Experimental rain attenuation statistics estimated from radar measurements useful to design satellite communication systems for mobile terminals

Abstract

Based on experimental data, we have reported a reliable method to scale the cumulative time T/sub F,T/(A) that rain attenuation A (dB) is exceeded in a fixed satellite system to the time T/sub M,T/(A) that it is exceeded in a satellite system for mobile terminals. Zigzag routes and ring-roads simulated city patterns; straight routes simulated freeways. In all cases, T/sub M,T/(A) can be expressed as T/sub M,T/(A)=/spl xi/T/sub F,T/ (A) with a probability scaling factor /spl xi/ independent of A. The simulations have been made at 19.77 GHz with satellite elevation angle /spl theta/ of 30.6/spl deg/, 45/spl deg/, 60/spl deg/, 80/spl deg/, and 90/spl deg/. For the horizontal structure of rain, we have used a very large number of rain-rate maps of rain storms randomly observed in 1989-1992 by a meteorological radar placed at Spine d'Adda (northern Italy). The vehicle speed was modeled as a log-normal random variable. We found: (a) in zigzag routes, T/sub M,T/(A)<T/sub F,T/(A), i.e., /spl xi/<1, with results depending on vehicle speed modeling and starting conditions; (b) in a ring-road, there is no difference between fixed and mobile systems (/spl xi/=1); and (c) in straight freeways, T/sub M,T/(A)/spl Lt/T/sub F,T/(A)(/spl xi//spl Lt/1); T/sub M,T/(A) can change significantly in different straight lines and in opposite directions (anisotropy and asymmetry) for medium-large attenuation. When compared to zigzag routes or ring-roads, the performance in straight freeways is the most optimistic. For /spl theta/>30.6/spl deg/ and for the same pattern, /spl xi/ is fairly independent of /spl theta/. Since the radar rain maps are a reliable estimate of the horizontal structure of rain, the findings, which can be considered frequency-independent, stand as a very good prediction of the results obtainable by experiments.