-
对于FiWi网络而言,其前端点共有i个,格式类型包含:源地址、终端地址、帧的长度、校样请求码等。则节点i的帧长度[11]为:
$$ {F}_{i}={F}_{\rm long}\times i+{F}_{0}\text{,}i\in {S}_{M} $$ (1) 式中:F0表示除帧头部分与帧尾部分的数据开销;i表示无线端节点个数;Flong表示节点带宽请求的信号长度;SM表示节点集合,其中M表示节点编号。可以看出,在边缘云结构中,每一个节点对应的帧的长度是不同的,节点号与M相关,则公式(1)的最大值可以表示为:
$$ {F_i}\left| {_{\rm max }} \right. = {F_{\rm long}} \times {\rm max} \left\{ {\left( {M - 1} \right),\left( {N - M} \right)} \right\} + {F_0} $$ (2) 式中:M表示节点号;N表示最大光纤通信网络分束模块的最大分光比。设节点位置上载与传输的信息传输速率为Vupload和Vcom,则任意相邻两个节点之间的通信时延Tij为:
$$ {T_{ij}} = \frac{{{F_i}}}{{{V_{\rm com}}}} $$ (3) 由此可以通过迭代循环完成对最大传输时间进行求解,则其带宽有:
$$ {B_{\rm max }} = \frac{{\left[ {{\rm max} \left( {{T_i}} \right) - N{T_{ij}}} \right] \times {V_{\rm upload}}}}{{8N}} $$ (4) 由此根据每个节点带宽需求Bi,得到总带宽需求Btol=ΣBi。则通过对比Bi是否大于Bmax就能判断该节点的带宽需求是否超过带宽上限,从而确定该点是否存在传输时延。在计算得到所有节点的带宽后,为存在传输时延的节点释放任务,将传输路径转移至最近的非延迟节点,从而实现对边缘云数据通信路程的优化。
-
根据时序分析中对各个节点间数据通信的传输效果分析可知,在计算得到各个节点带宽占用率的基础上需要对边缘云中的光纤网络通信路径进行优化。对于邻近节点而言,对路径长度排序,路由的权重系数是根据跳数和跨域数决定[12],则i节点与j节点之间的权重系数Wi,j有:
$$ {{{W}}_{i,j}} = \frac{{{L_{i,j}}}}{{{L_{\max }}}} + \frac{{{D_{i,j}}}}{{{D_{\max }}}} $$ (5) 式中:Li,j为节点i到节点j的路径长度;Lmax表示路径最长的链路;Di,j为节点i和节点j所在的域;Dmax为包含节点i和节点j域的边缘云域。由此可以通过边缘云限定大幅降低路径优化算法的计算量。
-
根据以上时序分析与路径优化原则,可以给出基于边缘云计算的时序优化算法的具体实现步骤如下:
(1) 根据通信网络范围确定边缘云区域,并对边缘云覆盖区域内所有的光纤网络节点编号1~M,完成FiWi网络基础数据采集,包含地址信息、帧长度等;
(2) 计算节点i的帧长度Fi,并在对每一个节点的帧长度进行排序,获得Fi|max,将数据集合Fi与对比参数Fi|max保存至运算数据库中;
(3) 测试节点位置上载速率Vupload与现有网络中的传输速率Vcom,从而计算通信时延;
(4) 根据光纤网络参数计算最大传输带宽Bmax,并且将其作为分类依据对所有节点带宽进行迭代判断,将大于和小于该值的测试数据分别构建新的数组S1和S2;
(5) 通过S1创建路径优化子区域,完成对L参数的选择,通过S2计算重用路径的最优值,完成对D参数的选择,循环所有节点数据后完成路径优化。程序流程如图2所示。
Optimal design of optical fiber wireless network based on edge cloud computing
-
摘要: 为了提升网络边缘数据处理能力,满足终端大带宽和低时延的要求,构建了基于边缘基础设施的云计算平台,设计了具有动态带宽调整的光纤网络模型。提出了一种基于边缘云计算的时序优化算法,并将其应用于光纤无线网络。通过OPNET软件仿真分析了时序优化算法的传输时延均值,结果显示,优化后最大时延为43.1 ms,仅为传统方法的34.2%。实验对局域网内多个终端之间的数据通信进行分析,讨论了三种算法的传输能效、光纤信道利用率及传输能耗。实验结果显示,采用时序优化算法的测试结果具有明显改善,其传输能效提升了近1倍,边缘云数据传输时延均值信道利用率提升了约6.2%,网络传输能耗均值最优。该光纤无线网络模型及其优化算法在传输时延、信道利用率以及网络能耗方面具有明显提升。其在提升光纤通信链路选择及边缘端数据交互中具有一定的优势。Abstract: In order to improve the data processing capabilities of network edge and meet the requirements of large bandwidth and low latency of the terminal, a cloud computing platform based on edge infrastructure is built, and an optical fiber network model is designed with dynamic bandwidth adjustment. A timing optimization algorithm based on edge cloud computing is proposed and applied to optical fiber wireless networks. The average transmission delay of the timing optimization algorithm is simulated and analyzed through OPNET software, and the results show that the maximum delay after optimization is 43.1 ms, which is only 34.2% of the traditional method. The experiment analyzes the data communication between multiple terminals in the local area network. The transmission efficiency, optical fiber channel utilization and transmission energy consumption of the three algorithms have been discussed. The experimental results show that the test results using the timing optimization algorithm have been significantly improved. Its transmission efficiency has increased by nearly double, the channel utilization rate of edge cloud data average transmission delay has increased by about 6.2%, and the network transmission energy consumption has the best value. The optical fiber wireless network model and its optimization algorithm have significant improvements in transmission delay, channel utilization, and network energy consumption. It has certain advantages in improving the selection of optical fiber communication links and data interaction at the edge.
