Quasioptimal Methods of Correlational Reception of Reed–Solomon Codes

The article poses the problem of developing quasi-optimal methods of decoding Reed–Solomon and Forney codes that almost realize their potential properties in operation in a Gaussian channel, but that are less complicated than parallel correlational reception. The following methods are investigated: inspection and comparison in Euclidean metric of $C_n^{d-1}$ erasures with $d-1$ symbols in each erasure; a new “sliding subgraph” method, and three versions of list subsets of the Viterbi algorithm. Analytic estimates are given for the complexity and noise stability of the first two methods, and results of statistical simulation of subsets of the Viterbi algorithm are given.