JP5326715B2 - An authentication medium for detecting failure-use attacks on authentication processing using elliptic curve cryptography - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a technique for fault attack detection in IC card authentication using an elliptic curve cryptosystem method, by overcoming the problem that any fault attack can not be detected in many cases because a conventional technique utilizes a decoding message as a point on an elliptic curve. <P>SOLUTION: In IC card, a cipher input from an authentication device is decrypted by a secret key to obtain a decrypted message, and then the correspondence between the decrypted message and the cipher is verified by using a degenerated pairing function on an elliptic curve, thus detecting the fact of fault attack. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to an authentication medium for detecting a failure use attack against authentication processing using elliptic curve cryptography.

  In authentication using an authentication medium such as an IC card, a public key cryptosystem is generally used. The IC card is equipped with a public key cryptosystem private key and an encryption (decryption) circuit, which makes it possible to perform authentication with cryptographically secure security.

  Elliptic curve cryptography is known as a public key cryptosystem usually used in IC card authentication. Elliptic curve cryptography is characterized in that it can secure the same level of security with a shorter key length than the RSA cryptosystem, which is also one of public key cryptosystems. For example, it is known that the security level of elliptic curve cryptography having a key length of 160 bits and RSA cryptography having a key length of 1024 bits are substantially the same. For this reason, elliptic curve cryptography is suitable for mounting in a small device such as a memory, an IC card with a small processing power, or a mobile phone.

  An example of an authentication process in an IC card authentication system using an elliptical Elgaral (ElGamal) encryption, which is one of elliptic curve encryptions, will be described. Let p be a prime number and G be a base point in an elliptic curve E on a finite field GF (p) (details of each symbol will be described later). In the IC card, the secret key d of the elliptic curve encryption based on the elliptic curve is secretly stored. Further, the authentication apparatus stores the public key Y = [d] G of the elliptic curve cryptography (here, the symbol [] indicates scalar multiplication).

  Based on FIG. 11, a processing flow for the authentication apparatus to authenticate an IC card (more precisely, a user holding the IC card) will be described. The authentication device detects that an IC card has been input (step S701). Next, the authentication device generates an appropriate message M and stores it in the storage device, and generates a ciphertext C = (C1, C2) by encrypting the message M using a public key (step S702). ). Here, C1 = [r] G and C2 = M + [r] Y (r is an appropriately selected random number). Then, the authentication device transmits the generated ciphertext C to the IC card (step S703).

  Next, the IC card receives the ciphertext C (step S705). The IC card obtains an intermediate message K by decrypting C1 using the secret key (step S707). Specifically, the IC card calculates K = [d] C1. Further, the IC card decrypts C2 and obtains a decrypted message M ′ (step S709). Specifically, the IC card calculates M ′ = C2-K. Then, the IC card transmits M ′ to the authentication device (step S711).

  The authentication apparatus receives M ′ (step S713). The authentication device reads M from the storage device, and compares M with the received M ′ (step S715). The authentication apparatus determines that “authentication is successful” when M and M ′ match (step S717). On the other hand, if M and M ′ do not match, the authentication apparatus determines “authentication failure” (step S719). The security of this authentication system is based on the fact that the IC card cannot calculate M ′ in which M ′ = M based on C in step 3 unless the IC card has a secret key.

  In the IC card, since the secret key is held in a tamper resistant manner, it is unlikely that the secret key is directly read from the outside, but there are several attacks that indirectly guess this. One such attack is a failure use attack. In the failure use attack, the processing result of the IC card is obtained by giving a physical change such as suddenly changing the supply voltage from the outside while the IC card is encrypting or decrypting using the secret key. Is intentionally mistaken. The attacker can guess the secret key under a predetermined condition by obtaining the erroneous processing result on the authentication device.

  That is, in steps S707 and S708, the IC card calculates a decryption message from the ciphertext C received from the authentication device. At that time, the attacker gives a physical operation such as a high voltage to the IC card. At this time, even if the IC card is valid (having a secret key corresponding to the public key on the authentication device), the decrypted message is not correctly calculated (M ′ in which M ′ ≠ M is calculated). The IC card transmits an incorrect result to the authentication device. The attacker can guess the secret key held in the IC card by obtaining the incorrect output and analyzing the output.

  As in the above example, if an authentication system using elliptic curve cryptography on an IC card is subjected to a failure use attack, the secret key of the IC card may be specified by the attacker. That is, since a failure use attack leads to leakage of a secret key, countermeasures are strongly desired.

As a conventional technique for detecting a failure use attack, there is a method that uses the property that when an IC card is subjected to a failure use attack, the calculation result may not be the point of the elliptic curve used in elliptic curve cryptography. (For example, Patent Document 1). In the prior art, the elliptic curve E relating to the IC card is expressed by the equation E: y ² + a ₁ xy + a ₃ y = x ³ + a ₂ x ² + a ₄ x + a ₆
The function Is_on_EC (x, y) is expressed as Is_on_EC (x, y) = x ³ + a ₂ x ² + a ₄ x + a ₆ − (y ² + a ₁ xy + a ₃ y)
It is determined. After obtaining the intermediate message K, the IC card determines whether K is an elliptic curve E point. Here, the reason for verifying the value of K, not M ′, is that the attacker wants to estimate the secret key d, so that it is considered that a failure use attack is performed when calculating K = [d] C1. On the other hand, since M ′ = C2−K, that is, d is not used in the calculation of M ′, it does not make sense to perform a failure use attack only during the calculation of M ′.

