一种基于业务识别的下一代网络资源分配系统研究

Study on the resource distribution system for next generation network based on service identification

基于业务识别以MPLS协议为基础提出了一种下一代网络资源优化分配系统,并研究可行的实现方法。系统的业务识别模块采用领域模型和改造的抽象工厂模式进行构造,能够做到各种识别策略的无缝部署和多层树形扩展。系统中下一代网络优化资源分配模型(ORDMNGN),根据下一代网络多业务类的特点进行资源分配优化建模与博弈分析,旨在提高全网效应值同时实现资源的优化分配和实现业务流的服务质量(QoS)保证;模型的实现采用了双链式结构,提高了系统的可扩展性。最后在合作博弈框架下对提出的系统进行了仿真验证,结果表明了方法的有效性和实用性。

Based on service identification and MPLS protocol, we present an optimal resource distribution system for Next Generation Network （NGN） and study its feasible implementation approaches. The service identification model is constructed by Domain Model and by improved Abstract Factory Model, so it can achieve tree extension in hierarchies and different identification policies can be added freely without exchanging existing models. The Optimal Resource Distribution Model of Next Generation Network （ORDMNGN） in this system models diverse service styles＇ resource distribution synthetically by game theory, aiming at increasing the efficiency of whole network and implementing service flows＇ QoS （Quality of Service） guarantee. ORDMNGN uses double chain framework, so it improves system＇s extensibility. Finally, simulation results under cooperative game environment show system＇s validity and practibility.