SHARE GENERATING DEVICE, RECONSTRUCTING DEVICE, SECURE COMPUTATION SYSTEM, SHARE GENERATION METHOD, RECONSTRUCTION METHOD, PROGRAM, AND RECORDING MEDIUM

Abstract
A share [x]i of plaintext x in accordance with Shamir's secret sharing scheme is expressed by N shares [x0]i, . . . , [xN−1]i, and each share generating device Ai obtains a function value ri=Pm(i(−))(si) of a seed si, obtains a first calculated value ζi=λ(i, i(−))[xi(−)]i+ri using a Lagrange coefficient λ(i, i(−)), a share [xi(−)]i, and the function value ri, and outputs the first calculated value ζi to a share generating device Ai(−). Each share generating device Ai accepts a second calculated value ζi(+), obtains a third calculated value zi=λ(i, i(+))[xi]i+ζi(+) using a Lagrange coefficient λ(i, i(+)), a share [xi]i, and the second calculated value ζi(+), and obtains information containing the seed si and the third calculated value zi as a share SSi of the plaintext x in secret sharing and outputs the share SSi.
Description
TECHNICAL FIELD

The present invention relates to cryptographic technologies and, in particular, relates to a secure computation technique.


BACKGROUND ART

One of secret sharing schemes is Shamir's secret sharing scheme (see, for example, Non-patent Literature 1 and so forth).


PRIOR ART LITERATURE
Non-Patent Literature

Non-patent Literature 1: A. Shamir, “How to Share a Secret”, Communications of the ACM, November 1979, Volume 22, Number 11, pp. 612-613.


SUMMARY OF THE INVENTION
Problems to be Solved by the Invention

Secure computation devices can perform secure computation using shares obtained in accordance with Shamir's secret sharing scheme without changing them. However, when shares are distributed among N secure computation devices in accordance with the secret sharing scheme and secure computation is performed, the total amount of data of shares, which indicate the secure computation result, is N times the amount of data of plaintext. Thus, if these shares are transmitted without being changed, the total amount of communication data is also N times the amount of data of the plaintext.


An object of the present invention is to provide a technique for generating shares whose total amount of communication data is small from shares in accordance with Shamir's secret sharing scheme.


Means to Solve the Problems

Each share generating device Ai included in N share generating devices A0, . . . , AN−1 executes the following. It is to be noted that N is an integer greater than or equal to 2, in is an integer greater than or equal to 1, m(i) is an integer greater than or equal to 0 and less than m, i=0, . . . , N−1 holds, j=0, . . . , N−1 holds, i(+)=i+1 mod N holds, i(−)=i−1 mod N holds, Pm(i) is a function, and the range of the function Pm(i) belongs to a set Fm(i) whose members are sequences of elements of m(i) fields F. The sum of a value which is obtained by multiplying a share [α]i in accordance with Shamir's secret sharing scheme by a Lagrange coefficient λ(i, i(−)) and a value which is obtained by multiplying a share [α]i(−) in accordance with Shamir's secret sharing scheme by a Lagrange coefficient λ(i(−), i) is a reconstructed value α. The sum of a value which is obtained by multiplying the share [α]i by a Lagrange coefficient λ(i, i(+)) and a value which is obtained by multiplying a share [α]i(+) in accordance with Shamir's secret sharing scheme by a Lagrange coefficient λ(i(+), i) is also the reconstructed value α.


A share [x]i∈Fm of plaintext x in accordance with Shamir's secret sharing scheme is expressed by N shares [x0]i, . . . , [xN−1]i, which satisfy m=m(0)+ . . . +m(N−1) and [xi]i∈Fm(j), by secure computation. Each share generating device Ai obtains a function value ri=Pm(i(−))(si)∈Fm(i(−)) of a seed si, obtains a first calculated value ζi=λ(i, i(−))[xi(−)]i+ri∈Fm(i(−)) using the Lagrange coefficient λ(i, i(−)), a share [xi(−)]i, and the function value ri, and outputs the first calculated value ζi to a share generating device Ai(−). Each share generating device Ai accepts a second calculated value ζi(+)∈Fm(i), obtains a third calculated value zi=λ(i, i(+))[xi]ii(+)∈Fm(i) using the Lagrange coefficient λ(i, i(+)), a share [xi]i, and the second calculated value ζi(+), and obtains information containing the seed si and the third calculated value zi as a share SSi of the plaintext x in secret sharing and outputs the share SSi.


Effects of the Invention

In the present invention, it is possible to generate shares whose total amount of communication data is small from shares in accordance with Shamir's secret sharing scheme.





BRIEF DESCRIPTION OF THE DRAWINGS


FIG. 1 is a block diagram illustrating the configuration of a secure computation system of an embodiment.



FIG. 2 is a block diagram illustrating the functional configuration of a share generating device of the embodiment.



FIG. 3 is a block diagram illustrating the functional configuration of a reconstructing device of the embodiment.



FIG. 4 is a flow diagram for explaining a share generation method of the embodiment.



FIG. 5 is a flow diagram for explaining a reconstruction method of the embodiment.





DETAILED DESCRIPTION OF THE EMBODIMENTS

Hereinafter, embodiments of the present invention will be described with reference to the drawings.


First Embodiment

First, a first embodiment will be described.


<Configuration>


As illustrated in FIG. 1, a secure computation system 1 of the present embodiment includes N share generating devices 11-A0, . . . , 11-AN−1 and a reconstructing device 12, and these devices can communicate with one another through a network such as the Internet. It is to be noted that N is an integer greater than or equal to 2. For example, N is an integer greater than or equal to 3. One example of N is N=3. Although FIG. 1 illustrates one reconstructing device 12, the secure computation system 1 may include a plurality of reconstructing devices.


As illustrated in FIG. 2, each share generating device 11-Ai (i=0, . . . , N−1) of the present embodiment includes a division unit 111-Ai, a seed obtaining unit 112-Ai, a function operation unit 113-Ai, secure computation units 114-Ai and 115-Ai, a share generation unit 116-Ai, a storage 117-Ai, a communication unit 118-Ai, and a control unit 119-Ai and executes each processing under control of the control unit 119-Ai. Information output from each processing unit (a part that performs processing) of the share generating device 11-Ai is stored in the storage 117-Ai and is read therefrom when necessary and input to another processing unit.


As illustrated in FIG. 3, the reconstructing device 12 of the present embodiment includes a communication unit 121, an arithmetic unit 122, a concatenation unit 123, a storage 127, and a control unit 128 and executes each processing under control of the control unit 128. Information output from each processing unit of the reconstructing device 12 is stored in the storage 127 and is read therefrom when necessary and input to another processing unit.


<Share Generation Method>


A share generation method (a share conversion method) which is performed by each share generating device 11-Ai (i=0, . . . , N−1) (FIG. 2) of the present embodiment will be described using FIG. 4. It is assumed that a share [x]i∈Fm of plaintext x∈Fm in accordance with Shamir's secret sharing scheme is stored in the storage 117-Ai of each share generating device 11-Ai. A “share” means a fragment in secret sharing. Shamir's secret sharing scheme of the embodiment is a 2-out-of-N threshold secret sharing scheme, in which, given any two different shares, a reconstructed value (plaintext) can be reconstructed; however, given any one piece of share information, information on the reconstructed value cannot be obtained at all. The sum of a value which is obtained by multiplying a share [α]i in accordance with Shamir's secret sharing scheme by a Lagrange coefficient λ(i, i(−)) and a value which is obtained by multiplying a share [α]i(−) in accordance with Shamir's secret sharing scheme by a Lagrange coefficient λ(i(−), i) is a reconstructed value α, and the sum of a value which is obtained by multiplying the share [α]i by a Lagrange coefficient λ(i, i(+)) and a value which is obtained by multiplying a share [α]i(+) in accordance with Shamir's secret sharing scheme by a Lagrange coefficient λ(i(+), i) is also the reconstructed value α. That is,





α=λ(i,i(−))[α]i+λ(i(−),i)[α]i(−)





