Fuzzy-logic based QoS and fault diagnosis mechanisms for broadband satellite networks under dynamic weather conditions

In this thesis, two techniques were developed to overcome the problems associated with broadband satellite communication systems. The congestion adaptation system was designed to improve the quality of service for satellite channels under dynamic weather conditions. Different from traditional congestion control techniques, this system uses the rain/snow-fall rate to tune system parameters. Instead of using an algorithm-based tuning method, this system utilizes a knowledge-based method to achieve congestion control in the territorial IP network. The knowledge base was built upon the knowledge of experienced network engineers. Test results indicated robust transmission of high priority traffic under dynamic rain/snow-fall rates.;The knowledge-based terminal fault diagnosis system was designed to diagnose defective terminals. The fault sources of satellite ground terminals were analysed, and the operation conditions of ground terminals were classified into five categories. Both statistical approach and knowledge-based approach were used to diagnose ground terminals. The overall diagnosis accuracy is satisfactory, and the effectiveness of the system was approved by a satellite service provider.