I. Introduction
Inverse Synthetic Aperture Radar (ISAR) can image non co-operative targets of motion and has great significance in both military and civilian applications. It can get high range resolution by wide bandwidth signal pulse compression, and high cross-range resolution is achieved by coherent processing of the echo modulated by the relative rotation between the target and the radar. In the imaging process, it is usually assumed that the target is a rigid body [1]. After the target translation is compensated, the Doppler frequency difference of different cross-range position points is used to distinguish. ISAR imaging reflects the idea that the primary synthetic aperture radar uses the Doppler beam sharpening (DBS) to achieve angle high resolution. However, in many cases there are fast moving micro-motion parts on the target, such as propellers or turbo fans on the aircraft. Their high-speed fretting will generate additional frequency modulation, which will not only make parameter estimation and motion compensation difficult, but also interferes with the final imaging results [2]. Gardner R. E first proposed and studied the jet engine modulation (JEM) phenomenon [3]. Victor C. Chen introduced the micro-Doppler (m-D) concept for the first time in microwave radar, and verified its existence through simulation experiments [4].Several methods are proposed for the m-D signature extraction recently [5] –[9].