Calculation of the spectral density from the noise convolution integral covariance matrix

A problem which arises in system identification of continuous systems with nonuniform sampling is that of calculating the spectral density of a vector white noise process from a noise convolution integral covariance matrix. A numerical solution to this problem is presented which converges in one step.