In order for the IC card to determine whether K is a point of the elliptic curve E, specifically, if K is expressed as (x _k , y _k ), Is_on_EC (x _k , y _k ) is calculated. That's fine. If Is_on_EC (x _k , y _k ) = 0, K is an elliptic curve point, and if Is_on_EC (x _k , y _k ) ≠ 0, it can be determined that K is not an elliptic curve point. K should originally be a point of an elliptic curve E (details will be described later, but in elliptic curve cryptography, the original message, ciphertext, and decryption message are all points of an elliptic curve). Therefore, when K is not an elliptic curve point, the IC card can be determined to have been subjected to a failure use attack.
JP 2004-252433 A

  However, the prior art has a problem that there are many cases where it cannot be detected in the first place that the IC card is under a failure use attack. Specifically, the prior art according to Patent Document 1 cannot detect an attack when a calculation result of a failure use attack happens to be a point on an elliptic curve. In such a case, it is known that 1/2 of the entire ciphertext is applicable. This is because when the value of x is given, the probability that y satisfying the expression E can be obtained (that is, the equation E relating to y can be solved) is ½. That is, the conventional technique has a problem that failure detection of failure utilization attacks fails at a rate of once every two times.

  The above problem is due to the fact that the prior art uses only the intermediate message K to determine whether there is a failure use attack. There are many Ks that satisfy the above-mentioned conditions (must be points on the elliptic curve) that the intermediate message K should satisfy (the number of types of K is equal to the number of points on the elliptic curve E). Therefore, even if an incorrect K is calculated in response to a failure use attack, there is a high possibility that the incorrect K satisfies the above condition.

  In order to solve the above problem, the present invention is characterized in that the intermediate message K is verified using C1 which is a ciphertext of K. That is, a one-to-one correspondence relationship between the ciphertext C1 and the intermediate message K (an intermediate message K is certainly a decryption of the ciphertext C1) is an approach different from the decryption process. By solving this problem, the above-mentioned problem is solved.

In order to realize this, the present invention uses a function called non- degenerate pairing on an elliptic curve. That is, the correspondence between the ciphertext C1 and the intermediate message K is grasped by using non- degenerate pairing, and thereby a failure use attack is detected.

The pairing e on the elliptic curve is a function that outputs a certain value for two points of the elliptic curve and satisfies the following two conditions (where P1, P2, and P3 are points of the elliptic curve and To do).
(1) e (P1, P2 + P3) = e (P1, P2) · e (P1, P3)
(2) e (P1, P2) = e (P2, P1)
In addition to the above two conditions, e that satisfies the following conditions is called “ non- degenerate pairing”.
(3) e (P1, P1) ≠ 1
Note that non- degenerate pairing itself on an elliptic curve is a known technique. Conventionally, non- degenerate pairing is used in an attack technique for elliptic curve cryptography or in internal processing of the elliptic curve cryptography itself.

The IC card according to the present invention uses the non- degenerate pairing e to detect the presence or absence of a failure use attack. Specifically, this is performed as follows. The IC card calculates e1 = e (G, K) based on the base point G and K after obtaining the intermediate message K from C1 in the decryption processing of the ciphertext C = (C1, C2) described above. At the same time, e2 = e (Y, C1) is calculated based on the public key Y and the C1. Here, e is a non- degenerate pairing function. The IC card determines whether e1 = e2 holds. If e1 = e2, the IC card determines that no failure use attack has been performed. On the other hand, if e1 ≠ e2, the IC card determines that a failure use attack has been performed.

  According to the disclosed IC card, there is an effect that it is possible to detect a failure use attack with a higher probability than in the prior art.

Consider the case where no failure use attack has been performed. At this time, based on the nature of non- degenerate pairing, the e1 is
e1 = e (G, K) = e (G, [d] C1) = e (G, [d−1] C1 + C1)
= E (G, [d-1] C1) .e (G, C1)
= E (G, [d-2] C1 + C1) · e (G, C1)
= E (G, [d-2] C1) · e (G, C1) ²
And can be transformed. If the same formula modification is repeated, the relational expression e1 = e (G, C1) ^d can be derived after all. Similarly, for e2,
e2 = e (Y, C1) = e ([d] G, C1) = e (G, C1) ^d
In the end, the relational expression e2 = e (G, C1) ^d can be derived. That is, e1 = e2 is always established if no failure use attack is performed. From this kinematic pair, if e1 ≠ e
If it is 2, a failure use attack is performed with a probability of 100%.

On the other hand, consider a case where a failure use attack is performed. At this time, except for the case where K is accidentally calculated from C1 (which occurs with a random probability, it occurs with probability 1 / p), K ′ that is different from C1 is calculated. At this time, e1 = e (G, K ′) ≠ e (G, C1) ^d . That is, if a failure use attack is performed, e1 ≠ e2 is established with a probability of 1− (1 / p). Here, since p is usually a very large prime number, 1 / p has a negligible probability. Therefore, if e1 = e2, a failure-use attack is not performed with a probability close to 100%.

  That is, by determining whether or not e1 = e2 holds, the failure use attack can be detected with high accuracy. Therefore, according to the method of the present application, it is possible to detect the presence / absence of a failure use attack in the IC card at a rate of almost 100%, and to avoid leakage of the secret key due to the failure use attack.

Hereinafter, embodiments of the present invention will be described with reference to the drawings. The present invention is based on discrete mathematics such as elliptic curve cryptography and non- degenerate pairing on an elliptic curve, and there is an aspect in which the invention is not easy to understand. Therefore, in the description according to the embodiment of the present invention, on the premise that non- degenerate pairing on elliptic curve cryptography or elliptic curve is configured, an authentication system that enables detection of a failure use attack using them Will be described. The elliptic ElGamal encryption and the non- degenerate pairing function on the elliptic curve will be described last after the description according to each embodiment.

