ChirpLab is a collection of MATLAB routines for the detection of highly oscillatory and nonstationary signals---also known as chirps---from very noisy data. The methods in ChirpLab might be especially suitable for the detection of gravitational waves in data from laser interferometric detectors such as LIGO. The methods are based on wave snippets called chirplets and network optimization algorithms that link together local empirical correlations into a coherent and meaningful event. To know more about this, check "Detecting highly oscillatory signals by chirplet path pursuit".

The contributors to ChirpLab are Emmanuel Candes, Hannes Helgason and Philip Charlton.

To download this software, please register to sign a free license agreement and obtain a password. For installation guidelines or to simply get an idea of what the software is about, please read the user's guide. For questions and feedback, please email .

Chirp Detection

"Detecting highly oscillatory signals by chirplet path pursuit."
by: Emmanuel Candes, Philip Charlton and Hannes Helgason.
Submitted, 2006.

This paper considers the problem of detecting nonstationary phenomena, and chirps in particular, from very noisy data. Given a set of noisy measurements, we would like to test whether there is signal or whether the data is just noise. We introduce detection strategies which are very sensitive and more flexible than existing feature detectors. The idea is to use structured algorithms which exploit information in the so-called chirplet graph to chain chirplets together adaptively as to form chirps with polygonal instantaneous frequency. We then search for the path in the graph which provides the best trade-off between complexity and goodness of fit. Underlying our methodology is the idea that while the signal may be extremely weak so that none of the individual empirical coefficients is statistically significant, one can still reliably detect by combining several coefficients into a coherent chain. This strategy is general and may be applied in many other detection problems.

Download (pdf)

Chirp Estimation

"Multiscale Chirplets and Near-Optimal Recovery of Chirps."
by: Emmanuel Candes
Technical Report, 2002.

This paper considers the problem of recovering chirp signals from noisy sampled measurements. The chirps are of the general form f(t)=A(t)cos(b p(t)) where b is a (large) base frequency, the phase function p(t) is time-varying and the amplitude A(t) is slowly varying. The methodology is adaptive in the sense that it does not require a-priori knowledge of the degree of smoothness the amplitude and phase and nearly attains the minimax risk over a range of chirp classes.

Download (pdf)

Links