Research On Mobile Agent Technology Used In Mobile Database

With the rapid development of network technology, the information can be accessed at any time and any place, which becomes the general requirement in 21st century. Mobile computing technology can make this demand be satisfied. Because of the differences between mobile computing environment and fixed computing environment, the traditional client/server model can not meet the needs of mobile computing. A flexible technology is needed to adapt to the mobile environment. As a result, mobile database based on mobile agent has aroused wide attention and has been studied extensively in recent years.Firstly, the paper proposes a structure model of mobile database based on multiple agents. Being different from traditional mobile database based on agent, the structure model that we researched in this paper begins with three-tier replication architecture, and then researches the agent applied technology used in mobile database from three aspects, including mobile database replication, caching, and broadcast. The research aims to improve treatment efficiency and reduce response time of mobile database system.Secondly, the key technologies used in the system are researched. A new communication mechanism among the Agents is designed, and the processes of agent migration and finding location are described in detail to solve the problem of communication failure. Simultaneously, the paper introduces a collaborative model of agents. A task scheduling algorithm is designed among agents based on Adaptive genetic algorithm. Matlab is adopted to do algorithm simulation. The result shows that the optimal solution of the improved algorithm is faster 5.3 percent than traditional solution.Finally, the paper establishes a cache structure based on mobile agent to solve the inconsistent problem. A cache management scheme is proposed based on the structure. The scheme uses agent technology to solve the problem of cache invalidation and to improve the utilization ration of cache. In local area, the cache management scheme based on agent is simulated on aglet platform and is compared with the traditional cache strategy based on broadcast invalidation report. The result shows that the cache performance of the prototype is better than traditional solution.