  FIG. 1 shows a functional block diagram of an authentication system according to the first embodiment of the present invention. The authentication system includes an IC card 100, which is an example of an authentication medium, and an authentication device 300.

  The IC card 100 includes a storage unit 101, a public information storage unit 103, a secret key storage unit 105, an attack detection unit 107, a decryption processing unit 109, a communication unit 111, and a control unit 113. The storage unit 101 stores and stores various calculation results and the like. The public information storage unit 103 stores and holds PKI (public key infrastructure) public information. Here, the public information must include at least the identifier of the user whose identity is guaranteed by the public key, the public key Y of the user, and the public key cryptosystem. A public key certificate defined in 509 can be used. In the present embodiment, elliptic El Gamal cryptography, which is one of elliptic curve cryptography, is used as public key cryptography. In the public key cryptosystem described above, in addition to using the elliptic Elgamal cryptosystem as the public key cryptosystem, the finite field GF (p), the elliptic curve E on the GF (p), and the base point G, which are public parameters, are used. (The details of each symbol will be described later). Here, p is a prime number, and p = 2 mod 3 is satisfied. That is, E is an elliptic curve of prime order. In many elliptic curve cryptography standards including FIPS 186-2, prime order elliptic curve cryptography is adopted. The secret key storage unit 105 stores a secret key d corresponding to the public key and holds it secretly. Here, it is assumed that the relationship Y = [d] G is established between the secret key d and the public key Y. The attack detection unit 107 detects a failure use attack. The decryption processing unit 109 decrypts the ciphertext and obtains a decrypted message. The communication unit 111 transmits and receives data such as a ciphertext and a decryption message between the authentication device 300 and the IC card 100. The control unit 113 controls the failure use attack detection unit 107, the decryption processing unit 109, and the communication unit 111.

  On the other hand, the authentication device 300 includes a storage unit 301, a public information storage unit 303, a random number generation unit 305, an authentication processing unit 307, an encryption processing unit 309, a communication unit 311, and a control unit 313. The storage unit 301 stores and stores various calculation results and the like. The public information storage unit 303 stores and holds PKI (public key infrastructure) public information. This public information is equivalent to information stored in the public information storage unit 103 of the IC card 100. The random number generation unit 305 generates a random number. The random number generated here may be a pseudo-random number generated by a mathematical method. The authentication processing unit 307 performs user authentication processing. The encryption processing unit 309 encrypts the message and generates a ciphertext. The communication unit 311 transmits and receives data such as ciphertext and decryption message between the authentication device 300 and the IC card 100. The control unit 313 controls the random number generation unit 305, the authentication processing unit 307, the encryption processing unit 309, and the communication unit 311.

  Based on FIG. 2, a processing flow of the authentication system according to the first embodiment will be described.

  First, the authentication device detects that an IC card has been input (step S1). Next, the encryption processing unit 309 of the authentication device 300 creates a ciphertext to be transmitted to the IC card 100 (step S2). This is done as follows. The encryption processing unit 309 appropriately selects the message M that is a point on the elliptic curve and stores it in the storage unit 301. Next, the random number generation unit 305 generates a random number r and stores it in the storage unit 301. For the message M, the random number generation unit 305 can generate a random number r ′ different from r, and the encryption processing unit 309 can generate it based on the r ′. Next, the encryption processing unit 309 reads the base point G from the public information storage unit 303, reads r from the storage unit 301, calculates the point [r] G using the read G and r, The calculation result is stored in the variable C1 on the storage unit 301. Further, the encryption processing unit 309 reads the public key Y from the public information storage unit 303, reads r and M from the storage unit 301, and uses the read Y, M, and r to obtain a point M + [r]. Y is calculated, and the calculation result is stored in the variable C2 on the storage unit 301. Let C = (C1, C2) be a ciphertext.

  Next, the communication unit 313 of the authentication device 300 transmits the created ciphertext C = (C1, C2) to the IC card 100 (step S3). The communication unit 113 of the IC card 100 receives the transmitted ciphertext C and stores it in the storage unit 101 (step S5). Next, the decryption processing unit 109 of the IC card reads the C1 from the storage unit 101, and decrypts the read C1 (step S7). That is, the decryption processing unit 109 reads the secret key d from the secret key storage unit 105, calculates the point [d] C1 using the read d, and stores the calculation result in the variable K on the storage unit 101.

  Next, the attack detection unit 107 of the IC card 100 determines whether there is a failure use attack (step S9). Details of the determination process will be described later. Next, the attack detection unit 107 branches the process based on the determination result of step S9 (step S11). If the determination result is YES (attack is detected), the IC card control unit 111 performs predetermined abnormal processing as necessary (step S13). That is, detecting an attack means that a failure use attack has been performed on the IC card. Therefore, the success or failure of the authentication is meaningless, and when the IC card outputs a decryption message to the authentication device 300, the attacker eavesdrops on it. Therefore, the control unit 111 does not transmit a decryption message to the authentication device 300, but performs abnormal processing such as reporting to the administrator or leaving an error log. Then, the current authentication process ends.

  On the other hand, if the determination result in step S9 is NO (no attack detected), the decryption processing unit 109 of the IC card 100 decrypts the message M ′ (step S15). That is, the decoding processing unit 109 reads K and C2 from the storage unit, calculates a point C2-K based on the read K and C2, and stores the calculation result in a variable M ′ on the storage unit 101.