α=λ(i,i(+))[α]i+λ(i(+),i)[α]i(+)


are satisfied. In other words, λ(i, i(−)) is a coefficient by which the share [α]i is multiplied when the reconstructed value α is reconstructed from the share [α]i and the share [α]i(−) and λ(i(−), i) is a coefficient by which the share [α]i(−) is multiplied at the time of this reconstruction. λ(i, i(+)) is a coefficient by which the share [α]i is multiplied when the reconstructed value α is reconstructed from the share [α]i and the share [α]i+) and λ(i(+), i) is a coefficient by which the share [α]i(+) is multiplied at the time of this reconstruction. It is to be noted that i(+)=i+1 mod N and i(−)=i−1 mod N hold. These Lagrange coefficients λ(i, i(−)), (i(−), i), λ(i, i(+)), and λ(i(+), i) can be easily determined by publicly known Lagrange's interpolation formula. Fm means a set whose members are sequences of elements of in fields F. One example of the set Fm is in-dimensional vectors, whose members are elements of in fields F. α∈β means that a is a member of β. m is an integer greater than or equal to 1. For instance, m is an integer greater than or equal to 2 or an integer greater than or equal to 3. An example of the field F is a set of remainders modulo a prime number p (α mod p, where α is any number), and the operation result in the field F in this case is obtained as a remainder modulo a prime number p. p≥3 holds and, for instance, p=261−1 holds. Moreover, the share [x]i may be a share transmitted from another device or a share obtained by performing secure computation on a share transmitted from another device.


The division unit 111-Ai (a first division unit) reads the share [x]i∈Fm of the plaintext x in accordance with Shamir's secret sharing scheme described above from the storage 117-Ai. The division unit 111-Ai divides the share [x]i into N shares [x0]i, . . . , [xN−1]i by secure computation and outputs the shares [x0]i, . . . , [xN−1]i. It is to be noted that m(i) is an integer greater than or equal to 0 and less than in (for example, m(i) is an integer greater than or equal to 1), in =m(0)+ . . . +m(N−1) is satisfied, and, for i=0, . . . , N−1 and j=0, . . . , N−1, [xi]i∈Fm(j) is satisfied. It is also possible to make m(0)= . . . =m(N−1)=m/N hold if m is a multiple of N. However, irrespective of whether or not in is a multiple of N, all of m(0), . . . , m(N−1) may not be identical with one another. For example, at least part of m(0), . . . , m(N−1) may be 0. It is to be noted that γ∈F0 represents a null value. If m(i)=0, [xi]i∈Fm(i) is a null value. [xi]i is a share of plaintext xj in accordance with Shamir's secret sharing scheme described above. The plaintext x is expressed by N x0∈Fm(0), . . . , xN−1∈Fm(N−1). For example, the plaintext x is expressed as a sequence x0| . . . |xN−1 obtained by arranging N x0∈Fm(0) . . . , xN−1∈Fm(N−1). To divide the share [x]i into the N shares [x0]i, . . . , [xN−1]i by secure computation, it is only necessary to, for instance, separate and divide the share [x]i directly into the N shares [x0]i∈Fm(0), . . . , [xN−1]i∈Fm(N−1). It is to be noted that, if m=1, only one of m(0), . . . , m(N−1) is 1 and the others are 0. In this case, the division unit 111-Ai does not have to divide the share [x]i and only has to output any one of the shares [x0]i, . . . , [xN−1]i as [x]i and output all of the other shares as null values. As described above, the share [x]i is expressed by the N shares [x0]i, . . . , [xN−1]i. For instance, the share [x]i is expressed as a sequence [x0]i| . . . |[xN−1]i obtained by arranging the N shares [x0]i, . . . , [xN−1]i (Step S111-Ai).


The seed obtaining unit 112-Ai obtains a seed si and outputs the seed si. There is no limitation on the data format of the seed si and any value can be used as the seed si. One example of the seed si is an element of a set Fw(i) whose members are sequences of elements of w(i) fields F (si∈Fw(i)). It is to be noted that w(i) is an integer greater than or equal to 1 and less than m. The seed si (i=0, . . . , N−1) may be an arbitrary value or an output value obtained by other processing. The “arbitrary value” may be a random number (a pseudo random number or a true random number), a value selected from a plurality of preset values, or a value obtained by other processing. It is to be noted that the seed si corresponding to the share [xj]i which is a null value may be set at a null value (Step S112-Ai).


The function operation unit 113-Ai obtains a function value ri=Pm(i(−))(si)∈Fm(−)) (for example, a pseudo random number) using the seed si as input and outputs the function value ri. It is to be noted that Pm(i(−))(si) is a function value which is obtained by operating a function Pm(i(−)) on the seed si. Pm(i):Fw(i(+))→Fm(i) is a function and the range of a function Pm(i) belongs to a set Fm(i) whose members are sequences of elements of m(i) fields F. One example of the set Fm(i) is a set of m(i)-dimensional vectors, whose members are elements of m(i) fields F. The domain of definition of the function Pm(i) may be any domain of definition. For instance, the domain of definition of the function Pm(i) belongs to a set Fw(i(+)) whose members are sequences of elements of w(i(+)) fields F. Preferably, w(i(+))<m(i) is satisfied (the amount of data of the seed si is smaller than the amount of data of the function value ri). An example of Pm(i) is a pseudo random number generating function. That is, Pm(i(−)) is a function and the range of the function Pm(i(−)) belongs to a set Fm(i(−)) whose members are sequences of elements of m(i(−)) fields F. For example, the domain of definition of the function Pm(i(−)) belongs to the set Fw(i) whose members are sequences of elements of w(i) fields F. Preferably, w(i)<m(i(−)) is satisfied. It is to be noted that the function value ri corresponding to the share [xj]i which is a null value may be set at a null value (Step S113-Ai).


The secure computation unit 114-Ai (a first secure computation unit) obtains, using a share [xi(−)]i and the function value ri as input, a calculated value (a first calculated value) ζi=λ(i, i(−))[xi(−)]i+ri∈Fm(i(−)) using the Lagrange coefficient λ(i, i(−)), the share [xi(−)]i, and the function value ri and outputs the calculated value ζi. It is to be noted that ζi=λ(i, i(−))[xi(−)]i+ri corresponding to the share [xi(−)]i which is a null value is set at a null value (Step S114-Ai).


The calculated value ζi is input to the communication unit 118-Ai. The communication unit 118-Ai (an output unit) outputs the calculated value ζi to a share generating device 11-Ai(−). The output calculated value ζi is transmitted to the share generating device 11-Ai(−) through the network (Step S1181-Ai).


A calculated value (a second calculated value) ζi(+)=λ(i(+), i)[xi]i(+)+ri(+)∈Fm(i) (ri(+)=Pm(i)(si(+))) output from a share generating device 11-Ai(+) is transmitted to the share generating device 11-Ai through the network in a similar manner. It is to be noted that ζi(+)=λ(i(+), i)[xi]i(+)+ri(+) corresponding to the share [xi]i(+) which is a null value is set at a null value. The calculated value ζi(+) is received (accepted) by the communication unit 118-Ai (an input unit) and output (Step S1182-Ai).


The secure computation unit 115-Ai (a second secure computation unit) obtains, using a share [xi]i and the calculated value i(+) as input, a calculated value (a third calculated value) zi=λ(i, i(+))[xi]ii(+)∈Fm(i) using the Lagrange coefficient λ(i, i(+)), the share [xi]i, and the calculated value ζi(+) and outputs the calculated value zi. Since ζi(+)=λ(i(+), i)[xi]i(+)+ri(+), zi=λ(i, i(+))[xi]i+λ(i(+), i)[xi]i(+)+ri(+) holds. Moreover, since xi=λ(i, i(+))[xi]i+λ(i(+), i)[xi]i(+), zi=xi+ri(+) holds. It is to be noted that zi corresponding to the share [xi]; which is a null value is set at a null value (Step S115-Ai).


