Signal Interpolator Design Using Weighted-Least-Squares Method

Authors

  • Noboru Ito and Wei Qin Author

Keywords:

Signal processing, signal reconstruction, sinc function, re-sampling, signal interpolation, interpolator design.

Abstract

This paper shows that the bandwidth of a cubic 
interpolator can be designed by using a weighted least-squares 
(WLS) method in the frequency-domain. We first show how to 
construct a length-6 cubic whose support is 6 by connecting 
three piecewise polynomials of the third degree, and then 
explain how to find its optimal coefficients through minimizing 
the weighted squared error between the ideal and actual 
frequency responses of the length-6 cubic. By adjusting the 
weighting functions on different frequency bands, we can easily 
control the bandwidth of interest. In practical applications, 
according to the distribution of the signal to be interpolated, we 
can adjust the weighting function in the interpolator design and 
thus obtain very accurate frequency bands. As a consequence, 
the original band-limited analog signal can be accurately 
reconstructed, which also leads to high-accuracy signal 
interpolation. A narrow-band design example is given to 
illustrate that the length-6 cubic can achieve much higher 
accuracy frequency response than other existing interpolators. 
Moreover, we also detail how to choose the weighting function 
including ``don't care'' frequency bands.  

Downloads

Download data is not yet available.

Downloads

Published

15.05.2013

How to Cite

Signal Interpolator Design Using Weighted-Least-Squares Method . (2013). International Journal of Information and Electronics Engineering, 3(3), 299-303. https://www.ijiee.org/index.php/ijiee/article/view/685