复杂光纤网络流量数据快速调度模型分析
|
摘要
当前光纤网络调度流量数据时,调度后流量数据在光纤网络中的传输速度较低,不能合理的利用光纤网络的链路,存在流量数据平均传输时延高、链路利用率低的问题。提出复杂光纤网络流量数据快速调度模型分析方法,通过令牌桶算法平滑复杂光纤网络中存在的流量数据,利用延迟速率服务模型计算光纤网络中会话的延迟,依据缓冲器的大小得到忙期中会话的分组丢失数,根据分组丢失数计算分组在复杂光纤网络中的最大丢失率,结合最大丢失数和丢失率得到光纤网络中会话的延迟。在邻居节点滑动窗口中通过数据稀有性和数据需求信息计算流量数据在光纤网络中的优先级,根据节点在光纤网络中下载和上传的流量数据计算节点的贡献率。结合会话延迟、流量数据优先级和节点贡献率构建复杂光纤网络流量数据快速调度模型,并采用粒子群算法求解调度模型,输出最优解,完成复杂光纤网络中流量数据的快速调度。实验结果表明,所提方法的流量数据平均传输时延低、链路利用率高。
When the current fiber network schedules traffic data,the transmission speed of the scheduled traffic data in the fiber network is low,and the link of the fiber network cannot be reasonably utilized. The average transmission time of the traffic data is high and the link utilization rate is low. Therefore,a fast scheduling model analysis method for complex fiber network traffic data is proposed. The token bucket algorithm is used to smooth the traffic data existing in the complex fiber network. The delay rate service model is used to calculate the delay of the session in the fiber network. The number of packet loss of the session in the busy period is obtained according to the size of the buffer. The maximum loss rate of the packet in the complex optical network is calculated according to the packet loss number,and the delay of the session in the optical network is obtained by combining the maximum loss number and the loss rate.The priority of the traffic data in the optical network is calculated by the data rarity and the data requirement information in the sliding window of the neighbor node,and the contribution rate of the node is calculated according to the traffic data downloaded and uploaded by the node in the optical network. Combining session delay with traffic data priority and node contribution rate,a fast scheduling model of complex fiber network traffic data is constructed. The particle swarm optimization algorithm is used to solve the scheduling model,and the optimal solution is output to complete the fast scheduling of traffic data in complex optical networks. The experimental results show that the traffic data of the proposed method has low average transmission delay and high link utilization.
When the current fiber network schedules traffic data,the transmission speed of the scheduled traffic data in the fiber network is low,and the link of the fiber network cannot be reasonably utilized. The average transmission time of the traffic data is high and the link utilization rate is low. Therefore,a fast scheduling model analysis method for complex fiber network traffic data is proposed. The token bucket algorithm is used to smooth the traffic data existing in the complex fiber network. The delay rate service model is used to calculate the delay of the session in the fiber network. The number of packet loss of the session in the busy period is obtained according to the size of the buffer. The maximum loss rate of the packet in the complex optical network is calculated according to the packet loss number,and the delay of the session in the optical network is obtained by combining the maximum loss number and the loss rate.The priority of the traffic data in the optical network is calculated by the data rarity and the data requirement information in the sliding window of the neighbor node,and the contribution rate of the node is calculated according to the traffic data downloaded and uploaded by the node in the optical network. Combining session delay with traffic data priority and node contribution rate,a fast scheduling model of complex fiber network traffic data is constructed. The particle swarm optimization algorithm is used to solve the scheduling model,and the optimal solution is output to complete the fast scheduling of traffic data in complex optical networks. The experimental results show that the traffic data of the proposed method has low average transmission delay and high link utilization.
引文
[1]熊余,高玉鹏,王汝言.弹性光网络中负载均衡的优先共享通路保护[J].上海交通大学学报,2017,51(5):610-618.
[2]李炳乾,王勇,吕颖,等.光纤通道信用更新驱动的流量与差错控制机制[J].计算机工程,2016,42(7):49-53.
[3]刘春桐,张正义,李洪才,等.基于FBG的液压系统中流量/压力/温度同时测量技术[J].光子学报,2016,45(11):10-16.
