Metaheuristic Optimization of Time Series Models for Predicting Networks Traffic

作     者：Reem Alkanhel El-Sayed M.El-kenawy D.L.Elsheweikh Abdelaziz A.Abdelhamid Abdelhameed Ibrahim Doaa Sami Khafaga 

作者机构：Department of Information TechnologyCollege of Computer and Information SciencesPrincess Nourah Bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia Department of Communications and ElectronicsDelta Higher Institute of Engineering and TechnologyMansoura35111Egypt Faculty of Artificial IntelligenceDelta University for Science and TechnologyMansoura35712Egypt Department of Computer ScienceFaculty of Specific EducationMansoura UniversityEgypt Department of Computer ScienceFaculty of Computer and Information SciencesAin Shams UniversityCairo11566Egypt Department of Computer ScienceCollege of Computing and Information TechnologyShaqra University11961Saudi Arabia Computer Engineering and Control Systems DepartmentFaculty of EngineeringMansoura UniversityMansoura35516Egypt Department of Computer SciencesCollege of Computer and Information SciencesPrincess Nourah bint Abdulrahman UniversityP.O.Box 84428Riyadh 11671Saudi Arabia 

出 版 物：《Computers, Materials & Continua》 (计算机、材料和连续体（英文）)

年 卷 期：2023年第75卷第4期

页      面：427-442页

核心收录：

学科分类：12[管理学] 1201[管理学-管理科学与工程(可授管理学、工学学位)] 081104[工学-模式识别与智能系统] 08[工学] 0835[工学-软件工程] 0811[工学-控制科学与工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：Princess Nourah bint Abdulrahman University Researchers Supporting Project Number (PNURSP2022R323) Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia 

主　　题：Network traffic soft computing LSTM metaheuristic optimization 

摘      要：Traffic prediction of wireless networks attracted many researchersand practitioners during the past decades. However, wireless traffic frequentlyexhibits strong nonlinearities and complicated patterns, which makes it challengingto be predicted accurately. Many of the existing approaches forpredicting wireless network traffic are unable to produce accurate predictionsbecause they lack the ability to describe the dynamic spatial-temporalcorrelations of wireless network traffic data. In this paper, we proposed anovel meta-heuristic optimization approach based on fitness grey wolf anddipper throated optimization algorithms for boosting the prediction accuracyof traffic volume. The proposed algorithm is employed to optimize the hyperparametersof long short-term memory (LSTM) network as an efficient timeseries modeling approach which is widely used in sequence prediction *** prove the superiority of the proposed algorithm, four other optimizationalgorithms were employed to optimize LSTM, and the results were *** evaluation results confirmed the effectiveness of the proposed approachin predicting the traffic of wireless networks accurately. On the other hand,a statistical analysis is performed to emphasize the stability of the proposedapproach.