一种基于QoS和动态负载均衡的路由策略  被引量:4

A Routing Strategy Based on QoS and Dynamic Load Balancing

在线阅读下载全文

作  者:孙杰[1] 李莉[1] 沈苏彬[1] 

机构地区:[1]南京邮电大学计算机学院软件学院,江苏南京210003

出  处:《计算机技术与发展》2016年第11期188-194,共7页Computer Technology and Development

基  金:江苏省未来网络前瞻性研究资助项目(BY2013095-1-08)

摘  要:为了提高SDN网络负载均衡调度的针对性和准确性,设计了SAS(Scheduling According to Stickiness)算法。该算法提出了链路粘值的概念,用链路粘值预估调度对流性能影响的大小,通过优先调度粘值低的流来减小调度对流性能的影响,达到优化调度的目的。在重路由过程中,该算法兼顾了流的QoS需求,并用近似算法进行多QoS约束的最优路径选择。基于Floodlight开源控制平台设计和实现了相应的原型系统,并根据负载均衡效果和负载均衡生成流的QoS两个测试目标,设计和实现了测试的方案。实验测试结果表明,在不同QoS需求的数据流竞争网络资源、导致网络负载偏离均衡状态的情况下,相较DLB、LABERIO机制,提出的技术方案在提高带宽利用率的同时可以兼顾分组流的QoS需求,并且可以降低对流性能的影响。In order to improve pertinence and accuracy of network load balancing scheduling for SDN (Software-Defined Networking), a SAS algorithm is designed. This algorithm puts forward the concept of link stickiness, and applies link stickiness to estimate the impacts of scheduling on flow performance,by preferentially scheduling flow with low stickiness to reduce the impacts of scheduling on flow performance, so as to achieve the purpose of optimized scheduling. In the process of re-routing, it takes QoS needs of flow into consideration,and applies approximation algorithm to calculate the optimal route under multi-QoS constraints. Then a prototype system is designed and implemented based on Floodlight,an open control platform. Aiming at two test targets, including the load_balancing effect and its generated flows' QoS, the test approach is designed and implemented. Experimental test results show that in case of different QoS requirement flows competing for limited network resources and resulting in network load deviations from the balance state ,compared with DLB and LABERIO,the technology approach proposed can improve the bandwidth utilization while fulfilling the requirements of QoS on flow, and reduce the influence on flow' s performance.

关 键 词:软件定义网络 路由策略 动态负载均衡 链路粘值 

分 类 号:TP392[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