An improved algorithm based on ACO for cloud service PDTs scheduling

作者：Xue Shengjun; Zhang Jie^*; Xu Xiaolong

来源：Advances in Information Sciences and Service Sciences, 2012, 4(18): 340-348.

DOI：10.4156/AISS.vol4.issue18.41

• Cloud services
• Colony size
• Computing technology
• Dependent tasks
• High costs
• Improved ant colony optimization
• Independent tasks
• PDTs
• PIACO
• QoS (quality of service)
• Scheduling models
• Time requirements
• Utilization efficiency

With the increasing popularity of cloud computing technology, a great many applications are submitted to the cloud platforms for execution. Faced with the low utilization efficiencies and high cost of resources, users need cloud services with "lowest pay and highest efficiency". In this paper, we define a new scheduling model for Partly Dependent Tasks (PDTs) which merge workflow tasks and independent tasks together. Moreover, the scheduling model also takes the time requirements of QoS (Quality of Service) constrains and cost reduction into consideration. In order to meet these requirements, an improved ant colony optimization algorithm also put forward to schedule the PDTs (PIACO). Furthermore, we experiment with a set of PDTs by varying many related parameters such as the iterations, colony size and different deadline settings. Finally, the proposed algorithm PIACO proves to be efficient and suitable for the cloud PDTs scheduling.

• 出版日期2012
• 单位南京大学