A modified state reduction algorithm for computing weight enumerators for convolutional codes

Abstract

The weight enumerator of a convolutional code is an important function that characterizes the codeword distance distribution and allows the error probability bounds of the code to be conveniently computed. An efficient state reduction algorithm to compute weight enumerators by iteratively modifying the symbolic adjacency matrix associated with the code was recently introduced. In this paper, we propose a dynamic elimination ordering technique that exploits the state diagram structure of the convolutional encoder for improving the efficiency of the state reduction algorithm.