Auto-calibration and Tomographic Reconstruction of Permittivity using a High Resolution Radar System.

This paper proposes an auto-calibration and tomography method to characterize low permittivity materials in a region of interest using time-of-flight measurements from a high resolution millimeter-wave radar system. In contrary to tomography methods with offline calibration, the proposed method uses the same set of time-of-flight measurements to perform simultaneous calibration and tomographic reconstruction. The proposed method uses the variable projection approach to separate and estimate the calibration parameters and tomography parameters independently from each other. Furthermore, the proposed method can also incorporate the prior information about the calibration parameters into the inherent least squares cost function. The incorporation of the prior information can limit the search space of the inherent least squares optimization. Thus, it helps the proposed method to efficiently find the optimum values for both the tomography and calibration parameters. The proposed method is evaluated using both the measurement data and simulation data. In measurement and simulation results, the auto-calibration and tomography method shows significantly better performance in terms of permittivity estimation accuracy compared to the least squares based approaches without calibration and online calibration and tomography approaches without constraints.