WOBAN中基于令牌桶算法的自适应流量整形策略

摘要

光无线融合网络中光域和无线域存在速率不匹配以及业务分类不同的问题.为提高融合网络的QoS及充分利用网络资源,该文提出一种基于令牌桶算法的流量整形策略.根据EPON与WiMAX业务的QoS映射以及调度策略,以业务在整个网络中的相对优先级来决定令牌桶参数的选择,并结合更符合实际的自相似网络流量推导出最佳令牌桶参数计算公式,并在此基础上,提出基于优先级的共享缓存策略,进而,得到通过QoS指标的变化而自适应动态调整的流量整形策略.仿真结果表明,所提流量整形策略能够有效降低网络设备压力,同时提高整个融合网络的QoS.%There are mismatches of transmission rates and different service classifications between optical domain and wireless domain in Wireless-Optical Broadband Access Network (WOBAN). To improve the QoS of the integration network and make full use of network resources, a new algorithm based on token bucket traffic shaping policy is proposed. According to the service mapping rules and scheduling of EPON and WiMAX, the selection of token bucket parameters B determined based on the relative service priority of the entire network. This paper also combines with more realistic self-similar network traffic to derive the best formula for token bucket parameters and then proposes a shared buffer strategy on the basis of the service priority. Furthermore, an adaptively dynamically adjust traffic shaping policy is obtained through the change of QoS metrics. The simulation results show that the proposed strategy can effectively reduce traffic shaping network equipment pressure while improving the QoS of whole converged networks.