基于循环神经网络的电信行业容量数据预测方法

doi:10.11772/j.issn.1001-9081.2020101677

计算机应用 ›› 2021, Vol. 41 ›› Issue (8): 2373-2378.DOI: 10.11772/j.issn.1001-9081.2020101677

所属专题：前沿与综合应用

基于循环神经网络的电信行业容量数据预测方法

丁尹^1,2, 桑楠¹, 李晓瑜¹, 吴飞舟²

1. 电子科技大学信息与软件工程学院, 成都 610054;
2. 北京思特奇信息技术股份有限公司, 北京 100046

收稿日期:2020-10-29 修回日期:2021-01-05 出版日期:2021-08-10 发布日期:2021-01-27
通讯作者: 李晓瑜
作者简介:丁尹(1997-),女,四川宜宾人,硕士研究生,主要研究方向:数据分析、数据挖掘;桑楠(1964-),男,四川营山人,教授,硕士,主要研究方向:嵌入式实时高可信技术、嵌入式软件工程、中间件;李晓瑜(1984-),女,山东菏泽人,副教授,博士,CCF会员,主要研究方向:量子计算、机器学习、大数据分析;吴飞舟(1962-),男,重庆人,硕士,主要研究方向:软件工程、智慧运营。
基金资助:
四川省科技计划项目（18KJFWSF0388）。

Prediction method of capacity data in telecom industry based on recurrent neural network

DING Yin^1,2, SANG Nan¹, LI Xiaoyu¹, WU Feizhou²

1. School of Information and Software Engineering, University of Electronic Science and Technology of China, Chengdu Sichuan 610054, China;
2. SI-TECH Information Technology Company Limited, Beijing 100046, China

Received:2020-10-29 Revised:2021-01-05 Online:2021-08-10 Published:2021-01-27
Supported by:
This work is partially supported by the Sichuan Science and Technology Program (18KJFWSF0388).

摘要/Abstract

摘要： 在电信运维的容量预测过程中，存在容量指标和部署业务种类繁多的问题。现有研究未考虑指标数据类型的差异，对所有类型的数据使用同种预测方法，使得预测效果参差不齐。为了提升指标预测效率，提出一种指标数据类型分类方法，利用该方法将数据类型分为趋势型、周期型和不规则型。针对其中的周期型数据预测，提出基于双向循环神经网络（BiRNN）的周期型容量指标预测模型，记作BiRNN-BiLSTM-BI。首先，为分析容量数据的周期特征，提出一种忙闲分布分析算法；其次，搭建循环神经网络（RNN）模型，该模型包含一层BiRNN和一层双向长短时记忆网络（BiLSTM）；最后，充分利用系统忙闲分布信息，对BiRNN输出的结果进行优化。与传统的三次指数平滑、差分自回归移动平均（ARIMA）模型和反向传播（BP）神经网络模型进行比较的实验结果表明，在统一日志数据集和分布式缓存数据集上，提出的BiRNN-BiLSTM-BI模型的均方误差（MSE）分别比对比模型中表现最优的模型降低了15.16%和45.67%，可见预测准确率得到了很大程度的提升。

关键词: 双向循环神经网络, 长短时记忆网络, 容量预测, 忙闲分布, 智能运维

Abstract: In the capacity prediction process of telecom operation and maintenance, there are problems of too many capacity indicators and deployed business classes. Most of the existing researches do not consider the difference of indicator data types, and use the same prediction method for all types of data, which results in both good and bad prediction effects. In order to improve the efficiency of indicator prediction, a classification method of data type was proposed, and the data types were divided into trend type, periodic type and irregular type. Aiming at the prediction of periodical data, a periodic capacity indicator prediction model based on Bi-directional Recurrent Neural Network (BiRNN), called BiRNN-BiLSTM-BI, was proposed. Firstly, In order to analyze the periodic characteristics of capacity data, a busy and idle distribution analysis algorithm was proposed. Secondly, a Recurrent Neural Network (RNN) model was built, which included a layer of BiRNN and a layer of Bi-directional Long Short-Term Memory network (BiLSTM). Finally, the output of BiRNN was optimized by the system's busy and idle distribution information. Experimental results compared with the best one among Holt-Winters, AutoRregressive Integrated Moving Average (ARIMA) model and Back Propagation (BP) neural network model show that, the proposed BiRNN-BiLSTM-BI model has the Mean Square Error (MSE) reduced by 15.16% and 45.67% on the unified log dataset and the distributed cache service dataset respectively, showing that the prediction accuracy is greatly improved.