[4]李周,许红升,叶彬,等.电力通信网结构优化及拓扑生成算法[J].电气自动化,2017,39(5):20-23.
[5]王玉宝,刘亚冲.基于粗时分复用技术的光纤光栅传感系统研究[J].光学技术,2016,42(5):441-444.
[6]吕华辉,林志达,黄宏聪.业务优先级感知的网络流量调度机制[J].电信科学,2017,33(7):151-158.
[7]林晚晴,李峭,赵露茜,等.基于Crossbar的混合关键性流量调度算法[J].电光与控制,2016,23(7):6-10.
[8]刘焕淋,徐一帆,陈勇.基于频谱感知的业务分割-合并的弹性光网络资源分配策略[J].电子与信息学报,2016,38(4):892-898.
[9]袁小虎,李春文.光纤量子通信网络路由选择协议[J].控制理论与应用,2017,34(11):1522-1527.
[10]谭小虎,王勇,褚文奎,等.时间触发光纤通道调度算法确定性分析[J].计算机工程与应用,2018,54(15):97-101.
[11]丁永强,王勇,谭小虎,等.光纤通道调度算法研究[J].火力与指挥控制,2017,42(11):77-81.
[12]刘安,谭小虎,褚文奎,等.实时任务调度光纤通道交换机的设计与实现[J].光通信技术,2017,41(8):40-43.
[13]熊余,蒋婧,张振振,等.TWDM-PON中带有业务区分的混合资源调度算法[J].电子学报,2017,45(6):1490-1497.
[14]许鹏,张继栋.基于改进遗传算法的光纤网络异常数据检测[J].科技通报,2016,32(7):163-166.
[15]田鸿.强电磁干扰下光纤网络的数据传输系统改进设计[J].现代电子技术,2016,39(20):53-56.
[16]邓伟,周桂平,范军丽,等.基于软件定义网络(SDN)的电力光纤到户带宽分配技术[J].电力建设,2017,38(11):81-86.
[2]李炳乾,王勇,吕颖,等.光纤通道信用更新驱动的流量与差错控制机制[J].计算机工程,2016,42(7):49-53.
[3]刘春桐,张正义,李洪才,等.基于FBG的液压系统中流量/压力/温度同时测量技术[J].光子学报,2016,45(11):10-16.
[4]李周,许红升,叶彬,等.电力通信网结构优化及拓扑生成算法[J].电气自动化,2017,39(5):20-23.
[5]王玉宝,刘亚冲.基于粗时分复用技术的光纤光栅传感系统研究[J].光学技术,2016,42(5):441-444.
[6]吕华辉,林志达,黄宏聪.业务优先级感知的网络流量调度机制[J].电信科学,2017,33(7):151-158.
[7]林晚晴,李峭,赵露茜,等.基于Crossbar的混合关键性流量调度算法[J].电光与控制,2016,23(7):6-10.
[8]刘焕淋,徐一帆,陈勇.基于频谱感知的业务分割-合并的弹性光网络资源分配策略[J].电子与信息学报,2016,38(4):892-898.
[9]袁小虎,李春文.光纤量子通信网络路由选择协议[J].控制理论与应用,2017,34(11):1522-1527.
[10]谭小虎,王勇,褚文奎,等.时间触发光纤通道调度算法确定性分析[J].计算机工程与应用,2018,54(15):97-101.
[11]丁永强,王勇,谭小虎,等.光纤通道调度算法研究[J].火力与指挥控制,2017,42(11):77-81.
[12]刘安,谭小虎,褚文奎,等.实时任务调度光纤通道交换机的设计与实现[J].光通信技术,2017,41(8):40-43.
[13]熊余,蒋婧,张振振,等.TWDM-PON中带有业务区分的混合资源调度算法[J].电子学报,2017,45(6):1490-1497.
[14]许鹏,张继栋.基于改进遗传算法的光纤网络异常数据检测[J].科技通报,2016,32(7):163-166.
[15]田鸿.强电磁干扰下光纤网络的数据传输系统改进设计[J].现代电子技术,2016,39(20):53-56.
[16]邓伟,周桂平,范军丽,等.基于软件定义网络(SDN)的电力光纤到户带宽分配技术[J].电力建设,2017,38(11):81-86.