Are blind equalizers assuming white data applicable to block encoded sequences?

Mannerkoski, J., Koivunen, V.

Are blind equalizers assuming white data applicable to block encoded sequences?
Several blind channel equalization/identification algorithms [Papadias 99, Giannakis 97, Moulines 95] assume the transmitted information sequence to be white. In real communication systems, redundancy is added to the source sequence in order to be able to detect and correct symbol errors in the receiver. It is not obvious how channel encoding will affect the assumption of whiteness. We analyze the autocorrelation sequence of linear block codes. The codes are presented in terms of a Markov model, for which the autocorrelation can be analytically obtained. The encoded data is used in the prediction-based blind equalizer of [Papadias99] and the performance is empirically compared to the case of unencoded data.