UAV Attack Dataset

Simulation of multiple UAV platforms with both benign flights and malicious flights where UAV experiences GPS spoofing attack. Simulation environment follows standard Gazebo/PX4 setup (https://dev.px4.io/v1.9.0/en/simulation/gazebo.html). When the attack starts, normal GPS messages are stopped and injected messages are created and sent from the Gazebo environment to the autopilots GPS sensor for 30 seconds. An autonomous survey flight is conducted over the University (~20min flight time). Full flight plan is located in the dataset package as "OTU-Survey.plan". The following UAV platforms are used:3DR IRIS+ - Quadcopter, SITLHolybro S500 - Quadcopter, HITLYuneec H480 - Hexacopter, SITLDeltaQuad VTOL - VTOL, SITLPX4 Standard Tailsitter - Tailsitter, SITLPX4 Standard Plane - Plane, SITL

本数据集面向多无人机（UAV）平台开展仿真，涵盖正常飞行与恶意飞行两类场景，其中恶意飞行场景下无人机将遭受GPS欺骗攻击（GPS spoofing attack）。本次仿真环境采用标准Gazebo/PX4搭建方案，具体搭建指南可参考链接：https://dev.px4.io/v1.9.0/en/simulation/gazebo.html。当攻击触发后，正常GPS报文传输将被中断，Gazebo仿真环境将持续30秒向无人机自动驾驶仪的GPS传感器注入伪造的GPS报文。数据集包含一段在大学校园上空执行的自主测绘飞行任务，单段飞行时长约20分钟。完整飞行计划已存储于数据集包内的"OTU-Survey.plan"文件中。本次仿真所使用的无人机平台如下：3DR IRIS+ 四旋翼无人机、SITLHolybro S500 四旋翼无人机、HITLYuneec H480 六旋翼无人机、SITLDeltaQuad VTOL 垂直起降（VTOL）无人机、SITLPX4 Standard Tailsitter 倾转旋翼机、SITLPX4 Standard Plane 固定翼飞机，以及SITL平台。

搜集汇总

背景与挑战

背景概述

该数据集专注于无人机（UAV）攻击检测研究，包含模拟和真实飞行日志，覆盖GPS欺骗、干扰和Ping DoS等多种攻击类型，以及正常飞行数据。数据以CSV和ULOG格式提供，适用于机器学习安全应用，旨在支持入侵检测算法的开发和评估。