Convolutional coding for bandlimited channels

Abstract

The application of convolutional codes for finite intersymbol interference (ISI) channels is considered. The performance of a bandlimited communication system in the presence of additive white Gaussian noise depends on the minimum Euclidean distance dmin between different signal sequences. To further increase dmin, convolutional or trellis coding could be used. Search results indicate that the encoder with the best free distance does not always give the best coded distance when combined with an ISI channel. It has been seen from the results for several ISI channels that moderate coding gains can be obtained if the channel and the encoder are properly matched. Furthermore, it has been observed that if the channel is a maximum distance one, then the resulting coded distance is usually higher than that for a non-maximum distance channel.