论文部分内容阅读
针对命名数据网络(NDN)难以保障多样化业务服务需求问题,提出了多态路由承载的内容分发模型(CDMPR)。该模型通过内容请求、路由计算、内容查表和内容缓存四种类型的重构模块组合来承载不同特征需求的业务传输,详细设计了面向实时业务、非实时流媒体业务和用户自产生业务的多态路由算法,分别为改进蚁群优化路由算法、循迹路由策略和基于导向性副本通告的捷径路由机制。仿真结果表明,相比基于业务类型的多样化内容分发机制(DCDS)算法,CDMPR将网络节点平均缓存命中率提高了4.45%~13.8%;相比内容中心网络(CCN)算法,三种业务的平均响应时延分别缩短了29.17%~53.02%。CDMPR实现了对多样化业务差异化服务需求的支持。
Aiming at the problem that it is difficult for NDN to guarantee diversified service demand, a Content Delivery Model (CDMPR) based on multi-state routing is proposed. The model carries four different types of refactoring modules such as content request, route calculation, content look-up table and content cache to carry the service transmission with different features and needs. The model is designed in detail for real-time service, non-real-time streaming media service and user self-generated service Polymorphic routing algorithm, respectively, for the improvement of ant colony optimization routing algorithm, tracking routing strategy and guided copy-based notification shortcuts routing mechanism. The simulation results show that CDMPR improves the average cache hit rate of network nodes by 4.45% ~ 13.8% compared with the DCDS algorithm based on service types. Compared with the Content Centric Network (CCN) algorithm, The average response delay was shortened 29.17% ~ 53.02%. CDMPR enables support for diverse business differentiated service needs.