Evaluation codes from order domain theory (Q2469475): Difference between revisions

In this paper the authors give a lower bound on the minimum distance of codes defined by means of their generator matrices. This new bound is closely related to the well-known Feng-Rao bound, [\textit{G.-L. Feng, T. R. N. Rao}, IEEE Trans. Inform. Theory 41, No. 6, Pt. 1, 1678--1693 (1995; Zbl 0866.94025)], this last for codes defined by means of their parity check matrices. As the Feng-Rao bound the new bound easily applies to the context of order domains, algebraic structures introduced by \textit{T. Høholdt}, \textit{J. H. van Lint} and \textit{R. Pellikaan} [ in: V. S. Pless (ed.) et al., Handbook of coding theory. Vol. 1, 871--961 (1998; Zbl 0922.94015)], which allow to deal with one-point algebraic geometry codes and their duals. In Section 2, to motivate the ideas that will be used later, the authors give a new (``nontraditional'') proof of the character MDS of Reed-Solomon codes. The proposed bound is described in Section 3 for general linear codes using the set-up of \textit{R. Matsumoto} and \textit{S. Miura} [IEICE Trans. Fundam. Electron. Commun. Comput. Sci. E 83-A (5), 923--927 (2000)], which only uses the concepts of linear algebra and the componentwise product of vectors. Theorem 8 proves the new bound for the minimum distance \(d\)\, of the code \(C(B,G)\)\, spanned by \(G\),\, subset of a base \(B\)\, of \(F_q^n\),\, while theorem 10 extends it to any generalized Hamming weight \(d_t\)\, of \(C(B,G)\),\, \(t=1, \dots, k\). Given \(\delta\leq n\)\, it is then possible to choose \(G\)\, such that \(C(B,G)\)\, has minimum distance at least \(\delta\)\, and as large dimension as possible (theorem 13). Section 4 and 5 show connections of the proposed bound to the Feng-Rao bound and to the Shibuya and Sakaniwa bound [\textit{T. Shibuya} and \textit{K. Sakaniwa}, IEICE Trans. Fundam. Electron. Commun. Comput. Sci. E84-A (2), 647--652 (2001)], respectively. In particular the authors show how the Shibuya and Sakaniwa bound can be derived from their theorem 10 (proposition 19). Section 6 translates the results of previous Sections into the set-up of order domains and Section 7 deals with one-point algebraic geometry codes. In this case the authors show (proposition 37) that the new bound is an improvement to the usual bound from algebraic geometry. Section 8 applies the machinery of Gröbner basis to the implementation of the method in finitely generated order domains, Section 9 shows in detail four concrete examples and in the Conclusions (Section 10) the authors point out some open problems.