Linear processing of nonuniformly sampled signals

The authors describe a novel approach, based on a coordinate transformation, that can process nonuniformly sampled signals efficiently, in the sense that the number of computations for processing the nonuniformly spaced data will be the same as in the uniform sampling case for the same number of samples in both cases. In situations where the nonuniform sampling results in fewer samples compared to the uniform sampling case, a reduction in total computations may be realized. Numerical examples are examined which confirm that the technique is valid.<
>