Squircle-Based Calibration Algorithm for Six-Port Radar

Abstract

This article presents a new approach to calibrate six-port radar systems without the need for well-known target positions or calibration standards. Due to an extension of an ellipse-based calibration, the detectors’ nonlinearities can also be included into this one-step calibration. A so-called squircle is used to approximate measured $I/Q$ data of arbitrary and possibly unknown target positions. It is demonstrated with measurement data, that the implicit function of the squircle is suitable to model errors caused by power imbalance, crosstalk or parasitic reflections as well as detector nonlinearities. Furthermore, a method for mapping the impaired measurement data to an ideal circle is presented. This enables the accurate reconstruction of a target’s displacement.