基于GRU-LSTM组合模型的云计算资源负载预测研究

doi:10.19678/j.issn.1000-3428.0062452

摘要/Abstract

摘要： 日益增多的应用部署在云端使得云数据中心的功耗波动剧烈，从而导致云数据中心资源利用率不平衡，高效的负载预测是解决该问题的关键技术。针对目前负载预测模型预测精度低、预测时间长的问题，建立一种基于门控循环单元（GRU）与长短期记忆（LSTM）网络的组合预测模型GRU-LSTM。该模型的网络结构包括3层，第一层采用GRU，利用GRU参数少、易收敛的特点减少模型训练时间，第二、第三层采用LSTM，结合LSTM参数多的优势提高模型的预测精度。在此基础上，对数据集作缺失值处理和标准化处理，使用随机森林算法对原始序列进行特征选择后得到一组新的序列值，将该序列值作为GRU-LSTM组合预测模型的输入，以对云计算资源进行高效预测。在集群公开数据集Cluster-trace-v2018上进行实验，结果表明，与传统的单一预测模型ARIMA、LSTM、GRU以及现有的组合预测模型ARIMA-LSTM、Refined LSTM等相比，GRU-LSTM模型预测结果的均方误差减少6~9，预测时间平均缩短约10%。

关键词: 云计算, 负载预测, 预测模型, 门控循环单元, 长短期记忆网络

Abstract: Increasingly more applications are deployed in the cloud, causing the violent power fluctuation of cloud data consumption and unbalancing resource usage in cloud data centers.Efficient load forecasting is an essential technology for solving these problems.Targeting the low prediction accuracy and long prediction time of current load prediction models, a combined prediction GRU-LSTM model based on Gated Recurrent Unit(GRU) and Long-Short Term Memory (LSTM) network is established.The network structure of the model includes three layers.The first layer adopts GRU, which reduces the training time of the model using the advantages of having few GRU parameters and easy convergence.The second and third layers adopt LSTM, which improves the prediction accuracy of the model by combining the advantages of many LSTM parameters.On this basis, the missing values and standardization of the dataset are processed.After the feature selection of the original sequence using the random forest algorithm, a set of new sequence values are obtained. The sequence values are used as the input of the GRU-LSTM combined prediction model to efficiently predict the cloud computing resources.Experiments are performed on the Cluster-trace-v2018 public dataset.Using the proposed GRU-LSTM model, the Mean Square Error(MSE) of the prediction results and average prediction time reduced by 6~9 and shortened by approximately 10%, respectively, compared with the traditional single prediction ARIMA, LSTM, and GRU models and existing combined prediction ARIMA-LSTM and Refined LSTM models.

Key words: cloud computing, load forecasting, forecasting model, Gated Recurrent Unit(GRU), Long-Short Term Memory(LSTM) network

中图分类号:

TP391

贺小伟, 徐靖杰, 王宾, 吴昊, 张博文. 基于GRU-LSTM组合模型的云计算资源负载预测研究[J]. 计算机工程, 2022, 48(5): 11-17,34.

HE Xiaowei, XU Jingjie, WANG Bin, WU Hao, ZHANG Bowen. Research on Cloud Computing Resource Load Forecasting Based on GRU-LSTM Combination Model[J]. Computer Engineering, 2022, 48(5): 11-17,34.

