Through-the-wall radar imaging for heterogeneous walls using compressive sensing

Front wall reflections are considered one of the main challenges in sensing through walls using radar. This is especially true under sparse time-space or frequency-space sampling of radar returns which may be required for fast and efficient data acquisition. Unlike homogeneous walls, heterogeneous walls have frequency and space varying characteristics which violate the smooth surface assumption and cause significant residuals under commonly used wall clutter mitigation techniques. In the proposed approach, the phase shift and the amplitude of the wall reflections are estimated from the compressive measurements using a Maximum Likelihood Estimation (MLE) procedure. The estimated parameters are used to model electromagnetic (EM) wall returns, which are subsequently subtracted from the total radar returns, rendering wall-reduced and wall-free signals. Simulation results are provided, demonstrating the effectiveness of the proposed technique and showing its superiority over existing methods.