Maximum Likelihood BSC Parameter Estimation for the Slepian-Wolf Problem

In the context of Distributed Source Coding, we propose a low complexity algorithm for the estimation of the cross-over probability p of the Binary Symmetric Channel (BSC) modeling the correlation between two binary sources. The coding is done with linear block codes. We propose a novel method to estimate p prior to decoding and show that it is the Maximum Likelihood estimator of p with respect to the syndromes of the correlated sources. The method can be utilized for the parameter estimation for channel coding of binary sources over the BSC.