Radar signature prediction using moment method codes via frequency extrapolation technique

A frequency extrapolation scheme is developed to efficiently predict radar signatures using moment method codes. The approach is to parameterize the induced current on the target based on a multipath excitation model via the ESPRIT superresolution algorithm. The multiple scattering mechanisms at high frequencies are included in the model to ensure the accuracy of the algorithm. The range profiles for several test targets are calculated to demonstrate the performance of the algorithm. A numerical experiment is conducted to find the possible error sources of the algorithm and determine an upper limit on the frequency extrapolation bandwidth. The algorithm is also extended to generate ISAR images by using the frequency extrapolation in conjunction with the bistatic approximation in the angular dimension. The ISAR image of the benchmark VFY-218 airplane is predicted at 400 MHz and compared against the chamber measurement result. The main scattering features in the measured image are successfully predicted using this approach with very modest computational resources.