Abstract
Detection and error capabilities are preserved when applying to a linear code an isomorphism which preserves Hamming distance. We study here two such isomorphisms: permutation isometries and monomial isometries.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Fripertinger, H.: Enumeration of the semilinear isometry classes of linear codes. Bayrether Mathematische Schriften 74, 100–122 (2005)
García-Planas, M.I., Magret, M.D.: Eigenvalues and eigenvectors of permutation matrices. Adv. Pure Math. 5, 390–394 (2015)
Sendrier, N., Simos, D.E.: How easy is code equivalence over \(\mathbb{F}_{q}\)? WCC 2013-International Workshop on Coding and Cryptography, Bergen (2013)
Sendrier, N., Simos, D.E.: The Hardness of Code Equivalence over \(\mathbb{F}_{q}\) and Its Application to Code-Based Cryptography. Post-Quantum Cryptography, pp. 203–216 (2013)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Appendix: Minimal Annihilating Polynomial of Permutation and Monomial Isometries
Appendix: Minimal Annihilating Polynomial of Permutation and Monomial Isometries
The minimal annihilating polynomial of a permutation isometry can be determined by the decomposition of the permutation in disjoint cycles. More concretely, if P 1, P 2 are two permutation matrices associated to two permutations \(\sigma _{2}\), \(\sigma _{2}\) with the same cycle type (conjugate in the symmetric group), the minimal annihilating polynomials \(M_{P_{1}}(t)\) and \(M_{P_{2}}(t)\) coincide.
Proposition 1
Let P be a permutation matrix associated to a permutation with cycle type \(1+\mathop{\ldots }\limits^{\mathop{ \smile }\limits^{ m_{1}}} + 1 + 2+\mathop{\ldots }\limits^{\mathop{ \smile }\limits^{ m_{2}}} + 2 + 3+\mathop{\ldots }\limits^{\mathop{ \smile }\limits^{ m_{3}}} + 3 +\ldots +r+\mathop{\ldots }\limits^{\mathop{ \smile }\limits^{ m_{r}}} + r\) . Then
where n i = 0 if m i = 0 and n i = 1 if m i > 0, 1 ≤ i ≤ r (see [2]).
Example 11
Let us consider P = P(2, 3, 4, 1, 6, 5). Then 234165 = (2, 3, 4, 1)(6, 5) has cycle type 2 + 4 and
For any monomial matrix, \(M = M(a_{1},\ldots,a_{n};i_{1}\ldots i_{n})\), the characteristic polynomial of the isomorphism g M can be obtained from the coefficients \(a_{1},\ldots,a_{n}\) and the cycle type of the permutation \(i_{1}\ldots i_{n}\).
Lemma 4
-
(a)
The characteristic polynomial of M is a product of factors, each of them corresponding to one of the disjoint cycles in the decomposition of \(i_{1}\ldots i_{k}\) .
-
(b)
Given a matrix \(M = M(a_{1},\ldots,a_{k};j_{1},\ldots,j_{k})\) with \(j_{1},\ldots,j_{k} = (\,j_{1},\ldots,j_{k})\) a cycle of length k, the characteristic polynomial of M is \(t^{k} - a_{1}\cdots a_{k}\) .
Example 12
2 3 1 5 6 7 4 = (2, 3, 1)(5, 6, 7, 4) and then \(Q_{M}(t) = (t^{3} - 2)(t^{4} - 16)\).
-
In \(\mathbb{F}_{5}[t]\): \(Q_{M}(t) = (t^{3} - 2)(t^{4} - 1)\).
-
In \(\mathbb{F}_{7}[t]\): \(Q_{M}(t) = (t^{3} - 2)(t^{4} - 2)\).
-
In \(\mathbb{F}_{11}[t]\): \(Q_{M}(t) = (t^{3} - 2)(t^{4} - 5)\).
Let us denote by G k (t) the GCD of all factors in the characteristic polynomial of \(M = M(a_{1},\ldots,a_{n};i_{1},\ldots,i_{n})\) of degree k corresponding to cycles of length k, 1 ≤ k ≤ n.
