Abstract
The unbalanced oil and vinegar signature scheme (UOV) is a multivariate signature scheme that has essentially not been broken for over 20 years. However, it requires the use of a large public key; thus, various methods have been proposed to reduce its size. In this paper, we propose a new variant of UOV with a public key represented by block matrices whose components correspond to an element of a quotient ring. We discuss how it affects the security of our proposed scheme whether or not the quotient ring is a field. Furthermore, we discuss their security against currently known and newly possible attacks and propose parameters for our scheme. We demonstrate that our proposed scheme can achieve a small public key size without significantly increasing the signature size compared with other UOV variants. For example, the public key size of our proposed scheme is 85.8 KB for NIST’s Post-Quantum Cryptography Project (security level 3), whereas that of compressed Rainbow is 252.3 KB, where Rainbow is a variant of UOV and is one of the third-round finalists of the NIST PQC project.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Bardet, M.: Étude des systèms algébriques surdéterminés. Applications aux codes correcteurs et à la cryptographie. Ph.D. thesis, Université Pierre et Marie Curie-Paris VI (2004)
Bardet, M., Faugère, J.-C., Salvy, B.: Complexity of Gröbner basis computation for semi-regular overdetermined sequences over \(\mathbb{F}_2\) with solutions in \(\mathbb{F}_2\). Research Report, INRIA (2003)
Bardet, M., Faugère, J.-C., Salvy, B., Yang, B.-Y.: Asymptotic behavior of the index of regularity of quadratic semi-regular polynomial systems. In: 8th International Symposium on Effective Methods in Algebraic Geometry (2005)
Bettale, L., Faugère, J.-C., Perret, L.: Hybrid approach for solving multivariate systems over finite fields. J. Math. Cryptol. 3, 177–197 (2009)
Beullens, W.: Improved cryptanalysis of UOV and rainbow. In: Canteaut, A., Standaert, F.-X. (eds.) EUROCRYPT 2021. LNCS, vol. 12696, pp. 348–373. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-77870-5_13
Beullens, W., Preneel, B.: Field lifting for smaller UOV public keys. In: Patra, A., Smart, N.P. (eds.) INDOCRYPT 2017. LNCS, vol. 10698, pp. 227–246. Springer, Cham (2017). https://doi.org/10.1007/978-3-319-71667-1_12
Bosma, W., Cannon, J., Playoust, C.: The Magma algebra system. I. The user language. J. Symbol. Comput. 24(3–4), 235–265 (1997)
Brakerski, Z., Gentry, C., Vaikuntanathan, V.: (Leveled) fully homomorphic encryption without bootstrapping. In: ITCS 2012, pp. 309–325. ACM (2012)
Buchberger, B.: Ein algorithmus zum auffinden der basiselemente des restklassenringes nach einem nulldimensionalen polynomideal. Ph.D. thesis, Universität Innsbruck (1965)
Courtois, N., Klimov, A., Patarin, J., Shamir, A.: Efficient algorithms for solving overdefined systems of multivariate polynomial equations. In: Preneel, B. (ed.) EUROCRYPT 2000. LNCS, vol. 1807, pp. 392–407. Springer, Heidelberg (2000). https://doi.org/10.1007/3-540-45539-6_27
Czypek, P., Heyse, S., Thomae, E.: Efficient implementations of MQPKS on constrained devices. In: Prouff, E., Schaumont, P. (eds.) CHES 2012. LNCS, vol. 7428, pp. 374–389. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-33027-8_22
Ding, J., et al.: Rainbow signature schemes proposal for NIST PQC project (round 3 version)
Ding, J., Schmidt, D.: Rainbow, a new multivariable polynomial signature scheme. In: Ioannidis, J., Keromytis, A., Yung, M. (eds.) ACNS 2005. LNCS, vol. 3531, pp. 164–175. Springer, Heidelberg (2005). https://doi.org/10.1007/11496137_12
Ding, J., Yang, B.-Y., Chen, C.-H.O., Chen, M.-S., Cheng, C.-M.: New differential-algebraic attacks and reparametrization of rainbow. In: Bellovin, S.M., Gennaro, R., Keromytis, A., Yung, M. (eds.) ACNS 2008. LNCS, vol. 5037, pp. 242–257. Springer, Heidelberg (2008). https://doi.org/10.1007/978-3-540-68914-0_15
Ding, J., Zhang, Z., Deaton, J., Schmidt, K., Vishakha, FNU.: New attacks on lifted unbalanced oil vinegar. In: Second PQC Standardization Conference 2019, NIST (2019)
