泛在末梢环境下基于动态备用的业务应急机制
doi: 10.3724/SP.J.1146.2012.01460
A Dynamic Backup-based Service Contingency Mechanism in Ubiquitous Stub Environments
-
摘要: 在移动自组织网络(MANET)组成的泛在末梢环境下,由于终端能力有限及网络拓扑的动态变化,无法保证持续高质量的异构业务。该文提出一种基于动态备用的业务应急机制,实现备用指标函数的构建和业务修复策略的设计,包括业务发现、业务选择和业务维护3个阶段。首先,该文综合终端个体能力、协同能力及终端移动性预测,抽象出备用指标函数模型。其次,针对业务选择阶段提出基于时间预测算法(TPA)来构建稳健的业务执行及备用的终端集合。再次,针对业务维护过程的失效,提出全局备用算法(GBA)来动态维持业务的持续性。最后,通过仿真表明在该机制作用下,业务中断次数少于基于动态监督器的业务恢复方法(DMBSRS)和冷备用业务替代策略(CBSRS)约30%~45%和14%~28%,在效益方面分别高于这两者约10%~40%和0%~25%,同时具有最小重启时间,由此证明该机制具有较优的效率。
-
关键词:
- 移动自组织网络(MANET) /
- 基于动态备用的业务应急机制 /
- 移动性预测 /
- 全局备用
Abstract: Because devices limited ability and networks dynamic topology can not ensure persistent high-quality heterogeneous services in Mobile Ad hoc NETwork (MANET), a Dynamic Backup-based Service Contingency Mechanism (DBSCM) is proposed, which includes three stages: service request, select and maintenance. Firstly, the individual and cooperative capability is fully taken into consideration to model the backup index function. Secondly, Time-Prediction based Algorithm (TPA) is proposed to construct robust devices set in service selection. Then Global Backup Algorithm (GBA) is proposed to improve execution persistence in service maintenance. Finally, the simulation results demonstrate that the mechanism lowers 30%~45% and 14%~28% about interrupt numbers, improves 10%~40% and 0%~25% about utility and lowers restart-time than Dynamic Monitor Based Service Recovery Strategy (DMBSRS) and Cold Backup Service Replacement Strategy (CBSRS).
计量
- 文章访问数: 2180
- HTML全文浏览量: 96
- PDF下载量: 815
- 被引次数: 0