In this paper an l1-l1 minimization technique is applied on real Through-the-Wall imaging data. It allows adding side information into the compressed sensing minimization algorithm. This technique is evaluated and compared to the conventional CS approach and the common coherent background subtraction method