On two-weight codes (Q2659228): Difference between revisions
From MaRDI portal
Latest revision as of 20:29, 24 July 2024
scientific article
Language | Label | Description | Also known as |
---|---|---|---|
English | On two-weight codes |
scientific article |
Statements
On two-weight codes (English)
0 references
25 March 2021
0 references
The minimum Hamming distance \(d\) is a core parameter for error-correcting codes. In the study of such codes, one may be even more specific and not only require that all mutual distances between codewords be greater than or equal to \(d\), but that they belong to some given set \(D\) where \(d \in D\). Cases where \(|D|\) is small are of special interest. Codes with \(|D| = 1\) are called equidistant and have been studied extensively. The case of \(|D|=2\) has been considered for linear codes (two-weight codes), and that work is here extended to unrestricted codes -- that is, codes that are either linear or nonlinear. Let \(A_q(n,D)\) denote the maximum size of a \(q\)-ary unrestricted code with length \(n\) and mutual distances between codewords in the set \(D\). The authors require that all elements of \(D\) occur as distances (but do not follow this strictly in the tables). The reviewer would prefer not to have this requirement, to get \(A_q(n,D_1) \geq A_q(n,D_2)\) when \(D_2 \subseteq D_1\) and (for the classical function) \(A_q(n,d) = A_q(n,\{d,d+1,\ldots ,n\})\). The case of \(D = \{d,d+1\}\) has already been studied by the authors [Probl. Inf. Transm. 56, No. 1, 33--44 (2020; Zbl. 07214047)], so the focus is here on \(D = \{d,d+\delta\}\) for \(\delta \geq 2\). Upper and lower bounds are studied and tabulated for \(2 \leq q \leq 4\), \(2 \leq \delta \leq 6\) and small \(n\). The lower bounds come from randomly generated codes; many of those bounds could perhaps be improved with more advanced search methods. Some results on the existence of linear two-weight codes are also obtained.
0 references
bounds on codes
0 references
error-correcting codes
0 references
linear codes
0 references
linear programming bound
0 references
two-weight codes
0 references
unrestricted codes
0 references