Application of Compressed Sensing to Ocean Remote Sensing with an HF Radar

Compressed sensing (CS) is a growing field which allows the reconstruction of a signal with fewer data samples than the traditional methods. CS utilizes the fact that a signal can be considered sparse in a specific domain. Sparsity implies that the signal only consists of a few non-zero coefficients in its transformation domain. High-frequency (HF) radar signals scattered from the ocean surface are often characterized by a limited number of frequencies, which can be due to the near-periodic structure of ocean surface waves. Utilizing the sparsity of the ocean-backscattered radar signals in the Fourier domain, the CS approach has the ability to reconstruct the signal from a fewer backscattered samples. In this paper, we discuss the theory of compressed sensing (CS) and then demonstrate the sparsity of ocean surface backscattered HF signals. By exploiting this sparsity, we show the effectiveness of this technique for retrieving ocean surface parameters with less acquired data. This may also reduce the need for a local oscillator and the need for complicated hardware. Thus, the primary goal of this article is to extract statistical parameters of the ocean surface, including significant wave height, from backscattered time series data that have been sampled at lower rates. This method simplifies the processing of backscattered electrical signals from the ocean surface.