Title :
A method of using Simpson´s rule in the DFT
Author_Institution :
Lakehead University, Thunder Bay, Ont., Canada
fDate :
8/1/1981 12:00:00 AM
Abstract :
The method of taking a 2N-point DFT by means of an N-point DFT is modified such that Simpson´s rule replaces the trapezoidal rule in the approximation to the Fourier integral. The Simpson´s rule version is compared to the trapezoidal rule version on the basis of equal upper bounds on truncation error, and it is shown that Simpson´s rule has the advantage of requiring a lower sampling rate under these conditions.
Keywords :
Discrete Fourier transforms; Finite wordlength effects; Frequency; Integral equations; Lakes; Sampling methods; Uncertainty; Upper bound; Yield estimation;
Journal_Title :
Acoustics, Speech and Signal Processing, IEEE Transactions on
DOI :
10.1109/TASSP.1981.1163629