An overview of radar-based, automatic, noncooperative target recognition techniques

Abstract

Radar target recognition techniques tend to fall into two principle classes: those that exploit the radar characterization of a platform's physical shape and those that exploit the radar characterization of the dynamic characteristics of the moving parts of the target. The former are based on the platform's (essentially instantaneous) range (time)-amplitude radar signature and are exploited through generation and analysis of the platform's ultrahigh range resolution (UHRR) profile. The latter are based on the platform's frequency-amplitude radar signature as represented in the time evolution of its high-resolution Doppler signature. The methodologies applicable to automatic, noncooperative recognition of platforms based on both these classes of techniques are discussed. The choice and implications of radar parameters, signal processing techniques, and pattern recognition techniques are discussed, compared, and evaluated in terms of their impact on recognition system performance.<>