Implementation and error performance evaluation of an iterative decoding algorithm

Abstract

This paper presents new experimental results about the error correction performance of an iterative threshold decoder at relatively high signal to noise ratio. To accomplish this task, an accelerated characterization platform has been developed. Without this platform, it would take approximately 37.23 years in computational time with a software version of the decoder to get the error correction performances over an extended signal to noise range. An acceleration factor of 4812 is obtained by using the accelerated characterization platform. The platform constitutes a new way for characterizing quickly and efficiently a new error correction algorithm. The acceleration characterization platform has allowed verifying the error correction performance at high SNR which will require a prohibitive computational time on a conventional computer.