The seed si output from the seed obtaining unit 112-Ai and the calculated value zi output from the secure computation unit 115-Ai are input to the share generation unit 116-Ai. The share generation unit 116-Ai obtains information containing the seed si and the calculated value zi as a share SSi of the plaintext x in secret sharing and outputs the share SSi. The share SSi corresponding to the seed si and the calculated value zi which are null values is information in which the seed si and the calculated value zi are set at null values (Step S116-Ai).


The share SSi output from the share generation unit 116-Ai is input to the communication unit 118-Ai. The communication unit 118-Ai outputs the share SSi to the reconstructing device 12. The output share SSi is transmitted to the reconstructing device 12 through the network. The size of the seed si and N do not depend on m. In (2, N)-Shamir's secret sharing, the order of magnitude of the total share size (the total share size which is transmitted) that relates to the data size m of the plaintext x is O(Nm); in the present embodiment, the order of magnitude of the total share size that relates to the data size in of the plaintext x is just O(m). The size of each share is O(m/N). For example, the total amount of data of the share SSi which is transmitted from each share generating device 11-Ai (i=0, . . . , N−1) is less than N times the amount of data of the plaintext x. For instance, the amount of data of each share SSi is smaller than the amount of data of the plaintext x (Step S118-Ai).


<Reconstruction Method>


A reconstruction method which is performed by the reconstructing device 12 (FIG. 3) of the present embodiment will be described using FIG. 5.


N shares SS0, . . . , SSN−1 output from the share generating devices 11-A0, . . . , 11-AN−1 are received (accepted) by the communication unit 121 (an input unit) of the reconstructing device 12 (Step S121).


For each i=0, . . . , N−1, the calculated value zi contained in the share SSi and a seed si(+) contained in a share SSi(+) are input to the arithmetic unit 122. The arithmetic unit 122 obtains a member xi=zi−Pm(i)(si(+)∈Fm(i) using the calculated value zi and the seed si(+) and outputs the member xi. It is to be noted that Pm(i)(si(+)) is a function value which is obtained by operating the function Pm(i) on the seed si(+). The function Pm(i) is the same as the function defined in Step S113-Ai. As indicated in Step S115-Ai, zi=xi+ri(+) holds. Moreover, as indicated in Step S1182-Ai, ri(+)=Pm(i)(si(+)) holds. Thus, the member xi is obtained by an operation zi−Pm(i)(si(+)). It is to be noted that the member xi corresponding to the calculated value zi and the seed si(+) which are null values is set at a null value (Step S122).


The members x0, . . . , xN−1 obtained in the arithmetic unit 122 are input to the concatenation unit 123. The concatenation unit 123 obtains a reconstructed value x obtained by concatenating (joining) the members x0, . . . , xN−1 to one another and outputs the reconstructed value x. For example, a sequence x0| . . . |xN−1 obtained by arranging x0∈Fm(0), . . . , xN−1∈Fm(N−1) is x (Step S123).


Features of the Present Embodiment

Each share generating device 11-Ai (i=0, . . . , N−1) obtains the share SSi containing the seed si and the calculated value zi∈Fm(i) from the share [x]i in accordance with Shamir's secret sharing scheme and outputs the share SSi. Here, the size of the seed si and N do not depend on m. In (2, N)-Shamir's secret sharing, the order of magnitude of the total share size (the total share size which is transmitted) that relates to the data size m of the plaintext x is O(Nm); in the present embodiment, the order of magnitude of the total share size that relates to the data size m of the plaintext x is just O(m). The size of each share is O(m/N). For example, by setting m(i) and w(i) so that m(i)<m and w(i)<m, it is possible to make the total amount of data of the shares SS0, . . . , SSN−1 less than N times the amount of data of the plaintext x∈Fm. This makes it possible to make the total amount of communication data smaller than that of shares in accordance with Shamir's secret sharing scheme. Each share generating device 11-Ai can obtain the share SSi whose amount of data is small from the share [x]i in accordance with Shamir's secret sharing scheme without obtaining information on the plaintext x. The reconstructing device 12 can reconstruct the plaintext x using the N shares SS0, . . . , SSN−1.


Second Embodiment

A second embodiment is a modification of the first embodiment. In the present embodiment, a checksum corresponding to a share is generated at the time of share generation and the share is verified using the checksum at the time of reconstruction of plaintext. In the following description, an explanation of a matter that has already been described in the first embodiment is sometimes simplified, using the same reference character as that of the first embodiment.


<Configuration>


As illustrated in FIG. 1, a secure computation system 2 of the present embodiment includes N share generating devices 21-A0, . . . , 21-AN−1 and a reconstructing device 22, and these devices can communicate with one another through a network such as the Internet. It is to be noted that N of the present embodiment is an integer greater than or equal to 3. One example of N is N=3. Although FIG. 1 illustrates one reconstructing device 22, the secure computation system 2 may include a plurality of reconstructing devices.


As illustrated in FIG. 2, each share generating device 21-Ai (i=0, . . . , N−1) of the present embodiment includes a division unit 111-Ai, a seed obtaining unit 112-Ai, a function operation unit 113-Ai, secure computation units 114-Ai and 115-Ai, a joint possession unit 215-Ai, a share generation unit 216-Ai, a division unit 217-Ai, checksum generation units 218-Ai and 219-Ai, a storage 117-Ai, a communication unit 118-Ai, and a control unit 119-Ai and executes each processing under control of the control unit 119-Ai. Information output from each processing unit (a part that performs processing) of the share generating device 21-Ai is stored in the storage 117-Ai and is read therefrom when necessary and input to another processing unit.


As illustrated in FIG. 3, the reconstructing device 22 of the present embodiment includes a communication unit 221, an arithmetic unit 122, a concatenation unit 123, a reconstruction unit 224, a verification value generation unit 225, a determination unit 226, a division unit 229, a storage 127, and a control unit 128 and executes each processing under control of the control unit 128. Information output from each processing unit of the reconstructing device 22 is stored in the storage 127 and is read therefrom when necessary and input to another processing unit.


<Share Generation Method>


A share generation method which is performed by the share generating device 21-Ai (FIG. 2) will be described using FIG. 4. First, in place of the share generating device 11-Ai, the share generating device 21-Ai executes the processing from Steps S111-Ai to S115-Ai described in the first embodiment. It is to be noted that, in the present embodiment, N and m are integers greater than or equal to 3.


Next, the joint possession unit 215-Ai of each share generating device 21-Ai possesses an arbitrary value ti(−−)∈Fv jointly with a joint possession unit 215-Ai(−−) of a share generating device 21-Ai(−−), possesses an arbitrary value ti∈Fv jointly with a joint possession unit 215-Ai(++) of a share generating device 21-Ai(++), possesses an arbitrary value ui(−−)∈Fv jointly with a joint possession unit 215-Ai(+) of a share generating device 21-Ai(+), and possesses an arbitrary value ui(−)∈Fv jointly with a joint possession unit 215-Ai(−−) of a share generating device 21-Ai(−−). That is, the joint possession unit 215-Ai and the joint possession unit 215-Ai(−−) obtain the same arbitrary value ti(−−), the joint possession unit 215-Ai and the joint possession unit 215-Ai(++) obtain the same arbitrary value ti, the joint possession unit 215-Ai(−) and the joint possession unit 215-Ai(+) obtain the same arbitrary value ui(−−), and the joint possession unit 215-Ai and the joint possession unit 215-Ai(−) obtain the same arbitrary value ui(−). It is to be noted that v is an integer greater than or equal to 1 and, for example, v=1 holds. A greater data amount reduction effect can be achieved if v is less than or equal to in (for instance, v is less than m). i(+)=i+1 mod N, i(−)=i−1 mod N, i(++)=i+2 mod N, and i(−−)=i−2 mod N hold. Joint possession of an arbitrary value γ between a joint possession unit β1 and a joint possession unit β2 may be performed by transmission of the arbitrary value γ or information for identification of the arbitrary value γ from the joint possession unit β1 to the joint possession unit β2 or may be performed by transmission of the arbitrary value γ or information for identification of the arbitrary value γ from the joint possession unit β2 to the joint possession unit β1, or the joint possession unit β1 and the joint possession unit β2 may jointly possess the arbitrary value γ by jointly possessing a common seed value and executing the same processing using the common seed value. The arbitrary values ti(−−), ti, ui(−−), and ui(−−) may be random numbers (pseudo random numbers or true random numbers), values obtained by other processing, or values selected from a plurality of predetermined values. Joint possession of the arbitrary values ti(−−), ti, ui(−−), and ui(−) may be performed when any one of the processing from Steps S111-Ai to S115-Ai is executed, in response to requests from the other joint possession units 215-Ai(−−), 215-Ai(++), 215-Ai(+), and 215-Ai(−), or in response to other events, or may be performed in advance. The joint possession unit 215-Ai outputs the obtained arbitrary values ti(−−), ti, ui(−−), and ui(−) (Step S215-Ai).


