Efficiently List-Decodable Punctured Reed-Muller Codes

Venkatesan Guruswami,Lingfei Jin,Chaoping Xing

DOI: https://doi.org/10.1109/tit.2017.2692212

IF: 2.5

2017-01-01

IEEE Transactions on Information Theory

Abstract:The Reed–Muller (RM) code, encoding $n$ -variate degree- $d$ polynomials over $ \mathbb {F}_{q}$ for $d < q$ , with its evaluation on $ \mathbb {F}_{q}^{n}$ , has a relative distance $1-d/q$ and can be list decoded from a $1-O(\sqrt {d/q})$ fraction of errors. In this paper, for $d \ll q$ , we give a length-efficient puncturing of such codes, which (almost) retains the distance and list decodability properties of the RM code, but has a much better rate. Specifically, when $q =\Omega ( d^{2}/ \varepsilon ^{2})$ , we give an explicit rate $\Omega \left ({\frac { \varepsilon }{d!}}\right )$ puncturing of RM codes, which have a relative distance at least $(1- \varepsilon )$ and efficient list decoding up to $(1-\sqrt { \varepsilon })$ error fraction. This almost matches the performance of random puncturings, which work with the weaker field size requirement $q= \Omega ( d/ \varepsilon ^{2})$ . We can also improve the field size requirement to the optimal (up to constant factors) $q =\Omega ( d/ \varepsilon )$ , at the expense of a worse list decoding radius of $1- \varepsilon ^{1/3}$ and rate $\Omega \left ({\frac { \varepsilon ^{2}}{d!}}\right )$ . The first of the above tradeoffs is obtained by substituting for the variables functions with carefully chosen pole orders from an algebraic function field; this leads to a puncturing for which the RM code is a subcode of a certain algebraic-geometric code (which is known to be efficiently list decodable). The second tradeoff is obtained by concatenating this construction with a Reed–Solomon-based multiplication friendly pair, and using the list recovery property of algebraic-geometric codes.