Published in IET Information Security
Received on 2nd February 2012
Revised on 18th October 2012
Accepted on 5th November 2012
doi: 10.1049/iet-ifs.2012.0035
ISSN 1751-8709
New multivariate hash function quadratic polynomials
multiplying linear polynomials
Youjiao Zou
1,2
, Wenping Ma
1
, Zhanjun Ran
2
, Shangping Wang
2
1
State Key Laboratory of Integrated Service Networks, Xidian University, 710071 Xi’an, People’s Republic of China
2
Department of Mathematics, Xi’an University of Technology, College of Science, 710048 Xi’an, People’s Republic of China
E-mail: youjiaozou@gmail.com
Abstract: In this study the authors propose a new multivariate hash function with HAsh Iterative FrAmework framework which
we call the hash function quadratic polynomials multiplying linear polynomials (QML). The new hash function is made of cubic
polynomials which are the products of quadratic polynomials and linear polynomials. The authors design the quadratic-
polynomial part of the compression function based on the centre map of the multivariate public key cryptosystem MIQ1 . The
hash function QML can keep the three cryptography properties and be immune to the pre-image attack, second pre-image
attack, collision attack, differential attack and algebraic attack. The required memory storage is about 50% of the one which is
built of the cubic polynomials and their coefficients are random. On the avalanche effect, by experiments the authors get the
result that about one half of the output bits are different when one input bit is changed randomly. The one-round diffusion of
the hash function QML is twice of that of Blake. Also the authors simplify the matrixes of the new hash function, analyse the
rationality and show the compa rable data. Finally, the authors give the advice to the parameters of the new hash function and
summarise the paper.
1 Introduction
The cryptographic hash function plays a very important role in
cryptography. Many protocols and signatures depend on the
cryptographic security of the hash functions. After Wang and
Wu [1] and Wang et al. [2, 3] claimed that she had attacked
the family of MD and SHA successfully, constructing secure
hash functions is one of the hot research topics within the
cryptographic community. In 2007, NIST announced a
public competition [4] to develop a new cryptographic hash
algorithm, which converts a variable length message into a
short ‘message digest’ that can be used in generating digital
signatures, message authentication codes and many other
security applications in the information infrastructure. The
public competition is similar to the development process for
the advanced encryption standard.
After two candidate Conferences, NIST has selected five
SHA-3 finalists – BLAKE [5], Grøstl [6], JH [7], Keccak
[8] and Skein[9] to advance to the third (and final) round of
the SHA-3 competition. The final SHA-3 Candidate
Conference is held in Washington DC in the spring of 2012
to dis cuss the public feedback on these candidates, and will
select the SHA-3 winner later in 2012. To these five
accepted algorithms, the idea is based on some block cipher
or ARX cryptosystems. No one is based on the multivariate
polynomials. It seems that the theory of multivariate
polynomials is a minor role in the cryptographic algorithms.
Therefore here we try to advance the safe multivariate hash
algorithm to discuss its application.
The compression function of the new multivariate hash
function quadratic polynomials multiplying linear
polynomials (QML) in this paper is built by multiplying the
quadratic polynomials with linear polynomials to get the
cubic multivariate polynomials. The part of the quadratic
polynomials is based on the idea of the multivariate public
key algorithm MI. The following section will show that the
security of the compression function is equivalent to the
MQ problem.
This paper is organised as follows. In Section 2, we show
the theory which the hash function QML is based on and the
notations which we will use in this paper. In Section 3, the
hash function QML is described in detail. In Section 4, we
discuss the security in theory and show the experimental
results on the performance. In Section 5, according to the
experimental result, we give an advice to choose th e
parameters based on finite field and matrix simplification.
Finally, the advantages about the new multivariate hash
function QML are summarised.
2 Theory basis MQ question and the
notations
The security of some cryptography algorithms depends on
some mathematic NP hard problem. For example, the
security of RSA is based on the difficulty of the large
integer factorization. The hash function which we build
depends on the difficulty of solving random systems of
www.ietdl.org
IET Inf. Secur., pp. 1–8 1
doi: 10.1049/iet-ifs.2012.0035
&
The Institution of Engineering and Technology 2013
Techset Composition Ltd, Salisbury
Doc: {IEE}IFS/Articles/Pagination/IFS20120035.3d
5
10
15
20
25
30
35
40
45
50
55
60
65
70
75
80
85
90
95
100
105
110
115
120
125
130