Code Optimization for the Choi-Williams Algorithm for ELINT Applications

This thesis investigates optimizing the speed of computation for computing the Choi-Williams distribution. The Choi-Williams distribution is a way of simultaneously representing a signal in both the time and frequency domains in a fashion that makes it possible to extract the waveform parameters of the
signal. The Choi-Williams distribution is particularly useful for analyzing low probability of intercept signals for electronic intelligence applications. The usefulness of the distribution is directly correlated to the speed of computation. This thesis examines methods in which the Choi-Williams distribution can be modified to increase the speed of computation while still maintaining its ability to provide a clear picture of the signal characteristics. By eliminating the computation of near zero terms of the Choi-Williams kernel function, the speed of computation can be increased dramatically while still preserving, and improving, the timefrequency characteristics. The optimizations developed in this thesis reduced the time to compute a 512
sample CWD from 6.9 seconds, to 0.0466 seconds on an Intel chip, Linux based PC—an increase in speed of 147X.