Longitudinal traces of platoon simulation with and without cyber-attacks and network latency

Longitudinal traces of platoon simulation with and without cyber-attacks and network latency This dataset contains simulation results from the FORESEEN project, focusing on Vehicle-to-Vehicle (V2V) and Vehicle-to-Edge (V2E) communication scenarios in platoon formations under various attack conditions. Overview The dataset consists of simulation data from Connected and Automated Vehicle (CAV) platoons using Cooperative Adaptive Cruise Control (CACC) systems. The simulations model a 10-vehicle platoon (1 leader + 9 followers) under different attack scenarios affecting both sensors and actuators. Directory Structure ├── README.md # This file ├── manual.txt # Detailed technical documentation ├── echarts.min.js # Visualization library ├── launch.sh # Batch processing script ├── utils/ # Analysis and visualization tools │ ├── build.sh # Installs the data vis. tools in the various folders │ ├── build_stats.py │ ├── graphs_template.html │ ├── install_graphs.sh │ └── make_index.py ├── v2v/ # Vehicle-to-Vehicle scenarios │ ├── Attack_First_Part-1/ # Sensor attacks on Car1 (modes 1,2,3) │ ├── Attack_First_Part-2/ # Actuator attacks on Car1 (modes 4,5) │ ├── Attack_Leader/ # Sensor attacks on Leader (modes 1,2,3) │ ├── Attack_Middle_Part-1/ # Sensor attacks on Car4 (modes 1,2,3) │ ├── Attack_Middle_Part-2/ # Actuator attacks on Car4 (modes 4,5) │ ├── No_Attack/ # Baseline scenarios without attacks │ └── stats.csv # Aggregated statistics across all scenarios └── v2e/ # Vehicle-to-Edge scenarios (centralized control via edge computing) ├── Attack_First_Part-1/ ├── Attack_First_Part-2/ ├── Attack_Leader/ ├── Attack_Middle_Part-1/ ├── Attack_Middle_Part-2/ ├── No_Attack/ └── stats.csv Simulation Scenarios Attack Types The dataset includes six main scenario categories: No_Attack: Baseline simulations with nominal system behavior Attack_Leader: Sensor attacks (modes 1-3) targeting the platoon leader Attack_First_Part-1: Sensor attacks (modes 1-3) on Car1 (first follower) Attack_First_Part-2: Actuator attacks (modes 4-5) on Car1 Attack_Middle_Part-1: Sensor attacks (modes 1-3) on Car4 (middle vehicle) Attack_Middle_Part-2: Actuator attacks (modes 4-5) on Car4 Attack Modes Mode 0: No attack Mode 1: Low-frequency sinusoidal attack on sensors Mode 2: High-frequency sinusoidal attack on sensors Mode 3: Combined low and high-frequency attacks on sensors Mode 4: Multiplicative attack on actuator (acceleration amplification) Mode 5: Additive attack on actuator (acceleration offset) Data Structure Individual Simulation Folders Each scenario folder contains multiple simulation runs with different parameter combinations. Each simulation folder includes: results.csv: Raw simulation data with time-series measurements config.mm.json: Simulation configuration parameters objectives.json: Simulation objectives and metrics index.html: Interactive visualization of results (when processed) CSV Data Format The results.csv files contain time-series data with the following columns: Time and Leader Data: - time: Simulation timestamp (seconds) - {Leader}.LeaderInstance.position_x: Leader X position (meters) - {Leader}.LeaderInstance.speed: Leader velocity (m/s) - {Leader}.LeaderInstance.acceleration: Leader acceleration (m/s²) - {Leader}.LeaderInstance.real_x: True leader position (unaffected by sensor attacks) - {Leader}.LeaderInstance.real_v: True leader velocity - {Leader}.LeaderInstance.real_a: True leader acceleration Follower Vehicle Data (Car1-Car9): For each follower vehicle i (1-9): - {Cari}.CarInstance_i.position_x: Reported X position (meters) - {Cari}.CarInstance_i.speed: Reported velocity (m/s) - {Cari}.CarInstance_i.acceleration: Reported acceleration (m/s²) - {Cari}.CarInstance_i.real_x: True X position (unaffected by sensor attacks) - {Cari}.CarInstance_i.real_v: True velocity - {Cari}.CarInstance_i.real_a: True acceleration Network/CACC Data: - {Network}.MecInstance.platoon_0_i_des_acc: Desired acceleration for vehicle i computed by CACC Statistical Analysis The stats.csv files contain aggregated safety metrics: Run name: Scenario identifier Folder: Path to simulation data d01-d89: Safety status for each vehicle pair (Leader-Car1, Car1-Car2, etc.) OVERALL: Worst-case safety status across all vehicles Safety status categories: - OK: Safe following distance maintained - TOO FAR: Distance > 18m (inefficient) - TOO CLOSE: Distance 10-12m (potentially unsafe) - VIOLATION: Distance < 10m (unsafe) - COLLISION: Distance ≤ 4m (collision) Key Parameters For more precise information check the manual.pdf. Vehicle Dynamics Platoon size: 10 vehicles (1 leader + 9 followers) Vehicle length: 4 meters each Target following distance: 14 meters Starting sequence: Each vehicle starts 4 seconds after the previous one Network Characteristics Uplink delay: 8 or 10ms (average) Downlink delay: 3 or 5ms (average) Communication protocol: V2V (distributed) or V2E (centralized edge computing) depending on scenario Leader Behavior Operational modes: Mode 0: Constant acceleration (10s) + sinusoidal Mode 1: Linear acceleration (8s) + sinusoidal Mode 2: Periodic sprint cycles Acceleration values: 1.7-2.0 m/s² Deceleration values: -1.5 to -1.7 m/s² Usage Viewing Individual Results Each batch of simulations is provided with a table of contents stored in the index.html, it’s useful to sort through the various parameters. Each processed simulation folder contains an index.html file that provides: - Interactive time-series plots of vehicle positions, velocities, and accelerations - Real vs. reported sensor values comparison - Following distance analysis - Configuration parameter tables To view the interactive visualizations (index.html), you must start a local web server using the provided launch.sh script. Opening the HTML files directly will prevent the JavaScript visualizations from working correctly due to browser security restrictions. You may consult the table of contents without starting the web server; although the hyperlinks will be broken. Run: ./launch.sh Then follow the instructions printed in the terminal to access the results in your browser. The aforementioned files have been generated through utils/build.sh. Technical Details For detailed technical specifications, control algorithms, and parameter descriptions, refer to manual.pdf. License The data and programs in this archive are released under the Creative Commons Attribution 4.0 International; except for the Apache ECharts which is released under the Apache License Version 2.0. --- Acknowledgment This study received funding from the European Union - Next- GenerationEU - National Recovery and Resilience Plan (NRRP) - MISSION 4 COMPONENT 2, INVESTMENT N. 1.1, CALL PRIN 2022 PNRR D.D. 1409 14-09-2022 - (FORESEEN) CUP N.P2022WYAEW and from the Italian Ministry of University and Research (MUR) in the framework of the Crosslab and FoReLab projects (Departments of Excellence).