| Title |
Deep Learning-based Beam and Blockage Prediction for Millimeter-wave Indoor Environment |
| Authors |
문상미(Sangmi Moon) ; 김현성(Hyeonsung Kim) ; 김진영(Jin Young Kim) ; 김대진(Intae Hwang) ; 황인태() |
| DOI |
https://doi.org/10.5573/ieie.2020.57.7.3 |
| Keywords |
blockage; deep learning; DNN; indoor; mmWave |
| Abstract |
In this paper, we propose the deep learning based beam and blockage prediction method for millimeter-wave (mmWave) indoor environment. The proposed method operates in two phases ? the offline learning phase and the online prediction phase. During the offline learning phase, a deep neural network (DNN) is designed to learn the mapping between the user positions along with data traffic demands and the corresponding optimal beam indices and blockage statuses. During a subsequent online prediction phase, the trained DNN is utilized to predict the optimal tunings of beams and blockages corresponding to the targeted user positions with the given data traffic demands. The system level simulation is based on 3rd generation partnership project (3GPP) new radio (NR) channel and blockage model. The simulation results reveal that the proposed scheme is capable of predicting mmWave blockages with an accuracy greater than 90%. Furthermore, these results confirm the viability of the proposed DNN model in predicting the optimal mmWave beams and spectral efficiencies. |