A DDS-Architecture Dynamic Evolution Framework Based On Reflection

In order to adapt themselves to nowadays opening environment of Internet and varying requirements of users, software systems should be able to adjust themselves frequently and dynamically. Owing to the merit of continuous usability, dynamic evolution has become a hotspot in the software engineering research areas. Software Architecture depicts system configuration states from a global perspective, so it can become a major gist and a driving factor for dynamic evolution.Because it provides control levels of low-latency, high-throughout and real-time performance, DDS has become a standard for Distributed Systems, and has been widely used in aerospace, defense, distributed simulation and other fields. Because of the real-time of DDS, the system based on DDS Architecture must be upgraded on-line. And the extensive application of DDS makes the research on Dynamic Evolution of DDS Architecture become a pressing demand.Firstly, this paper describes some related research on Dynamic Evolution of Architecture, and points out an important problem of the research, that is the lack of the high-level view of the field. To solve this problem, this paper introduces reflection and proposes a dynamic evolution model based on reflection. Through reflection, the information of the architecture can be acquired in time, and it can be showed to the users as an intuitive business view. Besides, the changes of the architecture can be showed on the business view.Secondly, based the formal analysis and description of the DDS Architecture, this paper introduces a DDS-Architecture Dynamic Evolution Framework(DDEF). This Framework is composed of System Information Collection Component, System Information Library, Running Monitor, Dynamic Evolution Algorithm Library, Dynamic Evolution Algorithm Generator and Dynamic Evolution Manager. The Algorithms in the Dynamic Evolution Algorithm Library is the core of this Framework, so this paper does some research on the design and correctness of the Dynamic Evolution Algorithms.Finally, the DDEF is applied to the deployment module of the Integrated Develop Platform to maintain the architecture of the Running Platform on-line. The functions of the deployment module include showing the architecture of the Platform, deleting a component, adding a component, updating a component and moving a component. Practice has proved that the dynamic evolution can be realized easily and the system can run correctly at the same time.