高级搜索

无线虚拟网络中基于自回归滑动平均预测的在线自适应虚拟资源分配算法

唐伦 杨希希 施颖洁 陈前斌

引用本文: 唐伦, 杨希希, 施颖洁, 陈前斌. 无线虚拟网络中基于自回归滑动平均预测的在线自适应虚拟资源分配算法[J]. 电子与信息学报, 2019, 41(1): 16-23. doi: 10.11999/JEIT180048 shu
Citation:  Lun TANG, Xixi YANG, Yingjie SHI, Qianbin CHEN. ARMA-prediction Based Online Adaptive Dynamic Resource Allocation in Wireless Virtualized Networks[J]. Journal of Electronics and Information Technology, 2019, 41(1): 16-23. doi: 10.11999/JEIT180048 shu

无线虚拟网络中基于自回归滑动平均预测的在线自适应虚拟资源分配算法

    作者简介: 唐伦: 男,1973年生,教授,博士生导师,主要研究方向为新一代无线通信网络、异构蜂窝网络、软件定义无线网络等;
    杨希希: 女,1992年生,硕士生,研究方向为网络虚拟化;
    施颖洁: 女,1993年生,硕士生,研究方向为网路切片;
    陈前斌: 男,1967年生,教授,博士生导师,主要研究方向为个人通信、多媒体信息处理与传输、下一代移动通信网络、异构蜂窝网络
    通讯作者: 杨希希,469519917@qq.com
  • 基金项目: 国家自然科学基金(61571073)

摘要: 该文针对无线虚拟化网络中业务的不确定和信息反馈的时延而引起虚拟资源分配不合理,提出一种基于自回归滑动平均(ARMA)预测的在线自适应虚拟资源分配算法。首先,该算法以保障虚拟网络队列上溢概率为目标对时频资源和缓存资源进行联合分配,并建立虚拟网络总成本最小化的理论分析模型。其次,考虑到虚拟网络对不同资源差异化的应用需求,设计了一种多时间尺度的资源动态调度机制,在长周期上基于ARMA模型的预测信息实现缓存资源的预留策略,在短周期上基于利用大偏差原理推导的队列上溢概率对虚拟网络优先级排序,并根据确定的优先级动态调度时频资源,从而满足各虚拟网络的业务需求。仿真结果表明,该算法可有效降低比特丢失率,同时提升物理资源的利用率。

English

    1. [1]

      AGIWAL M, ROY A, and SAXENA N. Next generation 5G wireless networks: A comprehensive survey[J]. IEEE Communications Surveys & Tutorials, 2017, 18(3): 1617–1655. doi: 10.1109/COMST.2016.2532458

    2. [2]

      KALIL M, AL-DWEIK A, SHARKH M A, et al. A framework for joint wireless network virtualization and cloud radio access networks for next generation wireless networks[J]. IEEE Access, 2017, 5: 20814–20827. doi: 10.1109/ACCESS.2017.2746666

    3. [3]

      ZHANG Haijun, LIU Na, CHU Xiaoli, et al. Network slicing based 5G and future mobile networks: Mobility, resource management, and challenges[J]. IEEE Communications Magazine, 2017, 55(8): 138–145. doi: 10.1109/MCOM.2017.1600940

    4. [4]

      RAHMAN M M, DESPINS C, and AFFERS S. Design optimization of wireless access virtualization based on cost & QoS trade-Off utility maximization[J]. IEEE Transactions on Wireless Communications, 2016, 15(9): 6146–6162. doi: 10.1109/TWC.2016.2580505

    5. [5]

      SALLENT O, PEREZ-ROMERO J, FERRUS R, et al. On radio access network slicing from a radio resource management perspective[J]. IEEE Wireless Communications, 2017, 24(5): 166–174. doi: 10.1109/MWC.2017.1600220WC

    6. [6]

      JIANG Menglan, CONDOLUCI M, and MAHMOODI T. Network slicing management & prioritization in 5G mobile systems[C]. The 22th European Wireless Conference, Oulu, Finland, 2016: 1–6.

    7. [7]

      ZHU Qixuan and ZHANG Xi. Game-theory based buffer-space and transmission-rate allocations for optimal energy-erfficiency over wireless virtual networks[C]. 2015 IEEE Global Communications Conference (GLOBECOM), San Diego, USA, 2015: 1–6.

