MPLS与DiffServ结合模型下的队列调度研究

Study on Queue Scheduling in MPLS and DiffServ-Combined Model

基于如何提高骨干网络的服务质量展开研究。首先介绍了MPLS与DiffServ结合模型;其次,结合该模型的特点深入研究了队列调度算法,并对WRR算法进行了改进,改进算法可以根据网络负载的实际状况动态的改变队列权值,实现带宽资源的动态分配,以提高算法的自适应性;最后,利用NS2仿真平台,在MPLS与DiffServ结合模型下对WRR算法和改进算法进行仿真,结果表明,改进算法能减少分组丢包率,提高网络资源的利用率。

The paper focuses on how to improve QoS in the Internet backbones. Firstly, the combined model of MPLS and DiffServ is presented, then the at-present queue scheduling algorithm based on the combined model of MPLS and DiffServ is analyzed, and the WRR algorithm improved. The improved algorithm could change the queuing weight dynamically according to the actual situation of network load, realize the dynamical assignment of bandwidth resource, and raise the self-adaptability of the algorithm. Finally, with NS2 simulation platform, WRR and the improved algorithm are simulated, the result indicates that the improved algorithm could reduce the probability of packet-loss ratio and raise the utilization of network resource.