Research On Immunology-Based Online Detection And Resolution Method For Feature Interaction

In the traditional telecommunication network, the interactions among features have been recognized as an important and serious problem and become an important research area. In the Next Generation Network, the feature interaction problem has not dispeared for the enhancement of the network capability and the rich of the network signals; on the contrary, the problem will become much more serious because of the individuation of the services and the service providing. Most of early researching works use formal methods to research feature interaction. The major shortage of this kind method is that internal informations of service logics are needed, but these informations are hardly obtained because of the privacy of service logic. Also there are some shortages in most of current online methods, such as the methods of negotiation between the agents are not compatible with the network architectures, and the method based on feature interaction management system can only support one or several kinds of feature interactions. They are applicable only under some specified conditions.In this thesis, we focus on the feature interaction problem in next generation network. Based on analyzing the similarities between feature interaction management system and immune system, we make deep researching about how to utilize the principles and mechanisms of immune system to online detect and resolve feature interactions.The main contributions are described in this thesis, which are following:1. An Immunology-based Multi-Protection Feature Interaction Management System (IMP-FIMS) model is proposed.After the similarities and mapping relations between feature interaction management system and immune system are analyzed, an IMP-FIMS is proposed.