Target detection for THz radar based on information geometry

Information geometry, a new discipline direction originated from differential geometry in the area of mathematics, studies the probability and information problems, and provides a new solution method to solve the problems in radar signal processing. In the current study, radar echoes were usually modeled as the complex multivariate Gaussian distribution. We propose a kind of CFAR detector for THz radar based on information geometry, and analyze its detection performance. Simulations show that the detection performance of the new geometry approach is better than the conventional detection method in THz band, which is helpful to further research on target recognition and parameter estimation in THz band based on information geometry.