-
[1] Chen Bo. Power-aware virtual optical network provisioning in exible bandwidth optical networks [J]. Photon Netw Commun, 2016, 32(2): 300-309. doi: 10.1007/s11107-016-0609-4 [2] Zhou Shenghai, Song Wei, Gu Ying, et al. Research on accurate transmission method of relay protection signal in optical fiber communication network [J]. Automation & Instrumentation, 2018(12): 15-18. (in Chinese) [3] Chatterjee B C, Sarma N, Oki E. Routing and spectrum allocation in elastic optical networks: A tutorial [J]. IEEE Commun Surveys Tuts, 2015, 17(30): 1776-1800. [4] Masoudi A, Pilgrim J A, Newson T P, et al. Subsea cable condition monitoring with distributed optical fiber vibration sensor [J]. J Light Technol, 2019, 37(1): 1352-1358. [5] Wei Heng, Lu Lin, Pu Tao, et al. Fiber time delay fluctuations measurement based on one-way transfer [J]. Infrared and Laser Engineering, 2020, 49(8): 20200018. (in Chinese) doi: 10.3788/IRLA20200018 [6] Chowdhury M, Steinbach E, Kellerer W, et al. Context-aware task migration for HART-centric collaboration over FiWi based tactile internet infrastructures [J]. IEEE Transactions on Parallel & Distributed Systems, 2018, 29(6): 1231-1246. [7] Ali K, Shaikh A. Energy and delay aware routing algorithm for fiber-wireless networks [J]. Wireless Networks, 2014, 20(6): 1313-1320. [8] Mamechaoui S, Senouci S M, Didi F, et al. Energy efficient management for wireless mesh networks with green routers [J]. Mobile Networks and Applications, 2015, 20(5): 567-582. doi: 10.1007/s11036-015-0601-8 [9] Liu Y, Yang Y H, Han P C, et al. Virtual network embedding in fiber-wireless access networks for resource-efficient IoT service provisioning [J]. IEEE Access, 2019, 7(10): 65506-65517. [10] He Rong, Xie Hui, Fang Xuming. Proportional fair and QoS-aware hierarchical dynamic bandwidth allocation scheme for fiber-wireless networks [J]. Journal of Electronics & Infor-mation Technology, 2017, 39(6): 1409-1416. (in Chinese) doi: 10.11999/JEIT160731 [11] Sarigiannidis A, Nikopolitidis P. Quality-of-service -aware fair bandwidth allocation scheme for fibre wireless networks [J]. IET Networks, 2016, 5(3): 56-63. doi: 10.1049/iet-net.2015.0090 [12] Wang Xiao, Gu Huaxi, Yang Yintang, et al. A highly scalable optical network-on-chip with small network diameter and deadlock freedom [J]. IEEE Trans Integr Syst, 2016, 24(12): 3424-3436. [13] Chen J, Li J. Passive optical network based mobile backhaul enabling ultra-low latency for communications among base stations [J]. Journal of Optical Communications & Network-ing, 2017, 9(10): 855-863. doi: https://doi.org/10.1364/JOCN.9.000855 [14] Li M, Chen H L. Energy-efficient traffic regulation and scheduling for video streaming services over LTE a networks [J]. IEEE Transactions on Mobile Computing, 2018, 18(2): 334-347. [15] Hossain A, Hossain A D, Kouar M, et al. Revisiting FiWi: On the merits of a distributed upstream resource allocation scheme [J]. Journal of Optical Communications & Networking, 2017, 9(9): 773-781. doi: https://doi.org/10.1364/JOCN.9.000773 [16] Chang W, Hu Y, Shou G, et al. An offloading scheme leveraging on neighboring node resources for edge computing over fiber-wireless (FiWi) access networks [J]. China Communications, 2019, 16(11): 107-119. doi: 10.23919/JCC.2019.11.009