SHAPE PARAMETER-ESTIMATION OF WAVE-FRONTS WITH KNOWN WAVE-FORM

Active sensor arrays are useful to parametrically estimate the wavefro nts in most applications where the excitation signal (waveform) is kno wn or can be consistently estimated (e.g., in remote sensing, radar an d communication systems). For the estimation of the shape parameters ( SPs) of the wavefronts several algorithms have been summarized from th e literature and compared, The algorithm proposed here is based on sin gular value decomposition (SVD) and has been proven to be effective in reducing coherent noise arising from interfering wavefronts. The prob lem of SP estimation is treated as a whole for plane and curved wavefr onts, for different geometries (1-D and 2-D array of sensors), and for wide and narrowband waveforms. Wavefield distortion due to propagatin g medium inhomogeneities is analyzed for increasing distortion. Compar ed with the methods adapted for SP estimation of narrowband wavefront, the maximum likelihood estimator (ML) and the method based on SVD sho w better resolution of the wavefront arrival time since they correlate with the known waveform. The Cramer-Rao bound is derived and some use ful properties are also established. Finally the paper contains a nume rical study of the SP estimators for incoherent and coherent noise. (C ) 1998 Elsevier Science B.V. All rights reserved.