An Improved ESPRIT Algorithm for Scaling Extrapolation

The traditional scale measurement method is strictly based on the scale transformation where the electrical dimension remains unchanged. However, it is difficult to implement due to the size of the darkroom quiet zone and the test frequency range. There is a scaling extrapolation method based on the scattering center model. Firstly, the scattering center parameters of the scale target are estimated by parameter estimation algorithms, then the parameters of the full-scale target are derived by parameter relation of the full-size and the scale target. Finally, the extrapolated RCS of the full-scale target is obtained by the scattering center model. However, this method is limited by the accuracy of scattering center estimation. In this paper, we propose an improved ESPRIT algorithm, which is suitable for use in extrapolation methods and effectively reduces the error of the extrapolation method.