The division unit 217-Ai divides, using the share [xj]i (j=0, . . . , N−1) output from the division unit 111-Ai as input, the share [xj]i into m(j) shares [(xj)0]i, . . . , [(xj)m(j)−1]i∈F by secure computation and outputs the shares [(xj)0]i, . . . , [(xj)m(j)−1]i. To divide the share [xj]i into m(j) shares [(xj)0]i, . . . , [(xj)m(j)−1]i∈F by secure computation, it is only necessary to, for instance, separate and divide the share [xj]i directly into m(j) shares [(xj)0]i, . . . , [(xj)m(j)−1]i∈F. It is to be noted that, if [xj]i∈Fm(i) is a null value, [(xj)0]i, . . . , [(xj)m(j)−1]i are also null values (Step S217-Ai).


The arbitrary values ti(−−) and ui(−−) and the shares [(xj)0]i, . . . , [(xj)m(j)−1]i (j=0, . . . , N−1) are input to the checksum generation unit 218-Ai. The checksum generation unit 218-Ai (a first checksum generation unit) obtains checksums ci(−−), i and di(−−), i in the following manner, using the arbitrary values ti(−−) and ui(−−) and the shares [(xj)0]i, . . . , [(xj)m(j)−1]i, and outputs the checksums ci(−−), i and di(−−), i.






c
i(−−),i==Σ0≤j<m′(i(−−)){ti(−−)j+1[(x′i(−−))j]i}+ti(−−)m′(i(−−))[x′i(−−))m′(i(−−))−1]i∈Fv






d
i(−−),i==Σ0≤j<m′(i(−−)){ui(−−)j+1[(x′i(−−))j]i}+ui(−−)m′(i(−−))[x′i(−−))m′(i(−−))−1]i∈Fv


It is to be noted that δ=0, . . . , N−1 holds, ceil is a ceiling function, and m′(i) is ceil(m(i)/v). [(x′δ)j]i is ([(xδ)vj]i, [(xδ)vj+1]i, . . . , [(xδ)v(j+1)−1)]i)∈Fv belonging to a set Fv. Moreover, for j=m′(i)−1, if the number of [(xδ)vj]i, [(xδ)vj+1]i, . . . , [(xδ)v(j+1)−1]i is less than v, it is assumed that [(x′δ)(m′(i)−1)]i=([(xδ)v(m′(i)−1)]i, [(xδ)v(m′(i)−1)+1]i, . . . , [(xδ)m(i)−1]i, 0, . . . , 0)∈Fv. The checksums ci(−−), i and di(−−), i may be calculated on an extension field of the order v over basic field F. Σ0≤j<m′(i(−−))j} represents β0+ . . . +βm′(i(−−)γ and βi(−−)γ represents (βi(−−))γ (Steps S2181-Ai and S2182-Ai).


The arbitrary values ti and ui(−) and the shares [(xj)0]i, . . . , [(xj)m(j)−1]i (j=0, . . . , N−1) are input to the checksum generation unit 219-Ai. The checksum generation unit 219-Ai (a second checksum generation unit) obtains checksums ci, i and di(−), i in the following manner, using the arbitrary values ti and ui(−) and the shares [(xj)0]i, . . . , [(xj)m(j)−1]i, and outputs the checksums ci, i and di(−), i.






c
i,i0≤j<m′(i){tij+1[(x′i)j]i}+tim′(i)+1[(x′i)m′(i)−1]i∈Fv






d
i,i0≤j<m′(i){uij+1[(x′i)j]i}+uim′(i)+1[(x′i)m′(i)−1]i∈Fv


It is to be noted that βiγ represents (βi)γ and βi(−)γ represents (βi(−))γ. The checksums ci, i and di(−), i may be calculated on an extension field of the order v over basic field F (Steps 2191-Ai and S2192-Ai).


The seed si output from the seed obtaining unit 112-Ai, the calculated value zi output from the secure computation unit 115-Ai, the arbitrary values ti(−−), ti, ui(−−), and ui(−) output from the joint possession unit 215-Ai, the checksums ci(−−), and di(−−), output from the checksum generation unit 218-Ai, and the checksums ci, i and di(−), i output from the checksum generation unit 219-Ai are input to the share generation unit 216-Aj. The share generation unit 216-Ai obtains information containing the seed si, the calculated value zi, the arbitrary values ti(−−), ti, ui(−−), and ui(−), and the checksums ci(−−), i, di(−−), i, ci, i, and di(−), i as a share SSi of the plaintext x in secret sharing and outputs the share SSi (Step S216-Ai).


The share SSi output from the share generation unit 216-Ai is input to the communication unit 118-Ai. The communication unit 118-Ai outputs the share SSi to the reconstructing device 22. The output share SSi is transmitted to the reconstructing device 22 through the network. The size of the seed si, N, and v do not depend on in. In (2, N)-Shamir's secret sharing, the order of magnitude of the total share size (the total share size which is transmitted) that relates to the data size in of the plaintext x is O(Nm); in the present embodiment, the order of magnitude of the total share size that relates to the data size m of the plaintext x is just O(m). The size of each share is O(m/N). For example, the total amount of data of the share SSi which is transmitted from each share generating device 21-Ai (i=0, . . . , N−1) is less than N times the amount of data of the plaintext x. For instance, the amount of data of each share SSi is smaller than the amount of data of the plaintext x (Step S218-Ai).


<Reconstruction Method>


A reconstruction method which is performed by the reconstructing device 22 (FIG. 3) of the present embodiment will be described using FIG. 5. N shares SS0, . . . , SSN−1 output from the share generating devices 21-A0, . . . , 21-AN−1 are received (accepted) by the communication unit 221 (an input unit) of the reconstructing device 22 and output (Step S221).


Next, as described in the first embodiment, the arithmetic unit 122 obtains a member xi=zi−Pm(i)(si(+))∈Fm(i) using the calculated value zi contained in the share SSi and a seed si(+) contained in a share SSi(+) and outputs the member xi (Step S122).


The checksums ci(−−), i, di(−−), i, ci, i, and di(−−), i contained in each share SSi (i=0, . . . , N−1) are input to the reconstruction unit 224. The reconstruction unit 224 reconstructs a checksum ci0≤j<m′(i){tij+1(x′i)j}+tim′(i)+1(x′i)m′(i)−1 using the checksum ci, i contained in the share SSi and a checksum ci, i(++) contained in a share SSi(++) as shares in accordance with Shamir's secret sharing scheme (the 2-out-of-N threshold secret sharing scheme) described above and outputs the checksum ci. Likewise, the reconstruction unit 224 reconstructs a checksum di0≤j<m′(i){uij+1(x′i)j}+uim′(i)+1(x′i)m′(i)−1 using a checksum di, i(+) contained in the share SSi(+) and a checksum di, i(++) contained in the share SSi(++) as shares in accordance with Shamir's secret sharing scheme described above and outputs the checksum di. Here, ci, i(++)0≤j<m′(i){tij+1[(x′i)j]i(++)}+tim(i)+1[(x′i)m′(i)−1]i(++) contained in the share SSi(++) and ci, i0≤j<m′(i){tij+1[(x′i)j]i}+tim(i)+1[(x′i)m′(i)−1]i contained in the share SSi are shares of the checksum ci0≤j<m′(i){tij+1(x′i)j}+tim′(i)+1(x′i)m′(i)−1 in accordance with Shamir's secret sharing scheme described above. Thus, the reconstruction unit 224 can reconstruct the checksum ci from ci, i(++), ci, i, and the Lagrange coefficients. Moreover, di, i(+)0≤j<m′(i){uij+1[(x′i)j]i(+)}+uim(i)+1[(x′i)m′(i)−1]i(+) contained in the share SSi(+) and di, i(++)0≤j<m′(i){uij+1[(x′i)j]i(++)}+uim′(i)+1[(x′i)m(i)−1]i(++) contained in the share SSi(++) are shares of the checksum di0≤j<m′(i){uij+1(x′i)j}+uim′(i)+1(x′i)m′(i)−1 in accordance with Shamir's secret sharing scheme described above. Thus, the reconstruction unit 224 can reconstruct the checksum di from di, i(+), di, i(++), and the Lagrange coefficients (Steps S2241 and S2242).


