Research On Workflow Scheduling Considering Security Factors In Cloud Computing Environment

With cloud computing technology gradually developing maturely, its application field is more and more widely, the workflow scheduling is an important content of cloud computing and information security issue remains one of the major challenges of cloud computing development. In the workflow scheduling of cloud environment, high-quality security services become more and more important. How to decrease the information security risk probability of scheduling and total execution time of all workflow instance while meet various constraint conditions, to reach equilibrium of workflow scheduling information security risk and efficiency, thereby enhancing the service capacity of scheduling system. And it has become a hot topic in theoretical research and practical application.The main contents of this paper have the following several aspects.(1) The model of workflow scheduling considering security factors in cloud computing environment is built, taking confidentiality, integrity, authentication of the three information security indicators into account. In addition, wing to the dynamic of cloud computing, when to measure the security risk probability of tasks of workflow instances that running on the virtual machine, in the case of no evaluation preference, it is difficult to give a suitable weight to every security indicator. Therefore entropy weight method is used to determine the weights of the three indicators for every task in workflow. In order to satisfy the needs of scheduling in the diversity of demand and the balance between different objectives, considering the minimum information security risk probability and all task execution time to conduct multi-objective optimization.(2) For the features of above model, improved NSGA-? algorithm is proposed to solve and opitimize above model. Two heuristic rules are blended in the algorithm. On the one hand, when the number of subsequent task have a big difference between different task, the priority algorithm of task rescheduling algorithm is improved, making the task that has the largest computing amount of subsequent task firstly be executed, consequently as for as possible decrease waiting time of task in workflow before performing and decrease vacancy rate of virtual machine. On the other hand, considering the different tasks having a different safety requirement for virtual machine's security level, virtual allocation rule considering security factor is proposed, to promote the decreasing of information security risk probability. Through simulation experiment, the results show that the method can well realize the balance of execution time and scheduling risk probability.The proposed multi-objective workflow scheduling method considering risk factors in cloud environment and solving algorithm of this paper,can further meet the diversity of workflow scheduling demand.