Research Of Task Scheduling Strategy On Cloud Cmputing Evironment

Cloud computing set off a wave of computing revolution in the information industry as a new technology,which organized the dynamically and extended resources can provide on-demand services for users through the Internet.It is a kind of three-tier architecture composed of infrastructure as a service, platform as a service, software as a service and involved scheduling problem. The so-called cloud scheduling refers to in special cloud computing environment, allocating and managing resources for user’s tasks according to certain resources usage rules. Cloud scheduling strategy can been divided into two categories:On the one hand, the algorithm pursuiting the time efficient instead of resource utilization and beneficial result. On the other hand:The existing algorithms aims to shortest completion time or cost spent goal,Lacking of comprehensive improve for the quality of service. With the rapid development of cloud computing technology, the efficient cloud task scheduling strategy has become the focus of Cloud computing research. Analysising the shortcomings of existing algorithms, The research article expanded for the followings:1. Task scheduling strategy based on the dynamic adaptive ant colony algorithm were proposed to solve the drawbacks like slow convergence, easily falling into local optimal which were existed in the ant colony algorithm. Chaos disrupted were introduced when selecting the resource node,the pheromone evaporation factor were adjusted adaptively based on nodes’pheromone and the pheromone were updated dynamically according to the solution’s performance for shortening the time span. The algorithm referenceed task properties to build mathematical model and solved the task scheduling problem effectively.2. Although the dynamic adaptive ant colony algorithm has made promoting for time span,it still doesn’t meet maximize economic efficiency and resource utilization purpose of the cloud computing. The third part of this commitment engaging in premising the algorithm time performance to maximize resource utilization,proposed single ant colony clustering Min-Min algorithm. This algorithm calculated the tasks’average similarity and combined the ant colony algorithm to cluster tasks.Then dispatched works by means of Min-Min algorithm according to task sets’load balancing matrix. The algorithm improved resource utilization while guaranting the basic time performance.