On the adaptive interpolation of discrete-time signals

Abstract In this paper, the problem of adaptive interpolation of discrete-time signals that could be modeled as autoregressive processes is studied. The interpolation of discrete-time signals has received considerable attention recently and a variety of interpolating techniques have been developed with varying degrees of success. A suboptimal iterative approach for the interpolation problem is developed and it is shown by simulation studies that this technique converges quite fast within a few iterations, and is computationally very attractive as it does not involve matrix inversion during interpolation. The effect of signal bandwidth and the center frequency of the pole on the accuracy of interpolation is also examined and some theoretical results for first and second order AR processes are derived. These theoretical results are verified using computer simulations.