  Next, the communication unit 113 of the IC card 100 transmits the decrypted message M ′ to the authentication device 300 (step S <b> 17). The communication unit 311 of the authentication device 300 receives the transmitted message M ′ and stores it in the storage unit 301 (step S <b> 19). Finally, the authentication processing unit 307 of the authentication device 300 determines whether authentication is successful (step S21). That is, the authentication processing unit 307 reads from the storage unit 301 and compares the message M selected and encrypted by the encryption processing unit 309 of the authentication apparatus 300 and the message M ′ received from the IC card. If M = M ′, the authentication processing unit 307 determines “authentication success” (step S <b> 23). If there is a correspondence between the public key Y stored in the public information storage unit 303 of the authentication device 300 and the private key d stored in the private key storage unit 105 of the IC card 100, the authentication device 300 performs encryption. This is because M = M ′ should be satisfied because the received message should be correctly decrypted by the IC card 100 and returned to the original message. On the other hand, if M ≠ M ′, the authentication processing unit 307 determines “authentication failure” (step S25). Regardless of whether the authentication is successful or unsuccessful, the control unit 311 of the authentication apparatus ends the current authentication process and waits for input.

Next, based on FIG. 3, the detailed processing flow of the process (step S9) in which the attack detection unit 107 of the IC card according to the first embodiment determines the presence or absence of a failure use attack will be described. The attack detection unit 107 reads out the base point G and K obtained by decrypting the ciphertext C1 from the storage unit 101, and calculates e1 = e (G, K) using the read G and K. (
Step S101). Here, e is a non- degenerate pairing function described later. Details of the processing in step S101 will be described later. Next, the attack detection unit 107 reads the public key Y from the public information storage unit 103, reads the ciphertext C1 from the storage unit 101, and uses the read Y and C1, e2 = e (Y, C1). Is calculated (step S102). Details of the processing in step 102 will also be described later. Next, the attack detection unit 107 confirms whether or not e1 = e2 is established (step S103). If e1 = e2 holds,
The attack detection unit 107 determines that no failure use attack has been received (step S104). If e1 = e2 is not established, the attack detection unit 107 determines that a failure use attack has been received (step S105).

  Next, details of the process of calculating e1 = e (G, K) in step S101 and the process of calculating e2 = e (Y, C1) in step S102 will be described. Since these two processes are almost the same except for a part, step S101 that is a process for calculating e1 will be described in detail below, and step S102 that is a process for calculating e2 is different from step S101. Briefly explained at the center.

  Details of the process of calculating e1 = e (G, K) (step S101) will be described based on the process flow of FIG. First, the attack detection unit 107 reads the base point G from the public information storage unit 103, reads the calculated K from the storage unit 101, and stores G in the variable P1 on the storage unit 101 and K in P2. (Step S201). Here, P1 and P2 are points on the plane, respectively, and can be expressed in a coordinate format such as (x, y). Then, the attack detection unit 107 calculates e (P1, P2) (step S202).

  Here, the details of the process of calculating e (P1, P2) (step S202) will be described based on the process flow of FIG. First, the attack detection unit 107 reads P2 from the storage unit 101, calculates Φ (P2), and stores the calculation result in P2 (step S401). Note that the mapping Φ is defined below.

Φ: (x, y) → (wx, y)
Here, w is a primitive third root of 1. Next, the attack detection unit 107 stores P1 in the variable S on the storage unit 101 (step S402). Then, the attack detection unit 107 derives a function f _S (Step S403).

Here, based on the process flow of FIG. 7, the details of the process of calculating f _S for the point S of the elliptic curve E will be described. First, the attack detection unit 107 calculates a binary expansion of the order p of the finite field and stores it in (n _t−1 ,..., N ₀ ) ₂ which is a t set of variables on the storage unit. (Step S501). Next, the attack detection unit 107 stores the natural numbers t−2 and 1 in the variables i and r on the storage unit 101, stores the function 1 in the variable f on the storage unit 101, and further stores the function 1 on the storage unit 101. S, which is the point of the elliptic curve E, is stored in the variable V (step S502).

Next, the attack detection unit 107 reads the variable i on the storage unit 101 and determines whether i <0 (step S503). When i <0 is not satisfied (NO in the branch), the attack detection unit 107 reads the variables V and r on the storage unit 101, obtains a straight line that passes through the point V of the elliptic curve E and touches the elliptic curve E, and calculates the straight line. It is assumed that l _{r, r} (x, y) = 0 (step S504). Next, the attack detection unit 107 calculates the point [2] V based on the read V and stores it in the variable V on the storage unit 101 (step S505). Further, the attack detection unit 107 reads the variables V and r on the storage unit 101, _{obtains a} straight line passing through V and the infinity point ∞ on the elliptic curve E, and _{obtains the} straight line as v _2r } (x, y) = 0. (Step S506). Then, the attack detection unit 107 reads f stored in the storage unit 101, and based on the obtained functions l _{r, r} (x, y) and v _2r (x, y), the function f ² · l _{r , r} (x, y) / v _2r (x, y) is calculated, and the calculated function is stored in the variable f on the storage unit 101 (step S507). The attack detection unit 107 calculates 2r based on the read r value, and stores the calculated value in the variable r on the storage unit 101 (step S508).