The member xi output from the arithmetic unit 122 is input to a division unit 229. The division unit 229 divides the member xi∈Fm(i) into m(i) sub-members (xi)0, . . . , (xi)m(i)−1∈F and outputs the m(i) sub-members (xi)0, . . . , (xi)m(i)−1. For example, the member xi is expressed as a sequence (xi)0| . . . |(xi)m(i)−1 obtained by arranging the m(i) sub-members (xi)0, . . . , (xi)m(i)−1, and the division unit 229 divides xi=(xi)0| . . . |(xi)m(i)−1 into the sub-members (xi)0, . . . , (xi)m(i)−1 (Step S229).


The arbitrary value ti contained in the share SSi or SSi(++), the arbitrary value ui contained in the share SSi(+) or SSi(++), and the sub-members (xi)0, . . . , (xi)m(i)−1 output from the division unit 229 are input to the verification value generation unit 225. The verification value generation unit 225 obtains verification values vci and vdi in the following manner, using the arbitrary values ti and ui and the sub-members (xi)0, . . . , (xi)m(i)−1, and outputs the verification values vci and vdi.






vc
i0≤j<m′(i){tij+1(x′i)j}+tim′(i)+1(x′i)m(i)−1∈Fv






vd
i0≤j<m′(i){uij+1(x′i)j}+uim′(i)+1(x′i)m(i)−1∈Fv


It is to be noted that m′(i) is ceil(m(i)/v), (x′i)j is ((xi)vj, (xi)vj+1, . . . , (xi)v(j+1)−1)∈Fv belonging to the set Fv, and βiγ represents (βi)γ (Step S2251, S2252).


The checksums ci and di output from the reconstruction unit 224 and the verification values vci and vdi output from the verification value generation unit 225 are input to the determination unit 226. The determination unit 226 determines whether ci=vci is satisfied for all i=0, . . . , N−1 (Step S2261). When a determination is made that ci≠vci for any i, the control unit 128 makes the processing terminate with an error message (Step S2263). On the other hand, when a determination is made that ci=vci is satisfied for all i=0, . . . , N−1, the determination unit 226 determines whether di=vdi is satisfied for all i=0, . . . , N−1 (Step S2262). When a determination is made that di≠vdi for any i, the control unit 128 makes the processing terminate with an error message (Step S2263). On the other hand, when a determination is made that di=vdi is satisfied for all i=0, . . . , N−1, the members x0, . . . , xN−1 obtained in the arithmetic unit 122 are input to the concatenation unit 123. In this case, as described in the first embodiment, the concatenation unit 123 obtains a reconstructed value x obtained by concatenating the members x0, . . . , xN−1 to one another and outputs the reconstructed value x (Step S123).


Features of the Present Embodiment

Also in the present embodiment, it is possible to make the total amount of communication data smaller than that of shares in accordance with Shamir's secret sharing scheme. Each share generating device 21-Ai can obtain the share SSi whose amount of data is small from the share [x]i in accordance with Shamir's secret sharing scheme without obtaining information on the plaintext x. The reconstructing device 22 can reconstruct the plaintext x using the N shares SS0, . . . , SSN−1.


Furthermore, in the present embodiment, the share SSi contains checksums and the reconstructing device 22 performs verification of the checksums. Thus, the reconstructing device 22 can detect unauthorized processing performed in any share generating device 21-Ai. Moreover, since the reconstructing device 22 verifies two types of checksums, the checksums ci and di, for each i=0, . . . , N−1, it is also possible to detect the generation of an unauthorized checksum in any share generating device 21-Ai.


First Modification of the Second Embodiment

In the second embodiment, when a determination is made in Step S2261 that ci=vci is satisfied, a determination as to whether di=vdi is satisfied is made in Step S2262. However, other determination processing may be performed, as long as processing in Step S123 is executed when ci=vci and di=vdi are satisfied for all i=0, . . . , N−1; otherwise Step S2263 is executed.


Second Modification of the Second Embodiment

In the second embodiment, the reconstructing device 22 verifies two types of checksums, the checksums ci and di. However, the reconstructing device 22 may be configured so as to verify only one of the checksums ci and di. Processing in this case is as follows.


<Share Generation Method>


After the processing from Steps S111-Ai to S115-Ai is executed, in place of the processing in Step S215-Ai, (Processing I) the joint possession unit 215-Ai of each share generating device 21-Ai possesses the arbitrary value ti(−−)∈F jointly with the joint possession unit 215-Ai(−−) of the share generating device 21-Ai(−−) and possesses the arbitrary value ti∈F jointly with the joint possession unit 215-Ai(++) of the share generating device 21-Ai(++) or (Processing 11) the joint possession unit 215-Ai of each share generating device 21-Ai possesses the arbitrary value ui−(−)∈F jointly with the joint possession unit 215-Ai(+) of the share generating device 21-Ai(+) and possesses the arbitrary value ui(−)∈F jointly with the joint possession unit 215-Ai(−) of the share generating device 21-Ai(−). Only Processing I is executed for all i=0, . . . , N−1 or only Processing II is executed for all i=0, . . . , N−1.


Next, the division unit 217-Ai executes the processing in Step S217-Ai.


Then, the following processing is executed depending on Processing I or Processing II. When joint possession of the arbitrary value ti(−−) has been performed by the joint possession unit 215-Ai (Processing I), the checksum generation unit 218-Ai performs Step S2181-Ai described above, and obtains the checksum ci(−−), i0≤j<m′i(−−){ti(−−)j+1[(x′i(−−))j]i}+ti(−−)m′(i(−−)+1[(x′i(−−))−1]i∈Fv and outputs the checksum ci(−−), i. In this case, the processing in Step S2182-Ai described above is not executed.


On the other hand, when joint possession of the arbitrary value ui(−−) has been performed by the joint possession unit 215-Ai (Processing II), the checksum generation unit 218-Ai performs Step S2182-Ai described above, and obtains the checksum di(−−), i0≤j<m′(i(−−){ui(−−)j+1[(x′j(−−))j]i}+ui(−−)m′(i(−−)+1[(x′i(−−))m′(i(−−))−1]i∈Fv and outputs the checksum di(−−), i. In this case, the processing in Step S2181-Ai described above is not executed.


Next, when joint possession of the arbitrary value ti has been performed by the joint possession unit 215-Ai (Processing I), the checksum generation unit 219-Ai performs Step S2191-Ai described above, and obtains the checksum ci, i0≤j<m′(i){tij+1[(x′i)j]i}+tim′(i)+1[(x′i)m′(i)−1]i∈Fv and outputs the checksum ci, i. In this case, the processing in Step S2192-Ai described above is not executed.


On the other hand, when joint possession of the arbitrary value ui(−) has been performed by the joint possession unit 215-Ai (Processing II), the checksum generation unit 219-Ai performs Step S2192-Ai described above, and obtains the checksum di(−), i0≤j<m′(i){ui(−)j+1[(x′i(−))j]i}+ui(−)m′(i(−))+1[(x′i(−))m′(i(−))−1]i∈Fv and outputs the checksum di(−), i. In this case, the processing in Step S2191-Ai described above is not executed.


