A comparison of five methods for computing the power spectrum of a random process using data segmentation

1 June 1977

journal article
Published by Institute of Electrical and Electronics Engineers (IEEE) in Proceedings of the IEEE

Vol. 65 (6), 984-986
https://doi.org/10.1109/PROC.1977.10600

Abstract

When we compute the power spectrum of a long input vector with limited core memory it is necessary to segment the data. Five alternative methods are possible. This paper presents a number of computed results comparing the resolution, stability, and overall quality of these methods.

Keywords

RANDOM PROCESSES
AUTOCORRELATION
FOURIER TRANSFORMS
RANDOM NUMBER GENERATION
STABILITY
FAST FOURIER TRANSFORMS
JOINING PROCESSES
WHITE NOISE
FILTERING
INFORMATION SCIENCE

Cited by 7 articles