Abstract: In the target identification based on the radar high resolution range profile(HRRP), data extrapolation changes the amplitude of HRRP, so the recognition performance of HRRP is deteriorated. In this paper, we first present a new HRRP preprocessing method-data lengthening, which can reduce the aligned error of HRRP. Then, for the nearest feature line(NFL) classifier, samples are expanded in a local region according to the distribution character of HRRP, so the NFL can achieve a better performance with reduced computation complexity. Finally the two methods are combined for identification of HRRP. Experiments based on measured and simulated data demonstrate the efficiency of the presented algorithm.

Key words: HRRP, data extrapolation, data lengthening, nearest feature line method