Proposition 2
The minimal annihilating polynomial of monomial matrix \(M = M(a_{1},\ldots,a_{n};i_{1},\ldots,i_{n})\) is:
Example 13
Let us consider
-
\(\begin{array}{l} M = M(2,3,1,1,3,4,1,1,4;2,3,1,5,4,7,6,8,9) \\ \quad = \left (\begin{array}{ccccccccc} 0&2&0&0&0&0&0&0&0\\ 0 &0 &3 &0 &0 &0 &0 &0 &0 \\ 1&0&0&0&0&0&0&0&0\\ 0 &0 &0 &0 &1 &0 &0 &0 &0 \\ 0&0&0&3&0&0&0&0&0\\ 0 &0 &0 &0 &0 &0 &1 &0 &0 \\ 0&0&0&0&0&1&0&0&0\\ 0 &0 &0 &0 &0 &0 &0 &1 &0 \\ 0&0&0&0&0&0&0&0&4\end{array} \right ) \in M_{9}(\mathbb{F}_{7}) \end{array}\)
-
2 3 1 5 4 7 6 8 9 = (2, 3, 1)(5, 4)(7, 6)(8)(9).
-
\(Q_{M}(t) = (t^{3} - 6)(t^{2} - 3)(t^{2} - 1)(t - 1)(t - 4)\).
-
\(G_{1}(t) = (t - 1)(t - 4)\), \(G_{2}(t) = (t^{2} - 3)(t^{2} - 1)\), \(G_{3}(t) = t^{3} - 6\).
-
\(P_{M}(t) = (t - 1)(t - 3)(t - 4)(t - 5)(t - 6)(t^{2} - 3)\).
An example of invariant subspaces are those spanned by eigenvectors. We can determine the set of eigenvectors of a permutation isomorphism from the decomposition of the permutation associated to it, into disjoint cycles and that of a monomial isomorphism from the decomposition into disjoint cycles and the coefficients in the matrix. The proofs are based on straightforward computations.
Proposition 3
Let P be a permutation matrix associated to a permutation which is a disjoint product of cycles, and \(\lambda \in \mathbb{F}_{p}\) an eigenvalue of P, \(\lambda\) an mth-root of unity, the vector \((\lambda ^{j_{k}},\ldots,\lambda ^{j_{1}})\) , where \(j_{1},\ldots,j_{k}\) is the result of re-ordering the indices of the cycle \((i_{1},\ldots,i_{k})\) in such a way that \(j_{1} \leq \ldots \leq j_{k}\) . Then \((\lambda _{j_{k}},\ldots,\lambda _{j_{1}})\) is the eigenvector associated to the eigenvalue \(\lambda\) .
Example 14
We will consider p = 5.
-
1.
Let us consider the 2-cycle (2, 1) and the 2 × 2-matrix associated to it. Then the eigenvector for the eigenvalue \(\lambda \in \mathbb{F}_{5}\), \(\lambda ^{2} = 1\), is \((\lambda,1)\). Since the roots of \(\lambda ^{2} = 1\) are 1 and 4, there are two linearly independent eigenvectors: (1, 1) and (4, 1).
-
2.
If the permutation 3 × 3-matrix is associated to the 2-cycle (2, 3, 1), the eigenvector corresponding to the eigenvalue \(\lambda \in \mathbb{F}_{5}\), \(\lambda ^{3} = 1\), is \((\lambda,\lambda ^{2},1)\). The equation \(\lambda ^{3} = 1\) has only one root, 1, and therefore there is an unique eigenvector is: (1, 1, 1).
If we consider the 3 × 3-permutation matrix is associated to the 2-cycle (3, 2, 1), there is also an unique eigenvector: (1, 1, 1).
-
3.
Let us consider now the case of 4 × 4-permutation matrices associated to 4-cycles. Let \(\lambda\) be a 4th-root of unity (there are four 4th-roots of unity: \(\lambda _{1} = 1\), \(\lambda _{2} = 2\),\(\lambda _{3} = 3\) and \(\lambda _{4} = 4\)).
