Energy-efficient Execution Scheduling Of Random Sequence Task In Mobile Cloud Computing

Mobile cloud computing is a new technology that combines mobile internet and cloud computing.To improve mobile terminal's weak processing capabilities,small storage space,and short battery life,users can migrate calculations and storage of data to the cloud.With the development of mobile cloud computing,users' demand for cloud services is increasing.It puts forward higher requirements on the quality of service and energy consumption.In mobile cloud computing,through the task's unloading and scheduling technology,the task within the terminal can be selectively scheduled and processed under the premise that the user can meet the task service quality requirements or the terminal performance requirements.Utilizing the computing resources of the local terminal or cloud server,it breaks through the limitations of mobile client processing tasks and improves user satisfaction.Based on the characteristics of the task,this paper studies the execution scheduling of random task sequences in mobile cloud computing.The main work of this dissertation is as follows:1.For mutually independent random sequence tasks,a mutually independent task queue model is established.For the unknown system with unpredictable random parameter distribution,the Liyapunov optimization theory is applied to the control of stochastic task scheduling optimization.The longer the data is waiting in the terminal for processing,the greater the amount of data backlog and the resulting instability of the system.Using the Lyapunov optimization algorithm,the offset plus penalty function is constructed on the premise that all queues are stable.By adding the amount of drift and the objective function to be optimized,and according to the constraints of the final expression,an appropriate penalty control threshold is found,which can satisfy the user's optimization goal and ensure the stability of the system.2.For mutually dependent mixed sequence tasks,a mutually dependent task queue model is established.For the randomness of tasks,a continuous time Markov decision process system model is established.Under the premise of meeting the average waiting time requirement of the task,the energy consumption generated by the mobile terminal during task processing or scheduling is used as an optimization goal.A continuous time Markov decision process system model is established.The model is solved using a strategy iterative algorithm.In order to find the optimal strategy,it can not only satisfy the time requirements but also reduce the energy consumption.The simulation validates the effectiveness of the algorithm.