Then, in place of the processing in Step S216-Ai, the following processing is executed depending on Processing I or Processing II. When joint possession of the arbitrary value ti(−−) and the arbitrary value ti has been performed by the joint possession unit 215-Ai (Processing I), the share generation unit 216-Ai obtains information containing the seed si, the calculated value zi, the arbitrary value ti(−−), the arbitrary value ti, the checksum ci(−−), i, and the checksum ci, i as the share SSi and outputs the share SSi. On the other hand, when joint possession of the arbitrary value ui(−−) and the arbitrary value ui(−) has been performed by the joint possession unit 215-Ai (Processing II), the share generation unit 216-Ai obtains information containing the seed si, the calculated value zi, the arbitrary value ui(−−), the arbitrary value ui(−), the checksum di(−−), i, and the checksum di(−), i as the share SSi and outputs the share SSi. Then, the processing in Step S218-Ai is executed.


<Reconstruction Method>


After the processing in Steps S221 and S122 is executed, the reconstruction unit 224 performs the following processing. When the share SSi contains the checksum ci, i and the share SSi(++) contains the checksum ci, i(++) (Processing 1), the reconstruction unit 224 executes Step S2241, and reconstructs the checksum ci using the checksum ci, i and the checksum ci, i(++) as shares in accordance with Shamir's secret sharing scheme and outputs the checksum ci. In this case, the processing in Step S2242 described above is not executed.


On the other hand, when the share SSi(+) contains the checksum di, i(+) and the share SSi(++) contains the checksum di, i(++) (Processing II), the reconstruction unit 224 executes Step S2242, and reconstructs the checksum di using the checksum di, i(+) and the checksum di, i(++) as shares in accordance with Shamir's secret sharing scheme and outputs the checksum di. In this case, the processing in Step S2241 described above is not executed.


Next, the division unit 229 executes the processing in Step S229.


Then, the following processing is executed depending on Processing I or Processing II. When the checksum ci has been reconstructed (Processing I), the verification value generation unit 225 executes the processing in Step S2251, and obtains the verification value vci0≤j<m′(i){tij+1(x′i)j}+tim′(i)+1(x′i)m′(i)−1∈Fv and outputs the verification value vci. In this case, the processing in Step S2252 described above is not executed. On the other hand, when the checksum di has been reconstructed (Processing II), the verification value generation unit 225 executes the processing in Step S2252, and obtains the verification value vdi0≤j<m′(i){uij+1(x′i)j}+uim′(i)+1(x′i)m′(i)−1∈Fv and outputs the verification value vdi. In this case, the processing in Step S2251 described above is not executed.


Then, the following processing is executed depending on Processing I or Processing II. If the checksum ci has been reconstructed (Processing I), the determination unit 226 determines in Step S2261 whether ci=vci is satisfied for all i=0, . . . , N−1. The processing in Step S2262 is not executed in this case. If a determination is made that ci≠vci for any i, the control unit 128 makes the processing terminate with an error message (Step S2263). On the other hand, if a determination is made that ci=vci is satisfied for all i=0, . . . , N−1, the procedure proceeds to Step S123. On the other hand, if the checksum di has been reconstructed (Processing II), the determination unit 226 determines in Step S2262 whether di=vdi is satisfied for all i=0, . . . , N−1. The processing in Step S2261 is not executed in this case. If a determination is made that di≠vdi for any i, the control unit 128 makes the processing terminate with an error message (Step S2263). On the other hand, if a determination is made that di=vdi is satisfied for all i=0, . . . , N−1, the procedure proceeds to Step S123.


[Other Modifications and so Forth]


It is to be noted that the present invention is not limited to the foregoing embodiments. For example, the above-described various kinds of processing may be executed, in addition to being executed in chronological order in accordance with the descriptions, in parallel or individually depending on the processing power of a device that executes the processing or when necessary. In addition, it goes without saying that changes may be made as appropriate without departing from the spirit of the present invention.


The above-described each device is embodied by execution of a predetermined program by a general- or special-purpose computer having a processor (hardware processor) such as a central processing unit (CPU), memories such as random-access memory (RAM) and read-only memory (ROM), and the like, for example. The computer may have one processor and one memory or have multiple processors and memories. The program may be installed on the computer or pre-recorded on the ROM and the like. Also, some or all of the processing units may be embodied using an electronic circuit that implements processing functions without using programs, rather than an electronic circuit (circuitry) that implements functional components by loading of programs like a CPU. An electronic circuit constituting a single device may include multiple CPUs.


When the above-described configurations are implemented by a computer, the processing details of the functions supposed to be provided in each device are described by a program. As a result of this program being executed by the computer, the above-described processing functions are implemented on the computer. The program describing the processing details can be recorded on a computer-readable recording medium. An example of the computer-readable recording medium is a non-transitory recording medium. Examples of such a recording medium include a magnetic recording device, an optical disk, a magneto-optical recording medium, and semiconductor memory.


The distribution of this program is performed by, for example, selling, transferring, or lending a portable recording medium such as a DVD or a CD-ROM on which the program is recorded. Furthermore, a configuration may be adopted in which this program is distributed by storing the program in a storage device of a server computer and transferring the program to other computers from the server computer via a network.


The computer that executes such a program first, for example, temporarily stores the program recorded on the portable recording medium or the program transferred from the server computer in a storage device thereof. At the time of execution of processing, the computer reads the program stored in the storage device thereof and executes the processing in accordance with the read program. As another mode of execution of this program, the computer may read the program directly from the portable recording medium and execute the processing in accordance with the program and, furthermore, every time the program is transferred to the computer from the server computer, the computer may sequentially execute the processing in accordance with the received program. A configuration may be adopted in which the transfer of a program to the computer from the server computer is not performed and the above-described processing is executed by so-called application service provider (ASP)-type service by which the processing functions are implemented only by an instruction for execution thereof and result acquisition.


Instead of executing a predetermined program on the computer to implement the processing functions of the present devices, at least some of the processing functions may be implemented by hardware.


