| Summary: | Rain precipitation along the path from one base station to another base station is not constant due to drop size distribution of the rainfall and variation rain intensities. The signal level that propagates through rain is decreasing especially when the frequency used is above 10GHz. Rain classification is an important factor in rain attenuation studies. Rain can be classified in two broad categories which are convective rain and stratiform rain. Both categories have different effect on rain attenuation values due to different drop size distribution and different rainfall rates. However, what previous studies have not discussed is the attenuation prediction result for both stratiform and convective events. Hence, this study attempts to achieve the classification of rain by using probability method, determining 0.01% rain rate for stratiform and convective events and determining the suitable rain model that fits stratiform and convective rain. In order to choose good rain attenuation models, it is necessary to consider the link type and the experimental region. For this project, the chosen link is terrestrial link and the experimental region is tropical region. Therefore, the suitable rain models for this project are Garcia model, ITU-R 530-16 and Mello Pontes model. The duration of rain collection used for rain classification procedure is from 1996 to 1999. The percentages of time from complementary cumulative distribution function (CCDF) are used to determine which rain models suits stratiform and convective events. The result of rain classification shows that the totals numbers of stratiform and convective events are 631 events and 211 events respectively. Finding indicated that when using combined data and convective data, Mello Pontes is the most appropriate rain model to predict attenuation at terrestrial link. In addition, ITU-R 530-16, Mello Pontes model and Garcia model show good performance when using stratiform data as the three have similar attenuation values.
|
|---|
