All Title Author
Keywords Abstract


基于DPSO负载可控的虚拟网络映射算法

DOI: 10.3969/j.issn.1005-3026.2014.01.003, PP. 10-14

Keywords: 网络虚拟化,映射算法,虚拟网络,整数线性规划,离散粒子群算法

Full-Text   Cite this paper   Add to My Lib

Abstract:

针对多租赁模式下的虚拟网络映射问题,以降低底层链路负载、加快映射速度、提高底层物理资源利用率为目标,将离散粒子群算法与虚拟节点映射规则相结合,提出了物理节点可复用、负载可控制的MLB-VNE-SDPSO算法.该算法在兼顾CPU等主机资源利用率的前提下节约了物理链路的带宽资源,缩短了虚拟链路的映射过程.仿真实验表明,在保证网络负载的前提下,获得了较好的物理节点利用率,提高了虚拟网络的收益成本比.

References

[1]  [1]Chowdhury N ,Boutaba R.A survey of network virtualization[J].Computer Networks,2010,54(5):862-876.
[2]  Muntasir R R ,Issam A,Raouf B.Survivable virtual network embedding[J].Lecture Notes in Computer Science,2010,6091:40-52.
[3]  Andy B ,Nick F,Mark H,et al.InVINIveritas:realistic and controlled network experimentation[C]//Special Interest Group on Data Communication 2006.Pisa:ACM,2006:3-14.
[4]  Mosharaf C ,Muntasir R R,Raouf B.ViNEYard:virtual network embedding algorithms with coordinated node and link mapping[J].Networking,2011,20(1):206-219.
[5]  Cheng X ,Su S,Zhang Z B,et al.Virtual network embedding through topology awareness and optimization [J].Computer Networks,2012,56(6):1797-1813.
[6]  Lu J ,Turner J.Efficient mapping of virtual networks onto a shared substrate[R].Washington D C:Washington University,2006.
[7]  Zhu Y ,Ammar M.Algorithms for assigning substrate network resources to virtual network components [C]//International Conferenceon Computer Communications 2006.Barcelona:IEEE,2006:1-12.
[8]  Chowdhury N ,Rahman M,Boutaba R.Virtual network embedding with coordinated node and link mapping [C]//International Conference on Computer Communications 2009.Riode Janeiro:IEEE,2009:783-791.
[9]  程祥 ,张忠宝,苏森,等.基于粒子群优化的虚拟网络映射算法[J].电子学报,2011,39(10):2240-2244.(Cheng Xiang,Zhang Zhong-bao,Su Sen,et al.Virtual network embedding based on particle swarm optimization[J].Acta Electronica Sinica,2011,39(10):2240-2244.)
[10]  Kennedy J ,Eberhart R C.Particle swarm optimization[C]//International Conference on Neural Networks 1995.Perth:IEEE,1995:1942-1948.(上接第4页)由图4可以看出,随着惩罚系数比值增大,人力资源分配数也变大.此外,随着耐心变化系数增大,系数比值对人力资源分配数的影响越来越小.4结论本文针对带有排队信息提示的呼叫中心系统,进行了人力资源分配问题的研究,在考虑顾客行为和耐心变化的基础上,通过排队理论给出了系统稳态情况下的性能指标计算公式,并给出了采用固定值算法的最小化人力资源的计算方法.通过数值实验结果的分析,给出了由排队信息提示导致的顾客耐心变化对系统人力资源配置影响趋势.以上研究成果,对现实中具有排队信息提示的呼叫中心人力资源管理具有一定的指导作用.

Full-Text

comments powered by Disqus