Improvement Of Service Processing Subsystem In IDP

Currently, mobile operators are changing from the mobile communication specialists to mobile information specialists. With the transform of mobile operators in the role of strategy, the entire value-added service market is changing. Mobile operators have already started to integrate market and probe self-supporting service. According to the present trend, an integrated data service platform will be the direction of the development of mobile data value-added service. Different types of data value-added services can be developed quickly in this platform. The platform is able to provide unified management for all kinds of resources (such as audio, photos, text messages, etc.), with a unified content management functions. And it also has the functions of operational role (as CP (Content Provider)) management. Under the current status of mobile data value-added service, we introduce and implement an Integrated Data-service Platform (IDP).IDP focuses on mobile user groups, fully integrates community resources, and provides many kinds of data valued-added service to mobile users such as life services, entertainment services and so on.IDP is an open, modular, standards-based platform. Downwards, it provides communications capability with ISMG (Internet Short Message Gateway), MMSC (Multimedia Message Service Center), BOSS (Business & Operation Support System), WAP (Wireless Application Protocol) Gateway and other functional entities. Upwards, it provides user authentication, SMS (Short Message Service), MMS (Multimedia Message Service), user management and other public functions. The main structure and the capability are on the basis of meeting existing data value-added service products. Service developers no longer need to concern about the underneath communication with other network entities, they can pay more attention to the development of service logic.As the data value-added service develops rapidly and IDP is widely used, we also find that IDP cannot meet more complex data value-add services'performance requirements in future. And the architecture also needs to be adjusted or IDP will be left behind the rapid developing trend of data value-added service. The efficiency of service processing is an important factor of system performance. Especially when there is no scaling plan, improving the efficiency is the only way to improve system performance, which can protect legacy investments and make a great economic benefits. However, each service has its own characteristics. Because of the long-term and expensive cost, we cannot optimize them one by one. But we can abstract common logics from IDP's different services to build a service processing subsystem, which is responsible for most common service processing functions. Then, we can optimize the service processing subsystem and system performance is also improved.This thesis focuses on the improvement of the service processing subsystem. In the first chapter, the present, future and substance of data value-added service is introduced, which describe the source and background of this article. In chapter 2, IDP is described in the points of functional, logical and physical view respectively. Based on analysis of IDP's disadvantages, we work out a two-step plan to improve it, "refactor firstly and profile later". The design and implementation of asynchronous general service processing subsystem in IDP is introduced in chapter 3. In this chapter, the asynchronous message bus, service interface, service persistence and service scheduling modules are described in details. In chapter 4, the development efficiency of original and improved IDP is compared. Then, four principal optimization methods used in the improving process is introduced, such as the improved dirty filtering algorithm, distributed caching, etc. In the last part of chapter 4, the IDP system-level testing method is introduced. And there is a detailed performance comparison of original and improved IDP in simple and complex data value-added service scenarios. The final chapter introduces the following work of IDP in four fields and the future of IDP is prospected.