An Iterative Interpolation Technique Using Finite Impulse Response Digital Filters.
Abstract
Interpolation operations have wide applicatons in signal processing systems. An efficient interpolation technique is developed and demonstrated. This is an iterative technique in which interpolation output at any stage is combined with the input to that stage so that the interleaved result forms the input to the succeeding stage. Sampling rate increases of high order are not difficult to achieve. Resampling of high-order interpolator outputs permits precise control of sampling rate conversion ratios. From this basic work in interpolation, a digital filter synthesis procedure is developed. The resulting class of finite impulse response filters is competitive in speed and storage with conventional designs. These filters have the property that the passband frequency function of a generic design can be scaled by any desired factor in frequency while maintainng stopband suppression to specified levels. This synthesis procedure uses a 'stretch-and-fill' iteration operation. The impulse response into a synthesis stage is stretched, and the vacated spaces are then filled from new data synthesized at that stage. The 'stretch' performs the frequency scaling function while the 'fill' operation eliminates image responses that would otherwise appear. Computation of the coefficients for these filters, of any order, is trivial. A FORTRAN computer program which uses the developed algorithm is provided for general interpolation service. Auxiliary use of this program for obtaining derived filter coefficients is described, and examples of filter synthesis are given. Special entry points are provided for data format mode conversion: complex-to-real and real-to-complex. Equivalent procedures for sampling rate reduction are also presented and discussed.
Document Details
- Document Type
- Technical Report
- Publication Date
- Feb 01, 1979
- Accession Number
- ADA067511
Entities
People
- J. P. Costas
Organizations
- General Electric