-
(i)
If the 4-cycle is (1, 2, 3, 4), the eigenvector corresponding to the eigenvalue \(\lambda \in \mathbb{F}_{5}\), is \((\lambda,\lambda ^{2},\lambda ^{3},1)\). That is to say, there are four linearly independent eigenvectors:
$$\displaystyle{(1,1,1,1),(2,4,3,1),(3,4,2,1),(4,1,4,1).}$$ -
(ii)
If the 4-cycle is (1, 3, 4, 2), the eigenvector corresponding to the eigenvalue \(\lambda \in \mathbb{F}_{5}\), \(\lambda ^{4} = 1\), is \((\lambda,\lambda ^{3},\lambda ^{2},1)\). That is to say, there are four linearly independent eigenvectors:
$$\displaystyle{(1,1,1,1),(2,3,4,1),(3,2,4,1),(4,4,1,1).}$$ -
(iii)
If the 4-cycle is (1, 2, 3, 4), the eigenvector corresponding to the eigenvalue \(\lambda \in \mathbb{F}_{5}\), \(\lambda ^{4} = 1\), is \((\lambda,1,\lambda ^{3},\lambda ^{2})\). That is to say, there are four linearly independent eigenvectors:
$$\displaystyle{(1,1,1,1),(2,1,3,4),(3,1,2,4),(4,1,4,1).}$$ -
(iv)
If the 4-cycle is (1, 3, 2, 4), the eigenvector corresponding to the eigenvalue \(\lambda \in \mathbb{F}_{5}\), \(\lambda ^{4} = 1\), is \((\lambda,\lambda ^{2},1,\lambda ^{3}\). That is to say, there are four linearly independent eigenvectors:
$$\displaystyle{(1,1,1,1),(2,4,1,3),(3,4,1,2),(4,1,1,4).}$$ -
(v)
If the 4-cycle is (1, 4, 2, 3), the eigenvector corresponding to the eigenvalue \(\lambda \in \mathbb{F}_{5}\), \(\lambda ^{4} = 1\), is \((\lambda,1,\lambda ^{2},\lambda ^{3})\). That is to say, there are four linearly independent eigenvectors:
$$\displaystyle{(1,1,1,1),(2,1,4,3),(3,1,4,2),(4,1,1,4).}$$ -
(vi)
If the 4-cycle is (1, 2, 4, 3), the eigenvector corresponding to the eigenvalue \(\lambda \in \mathbb{F}_{5}\), \(\lambda ^{4} = 1\), is \((\lambda,\lambda ^{3},1,\lambda ^{2})\). That is to say, there are four linearly independent eigenvectors:
$$\displaystyle{(1,1,1,1),(2,3,1,4),(3,2,1,4),(4,4,1,1).}$$
-
(i)
-
4.
Let us consider the permutation matrix associated to a cycle of type \(2 + 2 + 4 + 8\). Then the minimal annihilating polynomial is: \((t^{4} + 1)(t^{2} + 1)(t + 1)(t - 1)\). The eigenvalues in \(\mathbb{F}_{5}\) are: \(\lambda _{1} = 1\), \(\lambda _{2} = 4\), \(\lambda _{3} = 2\) and \(\lambda _{4} = 3\), being the algebraic multiplicities 4,4,1,1, respectively.
Let us assume, for example, that the 2-cycles are: (9, 16) and (13, 15), the 4-cycle is (1, 3, 5, 14) and the 8-cycle is (2, 4, 6, 7, 8, 10, 11, 12). Then the following linearly independent eigenvectors are obtained.
$$\displaystyle{\begin{array}{cc} \lambda _{1} = 1&(0,0,0,0,0,0,0,0,0,0,0,0,1,0,1,0) \\ &(0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,1) \\ &(1,0,1,0,1,0,0,0,0,0,0,0,0,1,0,0) \\ &(0,1,0,1,0,1,1,1,0,1,1,1,0,0,0,0) \\ \lambda _{2} = 4&(0,0,0,0,0,0,0,0,0,0,0,0,4,0,1,0) \\ &(0,0,0,0,0,0,0,0,4,0,0,0,0,0,0,1) \\ &(4,0,1,0,4,0,0,0,0,0,0,0,0,1,0,0) \\ &(0,4,0,1,0,4,1,4,0,1,4,1,0,0,0,0) \\ \lambda _{3} = 2&(2,0,4,0,3,0,0,0,0,0,0,0,0,1,0,0) \\ \lambda _{4} = 3&(3,0,4,0,2,0,0,0,0,0,0,0,0,1,0,0)\end{array} }$$
Let us consider now a monomial matrix Let \(M = M(a_{1},\ldots,a_{n};i_{1},\ldots,i_{n})\).
