Machine Learning-Based Channel State Estimators for 5G Wireless Communication Systems

作     者：Mohamed Hassan Essai Ali Fahad Alraddady Mo’ath Y.Al-Thunaibat Shaima Elnazer 

作者机构：Department of Electrical EngineeringFaculty of EngineeringAl-Azhar UniversityQena83513Egypt Department of Computer EngineeringCollege of Computers and Information TechnologyTaif UniversityTaif21944Saudi Arabia Department of Management Information SystemsTaif UniversityTaif21944Saudi Arabia 

出 版 物：《工程与科学中的计算机建模（英文）》 (Computer Modeling in Engineering & Sciences)

年 卷 期：2023年第135卷第4期

页      面：755-778页

核心收录：

学科分类：0710[理学-生物学] 0810[工学-信息与通信工程] 080904[工学-电磁场与微波技术] 0809[工学-电子科学与技术（可授工学、理学学位）] 08[工学] 080402[工学-测试计量技术及仪器] 0804[工学-仪器科学与技术] 0714[理学-统计学（可授理学、经济学学位）] 081001[工学-通信与信息系统] 0811[工学-控制科学与工程] 0701[理学-数学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：funded by Taif University Researchers Supporting Project No.(TURSP-2020/214) Taif University Taif Saudi Arabia 

主　　题：DLNNs channel state estimator 5G and beyond communication systems robust loss functions 

摘      要：For a 5G wireless communication system,a convolutional deep neural network(CNN)is employed to synthesize a robust channel state estimator(CSE).The proposed CSE extracts channel information from transmit-and-receive pairs through offline training to estimate the channel state ***,it utilizes pilots to offer more helpful information about the communication *** proposedCNN-CSE performance is compared with previously published results for Bidirectional/long short-term memory(BiLSTM/LSTM)NNs-based *** CNN-CSE achieves outstanding performance using sufficient pilots only and loses its functionality at limited pilots compared with BiLSTM and LSTM-based *** three different loss function-based classification layers and the Adam optimization algorithm,a comparative study was conducted to assess the performance of the presented DNNs-based *** BiLSTM-CSE outperforms LSTM,CNN,conventional least squares(LS),and minimum mean square error(MMSE)*** addition,the computational and learning time complexities for DNN-CSEs are *** estimators are promising for 5G and future communication systems because they can analyze large amounts of data,discover statistical dependencies,learn correlations between features,and generalize the gotten knowledge.