【摘要】 给出一种基于多属性信息综合评价的数据中心间数据传输调度方法。首先利用层次分析法,对多个属性之间的从属关系进行分析,然后根据属性值的分布差异由信息熵计算属性的相对权重,给出对于不同备选中转数据中心的综合评价。通过建立时间扩展图模型将基于多属性信息的数据中心间数据传输调度问题形式化为最小代价流问题。通过设置不同的参数,对一般存储转发方法、基于单属性信息的方法以及基于多属性的方法在不同属性上的性能优化差异做了实验对比。结果表明,相比于其他方法,该方法能够综合考虑多种属性信息,为数据中心间的数据传输选择综合评价最优的路径。更多还原

【Abstract】 A multi-attribute decision making method to schedule inter-datacenter bulk data was proposed.First,hierarchical analysis was used to analyze the dominant relationship among attributes.Based on the relative weights of attributes which were calculated via information entropy,the evaluation to each intermediate data center was provided.Then the multi-attribute aware scheduling problem was modeled on a time expanded graph and formulated as a minimal cost flow problem.Finally,the scheme with general store-and-forward mechanism and single attribute based approach under several different simulation settings was compared.The results demonstrate that our strategy can take multiple attributes into consideration,and can find the bulk transfer route with the best integrated evaluation.更多还原