Wireless Micro Cloud Load Tranfer And Multi-workflow Scheduling Algorithms

With the rapid development of mobile communications,the rate of wireless transmission is getting faster and faster,laying a foundation for the development of mobile intelligent terminals and cloud computing.Mobile intelligent terminals are getting more and more into the scenes of people's lives and carrying more and more functions.However,due to the limitations of the terminal's own hardware conditions,storage terminals,processing power and battery life will encounter bottlenecks.Therefore,mobile intelligent terminal cloud computing rise,in which data is stored and processed in the cloud data center.However,with the advent of 5G,the Internet of Things will also flourish and the amount of data will explode.The centralized cloud computing can not meet people's requirements for low latency more and more,and the micro-cloud computing comes into being.Micro-cloud computing is closer to users than cloud computing,which allows micro-cloud to store and process user data with lower latency.Do not have to pass all the data to the cloud computing center to deal with,also reduces the network load on the cloud.We consider the scenario in which a computing center is deployed in each area and all these computing centers form a micro-cloud computing system.However,since the user's arrival rate to different region is different,some areas will be blocked.In this paper,a joint optimization algorithm is proposed based on tabu search and artificial ant colony algorithms,which effectively reduces the congestion in the area.Further,this paper has proposed nine task scheduling algorithms,measured the performance of the algorithms in the presence of mixed workflows,and determined the most appropriate one for the deployment by performance comparison.First of all,for the micro-cloud system deployed over multiple areas,this paper defined the jump ar-rival rate as an important parameter of the computing center.Based on tabu search and artificial ant colony algorithms,a joint optimization algorithm is proposed.The function of the joint optimization algorithm is to transfer the users from the area where actual arrival rate of the user is higher than jump arrival rate to the area where the actual arrival rate of the user is lower than the jump arrival rate,to achieve a balanced distribution of users in different areas.At the same time,this paper designed a stochastic transfer strategy for theislnd area.Through the computer simulations,this paper test the algorithm's performance,which show that the congestion rate in each area can be greatly reduced.under proposed algorithm.Next,this studied the case of multiple workflows.First of all,this paper proposed the performance metrics which measure the performance of hybrid computing:mixed gain and integrated gain.These metrics can be used to measure the user's average time-to-reductions in computing as compared to a single workflow scenario for a scheduling algorithm in the presence of multiple workflows.This paper obtains nine algorithms by referring to or improving the single workflow scheduling algorithm in cloud computing:HNFT,PCP,minLOAD,minLoadPCP,EZDCP,MCP,SMD,ETF,DLS。Simulation results show that,PCP,EZDCP,MCP,and SMD are more suitable for multiple workflow deployment case.