Super-resolved time-frequency processing of wideband backscattered data

The time-frequency processing of wideband backscattered data from complex targets has recently been shown to be a powerful tool in analyzing and interpreting the scattering physics of targets. The basic concept behind the time-frequency processing is to process the data one small chunk at a time through the use of a sliding window function. In the two-dimensional time-frequency plane, scattering centers, natural resonances and dispersive phenomena can be simultaneously displayed to provide additional insights into the scattering mechanisms not available in either the time or the frequency domain. Unfortunately, the additional insights gained by the time-frequency display always come at a price in loss of resolution. The processing of the data within each time or frequency window using super-resolution techniques such as Prony's or ESPRIT algorithms appears quite attractive. It retains the advantage of simultaneous time-frequency display while overcoming the resolution issue. In this paper, we describe a super-resolution procedure for achieving parameter estimation of both scattering centers and natural resonances in the time-frequency plane.