Faugère, J.-C.: A new efficient algorithm for computing Gr\(\rm \ddot{o}\)bner bases (F4). J. Pure Appl. Algebra 139(1–3), 61–88 (1999)
Faugère, J.-C.: A new efficient algorithm for computing Gr\(\rm \ddot{o}\)bner bases without reduction to zero (F5). In: ISSAC 2002, pp. 75–83. ACM (2002)
Furue, H., Kinjo, K., Ikematsu, Y., Wang, Y., Takagi, T.: A structural attack on block-anti-circulant UOV at SAC 2019. In: Ding, J., Tillich, J.-P. (eds.) PQCrypto 2020. LNCS, vol. 12100, pp. 323–339. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-44223-1_18
Furue, H., Nakamura, S., Takagi, T.: Improving Thomae-Wolf algorithm for solving underdetermined multivariate quadratic polynomial problem. In: Cheon, J.H., Tillich, J.-P. (eds.) PQCrypto 2021 2021. LNCS, vol. 12841, pp. 65–78. Springer, Cham (2021). https://doi.org/10.1007/978-3-030-81293-5_4
Garey, M.-R., Johnson, D.-S.: Computers and Intractability: A Guide to the Theory of NP-completeness. Freeman, W.H, San Francisco (1979)
Grover, L.-K.: A fast quantum mechanical algorithm for database search. In: STOC 1996, pp. 212–219. ACM (1996)
Hashimoto, Y.: Minor improvements of algorithm to solve under-defined systems of multivariate quadratic equations. IACR Cryptology ePrint Archive: Report 2021/1045 (2021)
Kipnis, A., Patarin, J., Goubin, L.: Unbalanced oil and vinegar signature schemes. In: Stern, J. (ed.) EUROCRYPT 1999. LNCS, vol. 1592, pp. 206–222. Springer, Heidelberg (1999). https://doi.org/10.1007/3-540-48910-X_15
Kipnis, A., Shamir, A.: Cryptanalysis of the oil and vinegar signature scheme. In: Krawczyk, H. (ed.) CRYPTO 1998. LNCS, vol. 1462, pp. 257–266. Springer, Heidelberg (1998). https://doi.org/10.1007/BFb0055733
Lyubashevsky, V., et al.: CRYSTALS-DILITHIUM signature schemes proposal for NIST PQC project (round 2 version)
NIST: post-quantum cryptography CSRC. https://csrc.nist.gov/Projects/post-quantum-cryptography/post-quantum-cryptography-standardization
NIST: submission requirements and evaluation criteria for the post-quantum cryptography standardization process (2016). https://csrc.nist.gov/CSRC/media/Projects/Post-Quantum-Cryptography/documents/call-for-proposals-final-dec-2016.pdf
NIST: Status report on the first round of the NIST post-quantum cryptography standardization process. NIST Internal Report 8240, NIST (2019)
NIST: Status report on the second round of the NIST post-quantum cryptography standardization process. NIST Internal Report 8309, NIST (2020)
Petzoldt, A., Bulygin, S., Buchmann, J.: CyclicRainbow – a multivariate signature scheme with a partially cyclic public key. In: Gong, G., Gupta, K.C. (eds.) INDOCRYPT 2010. LNCS, vol. 6498, pp. 33–48. Springer, Heidelberg (2010). https://doi.org/10.1007/978-3-642-17401-8_4
The Rainbow Team: Response to recent paper by Ward Beullens (2020). https://troll.iis.sinica.edu.tw/by-publ/recent/response-ward.pdf
Regev, O.: On lattices, learning with errors, random linear codes, and cryptography. In: STOC 2005, pp. 84–93. ACM (2005)
Sakumoto, K., Shirai, T., Hiwatari, H.: On provable security of UOV and HFE signature schemes against chosen-message attack. In: Yang, B.-Y. (ed.) PQCrypto 2011. LNCS, vol. 7071, pp. 68–82. Springer, Heidelberg (2011). https://doi.org/10.1007/978-3-642-25405-5_5
Shor, P.W.: Polynomial-time algorithms for prime factorization and discrete logarithms on a quantum computer. SIAM J. Comput. 26(5), 1484–1509 (1997)
Szepieniec, A., Preneel, B.: Block-anti-circulant unbalanced oil and vinegar. In: Paterson, K.G., Stebila, D. (eds.) SAC 2019. LNCS, vol. 11959, pp. 574–588. Springer, Cham (2020). https://doi.org/10.1007/978-3-030-38471-5_23
Thomae, E., Wolf, C.: Solving underdetermined systems of multivariate quadratic equations revisited. In: Fischlin, M., Buchmann, J., Manulis, M. (eds.) PKC 2012. LNCS, vol. 7293, pp. 156–171. Springer, Heidelberg (2012). https://doi.org/10.1007/978-3-642-30057-8_10
Acknowledgments
This work was supported by JST CREST Grant Number JPMJCR14D6 and JPMJCR2113, Japan, and JSPS KAKENHI Grant Number JP21J20391 and JP19K20266, Japan.
