Sea Surface Roughness Estimation Using Signals from Multiple GNSS Satellites

Global Navigation Satellite System reflectometry (GNSS-R) has recently drawn significant attention since it can be employed in a range of applications including sea state, ocean altimetry, soil moisture measurement, and disaster (e.g. flooding, bushfire, and earthquake) monitoring. Although research on GNSS-R in the past two decades has made advances, there are no (or few) real applications of the technique. This is mainly because GNSSR is still not a mature technique, there are no satellite-borne missions, there are few systematic airborne experiments, and GNSS-R is not yet able to provide information with high enough resolution and reliability. The authors are investigating new approaches to improve the performance of the GNSSbased geophysical parameter estimation. This paper focuses on the sea surface roughness estimation through an analysis of the reflected signal power. In particular, the theoretical modelling of the sea wave and surface scattering is studied in detail. Solution-level and waveform-level combination techniques are proposed to improve the estimation accuracy through jointly processing the measurements of the reflected signals which are transmitted by multiple GNSS satellites. These combination methods also can be employed for other parameter estimation tasks. Airborne experiments were carried out by a UNSW-owned light aircraft over the sea off the coast of Sydney. The data were logged using the NordNav software receiver which has four front-ends so that the signals arriving at the LHCP (left hand circularly polarized) and RHCP (right hand circularly polarized) antennas could be recorded simultaneously. Both the delay waveforms and delay-Doppler waveforms were generated from the processing of the real data. Theoretical delay waveforms were also generated. The preliminary results confirm that the waveform-matching estimation method is not suited for scenarios where the flight height is below 500 m.