A note on short minimal codes from subgeometries
Sam Adriaensen, Peter Sziklai, Zsuzsa Weiner · 2025
Read A note on short minimal codes from subgeometries by Sam Adriaensen, Peter Sziklai, Zsuzsa Weiner (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-2511-15372.txt.
Introduction
Sam Adriaensen, Peter Sziklai, and Zsuzsa Weiner give a short proof of a result on minimal codes from subgeometries in projective space. A 2022 paper showed three subgeometries can union into a strong blocking set in a specific projective space. That construction proves certain linear minimal codes exist for every prime power. The codes have fully specified length and dimension parameters.
The new proof covers odd prime powers above nine. It uses the theory of small blocking sets in projective planes instead of heavier machinery. A shorter route clarifies why the construction works. Minimal codes matter for secret sharing and coding theory. Simplifications that widen understanding of existence proofs are valuable to finite geometers.
Catalog record
| CREATOR | Sam Adriaensen, Peter Sziklai, Zsuzsa Weiner |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 20 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-15372
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A note on short minimal codes from subgeometries Sam Adriaensen, Peter Sziklai, Zsuzsa Weiner Abstract In a 2022, Bartoli, Cossidente, Marino, and Pavese proved that in the projective space PG ( 3 , q 3 ) {\rm PG}(3,q^{3}) , one can find three 𝔽 q \mathbb{F}_{q} -subgeometries such that the union of their point sets is a strong blocking set. This proves the existence of linear minimal codes with parameters [ 3 ( q 2 + 1 ) ( q + 1 ) , 4 ] q 3 [3(q^{2}+1)(q+1),4]_{q^{3}} for every prime power q q . We give a short proof of this result for odd values of q > 9 q>9 , using the theory of small blocking sets in projective planes. Keywords. Minimal codes; Finite projective geometry; Blocking sets. MSC. 51E21, 94B27 1 Introduction Throughout the article, p denotes a prime number, q=p^{h} denotes a prime power, and the finite field containing q elements will be denoted as \mathbb{F}_{q} . The set of non-zero elements of \mathbb{F}_{q} is denoted as \mathbb{F}_{q}^{*} . A k -dimensional subspace C of \mathbb{F}_{q}^{n} is called a linear code with parameters [n,k]_{q} . Linear codes over finite fields play a central role in the theory of error-correction, and have applications in other branches of the mathematics of communication. Given a codeword c\in C , we define its support as the set \supp(c) of coordinate positions where c has a non-zero entry. The size of \supp(c) is called the (Hamming) weight of c . A codeword c\in C is called minimal if the only codewords w\in C with \supp(w)\subseteq\supp(c) are the scalar multiples of c . The concept of minimal codewords gained popularity after Massey [ Mas93 ] introduced a secret sharing scheme, constructed from a linear code C , where the minimal access structures correspond to the minimal codewords of C . It is in general a difficult problem to determine which codewords of a linear code are minimal. This motivated the study of linear codes in which all codewords are minimal, called minimal codes . Arguably, the central problem in the study of minimal codes is to study the shortest possible length n for which a minimal [n,k]_{q} code exists, in function of k and q . Let us denote this function as m(k,q) . A major breakthrough in the study of m(k,q) occurred quite recently by studying equivalent geometrical structures. We refer the reader to the recent survey by Scotti [ Sco24 ] for a more detailed account. Let \PG(k-1,q) denote the projective geometry consisting of the subspaces of the vector space \mathbb{F}_{q}^{k} . The projective dimension of a subspace of \PG(k-1,q) is one less than its vector space dimension. In particular, the points, lines, and planes of \PG(k-1,q) are the subspaces of \mathbb{F}_{q}^{k} of respective dimensions 1 , 2 , and 3 . The subspaces of \PG(k-1,q) of (projective) dimension k-2 are called hyperplanes . Definition 1 . A set B of points in \PG(k-1,q) is called a blocking set if each