An Explicit Construction of <tex>$q\text{-ary}$</tex> MDS Array Codes and Their Efficient Decoding

Jingjie Lv,Weijun Fang,Shu-Tao Xia,Hanxu Hou
DOI: https://doi.org/10.1109/ISIT57864.2024.10619119
2024-01-01
Abstract:In this short work, a new explicit construction of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$q-\mathbf{ary}$</tex> MDS array codes with multiple parities will be provided, whose code lengths can be up to <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$q^{m-1}$</tex> , where <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$m-1$</tex> is the size of subpackage. As far as we know, this may be the first explicit construction of practical MDS array codes with such long code lengths for general <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$q$</tex> . In addition, to demonstrate the applicability of our MDS array codes, by the LU factorization of Vandermonde matrices, we present an efficient decoding method aimed at the erased errors, whose computational complexity is <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$O(m^{2})$</tex> in total. Furthermore, if one stores a small number of polynomials in advance or computes the syndrome in a scheduled algorithm, the decoding efficiency of these MDS array codes can be further improved.
What problem does this paper attempt to address?