Optimal periodic sampling sequences for nearly-alias-free digital signal processing

Alias-free DSP (DASP) is a methodology of processing signals digitally inside bandwidths that are wider than the famous Nyquist limit of half of the sampling frequency. DASP is facilitated by suitable combination of nonuniform sampling and appropriate processing algorithms. In this paper we propose a new method of constructing sampling schemes for the needs of DASP. Unlike traditional approaches that rely on randomly selected sampling instants we use deterministic schemes. A method of optimizing such sequences aimed at minimization of aliasing is proposed. The approach is tested numerically in an experiment where an undersampled signal is processed using DASP; first to estimate the signal´s spectrum support function and then the spectrum itself. We demonstrate advantages of the proposed approach over those that use random sampling.