New matrix constant false alarm rate detectors for radar target detection

Abstract

Matrix constant false alarm rate (M-CFAR) detector based on information geometry provides a novel scheme to the detection of target in sea clutter environment for pulse Doppler radar with small bunch of pulses. The detection performance of the existing M-CFAR detectors is greatly influenced by the geometric measures on the matrix manifold. Different geometrical measures lead to different detection performance, some of which are not satisfactory. Motivated by this, the authors utilise Root-Euclidean, Power-Euclidean, Cholesky-Euclidean and new Riemannian measure to design several kinds of detection schemes, which improve the detection performance. Finally, simulation results verify that the designed detectors achieve a better detection performance with a relatively lower computational complexity.

Keywords

• Doppler radar
• object detection
• computational complexity
• radar signal processing
• radar detection
• radar clutter
• geometry
• new matrix constant false alarm rate detectors
• radar target detection
• matrix constant false alarm rate detector
• information geometry
• sea clutter environment
• pulse Doppler radar
• existing M-CFAR detectors
• geometric measures
• matrix manifold
• different geometrical measures
• different detection performance
• Riemannian measure
• detection schemes
• designed detectors