Modeling and simulation of bin-bin correlations in GNSS-R waveforms

Reflected signals from Global Navigation Satellite Systems (GNSS-R) can be used as a source of illumination for bistatic radar remote sensing of the ocean surface. When cross-correlated with a replica of the transmitted signal, the resulting waveform will have a shape dependent upon the slope distribution of the random rough surface. Substantial data from airborne experiments with GPS exist and have been used to demonstrate the feasibility of GPS reflectometry for ocean remote sensing. The available data from satellite receivers, or any experiments using signals with other modulations, such as Galileo, however, is much scarcer. In order to predict the performance of future spaceborne GNSS-R experiments, and to evaluate the utility of including BOC-modulated signals from Galileo and the modernized GPS in reflectometry measurements, a capability for generating synthetic GNSS-R waveforms has been under development. Individual waveforms are produced by cross-correlating a short block of reflected signal with a model signal over many delay-Doppler bins. As the receiver moves relative to the scattering surface, this cross-correlation is repeated in time, over subsequent blocks of reflected data. Noisy waveforms are thus correlated in time. Samples of the waveform at a given instant in time are also correlated with other samples at same time, but computed at different delays. A realistic simulator, producing synthetic data which accurately represents the statistics of observed signals, must properly account for these correlations. In this presentation, a model for the second effect, the correlation between waveform samples at different delays (or between ”bins”) is modeled and implemented into a simulator.