To bite or not to bite-a study of tail bits versus tail-biting

Abstract

Tail bits and tail-biting are the two commonly used methods for terminating the trellis of convolutional codes. In this paper, we evaluate the relative performance of these two options by comparing their performance in a variety of cases. Suboptimal decoding algorithms for the tail-biting case are proposed. Puncturing is proposed as an alternative method to prevent the rate loss when tail bits are used. The performance of several puncturing schemes are evaluated. In addition, we explore the bit error rate of these schemes as a function of position of bits in a block.