The Asymptotic Uniformity of the Output of Convolutional Codes Under Markov Inputs

Abstract

In this letter, we prove a published conjecture on the asymptotic uniformity of the outputs of a convolutional encoder under biased inputs. These results are interesting in light of recent research on joint source-channel coding as well as source coding using turbo codes in which the constituent encoders are convolutional codes. In particular, it is well-known that in many situations a good code should result in a uniform distribution on blocks of consecutive encoded symbols. The results presented here provide insights into the choice of encoders in such scenarios.