Researchers design algorithm to stop cyber attacks

Australian researchers from Charles Sturt University and the University of South Australia (UniSA) have designed an algorithm that can intercept a man-in-the-middle (MitM) cyber attack on an unmanned military robot and shut it down in seconds.

In an experiment using deep learning neural networks to simulate the behaviour of the human brain, artificial intelligence experts trained the robot’s operating system to learn the signature of a MitM eavesdropping cyber attack. This is where attackers interrupt an existing conversation or data transfer.

The algorithm, tested in real time on a United States army combat ground vehicle, was 99% successful in preventing a malicious attack. The system was validated by false positive rates of less than 2%, demonstrating its effectiveness.

The results were published in IEEE Transactions on Dependable and Secure Computing.

Professor Anthony Finn, UniSA autonomous systems researcher, said the proposed algorithm performs better than other cyber attack recognition techniques.

Finn and Dr Fendy Santoso, Charles Sturt Artificial Intelligence and Cyber Futures Institute, collaborated with the US Army Futures Command to replicate a man-in-the-middle cyber attack on a GVT-BOT ground vehicle and trained its operating system to recognise an attack.

According to Finn, the robot operating system (ROS) is extremely susceptible to data breaches and electronic hijacking because it is so highly networked. The Industry 4 advent has demanded that robots work collaboratively, where sensors, actuators and controllers need to communicate and exchange information with one another via cloud services.

This makes them highly vulnerable to cyber attacks; however, the speed of computing doubles every few years and it is now possible to develop and implement sophisticated AI algorithms to guard systems against digital attacks.

Santoso said that the robot operating system largely ignores security issues in its coding scheme due to encrypted network traffic data and limited integrity-checking capability, despite its benefits and widespread usage.

“Owing to the benefits of deep learning, our intrusion detection framework is robust and highly accurate,” Santoso said. “The system can handle large datasets suitable to safeguard large-scale and real-time data-driven systems such as ROS.”

Finn and Santoso plan to test their intrusion detection algorithm on different robotic platforms, such as drones, whose dynamics are faster and more complex compared to a ground robot.

October is Cyber Security Awareness Month.

Image credit: iStock.com/bymuratdeniz