Range-Spread Targets Detection in Unknown Doppler Shift via Semi-Definite Programming

Abstract

Based on the technique of generalized likelihood ratio test, we address detection schemes for Doppler-shifted range-spread targets in Gaussian noise. First, a detection scheme is derived by solving the maximization associated with the estimation of unknown Doppler frequency with semi-definite programming. To lower the computational complexity of the detector, we then consider a simplification of the detector by adopting maximization over a relaxed space. Both of the proposed detectors are shown to have constant false alarm rate via numerical or theoretical analysis. The detection performance of the proposed detector based on the semi-definite programming is shown to be almost the same as that of the conventional detector designed for known Doppler frequency.