The Hough transform for long chirp detection

The online detection of a very long and weak chirp signal is, studied. The signal has an extremely slowly-decreasing frequency and is corrupted by white Gaussian noise, and also possibly by powerful tones. Four methods (the Hough transform, multiple frequency tracker, Page's test and EM algorithm) are explored. It is found that the Hough transform (HT) detector appears to be most suitable given constraints on computational load and detectability. It is compared with the GLRT, which is assumed to be as "optimal" as possible. Applying a suitable threshold for the HT can increase the speed dramatically while preserving the performance. We have found that for the HT detector both dithering (taking varied frequency shifts for FFTs) and increasing the FFT length can reduce the minimum detectable frequency slope with nearly no additional computation.