Next, the attack detection unit 107 reads n _i on the storage unit 101 and determines whether the value is 1 (step S509). If n _i = 1 is not satisfied (NO at the branch), the attack detection unit performs a process of step S515 described later. On the other hand, if n _i = 1 (YES at branch), the attack detection unit 107 reads the variables V and r on the storage unit 101, obtains a straight line passing through the point V of the elliptic curve E and the point S, and The straight line is set to l _{r, 1} (x, y) = 0 (step S510). Next, the attack detection unit 107 calculates the point V + S based on the read V and S, and stores it in the variable V on the storage unit 101 (step S511). Further, the attack detection unit 107 reads the variables V and r on the storage unit 101, obtains a straight line passing through V and the infinity point ∞ on the elliptic curve E, and obtains the straight line as v _{r + 1} (x, y) = It is set to 0 (step S512). Then, the attack detection unit 107 reads f stored in the storage unit 101, and based on the obtained functions l _{r, 1} (x, y) and v _{r + 1} (x, y), the function f · l _{r, 1} (x, y) / v _{r + 1} (x, y) is calculated, and the calculated function is stored in the variable f on the storage unit 101 (step S513). Further, the attack detection unit 107 calculates r + 1 based on the read value of r, and stores the calculated value in the variable r on the storage unit 101 (step S514).

  The attack detection unit 107 reads i on the storage unit 101, calculates i + 1, and stores the calculated value in the variable i on the storage unit 101 (step S515). And the attack detection part 107 returns to step S503 and repeats a process.

On the other hand, if i <0 in the determination in step S503 (YES in branch), the attack detection unit 107 reads f from the storage unit 101 and stores it in the variable f _S on the storage unit 101 (step S516). . Thus, the derivation of the desired function f _S is completed.

Returning to the explanation of the processing in FIG. 6, the attack detection unit 107 stores the derived function f _S in f _P1 on the storage unit 101 (step S404). Further, the attack detection unit 107 stores P2 in the variable S on the storage unit 101 (step S405). Then, the attack detection unit 107 derives the function f _S again based on the processing flow of FIG. 7 described above (step S406). Then, the derived function f _S is stored in f _P2 on the storage unit 101 (step S407).

Next, the attack detection unit 107 arbitrarily selects two points on the elliptic curve E and stores them in U and V on the storage unit 101, respectively (step S408). The attack detection unit 107 reads P1, P2, U, and V from the storage unit 101, calculates two points V + P1 and U + P2 on the plane, and determines whether these two points are not ∞ (infinity point). Determination is made (step S409). If neither V + P1 nor U + P2 is ∞ (YES in branch), the attack detection unit 107 reads the function f _P2 and the function f _P2 from the storage unit 101, and e (P1, P2) based on the following formula: Is calculated (step S410).

e (P1, P2) = f _P1 (U + P2) / f _P1 (U) · f _P2 (P2) / f _P2 (V + P2)
On the other hand, if either V + P1 or U + P2 is ∞ in step S409 (NO in branch), the attack detection unit 107 returns to the process in step S304 and reselects U and V.

  Returning to the description of the processing in FIG. 4, the attack detection unit 107 stores the calculated e (P1, P2) in e1 of the storage unit 101 (step S203). Thus, the process of calculating e1 = e (G, K) is completed.

  Next, details of the process of calculating e2 = e (Y, C1) (step S102) will be described based on the process flow of FIG. First, the attack detection unit 107 reads the public key Y from the public information storage unit 103, reads the ciphertext C1 from the storage unit 101, and stores Y in the variable P1 on the storage unit 101 and C1 in P2. (Step S301). Then, the attack detection unit 107 calculates e (P1, P2) based on the processing flow of FIG. 6 described above (step S302). Finally, the attack detection unit 107 stores the calculated e (P1, P2) in e2 of the storage unit 101 (step S303). Thereby, the process of calculating e2 = e (Y, C1) ends.

  As described above, according to the authentication system according to the first embodiment of the present invention, it is possible to detect a failure use attack with a high probability. That is, by determining whether or not e1 = e2 holds, the failure use attack can be detected with high accuracy. Therefore, according to the method of the present application, it is possible to detect the presence / absence of a failure use attack in the IC card at a rate of almost 100%, and to avoid leakage of the secret key due to the failure use attack.

  Hereinafter, a second embodiment of the present invention will be described. The second embodiment is the same as the first embodiment in many parts, but the detailed processing flow of step S9, which is a process for determining the presence or absence of a failure use attack by the IC card, is different from FIG. Specifically, first, after performing the above-described conventional failure use attack detection process, the failure use attack detection process described in the first embodiment is performed.

Based on FIG. 8, the process flow of the process (step S9) which the attack detection part 107 of the IC card which concerns on 2nd Embodiment determines the presence or absence of a failure use attack is demonstrated. The elliptic curve E is the equation E: y ² + a ₁ xy + a ₃ y = x ³ + a ₂ x ² + a ₄ x + a ₆
The function Is_on_EC (x, y) is expressed as Is_on_EC (x, y) = x ³ + a ₂ x ² + a ₄ x + a ₆ − (y ² + a ₁ xy + a ₃ y)
It is determined.

First, the attack detection unit 107 reads K = (x _k , y _k ) calculated in step S7 from the storage device, and calculates Is_on_EC (x _k , y _k ) based on K (step S601). Then, the attack detection unit 107 determines whether Is_on_EC (x _k , y _k ) = 0 (step S602). If Is_on_EC (x _k , y _k ) ≠ 0 (NO in branch), the attack detection unit 107 determines that a failure use attack has been received (step S607). This is because if Is_on_EC (x _k , y _k ) ≠ 0, it can be determined that K is not a point on the elliptic curve E but has been subjected to a failure use attack.

