Coding Theory and Algebraic Geometry: an Interplay



One in a series of snapshots of applications of discrete mathematics to other parts of mathematics.



J. W. P. Hirschfeld
University of Sussex
United Kingdom

The Context

In 1981, Goppa derived a class of linear codes from algebraic curves over finite fields which (a) are quite general as codes, (b) have parameters circumscribed by the Riemann-Roch theorem, and (c) have asymptotic properties which improve the classical Gilbert-Varshamov bound. The discovery of these codes also gave renewed stimulus to investigations on the number of rational points on an algebraic curve for a particular genus as well as to asymptotic values of the ratio of the number of points to the genus. Manin showed that for small fields coding theory methods give better results for this ratio than the Hasse-Weil theorem.

The Problem

The three most important parameters of a linear code over the finite field are the length n which gives the speed of transmission, the dimension k which gives the number of words in the code, and the minimum distance d which gives the number of errors that can be corrected. ``Good'' codes have large information rate and relative distance . The relation between R and as n gets large is given by the Gilbert-Varshamov bound. Good codes can be constructed from an algebraic curve of genus g, and particular examples show that the G-V bound is not best possible. This brings to the fore the problem of determining the limit of for a sequence of curves. Here coding bounds give better values than methods of algebraic geometry.

The Construction

Let be an algebraic curve defined over . Let be an ordered set of rational point points of . Let with for , be the associated divisor. Let with be an -divisor such that for all i,j. Let be the space of functions associated to E; that is, consists of functions which have no pole of order greater than at for all j. Then the evaluation map at

is given by

and its image

is an algebraic geometry code.

The Hasse-Weil theorem says that and it is of considerable interest to find curves satisfying the upper bound. Although it follows from this bound that , it was shown by Drinfeld and Vladut that ; it is also known that equality holds for square q.

Further Research

What still has to be done is

  1. to investigate the behaviour of these codes on different types of curves (the codes are well understood for curves of genus 0 and 1);
  2. to find curves with large numbers of points;
  3. to find the value of for different sequences of curves and, in particular, the value of ;
  4. most importantly, to find more efficient decoding procedures.

For More Information

The IEEE Transactions in Information Theory is the journal where most articles on this topic can be found.

  1. J.W.P. Hirschfeld, Review of `` Algebraic Curves over Finite Fields by O. Moreno, Cambridge University Press, Cambridge, 1991'', in Bull. Amer. Math. Soc. 27 (1992), 327-332.
  2. H. Stichtenoth, Algebraic Function Fields and Codes, Springer-Verlag, Berlin, 1993.
  3. M.A. Tsfasman and S.G. Vladut, Algebraic-Geometric Codes, Kluwer, Dordrecht, 1991.

    Thu Sep 28 09:58:04 CDT 1995