The Study And Application Of Computation Partitioning Of Mobile Cloud Computing In Dynamic Environment

With the development of the mobile Internet, the mobile application in the cloud service system has produced a new service computing method, mobile cloud computing. Mobile terminal device itself has such as the calculation of the storage resources, battery capacity and ability of network communication limitations, mobile cloud computing can make mobile terminal in wireless networks, mobile application development, the use of cloud computing resources, according to need and easy expansion to obtain the required computing and storage resources. Mobile cloud computing make the part or all of the tasks required to migrate to the cloud, improve the quality of service(Qos) of mobile applications. At present, the mobile cloud computing method is less considered to adapt to the mobile client available resources and the environment with the change of network bandwidth to improve the performance of the case. Aiming at the above problems, this paper carried out the research about mobile cloud service application based on dynamic environment of mobile cloud computing of calculation of the segmentation method study in the dynamic environment under the influence, unit of computing tasks in dynamic decision mobile application migration to cloud computing or in the calculation. Mainly carry out the following work:Combined with the system framework of mobile cloud computing, a model of mobile data stream with state is constructed firstly. Definition of direction and problems of mobile cloud computing, including single frame data flow calculation of the segmentation scheme, task scheduling and executing efficiency. For multi frame data flow calculation segmentation decision and dynamic optimization problem defined analysis.On the basis of is studied in this paper single frame data flow calculation of the segmentation method, is proposed in the known calculation segmentation scheme under the condition of task data flow model and task scheduling, and analysis of the task scheduling method, and according to the scheduling method calculation data frame after the calculation of the segmentation of the execution time, through genetic algorithm to execution time shortest as fitness function, calculated the optimal calculation of the segmentation decision method.This paper analyses in single frame data flow calculation segmentation method based on multi frame data, research on multi frame data calculation segmentation decision and optimization adjustment and prediction based on changes in the future period of time is short of network bandwidth, by calculating the node calculation of the segmentation scheme optimization adjustment method, multi frame data calculation of the segmentation optimization and decision-making.In view of the above research work, this paper verifies the effectiveness of single frame data in the application of the data stream, and analyzes the performance of the method to optimize the adjustment of multi frame data. Finally, a mobile cloud computing platform prototype system "mobile cloud services face recognition" to verify the effectiveness of the method.The mobile cloud computing of calculation of the segmentation method, with 4G era of arrival and mobile terminal equipment performance improvement can be more used to the state of the data stream processing program such as two-dimensional bar code recognition, image recognition processing system, speech recognition system.