Invariants for sum-rank metric codes

Paolo Santonastaso, Ferdinando Zullo · 2025

Read Invariants for sum-rank metric codes by Paolo Santonastaso, Ferdinando Zullo (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-04158.txt.

Introduction

Paolo Santonastaso and Ferdinando Zullo forge new fingerprints for sum-rank codes in a 2025 paper. Coding theory asks when two codes are really the same code in disguise. That equivalence question feeds cryptography. The sum-rank metric, which blends two older metrics, breaks the classic fingerprints.

The authors define fresh invariants: generalized idealisers, a centraliser, a center, and a refined sense of linearity. From these grow nuclear parameters, borrowed from division algebra theory, where they excel at proving codes distinct. A computing framework on skew polynomials, mirroring the matrix setting, makes the parameters computable for known best codes. With them, the team studies equivalence among the largest known families of top-grade sum-rank codes. They offer the method where traditional tools fall short.

Catalog record

CREATORPaolo Santonastaso, Ferdinando Zullo
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE106 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-04158

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Invariants for sum-rank metric codes Paolo Santonastaso, Ferdinando Zullo Abstract . The code equivalence problem is central in coding theory and cryptography. While classical invariants are effective for Hamming and rank metrics, the sum-rank metric, which unifies both, introduces new challenges. This paper introduces new invariants for sum-rank metric codes: generalised idealisers, the centraliser, the center, and a refined notion of linearity. These lead to the definition of nuclear parameters, inspired by those used in division algebra theory, where they are crucial for proving inequivalence. We also develop a computational framework based on skew polynomials, which is isometric to the classical matrix setting but enables explicit computation of nuclear parameters for known MSRD (Maximum Sum-Rank Distance) codes. This yields a new and effective method to study the code equivalence problem where traditional tools fall short. In fact, using nuclear parameters, we can study the equivalence among the largest families of known MSRD codes. 2020 Mathematics Subject Classification 12E10; 16S36; 94B60 1. Introduction In coding theory, one of the fundamental questions is how to determine whether two linear codes are equivalent, that is, whether they can be transformed into one another by a certain isometry. This problem, known as code equivalence problem , plays a central role in the design and analysis of several cryptographic schemes. In particular, many code-based cryptosystems rely on the assumed hardness of code equivalence under various metrics (e.g., Hamming or rank metric) to hide the structure of secret codes and ensure security; see, for example, [ 5 , 41 , 47 , 9 ] . Classical invariants used to study code equivalence include the basic code parameters (length, dimension, and minimum distance), the weight distribution (i.e., the number of codewords of each weight), the structure of the automorphism group, and the supports of codewords, particularly those of minimum weight. These invariants are preserved under code equivalence and are therefore essential tools in both theoretical classification and cryptanalytic attacks on code-based cryptosystems; see [ 5 ] . This paper regards the study of invariants for sum-rank metric codes. The sum-rank metric unifies and generalizes both Hamming metric and rank metric, providing a flexible framework for applications in network coding, distributed storage, and cryptography; see e.g. [ 35 ] . The codewords of a sum-rank metric code are vectors of matrices (or vectors over extension fields), and the distance between two codewords is defined as the sum of the ranks of the differences of their corresponding blocks. Families such as Reed–Solomon codes and Gabidulin codes are examples or building blocks of optimal sum-rank metric codes. These codes have recently gained significant attention in both theoretical research and practical applications, particularly in designing resilient multi-shot network codes and secure distributed storage systems; see [ 33 , 34 ] . In recent years, a rich mathematical theory of sum-rank metric codes has been developed, primarily through the work of Martínez-Peñas [ 29 , 30 , 31 ] , and continues to grow with the emergence of new constructions of optimal codes tailored to various bounds; see [ 10 , 2 , 40 , 1 , 12 , 4 ] . Indeed, within the framework of skew polynomials, the known constructions of MSRD codes—those that attain equality in the Singleton-like bound—are as follows: • Linearized Reed-Solomon (LRS) codes, • (Additive) Twisted Linearized Reed-Solomon (ATLRS) codes, • Twisted Linearized Reed-Solomon (TLRS) codes of TZ-type. In most of the cases, these families provide examples of MSRD codes with the same parameters. This motivates the need to develop suitable invariants capable of distinguishing between such codes. Some of them have been recently developed for the sum-rank metric, such as supports [ 30 ] , an Overbeck-like invariant [ 21 ] , and latroids [ 19 ] ; however, they are generally very difficult to compute. A similar situation occurred with the rank metric. A key idea was first introduced by Liebhold and Nebe in [ 24 ] , and later developed by Lunardon, Trombetti, and Zhou in [ 28 ] . They adapted classical invariants from the theory of semifields and division algebras to the rank-metric setting. The core idea is to associate certain algebras related to a rank-metric code that correspond to subgroups of its automorphism group—subgroups that are, in some sense, much easier to compute. These algebras are known as idealisers . They have been widely used to prove the inequivalence of certain MRD codes, such as Gabidulin codes, twisted Gabidulin codes [ 55 ] , and codes arising from a broad class of scattered polynomials [ 26 ] . In the first part of this paper, we study and characterize the notion of nondegenerate rank-metric codes, which have been mainly considered in the previous papers only for a class of rank-metric codes which are known as {\mathbb{F}}_{q^{m}} -linear codes; see e.g. [ 45 , 3 ] . We now introduce several new invariants for the sum-rank metric: • generalised idealisers (Section 4.2 ), • the centraliser (Section 4.3 ), • the center (Section 4.3 ), • a refined notion of linearity (Section 4.4 ). Generalised idealisers can be seen as an extension of classical idealisers from the rank metric. The key difference is that, in the sum-rank setting, these become sets of t -tuples of matrices. Moreover, there can be more than two idealisers - specifically, up to 2^{t} , where t corresponds to the number of blocks of a sum-rank metric code. The notions of centraliser and center have been recently introduced in the rank-metric context as generalizations of the right nucleus and the center of semifields/division algebras (cf. [ 52 , 53 ] ). In this work, we further extend these concepts to the sum-rank metric framework, where we prove that they serve as invariants for codes in this setting. By analogy with the nuclear parameters of a semifield and those of a rank-metric code, we refer to the sizes of left and right idealisers, as well as those of the centraliser and the center, as the nuclear parameters of a sum-rank metric code, since they are invariants under code equivalence. Linearity is another important invariant. As shown in Example 4.20 , there exist equivalent sum-rank metric codes such that one is \mathbb{F}_{q^{m}} -linear and the other is only \mathbb{F}_{q} -linear, viewed as subspaces of the same ambient space. This implies that linearity alone is not sufficient to determine the code equivalence. To resolve this, we introduce a refined notion of linearity based on the left idealiser. This approach allows us to distinguish codes more effectively and avoid the issue described above. The second part of the paper is devoted to computing the nuclear parameters of the known families of MSRD codes. To this end, we introduce the framework of skew polynomials (see [ 36 ] ), which is isometric to the classical setting. However, this approach provides a novel and powerful method that allows the explicit computation of nuclear parameters of known MSRD code families. This perspective is new and opens up a completely different approach to proving the inequivalence of certain sum-rank metric codes — based not on traditional invariants, but on the structure and sizes of their nuclear parameters. Indeed, thanks to the nuclear parameters, we

Read the complete text