DATASET: IQ SAMPLES OF LTE, 5G NR, WI-FI, ITS-G5, AND C-V2X PC5

Dataset: IQ samples of LTE, 5G NR, WiFi, ITS-G5, and C-V2X PC5 The dataset comprises IQ samples captured from ITSG-5, C-V2X PC5, WiFi, LTE, 5G NR and Noise. In each dataset cluster, 7500 examples are collected from each considered technology. The dataset collected was used for technology recognition and traffic charchterization model. More details on the structure and performance of the model can be found here. Description of Dataset Collection LTE For LTE dataset collection, srsRAN, an open-source SDR platform, is used. This LTE SDR implementation is used to collect samples for the LTE dataset. The indoor testbed setup used to collect our dataset consists of one eNB host PC and one UE host PC. Each host PC is connected to a USRP X310 board, which is used as the RF front end. We use the latest srsRAN version 21.04, which is installed on each host PC. For the LTE dataset collection, the FDD mode with a 10 MHz bandwidth and a 5.9 GHz center frequency is used for the down-link traffic. For the LTE dataset, the traffic load was varied between 5 and 50 Mbps, and the MCS used was varied by manually configuring from MCS index 1 to 28 . 5G NR The OpenAirInterface SDR solution is used for 5G NR dataset collection. OpenAirInterface is an open source SDR platform that provides a 3GPP compliant implementation of eNB, UE, and EPC. The OpenAirInterface SDR solution also includes a 5G non-stand alone (NSA) mode which supports 5G networks by using existing 4G infrastructure. This SDR-based 5G network setup is used to collect the IQ samples for the 5G NR dataset. For the 5G NR dataset collection, a 1:1 static TDD configuration is used for up-link and down-link traffic in NSA mode. Numerology 1 is used at 10 MHz bandwidth and center frequency of 5.9 GHz. For the 5G NR dataset, the MCS used was varied by manually configuring different values ranging from MCS index 1 to 28, and the traffic load was varied between 5 and 50 Mbps. Wi-Fi For the WiFi dataset collection, the openwifi SDR solution is used. openwifi is an open-source full-stack IEEE 802.11 a/g/n SDR implementation based on the Xilinx Zynq System on Chip that includes a Field Programmable Gate Array (FPGA) and an ARM processor. For our dataset, we used an IEEE802.11n access point and a client connected to it. WiFi traffic generated cover a wide range of traffic loads, i.e., 10–200 packets per second (pps), with packet sizes ranging from 500 to 1500 bytes. The MCS used was varied by manually configuring the MCS index value between 0 and 7. ITS-G5 and C-V2X PC5 The CAMINO framework was used for the ITS-G5 and C-V2X PC5 dataset collection. CAMINO is a core framework for managing the V2X communication technologies, including ITS-G5, C-V2X PC5 and C-V2X Uu (5G/4G). The CAMINO framework is used to dynamically generate standardized C-ITS service packets, including Cooperative Awareness (CA), Decentralized Environmental Notification (DEN), and Infrastructure to Vehicle Information (IVI) message packets. The CAMINO software is implemented on the infrastructure deployed as part of the Belgian Smart Highway testbed. The Smart Highway is a testbed deployed by IMEC on the E313 highway, near Antwerp, Flanders. The Smart Highway testbed consists of eight RSUs and 2 OBUs. Each RSU and OBU includes a general purpose CPU running the CAMINO software, and Cohda MK5 and MK6c modules, which are COTS ITS-G5 and C-V2X modules respectively. In our dataset collection, RSU4 is used as a transmitter, and a USRP N310 connected to RSU3 is used to capture and store the samples. To represent a wide range of traffic characteristics, different packet sizes of 300B for CA, 300B and 600B for DEN, and 600B for IVI packets were used at inter-packet intervals of 20, 50, 100, and 200 seconds. The MCS used were varied by manually changing the MCS index value in the configuration files of the Cohda devices. The MCS index of ITS-G5 varied from 0 to 7. For the C-V2X PC5, the MCS is varied between 0 and 20.

