Stochastic Approach to a Rain Attenuation Time Series Synthesizer for Heavy Rain Regions

Abstrak

In this work, a new rain attenuation time series synthesizer based on the stochastic approach is presented. The model combines a well-known interest-rate prediction model in finance namely the Cox-Ingersoll-Ross (CIR) model, and a stochastic differential equation approach to generate a long-term gamma distributed rain attenuation time series, particularly appropriate for heavy rain regions. The model parameters were derived from maximum-likelihood estimation (MLE) and Ordinary Least Square (OLS) methods. The predicted statistics from the CIR model with the OLS method are in good agreement with the measurement data collected in equatorial Malaysia while the MLE method overestimated the result. The proposed stochastic model could provide radio engineers an alternative solution for the design of propagation impairment mitigation techniques (PIMTs) to improve the Quality of Service (QoS) of wireless communication systems such as 5G propagation channel, in particular in heavy rain regions.