Computation Migration Research And Resource Conflict Performance Optimization Based On Cloud Computing

With the rapid growth in the number of mobile intelligent devices and mobile applications running environment on the improvement of hardware performance requirements,and the resources of the mobile terminal(including calculation,memory capacity,battery power)limited problem increasingly prominent.This not only limits the battery life of mobile terminals,but also affects the user experience of mobile applications.Mobile computing migration is an effective way to solve the problem of mobile terminal resource limitation.However,the existing computational migration framework has not supported the agile development of mobile applications.The implementation and deployment of mobile computing migration application is too complex and cumbersome,and there is no research on resource scheduling conflict when the mobile terminal is widely accessed to calculate the server end.Therefore,based on the framework of calculating migration and resource scheduling conflict research,designed and implemented a prototype of the calculation of agile migration programming framework,and put forward the performance optimization algorithm for resource scheduling conflicts.The main work of this paper is as follows:First of all,in the discussion on the basis of the research framework for calculating migration progress,the key to explore the migration of mobile computing framework programming model and the technical implementation details,and analyses the advantages and disadvantages of the framework and algorithm.Secondly,according to the Android application,combined with Java annotation mechanism,APT compiler tool,design and implement the annotation-based agile computing migration programming framework prototype.According to the annotation mechanism provided by Java,the API used to mark the migration method is designed.Based on APT compiler,design and implement annotation code extraction,uninstall decision and communication code writing.At the same time,it provides some support for the application of Android application to calculate the problem of the execution environment of the server terminal component.Finally,this paper analyzes the performance bottleneck of resource scheduling during high concurrent access to the server and proposes a Nash balanced scheduling strategy,which reduces the conflict of resource scheduling in Shared cluster.In this paper,we compare the performance of multiple scheduling strategies in a Shared cluster,and the Nash equilibrium scheduling strategy shows the optimal performance.