多维信息感知数据集

本数据集可划分为三个部分，频谱态势感知实验数据、主被动CSI感知实验数据和多维联合感知实验数据。频谱态势预测数据集采用校园网络频谱公开数据集，该公开数据集使用全向天线并将扫描时间配置为 1 秒。一周的测量产生了约 335 000 次扫描的迹线（平均每次扫描需要 1.8 秒）。频段覆盖范围为20MHz-1.52GHz，带宽分辨率为200KHz。频谱感知数据集包含8个表格，分别对应了课题中相关实验仿真。主被动CSI数据集包括包含6个表格，分别对应6张基于ESPRIT的信道估计方案与对比信道估计方案的归一化均方误差曲线图。用户活跃状态和CSI联合感知数据集包含14个表格，分别对应了智能反射面辅助的活跃用户与信道状态信息联合感知和基于架构共生的活跃用户检测方案的参数设置、导频数据以及仿真数据。数据量为271MB。This dataset is divided into three parts: experimental data for spectrum situation awareness, experimental data for active-passive CSI perception, and experimental data for multi-dimensional joint perception. The spectrum situation prediction dataset adopts a public campus network spectrum dataset, which uses omnidirectional antennas and configures the scan interval as 1 second. Over one week of measurement, traces of approximately 335,000 scans were collected, with an average scan duration of 1.8 seconds. The frequency band coverage ranges from 20 MHz to 1.52 GHz, with a bandwidth resolution of 200 kHz. The spectrum perception dataset contains 8 tables, which respectively correspond to relevant experimental simulations in this research project. The active-passive CSI dataset includes 6 tables, each corresponding to 6 normalized mean square error (NMSE) curve plots of ESPRIT-based channel estimation schemes and their comparative channel estimation schemes. The joint perception dataset for user activity status and CSI contains 14 tables, which respectively correspond to the parameter settings, pilot data, and simulation data of two types of schemes: intelligent reflecting surface (IRS)-assisted joint perception of active users and channel state information, and architecture coexistence-based active user detection schemes. The total data volume of this dataset is 271 MB.

搜集汇总

背景与挑战

背景概述

该数据集由西南交通大学发布，包含频谱态势感知、主被动CSI感知和多维联合感知三部分实验数据，涉及多个表格和仿真数据，总数据量为271MB。它支持大维智能共生无线通信基础理论与技术研究，数据格式包括mat、txt和docx。