Subband Information Geometry Detection Method Based on Orthogonal Projection for Weak Radar Targets

Abstract

Herein, a novel and effective method for detecting radar targets with a low signal-to-clutter ratio is proposed based on the information geometry theory. In the proposed method, the target detection problem is converted to distinguishing the target from a clutter background on a manifold. However, this is challenging when dealing with small and weak targets embedded in a complex and strong clutter background, which limits the detection performance. Therefore, to address this issue, an orthogonal projection based subband information geometry detection method is proposed. In this method, the received radar signal undergoes subband decomposition by a designed filter bank, and the robust estimation of clutter signal subspace in each subband is implemented on the matrix manifold. Subsequently, the suppression of the strong clutter is achieved through orthogonal projection based on the manifold, thereby improving the discrimination between the target and the clutter. Finally, the effectiveness of the proposed method is evaluated using simulated and real sea clutter data. The experimental results confirm that the proposed method effectively suppresses strong clutter and exhibits excellent detection performance.

Keywords

• weak radar targets detection
• information geometry
• matrix manifold
• subband decomposition
• orthogonal projection
• clutter suppression