Coding, diversity and minimum distance

Low rate codes are sometimes used on channels that exhibit severe fading or interference. Such a low rate code may be obtained by designing a code with that rate and maximum minimum distance, or by using a higher rate code and diversity. The latter approach is usually suboptimal from a minimum distance point of view, but may have implementation advantages. This paper considers this trade-off between coding and diversity, in terms of minimum distance, for both block and convolutional codes. It is concluded that for many applications of convolutional codes, there is little performance loss in achieving the low rate code with diversity and coding. The same does not seem to hold true for block codes. This situation is quantified further in the paper.