Prediction and adaptation of satellite channels with weather-induced impairments

Abstract

Efficiency improvements using prediction and adaptation methods over satellite channels with weather-induced impairments are presented. Considering scintillation and rain attenuation as two dominant factors for signal fading over satellite-Earth paths over 10 GHz, we develop statistical and spectral analyses of these processes, and obtain simple linear estimators for received signal attenuation using autoregressive (AR) models. Using these estimators, we present results where we can predict the received signal attenuation within /spl plusmn/0.5 dB 1 second ahead and within /spl plusmn/1.0 dB 4 seconds ahead. For adaptation, we change signal transmission power, modulation symbol size, and/or code rate adaptively. In particular, we introduce a continuous power control and discrete rate control strategy, through which we build a set of modulation/code states, and discretely change the modulation symbol size and the code rate from state to state. Within each state, continuous power contral is implemented. The quantitative analysis of power consumption indicates that there is a substantial gain in performance with the adaptive schemes, e.g., as much as 13 dB on a lightly rainy day.