Contact us
 |
[email protected] |
 |
3275638434 |
 |
 |
| Paper Publishing WeChat |
Useful Links
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License
Article
Propagation Path Loss Models at 28 GHz Using K-Nearest Neighbor Algorithm
Author(s)
Vu Thanh Quang1, Dinh Van Linh1,2, and To Thi Thao3
Full-Text PDF
XML 1603 Views
DOI:10.17265/1548-7709/2022.01.001
Affiliation(s)
1. Hanoi University of Science and Technology, No. 1 Dai Co Viet road, Hanoi, Vietnam
2. Academy of Cryptography Techniques, 141 Chien Thang Road, Hanoi, Vietnam
3. Post and Telecommunications Institute, Hanoi, Vietnam, Km 10, Nguyen Trai Road, Hanoi, Vietnam
ABSTRACT
In this paper, we develop and apply K-Nearest Neighbor algorithm to propagation pathloss regression. The path loss models present the dependency of attenuation value on distance using machine learning algorithms based on the experimental data. The algorithm is performed by choosing k nearest points and training dataset to find the optimal k value. The proposed method is applied to impove and adjust pathloss model at 28 GHz in Keangnam area, Hanoi, Vietnam. The experiments in both line-of-sight and non-line-of-sight scenarios used many combinations of transmit and receive antennas at different transmit antenna heights and random locations of receive antenna have been carried out using Wireless Insite Software. The results have been compared with 3GPP and NYU Wireless Path Loss Models in order to verify the performance of the proposed approach.
KEYWORDS
K-nearest neighbor, regression, 5G, millimeter waves, path loss.
Cite this paper
References