Frequency-unknown matched field processing for frequency estimation and localization

Matched field processing (MFP) requires prior knowledge of the frequency for the source, which is often unknown, and must be estimated from the received signals. Extending conventional MFP, incoherent frequency-unknown matched field processing (FU-MFP) is presented by incorporating the inversion of transmission frequency and radial velocity. Motion compensation of the Doppler frequency shift is achieved through the inversion of the radial velocity, thereby extending the Fourier transform window. Coherent FU-MFP is studied by combining passive synthetic aperture processing, which provides robust estimation of frequency and radial velocity at low signal-to-noise ratio (SNR). The joint processing gain from synthetic aperture processing and motion compensation enables coherent FU-MFP to achieve localization by weak line spectrum signals, which are indistinguishable in low-frequency analysis and recording spectra or power spectra. Simulations and experimental results confirm the ability of FU-MFP to estimate source frequency and location in low SNR, significantly outperforming conventional MFP.