Skip to Main Content
Summary form only given. The effect of atmospheric propagation o is one of the relevant issues to be considered when designing Satellite communication systems operating at Ka and Q/V bands. At those frequencies the fixed margin solution employed in Ku band systems does not appear viable, considering also the physical limitations of on-board available mass and power. The adoption of fade mitigation techniques is now a commonly adopted design but it needs a detailed modelling of the propagation channels in terms of dynamics, spatial and frequency distribution. This issue is particularly relevant for: Satellite broadcast systems employing adaptive antenna patterns to counteract atmospheric attenuation on ground; Communication satellite systems using multiple antenna beams and adaptive coding and modulation technique to provide multimedia services on a continental region (e.g. Europe or Northern America). In both cases the need to implement a complex adaptive payload calls for models, data and tools for designing and simulating the system. The spatial scale of analysis required can range from intra-beam (e.g. for different users located inside one or overlapping beams of a TLC system) to a continental coverage region (e.g. for reallocating power from an area operating above nominal conditions to another area affected by severe atmospheric fading). The paper will present a modelling technique for the spatial distribution of rain, clouds and vapour for design of satellite telecommunication systems. This approach is based on the statistical dependence index ( ) that can be used to assess the spatial distribution from a large variety of meteorological databases such as long-term raingauge measurements, radar maps or Numerical Weather Predictions data. Such a parameter can be linked to the usual correlation factor by assuming a Gaussian bi-dimensional distribution. On those basis the modelling of joint-distribution of atmospheric attenuation in different places can be perfo- med by assuming that phenomena of smaller scale (embedded in a larger scale distribution) can be considered jointly independent.