Energy Saving Strategy And Performance Optimization Of Cloud Computing With Synchronous Semi-sleep Mechanism

With the widespread popularity of mobile devices and the rapid development of new5 G network applications,the requirements for large-scale device connections and computing processing capabilities have also increased.In order to meet the resource requirements of complex applications,the concept of mobile cloud computing has emerged.Taking into account the cloud task service quality and cloud system energy consumption level,introduce periodic semi-sleep mode and limit periodic semi-sleep mode in the cloud,and study cloud computing energy-saving strategies and performance optimization issues.First of all,in order to reduce the energy consumption level of the cloud system,for non-real-time network applications,a cloud computing energy-saving strategy integrating periodic semi-sleep mode is proposed.When a physical machine deployed on a cloud server has no cloud tasks,all virtual machines deployed on the physical machine enter the semi-sleep state from the normal state at the same time,and the virtual machines in the semi-sleep state can provide low-speed services for randomly arriving cloud tasks.When the semi-sleep timer expires,if the physical machine has no cloud tasks,all virtual machines deployed on the physical machine continue to remain in the semi-sleep state,and the semi-sleep timer is restarted.It can not only ensure the reduction of energy consumption of the cloud system,but also improve the response performance of cloud tasks.Secondly,in order to further improve the service quality of cloud tasks,on the basis of reducing the energy consumption level of the cloud system,for real-time network applications,a cloud computing energy-saving strategy that converges and limits the periodic semi-sleep mode is proposed.When all virtual machines deployed on a physical machine of the cloud server are in a semi-sleep state,and the number of consecutive sleeps reaches the limit,all virtual machines deployed on the physical machine enter the normal high-speed idle state,which is a cloud task that arrives at random Provide high-speed services instantly.Then,based on the proposed cloud computing energy-saving strategy,a system model is established and performance research is carried out.For the cloud computing energy-saving strategy that integrates the periodic semi-sleep mode,a queuing model with multiple service desks and multiple working vacations is constructed.To converge and limit the energy-saving strategy of cloud computing with periodic semi-sleep mode,construct a queuing model with multiple service desks synchronized and multi-level adaptive working vacation.Using the quasi birth-death process and the matrix geometric solution method,the average response time of the cloud task and the average power consumption of the cloud system in the steady state are obtained.Finally,based on the results of the system experiment,the impact of cloud task arrival rate and semi-sleep parameters on the performance of cloud computing energy-saving strategies is revealed,and the system performance of cloud computing energy-saving strategies is evaluated and optimized.We weigh the trade-off relationship between the average response time of cloud tasks and the average power of the cloud system,construct the system cost function,and establish the objective optimization problem of the minimum system cost function.The improved seagull optimization algorithm is used to obtain an intelligent optimization plan for cloud computing energy-saving strategies to minimize system costs.