https://www.ecice06.com/CN/Y2022/V48/I5/11

图/表 11

20220723172359

20220723172404

20220723172407

20220723172411

20220723172415

20220723172419

20220723172423

20220723172426

20220723172430

20220723172434

20220723172437

参考文献

[1] BOLIN J, YANG M K.Cloud computing:cost, security, and performance[C]//Proceedings of 2018 ACMSE Conference.New York, USA:ACM Press, 2018:1.
[2] MASDARI M, NABAVI S S, AHMADI V.An overview of virtual machine placement schemes in cloud computing[J].Journal of Network and Computer Applications, 2016, 66:106-127.
[3] MASDARI M, SALEHI F, JALALI M, et al.A survey of PSO-based scheduling algorithms in cloud computing[J].Journal of Network and Systems Management, 2017, 25(1):122-158.
[4] SINGH S, CHANA I.A survey on resource scheduling in cloud computing:issues and challenges[J].Journal of Grid Computing, 2016, 14(2):217-264.
[5] SHAO G L, CHEN J M.A load balancing strategy based on data correlation in cloud computing[C]//Proceedings of the 9th International Conference on Utility and Cloud Computing.New York, USA:ACM Press, 2016:364-368.
[6] MASDARI M, KHOSHNEVIS A.A survey and classifi-cation of the workload forecasting methods in cloud computing[J].Cluster Computing, 2020, 23(4):2399-2424.
[7] KHAIRALLA M A, NING X.Financial time series forecasting using hybridized support vector machines and ARIMA models[C]//Proceedings of 2017 International Conference on Wireless Communications, Networking and Applications.New York, USA:ACM Press, 2017:94-98.
[8] BI J, ZHANG L B, YUAN H T, et al.Hybrid task prediction based on wavelet decomposition and ARIMA model in cloud data center[C]//Proceedings of IEEE International Conference on Networking, Sensing and Control.Washington D.C., USA:IEEE Press, 2018:1-6.
[9] LIU C H, LIU C C, SHANG Y L, et al.An adaptive prediction approach based on workload pattern discrimination in the cloud[J].Journal of Network and Computer Applications, 2017, 80:35-44.
[10] ZHONG W, ZHUANG Y, SUN J, et al.The cloud computing load forecasting algorithm based on wavelet support vector machine[C]//Proceedings of the Australasian Computer Science Week Multiconference.New York, USA:ACM Press, 2017:1-5.
[11] GUPTA S, DILEEP A D, GONSALVES T A.Fractional difference based hybrid model for resource prediction in cloud network[C]//Proceedings of the 15th International Conference on Network, Communication and Computing.New York, USA:ACM Press, 2016:93-97.
[12] ZHANG Q C, YANG L T, YAN Z, et al.An efficient deep learning model to predict cloud workload for industry informatics[J].IEEE Transactions on Industrial Informatics, 2018, 14(7):3170-3178.
[13] ZHANG W S, LI B, ZHAO D H, et al.Workload prediction for cloud cluster using a recurrent neural network[C]//Proceedings of International Conference on Identification, Information and Knowledge in the Internet of Things.Washington D.C., USA:IEEE Press, 2016:104-109.
[14] SUDHAKAR C, KUMAR A R, SIDDARTHA N, et al.Workload prediction using ARIMA statistical model and long short-term memory recurrent neural networks[C]//Proceedings of International Conference on Computing, Power and Communication Technologies.Washington D.C., USA:IEEE Press, 2018:600-604.
[15] ZHANG Z H, ZHU W, ZHONG W, et al.Load forecasting model of mobile cloud computing based on glowworm swarm optimization LSTM network[C]//Proceedings of the 7th International Conference on Information Technology:IoT and Smart City.Washington D.C., USA:IEEE Press, 2019:113-119.
[16] 林涛, 冯竞凯, 郝章肖, 等.基于组合预测模型的云计算资源负载预测研究[J].计算机工程与科学, 2020, 42(7):1168-1173. LIN T, FENG J K, HAO Z X, et al.Cloud computing resource load prediction based on combined prediction model[J].Computer Engineering & Science, 2020, 42(7):1168-1173.(in Chinese)
[17] CHEN L, YU H H, TONG L, et al.Research on load forecasting method of distribution transformer based on deep learning[C]//Proceedings of the 7th IEEE International Conference on Cyber Security and Cloud Computing.Washington D.C., USA:IEEE Press, 2020:228-233.
[18] TANG D D, LI C, JI X H, et al.Power load forecasting using a Refined LSTM[C]//Proceedings of the 11th International Conference on Machine Learning and Computing.Washington D.C., USA:IEEE Press, 2019:104-108.
[19] YUAN C M, XIU T, LOU T Y.Probabilistic long-term load forecasting based on Stacked LSTM[C]//Proceedings of the 4th International Conference on Mathematics and Artificial Intelligence.Washington D.C., USA:IEEE Press, 2019:80-84.
[20] GUO J Y, WANG Z J, CHEN H W.On-line multi-step prediction of short term traffic flow based on GRU neural network[C]//Proceedings of the 2nd International Conference on Intelligent Information Processing.Washington D.C., USA:IEEE Press, 2017:1-6.
[21] CHEN Y T, KANG Y F, CHEN Y X, et al.Probabilistic forecasting with temporal convolutional neural network[J].Neurocomputing, 2020, 399:491-501.
[22] SHETTY J, SHOBHA G.An ensemble of automatic algorithms for forecasting resource utilization in cloud[C]//Proceedings of Future Technologies Conference.Washington D.C., USA:IEEE Press, 2016:301-306.
[23] DEVIAENE M, TESTELMANS D, BORZÉE P, et al.Feature selection algorithm based on random forest applied to sleep apnea detection[C]//Proceedings of the 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society.Washington D.C., USA:IEEE Press, 2019:2580-2583.
[24] CHENG Y, ANWAR A, DUAN X J.Analyzing Alibaba's co-located datacenter workloads[C]//Proceedings of IEEE International Conference on Big Data.Washington D.C., USA:IEEE Press, 2018:292-297.
[25] 蔡亮, 周泓岑, 白恒, 等.基于多层BiLSTM和改进粒子群算法的应用负载预测方法[J].浙江大学学报(工学版), 2020, 54(12):2414-2422. CAI L, ZHOU H C, BAI H, et al.Application load forecasting method based on multi-layer bidirectional LSTM and improved PSO algorithm[J].Journal of Zhejiang University(Engineering Science), 2020, 54(12):2414-2422.(in Chinese)
[26] 董兰芳, 张军挺.基于深度学习与随机森林的人脸年龄与性别分类研究[J].计算机工程, 2018, 44(5):246-251. DONG L F, ZHANG J T.Research on face age and gender classification based on deep learning and random forest[J].Computer Engineering, 2018, 44(5):246-251.(in Chinese)

选择文件类型/文献管理软件名称

选择包含的内容