Signature Extraction From Rotating Targets Based on a Fraction of HRRPs

Motion or structure signatures, extracted from radar echoes, are of great potential in target identification. Traditional techniques for signature extraction rely primarily on the assumption that the time series of the signal contains at least one oscillation or more during the illumination time. However, as many applications involve short duration signals or long duration signals, they are only partially useful in real-world scenarios. The conventional signature extraction algorithms may suffer from degraded precision. This paper presents a three-point model for signature extraction availing the characteristics of the high resolution range profiles (HRRPs) of rotating targets with stepped-frequency signal (SFS). The frequency and the length of rotating targets can be estimated accurately from a fraction of the HRRPs in slow-time-range plane. The accuracy and efficiency of the estimation are demonstrated by simulated and experimental trials.