Blind sea clutter suppression for spaceborne gnss-r target detection

• Published in: IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing
• Main Authors: Cheong, JW, Southwell, BJ, Dempster, AG
• Format: Article in Journal/Newspaper
• Language: unknown
• Published: Institute of Electrical and Electronics Engineers (IEEE) 2019
• Subjects: anzsrc-for: 0406 Physical Geography and Environmental Geoscience; anzsrc-for: 0801 Artificial Intelligence and Image Processing; anzsrc-for: 0909 Geomatic Engineering; Sea ice
• Online Access: http://hdl.handle.net/1959.4/unsworks_63973; https://unsworks.unsw.edu.au/bitstreams/42afbea1-f598-48d2-b3c6-91d9611c4003/download; https://doi.org/10.1109/JSTARS.2019.2956183

It has been demonstrated that delay-Doppler (DD) maps (DDMs) retrieved from spaceborne Global Navigation Satellite Systems Reflectometry missions contain sea clutter and also potential sea targets. Examples of sea targets include sea ice, oil rigs, and oil slicks. To distinguish sea targets from sea clutter, one can estimate the sea clutter and subtract it from the DDM, leaving behind only the sea target's signature on the DDM. This is computationally expensive and will not be robust when the sea clutter estimate is poor or if there are DD tracking errors. This article presents two blind methods for the suppression of sea clutter: first, an adaptive filter and second, an infinite impulse response low-pass filter. We use real DDMs collected by TechDemoSat-1 to show the autoregressive, hence temporally correlated nature of the DDM time series, and present evidence that two blind methods can suppress clutter at least as effectively as the existing method, albeit at a fraction of its computational cost.