工业系统信息安全异常检测数据集
收藏资源简介：
工业网数据记录了多种真实网络环境下攻击端向各类目标发起的多种网络攻击时的网络流量数据。涵盖 TCP、UDP、ICMP 等常见网络协议在不同网络节点间数据包交互过程中的流量数据包。涉及多类网络行为在如正常网络访问、遭受暴力 FTP、暴力 SSH、DoS、Heartbleed、Web 攻击、渗透、僵尸网络和 DDoS 等典型场景中的网络流量数据。进一步根据不同的攻击情况比例划分为7个比例的csv数据文件。 物联网数据包含23种不同场景的IoT流量，每个场景都从网络流量中捕获。这些场景包括20个来自被感染IoT设备的PCAP文件和三个真实世界的IoT网络流量捕获，可用于入侵检测中。包含9类标签，用于描述与恶意或可能的恶意活动相关的流之间的关系。并进一步根据不同的攻击情况比例划分为7个比例的csv数据文件。使用Python程序首先将数据集预处理并分为训练集和测试集，使用7:3的比例。为了验证模型在小规模数据集上的性能，进一步合成了七个不同比例的合成训练数据集，这些数据集包含正常流量和异常流量样本的不同百分比，数据量为364MB。
Industrial network data records network traffic data collected when attackers launch various cyber attacks against diverse targets across multiple real-world network environments. It covers traffic packets generated during packet exchange between different network nodes using common network protocols such as TCP, UDP, and ICMP. The dataset contains network traffic data corresponding to multiple network behaviors in typical scenarios including normal network access, brute-force FTP attacks, brute-force SSH attacks, DoS, Heartbleed, Web attacks, penetration testing, botnet activities, and DDoS attacks. Furthermore, it is split into 7 CSV data files with varying proportions of attack scenarios. IoT data includes IoT traffic from 23 distinct scenarios, all captured from network traffic. These scenarios comprise 20 PCAP files from compromised IoT devices and three real-world IoT network traffic captures, which are applicable for intrusion detection tasks. The dataset has 9 categories of labels used to describe the relationships between flows associated with malicious or potentially malicious activities. It is also divided into 7 CSV data files with varying attack scenario proportions. First, a Python program is utilized to preprocess the original dataset and split it into a training set and a test set at a 7:3 ratio. To validate the model's performance on small-scale datasets, seven synthetic training datasets with different proportions are additionally generated. These synthetic datasets contain different percentages of normal and anomalous traffic samples, with a total size of 364 MB.
