Team develops new ‘attacker’ tool to improve autonomous car safety

Modern cars and autonomous vehicles use millimeter wave (mmWave) radio frequencies to enable self-driving or assisted driving features that ensure the safety of passengers and pedestrians. However, this connectivity can also expose them to potential cyber attacks.

To help improve the safety and security of autonomous vehicles, researchers and collaborators in the lab of Dinesh Bhardia, an affiliate of the UC San Diego Qualcomm Institute (QI) and faculty member of the university’s Jacobs School of Engineering’s Department of Electrical and Computer Engineering from Northeastern University devised a novel algorithm designed to mimic an attacker’s device.

The algorithm, described in the paper “MMSpoof: Resilient Spoofing of Automotive Millimeter-Wave Radars Using Reflect Array,” lets researchers identify areas for improving autonomous vehicle safety.

“The invention of autonomous systems, like self-driving cars, was to enable the safety of humanity and prevent loss of life,” Bhardia said. “Such autonomous systems use sensors and sensing to provide autonomy. Therefore, safety and security depend on obtaining high-fidelity sensing information from sensors. Our team disclosed a radar sensor vulnerability and developed a solution did, which autonomous cars should strongly consider.”

defense against cyber attack

Autonomous cars detect obstacles and other potential hazards by sending out radio waves and recording their reflections as they bounce off surrounding objects. By measuring the time it takes for the signal to return, as well as the change in its frequency, the car can detect the distance and speed of other vehicles on the road.

However, like any wireless system, autonomous cars carry the risk of cyber attack. Attackers driving ahead of an autonomous unit may engage in “spoofing”, an activity that involves interfering with the vehicle’s return signal to help it register an obstacle in its path. After this, the vehicle may brake suddenly, which increases the risk of an accident.

To address this potential chink in the armor of autonomous cars, Venom and his colleagues devised a novel algorithm designed to mimic a spoofing attack. Previous attempts to develop an attacker’s device to test the cars’ resistance have had limited feasibility, assuming either that the attacker can synchronize with the victim’s radar signal to launch an attack, or assuming that both The cars are physically connected by a cable.

In their new paper presented by Venom at the IEEE Symposium on Security and Privacy in San Francisco on May 22, the team describes a new technique that uses the victim vehicle’s radar against itself. By subtly changing the parameters of the received signal to “lightspeed” before reflecting back, an attacker can hide their sabotage and make it much harder for the vehicle to filter out malicious behavior. All this can be done “on the go” and in real time without knowing anything about the victim’s radar.

“Automotive vehicles rely heavily on mmWave radar to enable real-time situational awareness and advanced features to promote safer driving,” Vennam said. “The security of these radars is paramount. We—mmSpoof—detect a serious security problem with mmWave radars and demonstrate a robust attack. The dangerous thing is that anyone using off-the-shelf hardware components Can build a prototype.

To counter this type of attack, Venom suggests, researchers seeking to improve the safety of autonomous vehicles could use a high-resolution radar capable of capturing multiple reflections from a car. So that the correct image can be correctly identified. Researchers can also create backup options for radar by including cameras and “light detecting and ranging” (LiDAR), which records the time it takes to hit an object and return to measure the area around it .

Alternatively, the team presents mmSpoof as a means of preventing dangerous tailgating. By placing a mmSpoof device behind their car, drivers can trick the tailgating car into entering a slower car in front of them and activating the brakes.