Proposition 4
The eigenvalues of M are the roots of the polynomials \(t^{k} - a_{j_{1}}\cdots a_{j_{k}}\) for each cycle \(j_{1}\ldots j_{k}\) of length k in the decomposition of the permutation \(i_{1}\ldots i_{n}\) into disjoint cycles, being \(a_{j_{1}},\ldots,a_{j_{k}}\) the coefficients of M in columns \(j_{1},\ldots,j_{k}\) .
Example 15
Let us consider
We have: 2 3 1 5 6 7 4 = (2, 3, 1)(5, 6, 7, 4) and therefore \(Q_{M}(t) = (t^{3} - 2)(t^{4} - 16)\).
-
In \(\mathbb{F}_{5}[t]\): \(Q_{M}(t) = (t^{3} - 2)(t^{4} - 1)\) and the eigenvalues of M are: 3(2), 1, 2, 4.
-
In \(\mathbb{F}_{7}[t]\): \(Q_{M}(t) = (t^{3} - 2)(t^{4} - 2)\) and the eigenvalues of M are: 2, 5.
-
In \(\mathbb{F}_{11}[t]\): \(Q_{M}(t) = (t^{3} - 2)(t^{4} - 5)\) and the eigenvalues of M are: 7, 2, 9.
Assume that the permutation \(i_{1},\ldots,i_{n}\) splits into m 1 cycles of length k 1,…, m l cycles of length k l . For any irreducible cycle ( j) of length 1, e j is an eigenvector. We can generalize this as follows.
Proposition 5
Let \(a_{j_{1}}\ldots a_{j_{k}}\) be a cycle of length k ≥ 2 in the decomposition of the characteristic polynomial of M into irreducible factors and \(t^{k} - a_{j_{1}}\ldots a_{j_{k}}\) the corresponding factor in Q M (t). For each root \(\lambda\) (in the case where there exists any) of this polynomial we obtain an eigenvector:
Example 16
\(\begin{array}{rl} M & = M(a_{1},a_{2},a_{3},a_{4},a_{5},a_{6},a_{7},a_{8},a_{9};2,3,4,1,6,7,8,9,5) \\ & = \left (\begin{array}{ccccccccc} 0 &a_{1}& 0 & 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 &a_{2}& 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 &a_{3}& 0 & 0 & 0 & 0 & 0 \\ a_{4}& 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\ 0 & 0 & 0 & 0 & 0 &a_{5}& 0 & 0 & 0 \\ 0 & 0 & 0 & 0 & 0 & 0 &a_{6}& 0 & 0 \\ 0 & 0 & 0 & 0 & 0 & 0 & 0 &a_{7}& 0 \\ 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 &a_{8} \\ 0 & 0 & 0 & 0 &a_{9}& 0 & 0 & 0 & 0 \end{array} \right ) \end{array}\)
the eigenvectors are:
-
\((\lambda ^{3},a_{2}a_{3}a_{4},\lambda a_{3}a_{4},\lambda ^{2}a_{4},0,0,0,0,0)\).
-
\((\mu ^{4},a_{6}a_{7}a_{8}a_{9},\mu a_{7}a_{8}a_{9},\mu ^{2}a_{8}a_{9},\mu ^{3}a_{9})\).
For each root \(\lambda\) of \(t^{4} - a_{1}a_{2}a_{3}a_{4}\) and μ of \(t^{5} - a_{5}a_{6}a_{7}a_{8}a_{9}\) in \(\mathbb{F}_{p}\).
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this chapter
Cite this chapter
García-Planas, M.I., Magret, M.D. (2016). Isometries of the Hamming Space and Equivalence Relations of Linear Codes Over a Finite Field. In: Ortegón Gallego, F., Redondo Neble, M., Rodríguez Galván, J. (eds) Trends in Differential Equations and Applications. SEMA SIMAI Springer Series, vol 8. Springer, Cham. https://doi.org/10.1007/978-3-319-32013-7_12
Download citation
DOI: https://doi.org/10.1007/978-3-319-32013-7_12
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-32012-0
Online ISBN: 978-3-319-32013-7
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)