Misbehaviour detection in C-ITS

This PhD project aims to propose a solution to identify and possibly prevent any type of Cooperative Intelligent Transportation Systems (C-ITS) misbehaviour in the network at an early stage, which in turn will reduce any devastating effects to the connected vehicles.

C-ITS primarily depends on the data received from various nodes in the network to perform any action, such as route deviation during an accident. When the connected vehicles in C-ITS misbehave due to an attack, the results can be devastating, such as major incidents.

Participants

• Deakin University

Project background

The number of road users continues to grow rapidly and is being expected to reach more than two million by 2050 worldwide. It is believed that this growth contributes to road accidents and 95% of the accidents are caused by human error.

Furthermore, these accidents contribute to significant economic losses such as, treatment cost, loss of work hours, and higher fuel consumption. It is estimated by the World Health Organisation that 1.3 million people die each year of road traffic crashes with road traffic injuries being the leading cause of death for children and young adults aged 15-29 years.

RSU and OBU

In order to address these challenges, C-ITS has been proposed in prior research and has proven its efficacy for improving road safety. It enables cooperative communication between road users and infrastructure units to ensure a safe and efficient journey for road users. This is through the use of two main components, namely the Road-Side Unit (RSU) and On-Board Unit (OBU).

The RSU is an infrastructure unit placed along the roads that facilitates Vehicle-to-Infrastructure (V2I) intercommunication between the RSU and the OBU. The OBU is a device in the vehicle that facilitates Vehicle-to-Vehicle (V2V) intercommunication and V2I intercommunication.

Identifying and addressing misbehaviours

While C-ITS has shown promising results in reducing road accidents, it is highly prone to misbehaviours. Misbehaviour is a broad term and many research commonly address misbehaviour as attacks executed on a specific node. However, this research project defines misbehaviour as a node in the network that does not function as expected or causes other nodes in the network to not function as expected based on the information it perceives.

Misbehaviours can occur under two conditions in C-ITS, attacks directed towards the node or network to cause disruption of normal services or sensor failure. Sensor failure is described as a legitimate node transmitting incorrect information without malicious intent, generally caused by one or multiple sensor failure, such as incorrect position information.

Misbehaviour Detection (MD) is concerned with monitoring of data semantics of the Vehicle-to-Everything (V2X) messages to identify any misbehaving entities in the network. While many researchers have implemented misbehaviour detection in C-ITS, there are still some research gaps that prior research has not addressed. One such research gap is the detection of attacks towards RSU that cause misbehaviours in C-ITS.

This research project is undertaken to be able to detect these misbehaviours and sensor failure as soon as possible to reduce its devastating effects. A vehicle in C-ITS communicates with other vehicles or RSU using certain message types such as, Basic Safety Message (BSM) that provides kinematic information such as speed, position, and heading.

Under normal circumstances, all nodes in C-ITS work as expected, however if a malicious actor was able to take over the node transmitting the message or was able to intercept the message and inject data, then the receiver would believe the messages are true and act upon them causing road accidents, unnecessary route deviation, or even stall the receiver completely.

If a sensor failure were to occur, the information in the BSM would not be accurate, causing the receiver to act upon it, contributing to misbehaviours. Therefore, this research project is aimed at detecting the common types of attacks and sensor failure by analysing the communicating message to prevent or minimise misbehaviours, this is often referred to as a data-centric misbehaviour detection model.

Project objectives

The main objective of this research project is to design and implement a robust data-centric framework that can detect misbehaviours in C-ITS.

The following project objectives will be pursued in order to achieve this objective:

• Identification and formalisation of existing types of misbehaviours in C-ITS through the study of extant techniques.
• Design, development, and implementation of a framework for misbehaviour detection in C-ITS testbed environment to evaluate its feasibility and performance.
• Develop predictive algorithms for early detection of common misbehaviours in a C-ITS.
• Stress testing and optimisation of developed algorithms and technologies for misbehaviour detection in C-ITS.

Please note …

This page will be a living record of this project. As it matures, hits milestones, etc., we’ll continue to add information, links, images, interviews and more. Watch this space!