Research And Implementation Of Distributed Communication Consistency Based On Adaptive Evolution

"Network is computer" has become a reality, the rise and widely used of cloud computing and big data guides the application requirements of users,And it also brought new technology implementation means,the business model and profit model for the supplier. Software systems running on the open Internet environment,to continue to meet user's expectations,need to deal with its operating environment changes,so the adaptive ability is the foundation of this software system.One of the research topic of software research membres domestic and abroad is how to make the Internet software entities can sense the dynamic change of environment, on-demand adapter, dynamic assembly, continuous evolution, providing reliable and high quality service.Traditionally, the software system is designed to be an open loop system, and the adaptive system requirement is a closed loop system. In order to deal with variable resources, system failure, and adjust the user's priority dynamically, it is necessary to manage,optimize and adjust the runtime system,the adaptive software arises at the historic moment. With the development of artificial intelligence and distributed computing technology, the relative theory and technology based on the Agent and Multi-Agent, because of its high autonomy and flexibility, provides a new theoretical basis and technology support for the software adaptive evolution.In this thesis,first of all,based on system integration, combined with Agent technology and multi Agent system,this thesis proposes a model based on the Agent support framework of adaptive evolution, and expounds in detail the structure and working mechanism of the framework.And then, the Agent communication language communication mechanism,to run in the bottom of the frame,was studied, a distributed communication framework model is put forward. For further improving the capacity of distributed communication framework model of concurrent processing and distributed node communication consistency, distributed communication synchronization mechanism and asynchronous I/O model was studied.A communication mechanism based on IOCP and the distributed communication algorithm based on Paxos algorithm, were proposed in this thesis to improve the adapatability of the system of traffic network.This article finally develop an example of software adaptive evolution, the model of distributed communication framework based on Agent were tested, through the analysis of function and performance, the validity of the algorithm and the framework is verified.