Simulation of rain attenuation time series by ARIMA model

According to Auto regression moving average theory, the ARIMA model is established based on the E M-B rain time series method and ITU-R rain parameters. The rain attenuation time series are simulated by ARIMA model at Beijing and Haikou area, at 12.5GHz. The probability of the duration time statistical for rain attenuation series of 50 simulative events is analyzed. A comparion among the results of simulated with measured and ITU-R predicted is done at Beijing and Haikou. It is shown that when duration time is smaller than several hundred second, the statistical results simulated by ARIMA are basically in agreement with the results by measuring and ITU-R predicting, but for the longer duration time, the difference among them is obvious. This shows that the ARIMA(3,1,3) model can be applied to the simulation of rain attenuation time series at Beijing and Haikou, but for longer duration time it is required to improve.