Generalised likelihood ratio test detector for micro motion targets in synthetic aperture radar raw signals

Micro-motion targets are gaining an increased interest from the synthetic aperture radar (SAR) community. Their SAR returns can be modelled as sinusoidal frequency-modulated signals in the noise/clutter environment. A detector based on the generalised likelihood ratio test (GLRT) is proposed to detect such targets using SAR raw signals instead of SAR imagery. The GLRT detector entails searching three parameters for the maximum of a statistic, followed by comparison of the maximum with a threshold determined by the constant false alarm rate. It can also provide estimation of the micro-motion parameters. The detection performance is derived theoretically and demonstrated by both simulated and quasi-real data, showing favourable detection ability under low signal-to-noise/clutter ratios.