An Intelligent QoS Control System for Satellite Networks Based on Markovian Weather Prediction

Prediction of channel characteristics can be of immense value in improving the quality of signals in high frequency satellite communication channels. Making prediction of rainfall rate (RR) using Markovian theory is the object of this paper. The paper also describes an intelligent system (IS) that uses the predictions made from Markovian model to maintain the Quality of Service (QoS) in channels impacted by rain attenuation (RA). The paper describes the method of prediction of RR using weather correlated database supplied by International Telecommunication Union-Radiocommunications (ITU-R) and applying the predictions to gateway and ground terminal for optimal control of channel characteristics. The proposed method of predicting weather characteristics using Markovian theory supplies valuable data to develop an enhanced back propagation-learning algorithm to iteratively tune the intelligent system to adapt to changing weather conditions. The effectiveness of the algorithm was tested on a simulated model for activating the weighted modulation and codepoint control. It demonstrated marked improvements in channel parameter tuning and signal quality.