Recognition Of Midcourse Targets Based On Range Profile Matching

Interception of ballistic missile has important significance for national defense. Ballistic movement is divided into three segments: the boost segment, the midcourse segment, the terminal segment. Since the midcourse segment has a long period and rich features, it has become the main target stage in ballistic missile interception. To break through interception, ballistic missile usually carries baits and jamming devices. In the midcourse segment, with the separation of warhead and missile, there will be lots of fragments. These baits and fragments will form a huge threat targets group to follow the movement of warheads. Thus, to intercept the ballistic missile, warhead must be recognized first from the huge targets group. This paper focuses on the rough classification of warhead, fragments and bait ball.In this paper, range profiles of the midcourse segment targets are analyzed. With the continuous development of broadband signal theory, high-resolution radars are used more and more, which makes the target range profile resolution improved. High resolution range profile(HRRP) reflects the characteristics of the target structure in the direction of the radar line of sight, so it can be used to recognize different targets.For confidential consideration of national defense, HRRP of ballistic missile will not be published. Also our research needs large scale of HRRP. So based on the derivation of target echo, this paper uses electromagnetic simulation software CST to simulate range profiles of targets. In the midcourse segment, ballistic missile targets have track movement. There are also precession movement, tumbling and other micro-movements. Generally, warhead is axisymmetric cone, this paper uses aspect angle to describe the warhead’s aspect change to the radar. Range profiles are sensitive to targets’ aspect angle changing, the paper simulates one-dimensional range profiles of warhead with aspect angle changing from 0 ° to 180 ° to form the range profile database.The correlation coefficient of two signals reflects the similarity degree between the signal envelopes. Based on the maximum sliding correlation coefficient, this paper gives the definition of matching score between the one-dimensional range profiles. The greater the matching score, the more likely this target is warhead. This article attempts to recognize warhead target through a lot of experiments trying to get matching scores’ probability distribution. Truncated Gaussian distribution model is established. The experimental results are valued by the recognition rate and false alarm rate with matching threshold value changing. The results show that when SNR is relatively high and the matching score is set as appropriate value, the detection rate can be high, while the false alarm rate is relatively low. Thus, the recognition rate of the corresponding matching can be set to determine the threshold. When the target to be identified has a matching score larger than this threshold, it will be determined to be warhead, otherwise it is not a warhead.