Improving the Calibration by Power Balancing in Truncated Singular Value Decomposition for Wide-Band Scattering Parameter Data

Abstract

The fi eld of microwave and millimeter-wave imaging systems is experiencing signi fi cant growth in research, primarily attributed to their versatile interaction capabilities with various materials. In the course of image reconstruction, the inverse scattering problem is resolved through either qualitative or quantitative means. The inverse scattering problem can be resolved by utilizing the truncated singular value decomposition (TSVD) algorithm, which is a suitable inversion method that can be employed when there is a need to discard a portion of data. The utilization of multi-frequency multi-static inversion techniques yields results that exhibit greater robustness and clarity when contrasted with single-frequency mono/bi-static con fi gurations. However, it should be noted that the TSVD formulation may not be su ffi cient in practical experimental situations, as it fails to account for potential deviations in power among excitations at varying frequencies. As a result, it is imperative to perform a calibration process to standardize the power variations across di ff erent frequencies. The proposal outlines a calibration methodology that aims to achieve power equilibrium across various frequencies and facilitate the automation of imaging algorithms. The proposed calibration technique is evaluated and demonstrated through the use of a through-the-wall experiment.