Vector Radix FFT Error Analysis

In this correspondence expressions are derived for the multiplication roundoff and scaling error variances in a multidimensional vector radix FFT algorithm, for fixed-point one's complement/signmagnitude arithmetic. Results are given for base-2 and base-4 algorithms for both decimation-in-time and decimation-in-frequency methods. Comparisons of the error performance are made with the conventional methods. Copyright © 1982 by The Institute of Electrical and Electronics Engineers, Inc.