# 数据集：LTE、5G NR、WiFi、ITS-G5及C-V2X PC5的IQ样本
该数据集包含从ITS-G5、C-V2X PC5、WiFi、LTE、5G NR及噪声环境中采集的同相正交（IQ）采样样本。每个数据集簇中，每种目标技术均采集7500条样本。本数据集可用于技术识别与流量特征建模相关研究。有关该模型的结构与性能细节可参见此处。

## 数据集采集说明
### LTE数据集采集
本次LTE数据集采集采用开源软件定义无线电（Software Defined Radio, SDR）平台srsRAN。该LTE SDR实现方案用于采集LTE数据集的样本。本次采集采用的室内测试平台包含1台演进型基站（evolved NodeB, eNB）主机与1台用户设备（User Equipment, UE）主机，每台主机均连接通用软件无线电外设（Universal Software Radio Peripheral, USRP）X310板卡作为射频前端。我们在每台主机上均安装了最新版srsRAN 21.04。
在LTE数据集采集过程中，下行链路采用带宽10 MHz、中心频率5.9 GHz的频分双工（Frequency Division Duplex, FDD）模式。LTE数据集的流量负载设置在5~50 Mbps区间，调制编码方案（Modulation and Coding Scheme, MCS）索引通过手动配置从1到28变化。

### 5G NR数据集采集
本次5G NR数据集采集采用OpenAirInterface SDR解决方案。OpenAirInterface是一款开源SDR平台，可提供符合3GPP标准的eNB、UE及分组核心网（Evolved Packet Core, EPC）实现方案。该方案还包含5G非独立组网（Non-Standalone, NSA）模式，可通过现有4G基础设施支持5G网络。本次基于SDR的5G网络搭建方案用于采集5G NR数据集的IQ样本。
在5G NR数据集采集过程中，非独立组网模式下采用1:1静态时分双工（Time Division Duplex, TDD）配置进行上下行链路传输。采用5G参数集1（Numerology 1），带宽为10 MHz，中心频率为5.9 GHz。本次5G NR数据集的MCS索引通过手动配置从1到28的不同取值，流量负载设置在5~50 Mbps区间。

### Wi-Fi数据集采集
本次WiFi数据集采集采用openwifi SDR解决方案。openwifi是一款基于Xilinx Zynq片上系统（System on Chip, SoC）的开源全栈式IEEE 802.11 a/g/n实现方案，该平台包含现场可编程门阵列（Field Programmable Gate Array, FPGA）与ARM处理器。本次采集采用IEEE 802.11n无线接入点与连接至该接入点的客户端设备。生成的WiFi流量覆盖宽泛的负载范围：每秒10~200个数据包（packets per second, pps），数据包大小介于500~1500字节之间。MCS索引通过手动配置在0~7区间变化。

### ITS-G5与C-V2X PC5数据集采集
本次ITS-G5与C-V2X PC5数据集采集采用CAMINO框架。CAMINO是一款用于管理车联网通信技术的核心框架，涵盖ITS-G5、C-V2X PC5及C-V2X Uu（5G/4G）等技术。该框架可动态生成标准化的合作智能交通系统（Cooperative Intelligent Transport System, C-ITS）服务数据包，包括合作感知（Cooperative Awareness, CA）消息、分布式环境通知（Decentralized Environmental Notification, DEN）消息及基础设施到车辆信息（Infrastructure to Vehicle Information, IVI）消息数据包。
CAMINO软件部署于比利时智能高速公路测试平台的基础设施中。该智能高速公路测试平台由IMEC在法兰德斯安特卫普附近的E313高速公路上搭建，包含8个路边单元（Road Side Unit, RSU）与2个车载单元（On Board Unit, OBU）。每个RSU与OBU均搭载运行CAMINO软件的通用CPU，以及分别为商用现货（Commercial Off-The-Shelf, COTS）级ITS-G5与C-V2X模块的Cohda MK5与MK6c模块。
本次数据集采集过程中，以RSU4作为发射端，连接至RSU3的USRP N310板卡用于采集并存储样本。为覆盖多样的流量特征，本次采集采用不同大小的数据包：CA消息为300字节，DEN消息为300字节与600字节，IVI消息为600字节；数据包发送间隔设置为20、50、100及200秒。MCS索引通过修改Cohda设备的配置文件手动调整：ITS-G5的MCS索引范围为0~7，C-V2X PC5的MCS索引范围为0~20。

搜集汇总

背景与挑战

背景概述

该数据集包含多种无线通信技术（LTE、5G NR、WiFi、ITS-G5和C-V2X PC5）的IQ样本，每种技术有7500个样本，用于技术识别和流量特征分析。数据收集使用了不同的SDR平台和测试环境，覆盖了广泛的调制编码方案和流量负载条件。