    8. [8]

      AHMADI H, MACALUSO I, GOMEZ I, et al. Substitutability of spectrum and cloud-based antennas in virtualized wireless networks[J]. IEEE Wireless Communications, 2017, 24(2): 114–120. doi: 10.1109/MWC.2016.1500303WC

    9. [9]

      LEANH T, TRAN N, NGO D T, et al. Resource allocation for virtualized wireless networks with backhaul constraints[J]. IEEE Communications Letters, 2017, 21(1): 148–151. doi: 10.1109/LCOMM.2016.2617307

    10. [10]

      SCIANCALEPORE V, SAMDANIS K, COSTA-PEREZ X, et al. Mobile traffic forecasting for maximizing 5G network slicing resource utilization[C]. IEEE INFOCOM 2017-IEEE Conference on Computer Communications, Atlanta, USA, 2017: 1–9.

    11. [11]

      CHU Yenming, HUANG Nenfang, and LIN Shenghsiung. Quality of service provision in cloud-based storage system for multimedia delivery[J]. IEEE Systems Journal, 2014, 8(1): 292–303. doi: 10.1109/JSYST.2013.2257338

    12. [12]

      AMIRI M and MOHAMMAD-KHANLI L. Survey on prediction models of applications for resources provisioning in cloud[J]. Journal of Network & Computer Applications, 2017, 82: 93–113.

    13. [13]

      李捷, 刘先省, 韩志杰. 基于ARMA的无线传感器网络流量预测模型的研究[J]. 电子与信息学报, 2007, 29(5): 1224–1227.
      LI Jie, LIU Xianxing, and HAN Zhijie. Research on the ARMA based traffic prediction algorithm for wireless sensor network[J]. Journal of Electronics &Information Technology, 2007, 29(5): 1224–1227.

    14. [14]

      MANDJES M. Large Deviations for Gaussian Queues: Modelling Communication Networks[M]. Chichester: Wiley, 2007: 55–60.

    15. [15]

      DEMBO A and ZEITOUNI O. Large Deviations Techniques and Applications[M]. Berlin: Springer, 2010: 303–304.

    16. [16]

      YANG Jian, RAN Yongyi, CHEN Shuangwu, et al. Online source rate control for adaptive video streaming over HSPA and LTE-Style variable bit rate downlink channels[J]. IEEE Transactions on Vehicular Technology, 2016, 65(2): 643–657. doi: 10.1109/TVT.2015.2398515

    17. [17]

      GARDNER E Jr. Exponential smoothing: The state of the art—Part II[J]. International Journal of Forecasting, 2006, 22(4): 637–666. doi: 10.1016/j.ijforecast.2006.03.005

    1. [1]

      缪祥华, 单小撤. 基于密集连接卷积神经网络的入侵检测技术研究. 电子与信息学报, 2020, 41(0): 1-7.

    2. [2]

      游凌, 李伟浩, 张文林, 王科人. 基于深度神经网络的Morse码自动译码算法. 电子与信息学报, 2020, 41(0): 1-6.

    3. [3]

      向敏, 饶华阳, 张进进, 陈梦鑫. 基于GCN的软件定义电力通信网络路由控制策略. 电子与信息学报, 2020, 42(0): 1-8.

    4. [4]

      李劲松, 彭建华, 刘树新, 季新生. 一种基于线性规划的有向网络链路预测方法. 电子与信息学报, 2020, 41(0): 1-9.

