Reconstruction Convergence of an Impulse-doublet Sampled Random Processes

We show in this paper that first order aliased spectra, from sampling a band limited low pass random process at an average Nyquist rate, may be eliminated. Our approach is to use two arbitrarily closely spaced samples each sampling instant. A convergence proof is given for the random waveforms. Matlab simulations demonstrate reconstruction of a random waveform with low pass characteristics.