Research On Mobile Cloud Computing Environment Task Allocation Problems

With continuous emergence and development of grid computing, cloud computing and othernew technologies, how to allocate tasks to each node in the computing system to achieve thepurpose of the rational use of resources and load balancing has become the research hot spot.Mobile cloud computing, as an extension and expansion of cloud computing, is facing a series ofchallenges such as big differences between devices complexity and high overhead caused bymobile network situation task allocation research. However research on task allocation in mobilecloud computing environment is simple and lagging.Cloud for mobile devices model is currently the most common model.The elastic applicationmodel can allocate the task in the cloud and mobile terminal according to the state of the mobiledevice and bandwidth flexibility. Compared with a fixed task allocation model, elastic applicationmodel can effectively avoid wasting resources and reduce network overhead. Paper draws on theresearch methods of dual-processor task allocation, assign tasks into a network flow for theminimum cut problem, redesign way to build concrete steps, such as network capacity calculatedflow according to the elastic model.Paper introduces permission value to transfer multi-objectiveoptimization problem into a single objective optimization problem. The examples show that themethod is efficient to reduce execution time and network costs.Micro-cloud model is hot spot in the development of future mobile cloud computing.InMicro-cloud network the number of its nodes and computing capabilities vary, and there is nocentral node responsible for task allocation. Each node can only allocate task on the basis of theirown circumstances and feedback from systems,which is different from elastic model and traditionaldistributed systems. Heuristic algorithm can effectively solve such problems. Paper draws onresearch methods in grid computing task allocation problem, applies the ant colony algorithm totask allocation in micro-cloud model, and gives the steps of the method.Simulation experimentalresults show that compared to the random assignment,this method is efficient to reduce executiontime and make load balancing.Adopting reasonable task allocation method in different mode can reduce running time andtraffic cost of mobile applications and offer better using experience to the user.