1. FadeNet: Deep Learning-Based mm-Wave Large-Scale Channel Fading Prediction and its Applications
- Author
-
Young-Jin Kim, Bingwen Zhang, Vishnu V. Ratnam, Charlie Jianzhong Zhang, Cho Minsung, Sung-Rok Yoon, Sameer Pawar, Hao Chen, and Soonyoung Lee
- Subjects
convolutional networks ,mm-Wave ,Cell planning ,General Computer Science ,Mean squared error ,Computer science ,02 engineering and technology ,Base station ,0203 mechanical engineering ,0202 electrical engineering, electronic engineering, information engineering ,General Materials Science ,Fading ,business.industry ,Deep learning ,General Engineering ,deep learning ,020206 networking & telecommunications ,020302 automobile design & engineering ,Parallel processing (DSP implementation) ,channel modeling ,Cellular network ,large scale fading ,lcsh:Electrical engineering. Electronics. Nuclear engineering ,Artificial intelligence ,business ,lcsh:TK1-9971 ,Algorithm ,Communication channel - Abstract
Accurate prediction of the large-scale channel fading is fundamental to planning and optimization in 5G millimeter-wave cellular networks. The current prediction methods, which are either too computationally expensive or inaccurate, are unsuitable for city-scale cell planning and optimization. This paper presents FadeNet, a convolutional neural-network enabled alternative for predicting large-scale fading with high computation speed and accuracy. By using carefully designed input features and neural-network architecture to capture topographical information, FadeNet accurately predicts the large-scale fading from a base station to each location in its coverage area. Evaluations on realistic data, derived from millimeter-wave cells across multiple cities, suggest that FadeNet can achieve a prediction accuracy of 5.6 decibels in root mean square error. In addition, by leveraging the parallel processing capabilities of a graphics processing unit, FadeNet can reduce the prediction time by $40X-1000X$ in comparison to industry prevalent methods like ray-tracing. Generalizations of FadeNet, that can handle variable topographies and base station heights, and its use for optimal cell site selection are also explored.
- Published
- 2021
- Full Text
- View/download PDF