On the other hand, if Is_on_EC (x _k , y _k ) = 0 in step S601 (YES in branch)
), The attack detection unit 107 obtains K obtained by decrypting the base point G and the ciphertext C1.
Is read from the storage unit 101, and e1 = e (G, K) is calculated using the read G and K (step S603). Here, e is a non- degenerate pairing function. Next, the attack detection unit 107 reads the public key Y from the public information storage unit 103, reads the ciphertext C1 from the storage unit 101, and uses the read Y and C1, e2 = e (Y, C1). Is calculated (step S604). Here, the details of the processing in steps 603 and S604 may be performed in accordance with the processing flows of FIGS. 4 and 5, respectively. Next, the attack detection unit 107 confirms whether or not e1 = e2 is satisfied (step S605). If e1 = e2 holds, the attack detection unit 107 determines that the failure use attack has not been received (step S606). If e1 = e2 is not established, the attack detection unit 107 determines that a failure use attack has been received (step S607).

  As described above, according to the second embodiment, the same operations and effects as those of the first embodiment are achieved. That is, it becomes possible to detect a failure use attack with a high probability. Furthermore, in the second embodiment, the failure use attack cannot be detected in many cases, but by executing the conventional failure use attack detection method having excellent processing performance in the first stage, the average processing time of the detection process is reduced to about half. can do. Therefore, according to the second embodiment, there is a specific effect that the detection process can be speeded up as compared with the first embodiment.

The configuration of the authentication system according to the first and second embodiments has been described above. However, the present invention is not limited to the above embodiment, and various design changes can be made within the scope of the technical idea. For example, in the above-described embodiment, the elliptic Elgamal encryption is used as the elliptic curve encryption, but the same failure use detection can be performed using other elliptic curve encryption including the elliptic RSA encryption. Also
, It was used Weil Pairing as withdrawal of pairing, may be employed instead Tate pairing (Tate Pairing) and these variations. Further, the IC card 100 may be any portable information terminal, and may be, for example, a mobile phone or a PDA terminal.

  FIG. 9 shows a hardware configuration of the IC card 100 according to the authentication system according to the first and second embodiments. In the IC card 100 of FIG. 9, a central processing unit (CPU) 11, a read only memory (ROM) 12, a random access memory (RAM) 13, a nonvolatile memory 14, and an input / output interface 15 are connected by a bus 17. ing.

  The CPU 11 reads and executes programs and data stored in the ROM 12 described later. The ROM 12 is a storage device that cannot be overwritten, and stores programs and data. The RAM 13 is used by the CPU 11 as a work storage device. The nonvolatile memory 14 is an electrically writable storage device, and stores different data such as personal information for each IC card. The input / output interface 15 inputs and outputs data with an IC card reader described later.

  The above hardware configuration is associated with the functional blocks in FIG. 1 as follows. Each processing unit (attack detection unit 107, decoding processing unit 109, communication unit 111, and control unit 113) of the IC card 100 is stored in the ROM 12. The public information storage unit 103 and the secret key storage unit 105 are realized by the nonvolatile memory 14. The storage unit 101 is realized by the RAM 13. The CPU 11 reads each processing unit from the ROM 12 and executes it. At this time, the CPU 11 reads data from the nonvolatile memory 14 as necessary and reads / writes data from / to the RAM 13.

  The authentication device 300 according to the authentication system according to the first and second embodiments can also be realized by a program operating on a computer.

  FIG. 10 shows an example of the hardware configuration of an authentication device that is a computer that executes a program according to the present invention. As a hardware configuration of the authentication device 300, for example, a CPU 31, a main memory 32, an auxiliary storage device 33, an output interface 34, an input interface 35, and an IC card R / W interface 36 are connected by a bus 37.

  The CPU 31 executes a program stored in a main memory 32 described later. As the main memory 32, a RAM is usually used, and a program to be executed and data to be used are read and temporarily stored from an auxiliary storage device 33 to be described later. As the auxiliary storage device 33, a hard disk drive (HDD) is usually used, and programs and data are stored and saved as files. As the auxiliary storage device 33, an external storage medium such as a CD-ROM, a DVD-ROM, or a USB memory can be used.

  A monitor 38 as a display device is connected to the output interface 34 as one of output devices. Program execution results are output and displayed on the monitor. A keyboard 39 and a mouse 40 are connected to the input interface 35 as input devices, and data is input from these input devices. An IC card reader / writer 41 is connected to the IC card R / W interface 36, and data is input / output to / from the IC card that has accessed the IC card reader / writer 41.

  The above hardware configuration is associated with the functional blocks in FIG. 1 as follows. Programs (random number generation unit 305, authentication processing unit 307, encryption processing unit 309, communication unit 311, and control unit 313) for causing the computer to function as the authentication device 300, and data (storage unit 301, public information storage unit 303) ) In advance in the auxiliary storage device 13. When the program is activated, the program and data are first read into the main memory 12, and then the main memory 12 and the CPU 11 cooperate to execute the program. Various data stored in the storage unit 301 during program execution is normally written in the main memory, but may be saved in the auxiliary storage device 13 as necessary.

Finally, the ellipse ElGamal cryptography, which is a premise of the above embodiment, and the non- degenerate pairing function on the elliptic curve will be described in detail using mathematical expressions.

  Hereinafter, the elliptic El Gamal encryption will be described. First, public key cryptography, which is one of the encryption schemes, will be described, then elliptic curve cryptography, which is one scheme of public key cryptography, will be described, and finally, elliptical Elgamal cryptography, which is one scheme of elliptic curve cryptography, will be described.

  First, public key cryptography, which is one of cryptosystems, will be described. The encryption method includes a public key encryption method and a common key encryption method. The public key cryptosystem is a scheme that uses different keys for encryption and decryption. In a typical public key cryptosystem, a key (public key) for encryption is made public in advance and can be obtained by anyone. The receiver encrypts the plain text using the public key of the other party who wants to transmit, creates a cipher text, and transmits the cipher text to the receiver. A key (secret key) for decrypting the ciphertext is held as secret information known only to the receiver, and the receiver obtains a plaintext by decrypting the ciphertext using the secret key.

