Model Based Synchronization for Weak Signal Detection

In this paper we present the detection/estimation of weak signals in the presence of unknown chaotic clutter using model based chaotic synchronization. We achieve this task in three steps: (i) First, using Taken's delay embedding theorem and support vector machine (SVM) a model is built from a set of training data, (ii) then a coupled chaotic synchronization scheme is applied to this model to get an accurate estimate of the clutter, and (iii) finally this estimate is subtracted from the observations to get a residue signal on which we apply standard signal detection/estimation techniques. The efficiency of the new estimator is evaluated by computing the mean square error (MSE) of the estimation. We show that the performance of the proposed method is superior to the conventional scheme.