Statistical Stability of Ultrawideband Time-Reversal Imaging in Random Media

We carry out a study on the statistical stability of ultrawideband (UWB) time-reversal (TR) imaging in random media under different combinations of random medium parameters and interrogating signal properties. We examine conditions under which frequency decorrelation in random media provides a more effective "self-averaging" and, therefore, better statistical stability. We also present a new frequency-synthesized technique for UWB TR-based imaging. This technique is employed to construct TR-operator-decomposition and multiple-signal-classification images using either linear or full-aspect transceiver array configurations. The proposed technique automatically provides the best images of desired target(s) in terms of focusing resolution without the need for (synthetic) propagation of TR signals and ad hoc determination of the optimal focusing time instant. In addition, information about the imaging domain (background) can be stored and reused to reconstruct different targets.