Study Of The Overload Control Algorithm For SCP In IN And The SPIRITS Protocol

The overload control technologies in Intelligent Network (IN) are the pivotal technologies in IN's realization. It ensures the stability of the IN, which is of great importance for offering the IN service with high quality and efficiency. The core of overload control technologies in IN is the algorithm of overload control for SCP. The traditional overload control algorithms for SCP are designed for the single-service IN mostly. As the application of IN become in-depth, IN have evolved into multi-service IN from single-service IN. There are some drawbacks in efficiency and fairness when we use the overload control algorithms designed for single-service IN in multi-service IN. So it is very useful to design an efficient and fair overload control algorithm for SCP in multi-service IN. People always pay attention to the overload control for SCP and neglect the overload control for SSP. Sometimes SSP will become overload, especially in Integrated IN. At present, it is urgent to research an overload control algorithm, which can prevent SSP and SCP from overload, as IN in our country evolves into Integrated IN.The plan of SPIRITS advocated by IETF brings forward an idea of request a service in Internet from IN. It is advantageous to enlarge the application of IN and to provide service providers with new services. The research of SPIRITS protocol is an important work.Firstly, we generalize and analyze the research of overload control algorithm of SCP in IN and technologies of SPIRITS; secondly, we study the basic theory of IN; thirdly, we study three conventional mechanisms and algorithm of overload control in SCP (windows, ticket bank and ACG), then we summarize criteria for evaluation of overload control algorithm in SCP; fourthly, we study overload control of SCP in multi-service IN and put forward an overload control model for SCP in multi-service IN based on ticket bank. According to the model, we bring forward an adaptive ticket bank with centralized control algorithm (ATBCC), which can adaptively distribute SCP's processing capacity among different services. The theoretical analysis and experiment prove ATBCC's efficient, robust and fair; fifthly, we study overload control in Integrated IN. After define service weight and queue model of IN, we bring forward a model and an algorithm of overload control based on network, which prevent SCP and SSP from overload at the same time; Sixthly, we study the technology of SPIRITS and put forward a scheme of SPIRITS protocol by extended HTTP, Which is simple, convenient and rapid to realize SPIRITS service; seventhly, we design a new SPIRITS service (PSTN, Instant Messaging) and realize the servicewith the SPIRITS protocol advocated in the paper; the conclusion summarizes the whole paper and preview the further developments of the work we have done.