5G环境下10种典型应用流量数据

数据集为CSV格式，从真实5G环境中采集，包含10种典型应用。在搭建的真实5G实验网环境中，通过客户端访问不同的典型应用，并在关键网络节点镜像流量，进而获得原始PCAP文件，再使用python脚本从原始PCAP文件中依据源目的IP及端口和协议这五元组信息来划分流量，为每条流量分配流序号、标签，在每条流量内部分配包序号，再提取每条流量的包长、包到达时间、包方向信息。每种应用的流量抓620条左右的流量，共6191条。This dataset is stored in CSV format, collected from real 5G environments, and covers 10 typical application scenarios. In a constructed real 5G experimental network, traffic was mirrored at key network nodes while clients accessed the target applications, to obtain raw PCAP files. Subsequently, a Python script was used to split the traffic from the raw PCAP files based on the 5-tuple information including source IP, destination IP, source port, destination port and protocol. For each traffic flow, a flow serial number, label and packet sequence numbers within each individual flow were assigned. Additionally, packet length, packet arrival time and packet direction information were extracted for each traffic flow. Approximately 620 traffic flows were collected for each application, yielding a total of 6191 traffic flows in the entire dataset.

搜集汇总

背景与挑战

背景概述

该数据集为CSV格式，从真实5G实验网环境中采集，包含10种典型应用的流量，共计6191条。每条流量数据均通过处理原始PCAP文件提取，包含包长、包到达时间和包方向等信息，适用于加密流量识别和深度学习研究。