Yourself: Practical Solutions to Identification and Signature Problems", Aug. 11-15, 1986.|
|2||A. Fiat, et al., Advances in Cryptology—Crypto '86 Proceedings, pp. 186-194, "How to Prove Yourself: Practical Solutions to Identification and Signature Problems", Aug. 11-15, 1986.|
|3||Brandt et al., "Zero-Knowledge Authentication Scheme with Secret Key Exchange", Advances in Cyptology-Crypto '88 Proceedings, pp. 583-588, Aug. 21-25, 1988.|
|4||Brandt et al., "Zero-Knowledge Authentication Scheme with Secret Key Exchange", Advances in Cyptology—Crypto '88 Proceedings, pp. 583-588, Aug. 21-25, 1988.|
|5||Fiat et al., "How to Prove Yourself: Practical Solutions to Identification and Signature Problems", Advances in Cyptology-Crypto '86 Proceedings, pp. 1806-194, Aug. 11-16, 1986.|
|6||Fiat et al., "How to Prove Yourself: Practical Solutions to Identification and Signature Problems", Advances in Cyptology—Crypto '86 Proceedings, pp. 1806-194, Aug. 11-16, 1986.|
|7||Final Office Action mailed Jan. 18, 2006, for U.S. Appl. No. 09/889,557.|
|8||Final Office Action mailed May 27, 2005, for U.S. Appl. No. 09/889,557.|
|9||Guillou et al., "A Practical Zero-Knowledge Protocol Fitted to Security Microprocessor Minimizing Both Transmission and Memory", Advances in Cryptology-Eurocrypt '88 pp. 123-128, May 25-27, 1988.|
|10||Guillou et al., "A Practical Zero-Knowledge Protocol Fitted to Security Microprocessor Minimizing Both Transmission and Memory", Advances in Cryptology—Eurocrypt '88 pp. 123-128, May 25-27, 1988.|
|11||H-P. Konigs, IEEE Communications Magazine, vol. 29, No. 6, pp. 42-48, "Cryptographic Identification Methods for Smart Cards in the Process of Standardization", Jun. 1991.|
|12||International Preliminary Examination Report received by WIPO on Dec. 21, 2000, for PCT App. PCT/FR00/00174.|
|13||International Search Report mailed Feb. 25, 2000, for PCT App. PCT1FR00/00174.|
|14||J. Brandt, et al., Advances in Cryptology-Crypto '88 Proceedings, pp. 583-588, "Zero-Knowledge Authentication Scheme with Secret Key Exchange", Aug. 21-25, 1988.|
|15||J. Brandt, et al., Advances in Cryptology—Crypto '88 Proceedings, pp. 583-588, "Zero-Knowledge Authentication Scheme with Secret Key Exchange", Aug. 21-25, 1988.|
|16||J-J. Quisquater, et al., Electronics Letters, vol. 18, No. 21, pp. 905-907, "Fast Decipherment Algorithm for RSA Public-Key Cryptosystem", Oct. 14, 1982.|
|17||Königs, "Cryptographic Identification Methods for Smart Cards in the Process of Standardization", IEEE Communications Magazine, vol. 29, No. 6, pp. 42-48, Jun. 1991.|
|18||L. C. Guillou, et al, Advances in Cryptology-Eurocrypt '88, pp. 123-128, "A Practical Zero-Knowledge Protocol Fitted to Security Microprocessor Minimizing Both Transmission and Memory", May 25-27, 1988.|
|19||L. C. Guillou, et al, Advances in Cryptology—Eurocrypt '88, pp. 123-128, "A Practical Zero-Knowledge Protocol Fitted to Security Microprocessor Minimizing Both Transmission and Memory", May 25-27, 1988.|
|20||Notice of Allowance mailed Nov. 2, 2006, for U.S. Appl. No. 09/889,557.|
|21||Notice of Reasons for Rejection for Japanese App. 2000-596695 mailed Apr. 23, 2010.|
|22||Office Action for JP App. 2000-596695 dated Feb. 24, 2011, including Concise Explanation of Relevance (4 pgs.).|
|23||Office Action mailed Jan. 31, 2005, for U.S. Appl. No. 09/889,557.|
|24||Office Action mailed Sep. 27, 2005, for U.S. Appl. No. 09/889,557.|
|25||Preliminary Search Report dated Dec. 28, 1999 for FR App. 9900887.|
|26||Quisquater et al., "Fast Decipherment Algorithm for RSE Public-Key Cryptosystem", Electronics Letters, vol. 18, No. 21, pp. 905-907, Oct. 14, 1982.|
|27||*||Scheiner, Bruce, Applied Cryptography-Protocvols, Algorithms, and Source Code in C(Second Edition), John Wiley & Sons, Inc(1996): pp. 249-250.|
|28||*||Scheiner, Bruce, Applied Cryptography—Protocvols, Algorithms, and Source Code in C(Second Edition), John Wiley & Sons, Inc(1996): pp. 249-250.|
|29||*||Scheiner, Bruce. "Public-Key Algorithms: RSA." Applied Cryptography: Protocols, Algorithms and Source Code in C. New York: John Wiley & Sons, 1996. pp. 470.|
|30||*||Schenier, Bruce. "Mathematical Background: Number Theory." Applied Cryptography: Protocols, Algorithms and Source Code in C. New York: John Wiley & Sons, 1996. pp. 249-250.|
|31||Schneier, "Mathematical Background: Number Theory", Applied Cryptography- Protocols, Algorithms, and Source Code in C (Second Edition) New York: John Wiley & Sones, 1996 pp. 249-250.|
|32||Schneier, "Public Key Algorithms: RSA", Applied Cryptography-Protocols, Algorithms, and Source Code in C (Second Edition) New York: John Wiley & Sones, 1996 pp. 470.|
|33||Schneier, "Mathematical Background: Number Theory", Applied Cryptography— Protocols, Algorithms, and Source Code in C (Second Edition) New York: John Wiley & Sones, 1996 pp. 249-250.|
|34||Schneier, "Public Key Algorithms: RSA", Applied Cryptography—Protocols, Algorithms, and Source Code in C (Second Edition) New York: John Wiley & Sones, 1996 pp. 470.|
|35||*||Schneier, Bruce. Applied Cryptography, "Number Theory". New York, NY, 1996, 244-80.|
|U.S. Classification||380/30, 713/171, 380/282, 713/168|
|International Classification||H04L9/28, G09C1/00, H04L9/32|
|Cooperative Classification||H04L9/3218, H04L9/3247|
|European Classification||H04L9/32G, H04L9/32N|
|Aug 19, 2010||AS||Assignment|
Owner name: PHENTAM DIRE NV, LLC, DELAWARE
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:FRANCE TELECOM SA;REEL/FRAME:024858/0861
Effective date: 20081217
|Nov 22, 2011||CC||Certificate of correction|
|Jul 25, 2014||FPAY||Fee payment|
Year of fee payment: 8
|Dec 22, 2015||AS||Assignment|
Owner name: CALLAHAN CELLULAR L.L.C., DELAWARE
Free format text: MERGER;ASSIGNOR:PHENTAM DIRE NV, LLC;REEL/FRAME:037358/0797
Effective date: 20150827