Fade countermeasure modelling for Ka band digital satellite links
This thesis investigates the modelling of fade countermeasures (FCMs) for the design of geostationary Ka band digital satellite communication systems. The analysis focuses on a typical low-power low-rate very small aperture terminal application using adaptive forward error correction as a way of counteracting the high level of detected dynamic atmospheric fading. The management and performance of such systems is conditioned greatly by the ability of practical controllers at detecting the actual level of total signal attenuation. At 20 or 30 GHz, rain attenuation and tropospheric scintillation are the two major propagation effects of interest. Part of the solution relies on the consideration and integration of their random and dynamic nature in the design process. The finite response time of practical countermeasure systems is a source of performance degradation which can be minimised by the implementation of predictive control strategies. This is the focal point of this thesis. A novel on-line short-term predictor matched to the Ka band fading process is proposed. While the rain attenuation component is efficiently predicted, tropospheric scintillation is the source of the estimation error. To take this into account, a statistical model, based on an extension of the global fading model for rain and scintillation, is then developed so that long term performance of predictive countermeasures can be drawn. Two possible ways to compensate for scintillation-induced prediction errors, namely the fixed and variable detection margin approaches, are proposed, analysed and then compared. This is achieved by calculating the FCM utilisation factor, as well as the throughput and bit error rate performance of a typical Ka band system in the presence of dynamic fading within the context of predictive fade countermeasure control operations. In the last part of this thesis, the inclusion of instantaneous frequency scaling in the design of efficient FCM control schemes is investigated. This is applicable to systems using fade detection at a base frequency. In particular, a new statistical model, accounting for the impact of the stochastic temporal variations of rain drop size distribution on rain attenuation, is presented. This thesis further confirms that countermeasure systems are technologically viable. The consideration of more specific design problems does not change the overall validity of this statement. In this thesis, it is shown that a predictive FCM technique, based on readily available punctured convolutional codes, with their relatively modest coding gain, is sufficient to provide high link availability and user data throughput on a low-power low-rate in-bound VSAT link.