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Appendices
Appendix A: Transformation on Polynomial Matrix from a Reducible Polynomial
First, we discuss the case in which f is reducible and decomposed into distinct irreducible polynomials.
Theorem 4
Let \(f \in {\mathbb F}_q [x]\) be a reducible polynomial with \(\deg {f} = \ell \) and W be an invertible matrix such that every element of \(W A_f\) is a symmetric matrix. If \(f=f_1 \cdots f_k\) \((k \in \mathbb N)\), where \(f_1, \dots , f_k\) are distinct and irreducible, and \(\deg {f_1} \le \cdots \le \deg {f_k}\), then there exists an invertible matrix \(L \in {\mathbb F}_q ^{\ell \times \ell }\) and \(i \in \{ 1, \dots , \ell -1 \}\) such that for any \(X \in W A_f\),
Proof
We first prove that every element of \(A_f W^{-1}\) is symmetric. For any \(g \in {\mathbb F}_q [x] / (f)\),
Therefore, every element of \(A_f W^{-1}\) is symmetric.
As f is reducible, there exists \(a,b \in {\mathbb F}_q [x] / (f)\) such that \(a \cdot b =0\). Then, for any \(g \in {\mathbb F}_q [x] / (f)\),
We assume that \(L \in {\mathbb F}_q ^{\ell \times \ell }\) is designed such that the first i column vectors of L are chosen from the column vector space of \(\varPhi _a^f W^{-1}\), and the other \((\ell -i)\) column vectors of L are chosen from the column vector space of \(\varPhi _b^f W^{-1}\). Then, Eq. (14) explicitly holds from the above equation.
We next show that there exists an invertible such an invertible matrix L. We take \(a = f_1\) and \(b = f_2 \cdots f_k\) (here, \(f_1,\dots ,f_k\) are seen as elements of \({\mathbb F}_q [x] / (f)\).) and prove that \(\mathrm{rank}\,\varPhi _a^f = \deg {b}\) (\(\mathrm{rank}\,\varPhi _b^f = \deg {a}\)). We use the bijective map \(V_1\) used in the proof of Theorem 2. From Eq. (7), for any \(c \in {\mathbb F}_q [x] / (f)\),
As there is no \(c \in {\mathbb F}_q [x] / (f)\) such that \(a \cdot c = 0\) and \(\deg {c} < \deg {b}\), the first \(\deg {b}\) column vectors are linearly independent. Furthermore, as \(\varPhi _a^f \cdot V_1(b) = \mathbf {0},\) \(\varPhi _a^f \cdot V_1(xb) = \mathbf {0}, \dots , \varPhi _a^f \cdot V_1(x^{\deg {a}-1} b) = \mathbf {0}\), we have \(\mathrm{rank}\,\varPhi _a^f = \deg {b}\). Similarly, it is proved that \(\mathrm{rank}\,\varPhi _b^f = \deg {a}\).
Next, we design \(L \in {\mathbb F}_q ^{\ell \times \ell }\) such that the first \(\deg {b}\) column vectors of L are bases of the column vector space of \(\varPhi _a^f W^{-1}\) and the other \((\ell - \deg {b})\) \((= \deg {a})\) column vectors of L are bases of the column vector space of \(\varPhi _b^f W^{-1}\).
