Analysis of Sea Clutter Using Dynamic Mode Decomposition

In this paper, a novel method based on a dynamic mode decomposition (DMD) for sea clutter analysis is proposed. It extracts the temporal patterns and corresponding dynamic modes from the sea clutter simultaneously. Moreover, the temporal patterns display similar properties with traditional analysis using Doppler spectrum. The corresponding dynamic modes represent the cardinal feature within the sea clutter. To demonstrate the effectiveness of the proposed method, the measured sea clutter data collected by IPIX radar is analyzed. It is shown that DMD spectrum has the same frequency-shift and similar amplitude with the Doppler Spectrum. In addition, the Probability Density Function (PDF) of dynamic modes in high sea condition shows the long tail phenomenon. Hence, this data-driven decomposition method can provide an effective two-dimensional subspace (Spectrum-Mode) for the detection of targets on the sea surface.