Developing a ROS2 node that detects sudden jumps in position data from Extended Kalman Filters (EKFs) and simulate spoofing/jamming attacks on GPS inputs using the MOBSTA package.
We read the information about the provided bag with ros2 bag info. We find multiple topics:
| topic name | message type | utility |
|---|---|---|
| /clock | rosgraph_msgs/msg/Clock | provides the system time in the ROS network |
| diffbot_base_controller/cmd_vel_out | geometry_msgs/msg/TwistStamped | provides velocity information (linear and angular) with a timestamp to the robot's base controller |
| /diffbot_base_controller/cmd_vel_unstamped | geometry_msgs/msg/Twist | similar as the topic above but without a timestamp |
| /diffbot_base_controller/odom_corrected | nav_msgs/msg/Odometry | provides corrected odemetry information for the robot including the robot's position, orientation (pose) and velocity |
| /gps/fix | sensor_msgs/msg/NavSatFix | provides GPS information including latitude, longitude, altitude and position invariance. |
| /gps/reversed_navheading | sensor_msgs/msg/Imu | provides IMU data including orientation, angular velocity, and linear acceleration in a reversed format |
| /imu/data | sensor_msgs/msg/Imu | provides IMU data not reversed |
| /joy_vel | geometry_msgs/msg/Twist | provides velocity commands (linear and angular) derived from a joystick input. |
| /odometry/global | nav_msgs/msg/Odometry | provides global odometry information, providing the robot's pose and velocity in a global reference frame. |
| /odometry/gps | nav_msgs/msg/Odometry | provides odometry data derived from GPS information. It includes the robot's pose and velocity |
| /odometry/local | nav_msgs/msg/Odometry | provides local odometry information, providing the robot's pose and velocity in a local reference frame. |
| /tf | tf2_msgs/msg/TFMessage | provides the transformation tree, which includes all the coordinate frames of the robot and their relative positions and orientations. |
| /tf_static | tf2_msgs/msg/TFMessage | provides static transformations, which are transformations that do not change over time. |
We create a package for the jump detection task: ros2 pkg create jump_detection --build-type ament_python --dependencies rclpy
We write the node in the jump_detection.py file in the jump_detection package, documentation to the code can be found in the source .py file. The node currently monitors and detects jumps in the latitude and longitude data provided from the /gps/fix topic, as this is the topic that will be mutated by MOBSTA. The same code logic can be used to work on other position data topics such as /odometry/local and /odometry/global.
We write another node in the jump_detection package to test the jump detection node. The node is called jammer, reads GPS positional data from /gps/fix, randomly modifies it to simulate a jamming attack, and published the jammed data to /gps/jammed. The jump detection node will detect jumps using the /gps/jammed topic.
We play the data from the bag: ros2 bag play rosbag2_2024_04_24-12_23_04_0.db3
Here is a screenshot of the jump_detection topic when the jump detection and the jammer nodes are running.
We download the repo
we build the mobsta package, it fails at rosbag2_py, probably due to conflicting versions: Error message: error: invalid use of incomplete type ‘PyFrameObject’ {aka ‘struct _frame’} 482 | frame = frame->f_back;
We create the configuration folder for our test case:
We play the data from the bag: ros2 bag play rosbag2_2024_04_24-12_23_04_0.db3