Research On Reliable And Location Transparent Communication Method For Mobile Agent

Since it came to 1990's, with the popularization of Internet and the flourishing development of mobile computing technology, The research upsurge on mobile agent is launched not only in academe but also in industry. Mobile agent uses its own mobile characteristic to achieve user's tasks by searching appropriate resource on heterogeneous network.The location of mobile agent is not fastness since it is created, it can migrate according to its needs. This is just the difference between the mobile computing model and traditional distributed computing model. There are two primary problems in this computing model for the communication mechanism of agent. First is how to inform the current location of an agent to other agents when it migrates. Second is how to handle the information that is sent to an agent when it is in the procedure of migration. Many researchers have put forward some kind of solutions of the Agent Location Tracking and the message storage and transfer for the two problems presently. Such as location tracking method of tree shape domain, location tracing method of distributed computing model and Agent communication method by using the AgentTracer. Although these methods have their advantages respectively, there are some disadvantages in tracing efficiency, network flux control, communication security and subagent callback for these methods more or less.In this thesis, author firstly outlines the current research status at home and abroad, introduces the significance of the paper and some relative theories. Then, analyzing the advantages and disadvantages of current agent communication model, we present acommunication model base on mobile agent as a solution for these disadvantages. The communication model is a twp-tier architecture, the main function of superstratum (Name server) is tracing the domain which target agent belong to presently. The main function of substrate (domain-inside tracing machine) is the message transfer and denominating the new created agent. We can finish a Location Transparent Agent Communication by using this model and we introduce the concept of agent energy at the same time in order to control subagent callback.Secondly, the thesis describes those realization algorithms of some function of the model. We also compare agent communication model presented in this paper with existing agent communication model in order to show the predominance of our model. Then author makes some simple simulative experiments about our agent communication model. At last, the paper puts forward the summarizing of this thesis and next step's work.