Next, elliptic curve cryptography, which is one method of public key cryptography, will be described. Elliptic curve cryptography is one of the cryptosystems included in the public key cryptosystem, and is a general term for cryptosystems that use arithmetic rules between points of an algebraic curve called an elliptic curve. prime number p, when the a natural number m, and the finite field GF (p ^m) on the elliptic curve number of elements p ^m, the equation _{^{E: y 2 + a 1 xy}} + a 3 y = x 3 + a 2 x 2 + a 4 x + a 6
This is a set obtained by adding a point ∞ called a point at infinity to a set of points (x, y) satisfying. The infinity point ∞ may be expressed as 0. Here, the constants a ₁ , a ₂ , a ₃ , a ₄ , and a ₆ and the variables x and y are elements of a finite field GF (p ^m ). A point on an elliptic curve can be expressed in a coordinate format (affine coordinates) such as (x, y), but an infinite point ∞ is the only point that cannot be expressed in a coordinate format such as (x, y). Thus, in the affine coordinates, points other than the infinity point are expressed as a set of two elements of GF (p ^m ).

Between points on the elliptic curve E, an operation called point addition can be defined as follows. Let P be a point on the elliptic curve E on GF (p ^m ). First, the inverse element P of P is defined as follows.

(1) When P = ∞, -P = ∞
(2) When P ≠ ∞ (P = (x, y)), −P = (x, −y)
Next, let P1 and P2 be two points on the elliptic curve E on GF (p ^m ). At this time, the sum P1 + P2 of P1 and P2 is defined as follows.

(1) When P1 = ∞, P1 + P2 = P2
(2) When P2 = ∞, P1 + P2 = P1
(3) When P1 = −P2, P1 + P2 = ∞
(4) When other than the above _{(P1 = (x 1, y} 1), P2 = (x 2, y 2), P1 + P2 = a _{(x 3, y 3))} ,
x ₃ = λ ² + a ₁ λ−a ₂ −x ₁ −x ₂
y ₃ = (x ₁ −x ₂ ) λ−y ₁ −a ₁ x ₃ −a ₃
However,
λ = (y ₂ −y ₁ ) / (x ₂ −x ₁ ) (when P1 ≠ P2)
λ = (3 × ₁ ² + 2a ₂ x ₁ + a ₄ −a ₁ y ₁ ) / (2y ₁ + a ₁ x ₁ + a ₃ ) (when P1 = P2)
Note that calculating P1 + P2 when P1 ≠ P2 is called ellipse addition. Also, calculating P1 + P2 = [2] P1 when P1 = P2 is called elliptic doubling.

  When an elliptic curve and a point P thereon are given, calculating a point [d] P = P + P +... + P (d sums) for a certain integer d is called scalar multiplication. Here, the point P is called a base point. Contrary to scalar multiplication, when an elliptic curve, a point P above it and a scalar multiple [d] P are given, the problem of finding an integer (scalar) d is called an elliptic curve discrete logarithm problem. The security of elliptic curve cryptography depends on the fact that scalar multiplication can be easily calculated, but it is difficult to solve the elliptic curve discrete logarithm problem. For this reason, many elliptic curve cryptosystems often use the scalar d as a secret key and the scalar multiple [d] P as part of the public key.

Next, the elliptic El Gamal encryption, which is one method of elliptic curve encryption, will be described. It is assumed that the finite field GF (p ^m ), the elliptic curve E thereon, and the base point P are defined in advance and disclosed. At this time, if the recipient's private key in the elliptic El Gamal encryption is d, the public key can be defined as Y = [d] P.

The sender who wants to send the message M secretly to the recipient performs the following encryption process using the recipient's public key Y. Here, it is assumed that the message M is represented as a point of the elliptic curve E.
(1) Generate a (pseudo) random number r.
(2) The point [r] P is calculated and held as C1.
(3) The point M + [r] Y is calculated and held as C2.
(4) The pair (C1, C2) of C1 and C2 is sent to the receiver as ciphertext.
The recipient who has received the ciphertext (C1, C2) performs the following decryption process using his / her private key d.
(1) The point [d] C1 is calculated and held as K.
(2) The point C2-K is calculated and set as a decrypted message.

  The security of the elliptic Elgamal cryptosystem will be described. The goal of an attacker who intercepted the above communication is to decrypt the plaintext. First, the attacker can obtain the public key Y (= [d] P), but calculating the secret key d from this is difficult because it means solving the elliptic curve discrete logarithm problem. Next, the attacker can obtain (intercept) the ciphertext C = (C1, C2) = ([r] P, M + [r] Y). It is difficult to calculate the random number r from C1 and the base point P because it means solving the elliptic curve discrete logarithm problem. Therefore, since the random number r cannot be obtained for the attacker, it is difficult to calculate the scalar multiple [r] Y = [r] [d] P, and it is also difficult to obtain the plaintext M from C2. As described above, it is known that it is difficult to decipher the elliptic Elgamal code under the assumption that the elliptic curve discrete logarithm problem is difficult.

Hereinafter, a non- degenerate pairing function on an elliptic curve will be described. Pairing function on the elliptic curve is divided into two but a reduction retreat as a non-degenerate, nondegenerate pairing function can be configured using a regression of pairing function. Accordingly, it described or which is an embodiment of the withdrawal of pairing function not a Weil Pairing (Weil Pairing), described below how to configure the non-degenerate pairing function using the Weil pairing.

Let E be an elliptic curve over a finite field GF (p ^m ). The pairing e on the elliptic curve is a function that outputs a certain value for two points of the elliptic curve and satisfies the following properties.