Finally, we prove that the column vector spaces of \(\varPhi _a^f W^{-1}\) and \(\varPhi _b^f W^{-1}\) have no intersection, that is, the column vector spaces of \(\varPhi _a^f\) and \(\varPhi _b^f\) have no intersection. If this statement holds, then L constructed using this approach is invertible. We assume that the column vector spaces of \(\varPhi _a^f\) and \(\varPhi _b^f\) have an intersection. Then, there exist two vectors \(\mathbf {x}, \mathbf {y} \in {\mathbb F} _q ^\ell \) such that the last \((\ell - \deg {b})\) elements of \(\mathbf {x}\) and the last \((\ell - \deg {a})\) elements of \(\mathbf {y}\) are zero, and \(\varPhi _a^f \mathbf {x} = \varPhi _b^f \mathbf {y}\) because the first \(\deg {b}\) (\(\deg {a}\)) vectors of \(\varPhi _a^f\) (\(\varPhi _b^f\)) are linearly independent. From the definition of \(\varPhi _g^f\), \(a V_1 ^{-1} (\mathbf {x}) = b V_1 ^{-1} (\mathbf {y})\), \(\deg {(V_1 ^{-1} (\mathbf {x}))} < \deg {b}\), and \(\deg {(V_1 ^{-1} (\mathbf {y}))} < \deg {a}\). However, this contradicts that \(f_1, \dots , f_k\) are distinct and irreducible. Therefore, the column vector spaces of \(\varPhi _a^f\) and \(\varPhi _b^f\) have no intersections. \(\square \)
Next, we discuss another case where f is reducible.
Theorem 5
With the same notation as in Theorem 4, if there exists \(f' \in {\mathbb F}_q [x]\) such that \({f'}^2 \mid f\), there exists an invertible matrix \(L \in {\mathbb F}_q ^{\ell \times \ell }\) such that, for any \(X \in W A_f\),
Proof
From this assumption, there exists \(a \in {\mathbb F}_q [x] / (f)\) such that \(a ^2 = 0\). Therefore, for any \(g \in {\mathbb F}_q [x] / (f)\),
and \(\varPhi _a^f W^{-1}\) is symmetric. We suppose that \(L \in {\mathbb F}_q ^{\ell \times \ell }\) is an invertible matrix, wherein the \(\ell \)-th column vector is chosen from the column vectors of \(\varPhi _a^f W^{-1}\). From the above equation, the \((\ell ,\ell )\) component of \(L ^{\top } (W \varPhi _g^f) L\) is zero for any \(g \in {\mathbb F}_q [x] / (f)\). \(\square \)
Appendix B: Proof of Theorem 3 in Subsect. 5.3
Theorem 3. With the same notation as in Theorem 2,
-
(i) There exists an invertible matrix \(L \in \mathbb {F}_{q^\ell } ^{\ell \times \ell }\) such that \(L ^{-1} \varPhi _g ^f L\) is diagonal for any \(g \in {\mathbb F}_q [x] / (f)\).
-
(ii) The matrix L described in (i) satisfies the condition that \(L^{\top } X L\) is diagonal for any \(X \in W A_f\).
-
(iii) If there exists \(\mathbf {y} \in \mathbb {F}_{q ^{\ell }} ^{\ell }\) such that \(\mathbf {y}^{\top } X \mathbf {y} =0\) for any \(X \in W A_f\), then \(\mathbf{y} = \mathbf{0}\).
Proof
First, we prove statement 1. The characteristic polynomial of \( \varPhi _x ^ f \) is equal to f for \(x \in {\mathbb F}_q [x] / (f)\). As f is irreducible over \({\mathbb F}_q [x]\), f is separable, and its roots are distinct in \(\mathbb {F}_{q^\ell } [x]\). Therefore, the eigenvalues of \(\varPhi _x ^f\) are distinct in \(\mathbb {F}_{q^\ell }\), and there exists \(L \in \mathbb {F}_{q^\ell } ^{\ell \times \ell }\) such that \(L ^{-1} \varPhi _x ^f L\) is diagonal. Furthermore, \(\varPhi _{1} ^f\) is the identity matrix, and \(\varPhi _{x^i} ^f\) \((i = 2,\dots , \ell -1)\) can be diagonalized using L:
Then, for any \(g \in {\mathbb F}_q [x] / (f)\), \(L^{-1} \varPhi _g ^f L\) becomes diagonal because \(A_f\) is spanned by \(\{\varPhi _{1} ^f, \varPhi _{x} ^f, \dots , \varPhi _{x^{\ell -1}} ^f \}\) over \({\mathbb F}_q\).
Next, we prove statement 2 by using the following lemma.
Lemma 2
With the same notation as in Theorem 2, for \(L \in \mathbb {F}_{q^\ell } ^{\ell \times \ell }\) described in Theorem 3, \(L ^{\top } W L\) is diagonal.