  • 图 1  系统架构

    图 2  多时间尺度的资源配置示意图

    图 3  长周期上基于ARMA预测的缓存资源预留策略流程图

    图 4  不同方案平均资源成本

    图 5  不同方案平均资源利用率

    图 6  不同方案平均比特丢失率

    图 7  不同T对应的平均资源利用率

    图 8  不同T对应的平均比特丢失率

    图 9  各虚拟网络的平均负载实际值与预测值比较

    表 1  算法1:时频资源动态调度算法

     (1) 在短周期$t$上观察当前各虚拟网络队列状态${Q_k} \left( t \right)$、预留的    缓存资源大小${B_k} $
     (2) for $k = 1;k < K;k + + $ do
     (3)   计算${a_k} $,根据式(22)估计${{\widehat m}_k} $
     (4)   if ${{\widehat m}_k} \ge {a_k} $ then
     (5)    加入虚拟网络集合${{K}}_1 $,根据式(24)估计溢出剩余时间${T_k} $
     (6)   else
     (7)    加入虚拟网络集合${{K}}_2 $,执行黄金分割搜索算法估计      ${P_{\rm of}^k} \left( {t{\rm{ + }}T} \right)$
     (8)   end if
     (9) end for
     (10) while ${{K}}_1 \ne \varnothing $ do
     (11) 令$m = 1$,选择虚拟网络$k = {\arg \min }_{k \in {{K}}_1 } \left\{ {{T_k} } \right\}$
     (12) while ${A_k} \left( t \right) > {C_k} \left( t \right)$ do
     (13)  $\begin{aligned} & m \leftarrow m + 1,{C_k} \left( t \right) \leftarrow mr, \\ &N \leftarrow N - 1 \\ \end{aligned} $
     (14) end while
     (15) ${{K}}_1 = {{K}}_1 \backslash \left\{ k \right\}$
     (16) end while
     (17) while ${{K}}_2 \ne \varnothing$ do
     (18) 令$m = 1$,选择虚拟网络${k^*} = {\arg \max }_{{k^*} \in {{K}_2}} \left\{ {P_{\rm of}^{{k^*}}\left( {t{\rm{ + }}T} \right) - {\varepsilon _{{k^*}}}} \right\}$
     (19) 重复步骤(12)—步骤(14)
     (20) ${{K}}_2 = {{K}}_2 \backslash \left\{ {k^ * } \right\}$
     (21) end while
     (22) if $N \ne 0$ then
     (23) for $k = 1;k < K;k + + $ do
     (24)  if ${C_k} \left( t \right) < \left({Q_k} \left( t \right) + {A_k} \left( t \right)\right)$ then
     (25)  加入虚拟网络集合${{K}}_3 $
     (26)  end if
     (27) end for
     (28) while ${{K}}_3 \ne \varnothing $ and $N \ne 0$
     (29) 令$m = 1$,选择虚拟$k^{''} = {\arg \min }_{k^{''} \in {{K}}_3 } \left\{ {\alpha _{{k^{''}}}} \right\}$
     (30)  $ {{while}} \quad \left({Q_{{k^{''}}}} \left( t \right) + {A_{{k^{''}}}} \left( t \right)\right) > \left({{\bar C}_{{k^{''}}}} \left( t \right) + {C_{{k^{''}}}} \left( t \right)\right)\quad {{do}} $
     (31) $m \leftarrow m + 1,{{\bar C}_{{k^{''}}}} \left( t \right) \leftarrow mr,N \leftarrow N - 1$
     (32)  end while
     (33) ${{K}}_3 = {{K}}_3 \backslash \left\{ {k^{''} } \right\}$
     (34) end while
     (35) end if
    下载: 导出CSV

    表 2  仿真参数设置

    仿真参数仿真值
    虚拟网络数量2,3,4,5,6
    系统带宽10 MHz (50 RBs)
    短周期时长1 ms
    长周期时长300 ms
    负载到达过程泊松分布
    比特到达速率$\lambda = 58.7\ {\rm kbit} $/ms
    RB单价$\alpha $1.2, 2.0, 1.5 unit/RB
    缓存资源单价$\rho $8, 6, 4 unit/kbit
    队列上溢概率$\varepsilon $0.13, 0.05, 0.12
    滑动窗口大小${T_w} $60 ms
    平滑指数$\eta $0.7
    仿真时间6600 ms
    下载: 导出CSV
  • 加载中
图(9)表(2)
计量
  • PDF下载量:  45
  • 文章访问数:  518
  • HTML全文浏览量:  276
文章相关
  • 通讯作者:  杨希希, 469519917@qq.com
  • 收稿日期:  2018-01-15
  • 录用日期:  2018-09-26
  • 网络出版日期:  2018-10-19
  • 刊出日期:  2019-01-01
通讯作者: 陈斌, bchen63@163.com
  • 1. 

    沈阳化工大学材料科学与工程学院 沈阳 110142

  1. 本站搜索
  2. 百度学术搜索
  3. 万方数据库搜索
  4. CNKI搜索

/

返回文章