DESCRIPTION OF REFERENCE NUMERALS






    • 1 secure computation system


    • 11-Ai, 21-Ai share generating device


    • 12, 22 reconstructing device




Claims
  • 1: A share generating device that is a share generating device Ai included in N share generating devices A0, . . . , AN−1, wherein N is an integer greater than or equal to 2, m is an integer greater than or equal to 1, m(i) is an integer greater than or equal to 0 and less than m, i=0, . . . , N−1 holds, j=0, . . . , N−1 holds, i(+)=i+1 mod N holds, i(−)=i−1 mod N holds, Pm(i) is a function, and a range of the function Pm(i) belongs to a set Fm(i) whose members are sequences of elements of m(i) fields F,a sum of a value which is obtained by multiplying a share [α]i in accordance with Shamir's secret sharing scheme by a Lagrange coefficient λ(i, i(−)) and a value which is obtained by multiplying a share [α]i(−) in accordance with the Shamir's secret sharing scheme by a Lagrange coefficient λ(i(−), i) is a reconstructed value α and a sum of a value which is obtained by multiplying the share [α]i by a Lagrange coefficient λ(i, i(+)) and a value which is obtained by multiplying a share [α]i(+) in accordance with the Shamir's secret sharing scheme by a Lagrange coefficient λ(i(+), i) is the reconstructed value α,a share [x]i∈Fm of plaintext x in accordance with the Shamir's secret sharing scheme is expressed by N shares [x0]i, . . . , [xN−1]i which satisfy m=m(0)+ . . . +m(N−1) and [xj]i∈Fm(j), andthe share generating device includes a function operation unit that obtains a function value ri=Pm(i(−))(si)∈Fm(i(−)) of an input seed si,a first secure computation unit that obtains a first calculated value ζi=λ(i, i(−))[xi(−)]i+ri∈Fm(i(−)) using the Lagrange coefficient λ(i, i(−)), a share [xi(−)]i, and the function value ri,an output unit that outputs the first calculated value ζi to the share generating device Ai(−),an input unit that accepts a second calculated value ζi(+)∈Fm(i),a second secure computation unit that obtains a third calculated value zi=λ(i, i(+))[xi]i+ζi(+)∈Fm(i) using the Lagrange coefficient λ(i, i(+)), the share [xi]i, and the second calculated value ζi(+), and a share generation unit that obtains information containing the seed si and the third calculated value zi as a share SSi of the plaintext x in secret sharing and outputs the share SSi.
  • 2: The share generating device according to claim 1, wherein N is an integer greater than or equal to 3, i(++)=i+2 mod N holds, i(−−)=i−2 mod N holds, δ=0, . . . , N−1 holds, v is an integer greater than or equal to 1, ceil is a ceiling function, the share [xj]i can be divided into m(j) shares [(xj)0]i, . . . , [(xj)m(j)−1]i∈F, m′(i) is ceil(m(i)/v), and [(x′δ)j]i is ([(xδ)vj]i, [(xδ)vj+1]i, . . . , [(xδ)v(j+1)−1)i]i)∈Fv belonging to a set Fv,the share generating device includes a joint possession unit that possesses an arbitrary value ti(−−)∈Fv jointly with a share generating device Ai(−−) and possesses an arbitrary value ti∈Fv jointly with a share generating device Ai(++) and/or possesses an arbitrary value ui(−−)∈Fv jointly with a share generating device Ai(+) and possesses an arbitrary value ui(−)∈Fv jointly with the share generating device Ai(−),a first checksum generation unit that obtains a checksum ci(−−), i=Σ0≤j<m′(i(−−)){ti(−−)j+1[(x′i(−−))j]i}+ti(−−)m′(i(−−))[x′i(−−))m′(i(−−))−1]i∈Fv when joint possession of the arbitrary value ti(−−) has been performed by the joint possession unit and obtains a checksum di(−−), i=Σ0≤j<m′(i(−−)){ui(−−)j+1[(x′i(−−))j]i}+ui(−−)m′(i(−−))[x′i(−−))m′(i(−−))−1]i∈Fv when joint possession of the arbitrary value ui(−−) has been performed by the joint possession unit, anda second checksum generation unit that obtains a checksum ci, i=Σ0≤j<m′(i){tij+1[(x′i)j]i}+tim′(i)+1[(x′i)m′(i)−1]i∈Fv when joint possession of the arbitrary value ti has been performed by the joint possession unit and obtains a checksum di(−), i Σ0≤j<m′(i){uij+1[(x′i)j]i}+uim′(i)+1[(x′i)m′(i)−1]i∈Fv when joint possession of the arbitrary value ui(−) has been performed by the joint possession unit,when joint possession of the arbitrary value ti(−−) and the arbitrary value ti has been performed by the joint possession unit, the share SSi, which is generated by the share generation unit, further contains the arbitrary value ti(−−), the arbitrary value ti, the checksum ci(−−), i, and the checksum ci, i, andwhen joint possession of the arbitrary value ui(−−) and the arbitrary value ui(−) has been performed by the joint possession unit, the share SSi, which is generated by the share generation unit, further contains the arbitrary value ui(−−), the arbitrary value ui(−), the checksum di(−−), i, and the checksum di(−), i.
  • 3: The share generating device according to claim 1, wherein N is an integer greater than or equal to 3, i(++)=i+2 mod N holds, i(−−)=i−2 mod N holds, δ=0, . . . , N−1 holds, v is an integer greater than or equal to 1, ceil is a ceiling function, the share [xj]i can be divided into m(j) shares [(xj)0]i, . . . , [(xj)m(j)−1]i∈F, m′(i) is ceil(m(i)/v), and [(x′δ)j]i is ([(xδ)vj]i, [(xδ)vj+1]i, . . . , [(x)v(j+1)−1]i)∈Fv belonging to a set F,the share generating device includes a joint possession unit that possesses an arbitrary value ti(−−)∈Fv jointly with a share generating device Ai(−−), possesses an arbitrary value ti∈Fv jointly with a share generating device Ai(++), possesses an arbitrary value ui(−−)∈Fv jointly with a share generating device Ai(+), and possesses an arbitrary value ui(−)∈Fv jointly with the share generating device Ai(−),a first checksum generation unit that obtains a checksum ci(−−), i=Σ0≤j<m′(i(−−)){ti(−−)j+1[(x′i(−−))j]i}+ti(−−)m′(i(−−))[x′i(−−))m′(i(−−))−1]i∈Fv and obtains a checksum di(−−), i=Σ0≤j<m′(i(−−)){ui(−−)j+1[(x′i(−−))j]i}+ui(−−)m′(i(−−))[x′i(−−))m′(i(−−))−1]i∈Fv, anda second checksum generation unit that obtains a checksum ci, i=Σ0≤j<m′(i){tij+1[(x′i)j]i}+tim′(i)+1[(x′i)m′(i)−1]i∈Fv and obtains a checksum di(−), i=Σ0≤j<m′(i−)){uij+1[(x′i)j]i}+uim′(i)+1[(x′i)m′(i)−1]i∈Fv, andthe share SSi which is generated by the share generation unit, further contains the arbitrary value ti(−−), the arbitrary value ti, the arbitrary value ui(−−), the arbitrary value ui(−), the checksum ci(−−), i, the checksum ci, i, the checksum di(−−), i, and the checksum di(−), i.
  • 4: The share generating device according to any one of claims 1 to 3, wherein an amount of data of the seed si is smaller than an amount of data of the function value ri.
  • 5: The share generating device according to any one of claims 1 to 3, wherein a total amount of data of the shares SS0, . . . , SSN−1 is less than N times an amount of data of the plaintext x.
  • 6: A reconstructing device, wherein N is an integer greater than or equal to 2, m is an integer greater than or equal to 1, m(i) is an integer greater than or equal to 0 and less than m, i=0, . . . , N−1 holds, i(+)=i+1 mod N holds, Pm(i) is a function, and a range of the function Pm(i) belongs to a set Fm(i) whose members are sequences of elements of m(i) fields F, andthe reconstructing device includes an input unit that accepts N shares SS0, . . . , SSN−1,an arithmetic unit that obtains a member xi=zi−Pm(i)(si(+))∈Fm(i) using a third calculated value zi∈Fm(i) contained in a share SSi and a seed si(+) contained in a share SSi(+), anda concatenation unit that obtains a reconstructed value x∈Fm which is expressed by members x0, . . . , xN−1 obtained in the arithmetic unit.
  • 7: The reconstructing device according to claim 6, wherein N is an integer greater than or equal to 3, v is an integer greater than or equal to 1, ceil is a ceiling function, the member xi∈F(i) can be divided into m(i) sub-members (xi)0, . . . , (xi)m(i)−1∈F, m′(i) is ceil(m(i)/v), (x′i)j is ((xi)vj, (xi)vj+1, . . . , (xi)v(j+1)∈Fv belonging to a set Fv, and the share SSi further contains an arbitrary value ti(−−), an arbitrary value ti, a checksum ci(−−), i, and a checksum ci, i and/or an arbitrary value ui(−−), an arbitrary value ui), a checksum di(−−), i and a checksum di(−), i, andthe reconstructing device includes a reconstruction unit that reconstructs, when the share SSi contains the checksum ci, i and a share SSi(++) contains a checksum ci, i(++), a checksum ci using the checksum ci, i and the checksum ci, i(++) as shares in accordance with Shamir's secret sharing scheme and reconstructs, when the share SSi(+) contains a checksum di, i(+) and the share SSi(++) contains a checksum di, i(++), a checksum di using the checksum di, i(+) and the checksum di, i(++) as shares in accordance with the Shamir's secret sharing scheme,a verification value generation unit that obtains, when the checksum ci has been reconstructed, a verification value vci=Σ0≤j<m′(i){tij+1(x′i)j}+tim′(i)+1(x′i)m(i)−1∈Fv and obtains, when the checksum di has been reconstructed, a verification value vdi=Σ0≤j<m′(i){tij+1(x′i)j}+uim′(i)+1(x′i)m(i)−1∈Fv, anda determination unit that determines whether ci=vci is satisfied when the checksum ci has been reconstructed and determines whether di=vdi is satisfied when the checksum di has been reconstructed.
