用机器学习生成加拿大全域地貌图，提升毫米波传播预测精度
Development of a Canada-Wide Morphology Map for the ITU-R P. 1411 Propagation Model
- 基于机器学习自动划分住宅、低层和高层城区
- 实现300MHz至100GHz频段内更精准的路径损耗估算
- 为无线网络规划提供高精度地理环境数据支持
本文依据国际电信联盟无线电通信部门推荐标准ITU-R P.1411-12，构建了覆盖加拿大的地貌分类地图，将区域划分为住宅区、低层城市和高层城市三种环境类型。针对推荐中环境描述的定性问题，采用机器学习方法实现自动化分类。通过大量实验优化分类准确率，最终生成的加拿大全域地貌图显著提升了300 MHz至100 GHz频段下短距离室外传播的路径损耗预测精度。
原文摘要 · Abstract (English)
This paper outlines the development of a Canada-wide morphology map classifying regions into residential, urban low-rise, and urban high-rise environments, following the ITU-R P.1411-12 propagation model guidelines. To address the qualitative nature of the environment-type descriptors found in the Recommendation, a machine learning approach is employed to automate the classification process. Extensive experimentation optimized classification accuracy, resulting in a Canada-wide morphology map that ensures more accurate path loss estimations for outdoor short-range propagation at frequencies ranging from 300 MHz to 100 GHz.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。