(1) e (P1, P2 + P3) = e (P1, P2) · e (P1, P3)
(2) e (P1, P2) = e (P2, P1)
Then, the retraction of pairing, the above conditions (1), in addition to (2), the condition (3) e (P1,
The one that satisfies P1) = 1.

Examples of regression of pairing, Weil pairing is known. Below, we will explain veil pairing. For a natural number n, let E [n] be the total set of points P of the elliptic curve E that satisfy [n] P = ∞. Weil pairing e (P1, P2) for two points P1, P2 of the set E [n] is determined as follows.

e (P1, P2) = f _P1 (U + P2) / f _P1 (U) · f _P2 (V) / f _P2 (V + P1)
Here, f _P1 and f _P2 are functions on E each composed of the points P1 and P2 of the elliptic curve. For example, f _P1 (U) represents a value obtained by substituting the point U on E for f _P1 . The configuration of f _P1 and f _P2 will be described later. Points U and V are arbitrary points on the elliptic curve E where U + P2 and V + P1 do not become the infinity point ∞, respectively.

Next, an algorithm for deriving the function f _P for the point P of the elliptic curve will be described. First, the equation l _{i, j} (x, y) = 0 is defined as follows.

When i ≠ j, a straight line passing through the two points [i] P and [j] P on the elliptic curve E When i = j, a straight line passing through the point [i] P on the elliptic curve E and touching the elliptic curve E The equation v _i (x, y) = 0 is defined as a straight line passing through [i] P and the point infinity ∞. Next, define functions g _n by the following recurrence formula. However, q is a natural number.

g ₁ = 1
g _2q = l _{q, q} (x, y) / v _2q (x, y) · g _q ²
g _{2q + 1} = l _{2q, 1} (x, y) / v _{2q + 1} (x, y) · g _2q
At this time, it is known that f _P = g _p .

Next, non- degenerate pairing will be described. The non-degenerate pairing is defined as the pairing is not a withdrawal of, as an example, can be configured as follows.

Let p be a prime number and satisfy p = 2 mod 3. An elliptic curve E ′ on the finite field GF (p) is assumed to be y ² = x ³ +1. Here, as described above, the elliptic curve can be originally defined in GF (p ^m ), but it should be noted that it is limited to m = 1 here. Let w be the primitive cube root of 1 and map Φ: (x, y)-> (wx, y). At this time, f (x, y) is also a point of the elliptic curve E ′ with respect to the point (x, y) of the elliptic curve E ′.

As described above, the non- degenerate pairing e ′ can be configured by the following formula for the two points P1 and P2 of the elliptic curve.

e ′ (P1, P2) = e (P1, Φ (P2))
Here, e is veil pairing. Note that the map Φ used for the non- degenerate pairing configuration is called a distortion map.

It is a functional block diagram of the authentication system which concerns on the 1st Embodiment of this invention. It is the flowchart which showed the whole process sequence of the authentication system which concerns on the 1st Embodiment of this invention. It is the flowchart which showed the processing procedure of the attack detection of the authentication system which concerns on the 1st Embodiment of this invention. It is the flowchart which showed the process sequence of calculation of e1 in the attack detection process of the authentication system which concerns on the 1st Embodiment of this invention. It is the flowchart which showed the process sequence of calculation of e2 in the attack detection process of the authentication system which concerns on the 1st Embodiment of this invention. It is the flowchart which showed the process sequence of calculation of e (P1, P2) in the attack detection process of the authentication system which concerns on the 1st Embodiment of this invention. In the first attack detection process of the authentication system according to an embodiment of the present invention, it is a flowchart illustrating a processing procedure of calculating the f _S. It is the flowchart which showed the processing procedure of the attack detection of the authentication system which concerns on the 2nd Embodiment of this invention. It is the figure which showed the hardware constitutions of IC card. It is the figure which showed the hardware constitutions of the authentication apparatus which is a computer. It is the flowchart which showed the whole process sequence of the conventional common authentication system.

11 CPU
12 ROM
13 RAM
14 Non-volatile memory 15 Input / output interface 17 Bus 31 CPU
32 Main memory 33 Auxiliary storage device 34 Output interface 35 Input interface 36 IC card R / W interface 37 Bus 38 Monitor 39 Keyboard 40 Mouse 41 IC card reader / writer 100 IC card 101 Storage unit 103 Public information storage unit 105 Secret key storage unit 107 Attack detection unit 109 Decryption processing unit 111 Control unit 113 Communication unit 300 Authentication device 301 Storage unit 303 Public information storage unit 305 Random number generation unit 307 Authentication processing unit 309 Encryption processing unit 311 Communication unit 313 Control unit

Claims (2)

A secret key storage unit that secretly stores a secret key d of the elliptic ElGamal encryption based on a predetermined elliptic curve;
A public information storage unit storing the public key Y and the base point G of the elliptical Elgamal encryption;
An input unit that receives an input of ciphertext C = (C1, C2) encrypted by the elliptic Elgamal encryption;
Based on the secret key d stored in the storage unit, K = [d] C1 and M ′ = C
In an authentication medium comprising: a decryption unit that generates a decrypted message M ′ by sequentially calculating 2-K; and an output unit that outputs the decrypted message M ′.
A determination unit that determines whether the relational expression e (G, K) = e (Y, C1) is satisfied based on the non- degenerate pairing e;
In addition,
The output medium outputs the decrypted message M ′ when the relational expression is satisfied. The determination unit performs determination based on the relational expression when K calculated by the decoding unit is a point of the predetermined elliptic curve.
The authentication medium according to claim 1.

Images