Key words: Bi-directional Recurrent Neural Network (BiRNN), Long Short-Term Memory network (LSTM), capacity prediction, busy and idle distribution, Artificial Intelligence for IT Operations (AIOps)

中图分类号:

TP391

丁尹, 桑楠, 李晓瑜, 吴飞舟. 基于循环神经网络的电信行业容量数据预测方法[J]. 计算机应用, 2021, 41(8): 2373-2378.

DING Yin, SANG Nan, LI Xiaoyu, WU Feizhou. Prediction method of capacity data in telecom industry based on recurrent neural network[J]. Journal of Computer Applications, 2021, 41(8): 2373-2378.

参考文献

[1] 周艳庆. 软件系统容量管理探究[J]. 中国金融电脑,2013(11):60-63. (ZHOU Y Q. Exploration of software system capacity management[J]. Financial Computer of China,2013(11):60-63.)
[2] 常润梅, 孟利青, 刘万军. 电信企业云计算数据中心容量管理[J]. 辽宁工程技术大学学报(自然科学版),2013,32(8):1112-1117.(CHANG R M,MENG L Q,LIU W J. Telecom enterprise cloud computing data center capacity management[J]. Journal of Liaoning Technical University(Natural Science),2013,32(8):1112-1117.)
[3] FLIESS M,JOIN C,BEKCHEVA M,et al. Easily implementable time series forecasting techniques for resource provisioning in cloud computing[C]//Proceedings of the 6th International Conference on Control, Decision and Information Technologies. Piscataway:IEEE,2019:48-53,
[4] FARAHNAKIAN F,LILJEBERG P,PLOSILA J. LiRCUP:linear regression based CPU usage prediction algorithm for live migration of virtual machines in data centers[C]//Proceedings of the 39th Euromicro Conference Series on Software Engineering and Advanced Applications. Piscataway:IEEE,2013:357-364.
[5] BELOGLAZOV A,BUYYA R. OpenStack Neat:a framework for dynamic and energy-efficient consolidation of virtual machines in OpenStack clouds[J]. Concurrency and Computation:Practice and Experience,2015,27(5):1310-1333.
[6] 闻静. 基于CPU利用率预测的虚拟机动态部署方法研究[D]. 沈阳:东北大学,2014:26-33.(WEN J. Research on the method of dynamic deployment of virtual machine based on prediction of CPU utilization[D]. Shenyang:Northeastern University,2014:26-33)
[7] WANG J,YAN Y M,GUO J. Research on the prediction model of CPU utilization based on ARIMA-BP neural network[J]. MATEC Web of Conferences,2016,65:No. 03009.
[8] 马小淋. 一种基于负载特征预测的容器云弹性伸缩策略[J]. 信息安全研究,2019,5(3):236-241.(MA X L. A container cloud elastic scaling strategy based on load characteristics prediction[J]. Journal of Information Security Research,2019,5(3):236-241.)
[9] 吴俊伟, 姜春茂. 一种BP神经网络和遗传算法混合的云平台负载预测方法[J]. 软件,2017,38(8):18-24.(WU J W,JIANG C M. A load forecasting method of cloud platform based on BP neural network and genetic algorithm[J]. Computer Engineering & Software,2017,38(8):18-24.)
[10] MASON K,DUGGAN M,BARRETT E,et al. Predicting host CPU utilization in the cloud using evolutionary neural networks[J]. Future Generation Computer Systems,2018,86:162-173.
[11] RAO S N, SHOBHA G, PRABHU S, et al. Time series forecasting methods suitable for prediction of CPU usage[C]//Proceedings of the 4th International Conference on Computational Systems and Information Technology for Sustainable Solution. Piscataway:IEEE,2019:1-5.
[12] TRAN N,NGUYEN T,NGUYEN B M,et al. A multivariate fuzzy time series resource forecast model for clouds using LSTM and data correlation analysis[J]. Procedia Computer Science, 2018,126:636-645.
[13] 中国石油化工股份有限公司胜利油田分公司物探研究院. 一种基于双向循环神经网络的储层预测方法:中国, 201811632631. X[P]. 2019-05-24. (Geophysical Research Institute of Shengli Oilfield Branch of SINOPEC. A reservoir prediction method based on bidirectional cyclic neural network:China,201811632631. X[P]. 2019-05-24.)
[14] GUPTA S,DINESH D A. Resource usage prediction of cloud workloads using deep bidirectional long short term memory networks[C]//Proceedings of the 2017 IEEE International Conference on Advanced Networks and Telecommunications Systems. Piscataway:IEEE,2017:1-6.
[15] 杨海民, 潘志松, 白玮. 时间序列预测方法综述[J]. 计算机科学,2019,46(1):21-28.(YANG H M,PANG Z S,BAI W. Overview of time series forecasting methods[J]. Computer Science,2019,46(1):21-28.)
[16] KUMAR J,GOOMER R,SINGH A K. Long Short Term Memory Recurrent Neural Network (LSTM-RNN) based workload forecasting model for cloud datacenters[J]. Procedia Computer Science,2018,125:676-682.
[17] MOREL M,ACHARD C,KULPA R,et al. Time-series averaging using constrained dynamic time warping with tolerance[J]. Pattern Recognition,2018,74:77-89.
[18] DUGGAN M,MASON K,DUGGAN J,et al. Predicting host CPU utilization in cloud computing using recurrent neural networks[C]//Proceedings of the 12th International Conference for Internet Technology and Secured Transactions. Piscataway:IEEE,2017:67-72.

