认知网络中基于服务信息的队列管理算法

通过感知网络外部环境，认知网络能够自主地调节并分配网络资源，从而保证网络服务的正常运行。当发生网络拥塞时，当前普遍应用的主动队列管理算法只能从数据层面对数据包进行丢弃，不具备服务QoS对底层队列的调控。本文针对以上问题，提出了一种综合服务等级以及服务传输特点的队列管理算法，引入了服务对数据队列的参与调节模型，保证了在网络拥塞环境下特定服务的服务质量。使用Matlab对算法进行仿真，表明该算法在发生网络拥塞时，能够依据服务等级和服务传输特点对数据队列进行调节，从数据层面体现了服务层面的优先策略，从而保证了高级别服务的服务质量。

Active Queue Management Based on Service Information for Cognitive Network

By sensing the environment of the network, cognitive network could adjust and allocate resource automatically to guarantee service running without disturb. When congestion occurs, the queue management widely used could only drop packets at raw data level without adjusting at quality of service level. This paper proposed a active queue management where service grading and transport protocol are important factors for queue management. The paper also introduces an adjust model to guarantee the quality of speciifc service when congestion occurs and simulate the model using MATLAB. The result proves that the adjust model could adjust the queue by service grading and transport protocol of service. The strategy of service priority is relfected at raw data level and guarantees the quality of service with priority theoretically.