Add smoothing algorithms for GNSS measurements
Several studies on Orekit's GNSS orbit restitution have shown the presence of a sinusoidal signal in the evolution of the pseudorange measurement residuals [RD.1] [RD.2].
Example from [RD.1]:
![Measurement residuals [RD. 1]](/orekit/orekit/uploads/f858a6af7d73d1e1d4caf654d68b67e6/GPS_07_BLS.png)
Example from [RD.2]:
The sinusoidal signal in the residuals comes from the noise of the pseudorange measurement and in particular from multipath effects. Smoothing the measurements would considerably reduce the noise and multipath effects on the pseudorange measurements. Therefore, it will also reduce the sinusoidal signal in the evolution of the pseudorange measurement residuals.
Many studies have shown the positive impact of smoothing methods using low-pass filters to reduce the the sinusoidal signal in the evolution of the pseudorange measurement residuals [RD.3] [RD.4].
Therefore, adding smoothing methods for GNSS measurement can be a very good enhancement for Orekit. Some methods exist and can be implemented [RD.5] [RD.6]. A study to find other methods can also be performed.
[RD.1]: Maisonobe et al, 2018
[RD.2]: Cazabonne et al, 2020
[RD.3]: Zhang et al, 2014
[RD.4]: Geng et al, 2019