CDL Channel Dataset

该数据集包含了针对视距（LOS）和非视距（NLOS）环境的模拟信道数据，其中包括每个CDL信道生成的10,000个原始数据样本，总样本量达到50,000个。该数据集用于评估所提出的EMEV-IdNet架构在不同条件下识别不同CDL信道的能力，包括在有噪声存在的情况下的表现。数据集规模为总计50,000个样本，按65:15:20的比例分为训练集、验证集和测试集。该数据集的任务是用于在毫米波系统中进行信道识别。This dataset contains simulated channel data for line-of-sight (LOS) and non-line-of-sight (NLOS) environments. Specifically, 10,000 raw data samples are generated for each CDL channel, resulting in a total of 50,000 samples overall. This dataset is used to evaluate the ability of the proposed EMEV-IdNet architecture to identify different CDL channels under various conditions, including its performance in the presence of noise. The dataset has a total scale of 50,000 samples, which is divided into training, validation and test sets at a ratio of 65:15:20. The task of this dataset is channel identification in millimeter-wave systems.

搜集汇总

背景与挑战

背景概述

该数据集是GLOBECOM2022论文的开源代码，用于信道识别任务，主要使用MATLAB语言实现。代码仓库包含生成器工具，但未提供数据集的规模或具体内容描述。