Fast implementation of adaptive Chirplet decomposition

The adaptive Chirplet decomposition contains a lot of unknown parameters and has much difficulty in its implementation. In this conditions fast implementation of adaptive Chirplet decomposition is presented in this paper. By computing the quadratic phase (QP) function of the signal, it is shown that the QP function is concentrated along the frequency rate curve of the signal, and the peak of the QP function yields the estimates of the Chirplet rate, time center and the amplitude; Then the initial frequency and the time extent can be obtained by the dechirp technique. The scheme of the algorithm is shown in this paper, and the results of simulated data and echolocation signal from brown bat demonstrate the validity of the method proposed.