  • 8: The reconstructing device according to claim 6, wherein N is an integer greater than or equal to 3, v is an integer greater than or equal to 1, ceil is a ceiling function, the member xi∈Fm(i) can be divided into m(i) sub-members (xi)0, . . . , (xi)m(i)−1∈F, m′(i) is ceil(m(i)/v), (x′i)j is ((xi)vj, (xi)vj+1, . . . , (xi)v(j+1−1)∈Fv belonging to a set Fv, and the share SSi further contains an arbitrary value ti(−−), an arbitrary value ti, a checksum ci(−−), i, a checksum ci, i, an arbitrary value ui(−), an arbitrary value ui(−), a checksum di(−−), i, and a checksum di(−−), i, andthe reconstructing device includes a reconstruction unit that reconstructs a checksum ci using the checksum ci, i and the checksum ci, i(++) as shares in accordance with Shamir's secret sharing scheme and reconstructs a checksum di using the checksum di, i(+) and the checksum di, i(++) as shares in accordance with the Shamir's secret sharing scheme,a verification value generation unit that obtains a verification value vci=Σ0≤j<m′(i){tij+1(x′i)j}+tim′(i)+1(x′i)m(i)−1∈Fv and obtains a verification value vdi=Σ0≤j<m′(i){uij+1(x′i)j}+uim′(i)+1(x′i)m(i)−1∈Fv, anda determination unit that determines whether ci=vci is satisfied and determines whether di=vdi is satisfied.
  • 9: The reconstructing device according to any one of claims 6 to 8, wherein an amount of data of the seed si(+) is smaller than an amount of data of the Pm(i)(si(+)).
  • 10: The reconstructing device according to any one of claims 6 to 8, wherein a total amount of data of the shares SS0, . . . , SSN−1 is less than N times an amount of data of plaintext x.
  • 11: A secure computation system comprising: a reconstructing device; andN share generating devices A0, . . . , AN−1, whereinN is an integer greater than or equal to 2, m is an integer greater than or equal to 1, m(i) is an integer greater than or equal to 0 and less than m, i=0, . . . , N−1 holds, j=0, . . . , N−1 holds, i(+)=i+1 mod N holds, i(−)=i−1 mod N holds, Pm(i) is a function, and a range of the function Pm(i) belongs to a set Fm(i) whose members are sequences of elements of m(i) fields F,a sum of a value which is obtained by multiplying a share [α] in accordance with Shamir's secret sharing scheme by a Lagrange coefficient (i, i(−)) and a value which is obtained by multiplying a share [α]i(−) in accordance with the Shamir's secret sharing scheme by a Lagrange coefficient λ(i(−), i) is a reconstructed value α and a sum of a value which is obtained by multiplying the share [α]i by a Lagrange coefficient λ(i, i(+)) and a value which is obtained by multiplying a share [α]i(+) in accordance with the Shamir's secret sharing scheme by a Lagrange coefficient λ(i(+), i) is the reconstructed value α,a share [x]i∈Fm of plaintext x in accordance with the Shamir's secret sharing scheme is expressed by N shares [x0]i, . . . , [xN−1]i which satisfy m=m(0)+ . . . +m(N−1) and [xj]i∈Fm(j),a share generating device Ai included in the N share generating devices A0, . . . , AN−1 includes a function operation unit that obtains a function value ri=Pm(i(−))(si)∈Fm(i(−)) of an input seed si,a first secure computation unit that obtains a first calculated value ζi=λ(i, i(−))[xi(−)]i+ri∈Fm(i(−)) using the Lagrange coefficient λ(i, i(−)), a share [xi(−)]i, and the function value ri,an output unit that outputs the first calculated value ζi to the share generating device Ai(+),a first input unit that accepts a second calculated value ζi(+)∈Fm(i),a second secure computation unit that obtains a third calculated value zi=λ(i, i(+))[xi]i+ζi(+∈Fm(i) using the Lagrange coefficient λ(i, i(+)), the share [xi]i, and the second calculated value ζi(+), anda share generation unit that obtains information containing the seed si and the third calculated value zi as a share SSi of the plaintext x in secret sharing and outputs the share SSi, andthe reconstructing device includes a second input unit that accepts N shares SS0, . . . , SSN−1,an arithmetic unit that obtains a member xi=zi−Pm(i)(si(+))∈Fm(i) using the third calculated value zi∈Fm(i) contained in the share SSi and a seed si(+) contained in a share SSi(+), anda concatenation unit that obtains a reconstructed value x∈Fm which is expressed by members x0, . . . , xN−1 obtained in the arithmetic unit.
  • 12: A share generation method of a share generating device Ai included in N share generating devices A0, . . . , AN−1, wherein N is an integer greater than or equal to 2, m is an integer greater than or equal to 1, m(i) is an integer greater than or equal to 0 and less than m, i=0, . . . , N−1 holds, j=0, . . . , N−1 holds, i(+)=i+1 mod N holds, i(−)=i−1 mod N holds, Pm(i) is a function, and a range of the function Pm(i) belongs to a set F′(i) whose members are sequences of elements of m(i) fields F,a sum of a value which is obtained by multiplying a share [α]i in accordance with Shamir's secret sharing scheme by a Lagrange coefficient λ(i, i(−)) and a value which is obtained by multiplying a share [α]i(−) in accordance with the Shamir's secret sharing scheme by a Lagrange coefficient λ(i(−), i) is a reconstructed value α and a sum of a value which is obtained by multiplying the share [α]i by a Lagrange coefficient λ(i, i(+)) and a value which is obtained by multiplying a share [α]i(+) in accordance with the Shamir's secret sharing scheme by a Lagrange coefficient λ(i(+), i) is the reconstructed value α,a share [x]i∈Fm of plaintext x in accordance with the Shamir's secret sharing scheme is expressed by N shares [x0]i, . . . , [xN−1]i which satisfy m=m(0)+ . . . +m(N−1) and [xi]i∈Fm(j), andthe share generation method includes a function operation step in which a function operation unit obtains a function value ri=Pm(i(−))(si)∈Fm(i(−)) of an input seed si,a first secure computation step in which a first secure computation unit obtains a first calculated value ζi=λ(i, i(−))[xi(−)]i+ri∈Fm(i(−)) using the Lagrange coefficient λ(i, i(−)), a share [xi(−)]i, and the function value ri,an output step in which an output unit outputs the first calculated value ζi to the share generating device Ai(−),an input step in which an input unit accepts a second calculated value ζi(+)∈Fm(i),a second secure computation step in which a second secure computation unit obtains a third calculated value zi=λ(i, i(+))[xi]i+ζi(+)∈Fm(i) using the Lagrange coefficient λ(i, i(+)), the share [xi]i, and the second calculated value ζi(+), anda share generation step in which a share generation unit obtains information containing the seed si and the third calculated value zi as a share SSi of the plaintext x in secret sharing and outputs the share SSi.
  • 13: A reconstruction method, wherein N is an integer greater than or equal to 2, m is an integer greater than or equal to 1, m(i) is an integer greater than or equal to 0 and less than m, i=0, . . . , N−1 holds, i(+)=i+1 mod N holds, Pm(i) is a function, and a range of the function Pm(i) belongs to a set Fm(i) whose members are sequences of elements of m(i) fields F, andthe reconstruction method includes an input step in which an input unit accepts N shares SS0, . . . , SSN−1,an arithmetic step in which an arithmetic unit obtains a member xi=zi−Pm(i)(si(+))∈Fm(i) using a third calculated value zi∈Fm(i) contained in a share SSi and a seed si(+) contained in a share SSi(+), anda concatenation step in which a concatenation unit obtains a reconstructed value x∈Fm which is expressed by members x0, . . . , xN−1 obtained in the arithmetic unit.
  • 14. (canceled)
  • 15: A non-transitory computer-readable recording medium that stores a program for making a computer function as the share generating device according to any one of claims 1 to 3 or the reconstructing device according to any one of claims 6 to 8.
Priority Claims (1)
Number Date Country Kind
2017-159346 Aug 2017 JP national
PCT Information
Filing Document Filing Date Country Kind
PCT/JP2018/030442 8/16/2018 WO 00