摘要
针对云存储的收费机制和内容,在分析已有的用户贪婪、服务器贪婪等启发式解决算法的基础上,提出改进的启发式云存储静态内容分发遗传算法;综合考虑资源的访问热度、资源的缺乏程度,提出基于热点预测和经济模型的动态内容分发技术;进而全面考虑当前网络带宽、边缘云存储节点性能及历史访问价值,提出概率匹配自收敛的云存储中内容分发负载均衡技术,并将提出的算法分别在模拟器CloudSim上进行测试,同时和现有的内容分发算法、负载均衡算法进行对比.实验结果证明,本文提出的算法能够应用到云存储内容分发技术当中,并且能够在提高内容分发效率的同时有效降低分发成本.
Charging mechanism and cloud storage has been composed on the basis of content delivery network.In this paper,the user greedy and the server greedy algorithm have been analyzed based on heuristic,an improved genetic algorithm for static content delivery technology in cloud-storage is promoted.Comprehensively considering of resource access popularity,lack degree,a content delivery technique has been posed on the basis of hot resource and economic model.Considering the current network bandwidth,the edge cloud storage node＇s performance and historical visit value,aproportional matching self-convergent content delivery load balancing technique has been proposed.The proposed algorithm is implemented on the CloudSim simulator,and fully tested with existing content delivery algorithm,load balancing algorithm.The experimental results can validate the proposed algorithms are well applied to the cloud storage content delivery.They can effectively improve content delivery at the same time reduce the delivery costs.
出处
《西南师范大学学报（自然科学版）》
CAS
北大核心
2016年第6期111-118,共8页
Journal of Southwest China Normal University(Natural Science Edition)
