Feature extraction of precession target based on Doppler profile sequences by maximum likelihood estimation. Issue 9 (1st May 2019)

• Record Type: Journal Article
• Authors: Wang, Ning; Mo, Di; Song, Ziqi; Wang, Ran; Li, Guangzuo; Wu, Yirong
• Is Part Of: Electronics letters. Volume 55:Issue 9(2019)
• Page Start: 554
• Page End: 556
• Publication Date: 2019-05-01
• Subjects: military radar -- radar tracking -- target tracking -- Doppler radar -- radar computing -- ballistics -- radar resolution -- maximum likelihood estimation -- radar imaging -- radar signal processing -- Gaussian distribution -- radar target recognition -- optical radar -- feature extraction; modulation bandwidth -- utilising single‐frequency ladar -- precession parameter extraction algorithm -- high‐resolution Doppler profile sequences -- Doppler bandwidth -- cone‐shaped precession target -- maximum likelihood estimation method -- micromotion feature extraction -- missile ballistic defence -- high‐resolution range profile sequences -- Doppler profile sequences; Electronics -- Periodicals; 621.381
• DOI: 10.1049/el.2019.0253
• Languages: English
• ISSNs: 0013-5194

Abstract

Abstract : The micro‐motion feature extraction of precession target is an effective way for discriminating warhead from decoy in missile ballistic defence. Generally, traditional methods based on high‐resolution range profile (HRRP) sequences put forward high requirements on modulation bandwidth for the system, increasing the complexity of the system significantly. Utilising single‐frequency ladar without modulation, the authors proposed a precession parameter extraction algorithm of precession target based on high‐resolution Doppler profile (HRDP) sequences. Firstly, they model the Doppler bandwidth of the cone‐shaped precession target with the observation angle. Under the assuming that the measurement error obeys zero‐mean Gaussian distribution, the probability density function of Doppler bandwidth is established. Then, the precession parameters are estimated with measured Doppler profile sequences through maximum likelihood estimation method. Compared with the methods based on HRRPs, the proposed method based on the Doppler profile has lower requirements on the system and easier to implement. Laboratorial data has been used to validate the feasibility of the proposed method.