Time-reversal (TR) was originated in acoustics as a technique for re-focusing waves around their source location. Under certain conditions, the wave equation is invariant under TR, therefore, waves emanated from a source or scattered from a passive target, and recorded by a transceivers array, will retrace their forward path and automatically focus at the source/target location if back propagated in a time-reversed (last-in first-out) fashion from that array. Focusing resolution of time-reversed back propagation in rich scattering environments beats that in free space, yielding what is known as `superresolution’. Moreover, under ultrawideband (UWB) operation, TR exhibits the distinctive property of `statistical stability’, which makes it an attractive technique for imaging in disordered media whose characteristics are not known deterministically (random media). Over the past few years, TR has been exploited in a variety of electromagnetic sensing and imaging applications such as ground penetrating radar, breast cancer detection, nondestructive testing, and through-wall imaging. In addition, TR has been extensively applied in UWB wireless communication providing myriad of advantages including reduced receiver complexity, power saving, increased system capacity, and enhanced information secrecy.
In this work, we introduce new TR-based signal processing techniques for imaging, tracking, and communicating with targets/users embedded in rich scattering environments. We start by demonstrating, both numerically and experimentally, the statistical stability of UWB TR imaging in inhomogeneous 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. Then, we devise a technique for detecting and tracking multiple moving targets in cluttered environments based on differential TR. This technique provides real-time tracking and exhibits superior clutter rejection at minimal processing costs. It also exploits the distinctive features of time-reversal such as statistical stability and superresolution. Next, we develop an UWB inverse scattering technique for extended targets, with continuous permittivity and/or conductivity fluctuations, based on Bayesian compressive sensing. Bayesian inversion provides means for estimating the confidence level of the inversion, and for adaptively optimizing subsequent measurements. This technique is applied to a wide range of problems of practical interest such as underground crosshole sensing, medical imaging, and rough surface reconstruction. Finally, we develop new TR-based wireless communication techniques for UWB multiple-input single-output (MISO) and multiple-input multiple-output (MIMO) systems. We contrast relative strengths and limitations of those techniques for different scenarios of operation.