Target Recognition with Missing Stepped Frequency Backscatter

摘要

This paper examines target recognition using stepped frequency radars when there are missing data. The Gerchberg-Papoulis algorithm is used to interpolate the missing data and target recognition is then performed using features extracted from the reconstructed data. Missing section or sections of the stepped frequency backscatter could be due to system failure, a jammer, or could be intentional to expedite the target interrogation process or reduce the recorded data. The interpolation algorithm is first tested on a synthetic target, then implemented using the returns of real commercial aircraft models. The stepped frequency data on scale model commercial aircraft was recorded in a compact range at high SNR. Scenarios where the target azimuth is completely or partially known are examined.