Statistical analysis of linear filters driven by non-white noise

The problem of the statistical analysis of the multidimensional linear filters is considered on the basis of kinetic equations for the non-Markovian stochastic processes. The kinetic coefficients are calculated using the convolution integral instead of the system of differential equations. The advantage of this method is a reduction of the dimension of the problem. Two examples are given and some results of a statistical simulation are presented when the input is the binary Markovian process.