Channel modelling based on N-state Markov chains for satcom systems simulation

Current trends in telecommunications are driven by the demand for multimedia services which require high data rates. In the fixed satellite service, the frequency bandwidths required to carry such data rates are available only at frequency bands such as Ka-band (20-30 GHz), V-band (40-50 GHz) or EHF-band (20-45 GHz). A problem with these frequency bands is the influence of the atmosphere on radiowave propagation. As the operating frequency is increased, attenuation and scintillation effects of atmospheric gases, clouds, rain and the melting layer become more severe. To evaluate the behaviour and quality of service of communications systems, the impact of the propagation channel has to be taken into account. The conventional approach, which consists in predicting only the attenuation distribution with respect to time, is not sufficient and finer information about the propagation channel dynamics is needed. We present the description of a time series generator relying on N-state Markov chains. With such a model, it is possible to synthesise, on demand, events that are characterised by a given depth and a given duration. Another key point of such an approach is the fact that it can be applied anywhere because the process relies on information provided by world-wide radio-meteorological maps and propagation prediction models, especially from the new ITU-R recommendation on fade dynamics. A preliminary validation of the Markov chains has been carried out to check the convergence of the stochastic process.