基于循环神经网络的电信行业容量数据预测方法

Prediction method of capacity data in telecom industry based on recurrent neural network

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 12

编辑推荐

Metrics

[1]	张永斌, 常文欣, 孙连山, 张航. 基于字典的域名生成算法生成域名的检测方法[J]. 计算机应用, 2021, 41(9): 2609-2614.
[2]	赵小虎, 李晓. 基于多特征提取的图像语义描述算法[J]. 计算机应用, 2021, 41(6): 1640-1646.
[3]	邱宁佳, 王晓霞, 王鹏, 王艳春. 融合语法规则的双通道中文情感模型分析[J]. 计算机应用, 2021, 41(2): 318-323.
[4]	董永峰, 刘超, 王利琴, 李英双. 融合多跳关系路径信息的关系推理方法[J]. 计算机应用, 2021, 41(10): 2799-2805.
[5]	马停停, 冀天娇, 杨冠羽, 陈阳, 许文波, 刘宏图. 基于长短时记忆神经网络的手足口病发病趋势预测[J]. 计算机应用, 2021, 41(1): 265-269.
[6]	戎炜, 蒋哲远, 谢昭, 吴克伟. 基于聚类关联网络的群组行为识别[J]. 计算机应用, 2020, 40(9): 2507-2513.
[7]	傅洪亮, 雷沛之. 基于去噪自编码器和长短时记忆网络的语音测谎算法[J]. 计算机应用, 2020, 40(2): 589-594.
[8]	夏彬, 白宇轩, 殷俊杰. 基于生成对抗网络的系统日志级异常检测算法[J]. 计算机应用, 2020, 40(10): 2960-2966.
[9]	闻畅, 刘宇, 顾进广. 基于注意力机制的双向长短时记忆网络模型突发事件演化关系抽取[J]. 计算机应用, 2019, 39(6): 1646-1651.
[10]	郑毅, 李凤, 张丽, 刘守印. 基于长短时记忆网络的人体姿态检测方法[J]. 计算机应用, 2018, 38(6): 1568-1574.
[11]	李雅昆, 潘晴, Everett X. WANG. 基于改进的多层BLSTM的中文分词和标点预测[J]. 计算机应用, 2018, 38(5): 1278-1282.
[12]	冯伟, 易绵竹, 马延周. 基于TensorFlow的俄语词汇标音系统[J]. 计算机应用, 2018, 38(4): 971-977.