欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为解决网络系统复杂性所带来的服务路径上的节点失效问题,提出了一种基于Bellman动态规划的服务恢复方法.通过服务路径选择策略从当前所有路径中选择一条最优路径进行失效恢复,再通过对逻辑层网络自组织和设定效能函数将服务恢复决策问题转化为动态规划问题,最后通过Bellman动态规划算法实现最优相应路径选择,并按照问题的规模选择动态迁移或重建技术.实验结果表明,相比服务备份路径方法,Bellman动态规划恢复策略使得CPU负载峰值降低了20%,而服务响应率提高了30%,且恢复路径选择的时间复杂度由O（NM）降低为O（2MN2）.
英文摘要：
To resolve node failure in a service path caused by network system complexity,a service recovery method based upon Bellman dynamic programming was proposed in this paper.First,according to service path strategy,an optimal path selected from all existing ones served as the failure recovery foundation.Then,the service recovery decision was transferred to the dynamic programming problem by logical layer network self-organization and a utility function definition.Finally,the optimal path was calculated by using Bellman dynamic programming.Due to the scale of this problem,dynamic migration or reconstruction technology was applied to service recovery.Experimental results show that with the proposed Bellman dynamic programming method,the CPU load peak is reduced by 20%,the service response rate is increased by 30%,and the complexity of choosing a recovery path ID decreased from O（NM） to O（2MN2）,compared to the traditional service backup path method.
同期刊论文项目
同项目期刊论文
期刊信息
位置：