Alleviating the Practical Restrictions for MUSIC Algorithm in Actual Microwave Imaging Systems: Experimental Assessment.

Multiple signal classification (MUSIC) method uses the multi-static response matrix (MSR) to locate small inclusions in a microwave imaging (MI) system. The MSR cannot accommodate all available data in an actual MI system because practical restrictions make some receivers be inactive for every active transmitter. The current work presents a method for completing the MSR acquired in practical systems. Using the least squares method, the complete MSR matrix is formed by the estimation of its unknown elements, associated to the signal at inactive receivers. The estimation algorithm employs the signal of other active receivers to obtain inactive receiver signals. In this way, all available data contributes in imaging, which improves the localization performance. The merits of our proposal are demonstrated using the Institut Fresnel experimental data sets as well as various realistic synthetic data. [ABSTRACT FROM PUBLISHER]