Satellite Link-Budget Statistical Prediction From Weather Forecast Models: Verification With Hayabusa-2 Ka-Band Data

This work provides a weather-forecast-based technique (named RadioMetOP) for the stochastic optimization of Ka-band satellite links. The stochastic approach is given by a space-time ensemble obtained exploiting the temporal and spatial evolution of the weather-forecast model. This approach minimizes the doublepenalty error typically connected to weather-forecast models and allows to take into account the forecast uncertainty directly related to the multiplicity of possible meteorological scenarios. For the first time, the RadioMetOP technique can be verified thanks to the availability of Ka-band data from the Hayabusa-2 deep-space mission directed by the Japan aerospace exploration agency and supported by the European space agency. The comparison with measurements has proofed the strong advantage of the RadioMetOP model (with a signal-to-noise ratio gain even higher than 8 dB) and its accuracy (with validation results showing mean values of correlation, bias and error standard deviation of 0.88, 0.17 dB and 0.65 dB, respectively).