基于链路预测的SDN组播树故障恢复机制  被引量:1

Fault Recovery Mechanism of SDN Multicast Tree Based on Link Prediction

在线阅读下载全文

作  者:崔丽丽 曾学文[1,2] 朱小勇 CUI Lili;ZENG Xuewen;ZHU Xiaoyong(National Network New Media Engineering Research Center,Institute of Acoustics,Institute of Acoustics,Chinese Academy of Sciences,Beijing,100190,China;University of Chinese Academy of Sciences,Beijing,100049,China)

机构地区:[1]中国科学院声学研究所国家网络新媒体工程技术研究中心,北京100190 [2]中国科学院大学,北京100049

出  处:《网络新媒体技术》2022年第3期17-24,37,共9页Network New Media Technology

基  金:中国科学院战略性科技先导专项课题(XDC02070100)。

摘  要:软件定义网络架构下的组播数据按照组播树复制分发时,会出现链路故障的情况,而传统主动式恢复机制中,预先设置的备份组播树并不是动态更新,影响了组播数据传输的可靠性。针对备份组播树可能出现的链路拥塞问题,提出一种基于链路预测的恢复机制。基于链路负载变化的规律,使用最小二乘支持向量回归预测模型得到预测可用带宽,将链路预测模型与备份组播树周期更新机制相结合,实现组播数据的可靠传输。仿真结果显示,在同等数据发送速率下,该机制备份组播树的链路带宽占用率最低,可根据链路最新状况选择最优备份组播树,较好地防止故障恢复后的链路拥塞,达到全局负载均衡的效果。In software defined network,when the multicast data architecture is copied and distributed according to the multicast tree,there will be a link failure.However,in the traditional active recovery mechanism,the backup multicast tree is not updated dynamically,which affects the reliability of multicast data transmission.Aiming at the possible link congestion problem of backup multicast tree,a recovery mechanism based on link prediction is proposed.Based on the change law of link load,the least squares support vector regression prediction model is used to get the predicted available bandwidth.The link prediction model is combined with the periodic update mechanism of backup multicast tree to realize the reliable transmission of multicast data.The simulation results show that the bandwidth occupancy of the backup multicast tree is the lowest under the same data transmission rate.According to the latest status of the link,the optimal backup multicast tree can be selected,so as to better prevent the link congestion after fault recovery and achieve the effect of global load balancing.

关 键 词:链路预测 软件定义网络 组播树 故障恢复 链路拥塞 

分 类 号:TP393.02[自动化与计算机技术—计算机应用技术]

 

参考文献:

正在载入数据...

 

二级参考文献:

正在载入数据...

 

耦合文献:

正在载入数据...

 

引证文献:

正在载入数据...

 

二级引证文献:

正在载入数据...

 

同被引文献:

正在载入数据...

 

相关期刊文献:

正在载入数据...

相关的主题
相关的作者对象
相关的机构对象