Core Concepts
Fractional decoding algorithms for algebraic geometry codes over extension fields using virtual projections and interleaved codes.
Abstract
The paper studies algebraic geometry codes from curves over the extension field Fqℓ through their virtual projections, which are algebraic geometric codes over the base field Fq. The virtual projections are used to provide fractional decoding algorithms for the codes over Fqℓ.
Key highlights:
Fractional decoding seeks to perform error correction using a smaller fraction of Fq-symbols than a typical decoding algorithm.
The virtual projection of an algebraic geometry code can be seen as an algebraic geometry code itself, with the degree of a pole divisor of an annihilator function affecting the bound on the number of correctable errors.
Viewing the virtual projections as interleaved codes allows, with high probability, correcting more errors than anticipated.
The fractional decoding approaches can make use of advances in traditional decoding algorithms for algebraic geometry codes.
The paper provides a general framework for fractional decoding of algebraic geometry codes over extension fields, with particular instances for important families of codes like those from Kummer extensions or Castle curves.
Stats
The number of correctable errors is bounded by:
(n - (deg G + (l - m) max {deg(pt)∞: t ∈[m]})) / 2