Proof
Since \(W \varPhi _g ^f\) is symmetric,
Furthermore, because W is symmetric, we have
As \(L^{-1} \varPhi _g ^f L\) is symmetric,
Then, \(L^{\top } W L\) and \(L^{-1} \varPhi _g ^f L\) are commutative. As \(L^{-1} \varPhi _g ^f L\) is diagonal, and the diagonal components are distinct, \(L^{\top } W L\) is diagonal. \(\square \)
For any \(g \in {\mathbb F}_q [x] / (f)\), we can transform \(L^{\top } W \varPhi _g ^f L\):
From statement 1 and Lemma 2, \(L^{\top } W \varPhi _g ^f L\) are diagonal.
Finally, we prove statement 3. Let \(\mathbf {y} := L ^{-1} \mathbf {x}\); then,
If we define the diagonal components of \(L ^{-1} \varPhi _x ^f L\) as \(\theta _1, \dots , \theta _\ell \) (the roots of f in \(\mathbb {F}_{q^\ell }\)), the diagonal components of \(L ^{-1} \varPhi _g ^f L\) are equal to \(g(\theta _1), \dots , g(\theta _\ell )\). If ,
since \(L ^{\top } W L\) is diagonal.
Let \(g_1, \dots , g_\ell \) be the basis of \({\mathbb F}_q [x] / (f)\) over \({\mathbb F}_q\), then, satisfying Eq. (16) for any \(g \in {\mathbb F}_q [x] / (f)\) is equivalent to
In addition, \(g_1, \dots , g_\ell \) form the basis of \({\mathbb F}_{q ^\ell } [x] / (f)\) over \({\mathbb F}_{q ^\ell }\), and
Therefore, \(\left( g_i (\theta _1) \cdots g_i (\theta _\ell ) \right) \) \((i=1,\dots ,\ell )\) are linearly independent, and
\(\square \)
Appendix C: Performance in Magma
Here, we present the execution times for key generation, signature generation, and verification of the improved QR-UOV in Subsect. 4.2. All experiments were performed on a MacBook Pro with a 2.4-GHz quad-core, Intel Core i5 CPU, and the Magma algebra system (V2.24-82) [7]. Table 6 shows the average times for 100 runs using the improved QR-UOV scheme described in Subsect. 4.2 and our proposed parameters for levels I, III, and V of the NIST PQC project. All timings are in second. These are not optimized implementations.
In the key generation step, we first generate two 32-bit seeds (\(\mathbf {s}_{sk}\) and \(\mathbf {s}_{pk}\)) by using the Magma Random command. We then use the Magma SetSeed command as a pseudo-random number generator to generate part of the public and secret keys. (In Subsect. 6.2, we stated that the size of the two seeds is 256 bits; however, we use two 32-bit seeds because the size of the input for SetSeed is at most 32 bits.) Next, we generate a secret key using the method described in Subsect. 4.2. In the signature generation step, we recover the public and secret keys from the two seeds and perform the procedure explained in Subsect. 2.2. The signature is generated in the same manner as a signature is generated in the compressed Rainbow [12]. In the verification step, we generate the public key from the \(\mathbf {s}_{pk}\) seed and follow the procedure explained in Subsect. 2.1. In the signature generation and verification steps, we need to compute the product of a vector and matrices \(W \varPhi _g^f\) or \(\varPhi _g^f\), which is made more efficient using the structure of the polynomial matrix.
For example, in Table 6, the execution times of the key generation, signature generation, and verification steps of QR-UOV for level I are 0.06 s, 0.04 s, and 0.01 s, respectively. In most cases, our performance is approximately one order of magnitude slower than that of compressed Rainbow [12]. It should be noted that their implementation is in C, and ours is in Magma, and the signing and verification times of compressed Rainbow are dominated by the use of a cryptographic hash function which is not used in the implementation of QR-UOV.
Rights and permissions
Copyright information
© 2021 International Association for Cryptologic Research
About this paper
Cite this paper
Furue, H., Ikematsu, Y., Kiyomura, Y., Takagi, T. (2021). A New Variant of Unbalanced Oil and Vinegar Using Quotient Ring: QR-UOV. In: Tibouchi, M., Wang, H. (eds) Advances in Cryptology – ASIACRYPT 2021. ASIACRYPT 2021. Lecture Notes in Computer Science(), vol 13093. Springer, Cham. https://doi.org/10.1007/978-3-030-92068-5_7
Download citation
DOI: https://doi.org/10.1007/978-3-030-92068-5_7
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-92067-8
Online ISBN: 978-3-030-92068-5
eBook Packages: Computer ScienceComputer Science (R0)