Asymptotic Properties of a One-bit Estimator of Parametric Signals

This paper considers an estimator of the parameters in a linear-in-the-parameter signal, based on its binary quantization. It shows an analysis of the systematic errors for growing lengths of the data set. It also shows how to express the cost function, when the minimum is reached. Zero-mean additive noise is assumed in the model, having independent outcomes. Furthermore, a method is introduced to validate the estimation process. Monte Carlo results validate the proposed analyses.