hyperplane of \PG(k-1,q) contains a point of B . If in addition each hyperplane is spanned by its intersection with B , we call B a strong blocking set . We refer the reader to [ BSS11 ] for a survey on blocking sets. We mention that the smallest blocking sets of \PG(k-1,q) have size q+1 and consist of the points on a line, see e.g. [ BSS11 , Theorem 2.1] . Strong blocking sets were first defined by Davydov, Giulieti, Marcugini, and Pambianco [ DGMP11 ] in the study of so-called covering codes. They were introduced into the study of minimal codes by Bonini and Borello [ BB21 ] (actually, they were reinvented in [ BB21 ] under the name cutting blocking sets). Alfarano, Borello and Neri [ ABN22 ] , and independently Tang, Qiu, Liao, and Zhou [ TQLZ21 ] proved an equivalence between strong blocking sets and minimal codes. This equivalence works as follows. A generator matrix of a linear [n,k]_{q} code C is a matrix G\in\mathbb{F}_{q}^{k\times n} whose row space equals C . We will assume that G has no zero columns, such that each column of G corresponds to a point of \PG(k-1,q) . This yields a (multi)set of n points in \PG(k-1,q) , which we call a projective system associated to C . Note that if G is one generator matrix of C , then its other generator matrices are MG with M\in\GL(k,q) . Therefore, the projective systems associated to C form an orbit under the action of \PGL(k,q) . Result 2 ( [ ABN22 , TQLZ21 ] ) . A linear [n,k]_{q} code is minimal if and only if its associated projective systems are strong blocking sets. Therefore, m(k,q) can equivalently be defined as the size of the smallest strong blocking set in \PG(k-1,q) . This interpretation allowed Alfarano, Borello, Neri, and Ravagnani [ ABNR22 , Theorem 2.14] to prove that m(k,q)\geqslant(k-1)(q+1) . On the other hand, Héger and Nagy [ HN22 , Theorem 4.1] proved that m(k,q)\leqslant(2-o_{q}(1))(k-1)(q+1) using a probabilistic construction. Therefore, we are interested in constructing families of strong blocking sets in \PG(k-1,q) whose size is along the lines of c(k-1)q for a reasonably small constant c . Several of the geometric constructions of strong blocking sets consist of taking the union of several small blocking sets. Recall that the lines constitute the smallest blocking sets. Sets of lines whose union yields a strong blocking set are said to be in higgledy-piggledy arrangement , and their study was initiated by Fancsali and the second author [ FS14 ] , actually predating the link with minimal codes. Subsequently, several papers have investigated small sets of lines in \PG(k-1,q) whose union yields a strong blocking set, see e.g. [ ABDN24 , BCMP22 , Den22 ] and the aforementioned result by Héger and Nagy [ HN22 ] . If q\geqslant\left\lfloor\frac{3}{2}(k-1)\right\rfloor , then a set of lines in higgledy-piggledy arrangement in \PG(k-1,q) contains at least \left\lfloor\frac{3}{2}(k-1)\right\rfloor lines [ FS14 ] . The corresponding strong blocking set therefore has size approximately at least \frac{3}{2}(k-1)q . It is hence natural to wonder whether we can lower the constant \frac{3}{2} by taking unions of other small blocking sets than lines. Definition 3 . Consider in \PG(k-1,q^{m}) the subspaces of \mathbb{F}_{q^{m}}^{k} having a basis of vectors of \mathbb{F}_{q}^{k} . These subspaces form a projective geometry isomorphic to \PG(k-1,q) . This set of subspaces and any image of this set under the action of \PGL(k,q^{m}) is called a subgeometry of \PG(k-1,q^{m}) (or an \mathbb{F}_{q} -subgeometry in case we want to emphasize which subfield of \mathbb{F}_{q^{m}} is used to construct the subgeometry). The points of an \mathbb{F}_{q} -subgeometry of \PG(k-1,q^{k-1}) are known to be a blocking set of size \frac{q^{k}-1}{q-1} , see e.g. [ BSS11 , Proposition 2.10] . The size is close to q^{k-1}+1 , the number of points on a line. Moreover, the points of \PG(k-1,q^{k-1}) can be partitioned into disjoint \mathbb{F}_{q} -subgeometries, see e.g. [ Hir98 , §4.3] . In this note, we will explore how to choose a small number of subgeometries, such that the union of their point sets constitutes a strong blocking set. In \PG(k-1,q^{k-1}) , one would need at