Adaptive blind channel identification of FIR channels for Viterbi decoding

Relying upon the diversity introduced by fractional-sampling or by multiple sensors, two different approaches for blind adaptive channel identification are proposed based on existing batch algorithms. We show that the adaptive approach allows tracking of time-variations in the channel and provides robustness to nearly common channel roots. Existing second-order adaptive procedures typically employ suboptimal linear equalizers whereas the adaptive channel estimators we propose can be used in an optimal maximum likelihood sequence estimator. Simulations are provided that demonstrate the performance of the proposed channel estimators when used in the Viterbi algorithm for symbol detection.