ESPRIT-based feature extraction of helicopter acoustic signal

The helicopter acoustic signal is a narrow-band signal with harmonic characteristics, so the harmonic frequencies of the signal are considered as the features for helicopter identification. The algorithm of the estimation of signal parameters via rotational invariance techniques (ESPRIT) based on singular value decomposition (SVD) and total least square (TLS) was used to extract harmonic frequencies. The simulation results for several methods have shown that the ESPRIT can correctly extract harmonic frequencies under the condition of lower signal to noise ratio. The ESPRIT was applied to process data of limited length under the various flight conditions, and the results have indicated that the algorithm is effective.