论文部分内容阅读
以HFC网络为背景,针对嵌入式设备的存储资源有限性、网络带宽有限性以及宽带网络中多业务流的特点,提出一种缓冲区管理和分组调度联合算法.该算法通过控制业务流队列长度和设置业务流优先级方程的方式来管理缓冲区和带宽的分配,使内存管理和分组调度协调配合,从而为不同类型的网络数据提供区分服务.通过测试,在本文的仿真设置下,当网络资源紧张时,使用该算法后,可以达到实时业务流的超时数据比未使用该算法时的超时数据少89.6%的效果;而非实时业务流也可以达到平均丢包率比不使用该算法时的平均丢包率低90%的效果.
Based on the HFC network background, this paper proposes a joint algorithm of buffer management and packet scheduling in view of the limited storage resources of embedded devices, the limited network bandwidth and the characteristics of multi-service flows in broadband networks. By controlling the length of the service flow queue And set up the business flow priority equation way to manage the allocation of buffer and bandwidth, memory management and packet scheduling coordination, so as to provide different types of network data to provide differentiated services.Through testing, in this paper, the simulation settings, when the network When the resource is tight, the algorithm can achieve the effect that the time-out data of the real-time service flow is 89.6% less than the time-out data when the algorithm is not used, and the non-real-time traffic flow can also achieve the average packet loss ratio than when not using the algorithm The average packet loss rate of 90% lower effect.