Now a majority of new cars run with sophisticated computing technology. They are vulnerable to malicious attacks in their networks that could lead to harmful safety issues.
Researchers at UMBC (University of Maryland, Baltimore County) have tried to build different methods for improving the safety of vehicles with technology. Riadul Islam, assistant professor of computer science and electrical engineering, collaborated with the University of Maryland, Baltimore County, and the University of Michigan-Dearborn to design a simple and easily adapted method for detecting breaches in security. The research paper is published in the IEEE (Institution of Electrical and Electronic Engineers) publication Transactions on Intelligent Transportation Systems.
CAN or Controller Area Network is the most widely used intra-vehicular communications network in the automobile industry. CAN is simple to use, a network that makes it appealing for both consumers and manufacturers. But the main issue is that it is vulnerable to potential security threats. Any entity can read the messages coming from a car and possibly send conflicting messages as Controller Area Network is a broadcasting network. With the help of the Controller Area Network, it is possible to control the car from any other device. It’s both a characteristic and a bug, creating security concerns and enabling many innovations. Any person could take control over the network and can send new commands to a vehicle, for example, creating dangerous circumstances, such as disabling the breaks or causing engine failure. The first step to completely removing these threats is detecting them.
Identification of threat does not require any extensive technology but needs formulation of graph-based anomaly detection techniques that will show the complex relationship between data, commented by Islam. To detect the intruders or threads Islam’s team took the graphs that demonstrate the data on the network and conducted a simple statistical analysis. This relies on methods that are already well understood by statisticians and capable of functioning intuitively so it does not require any costly machinery. Detecting and addressing network vulnerabilities becomes essential since computerized cars become a reality. Islam and his team have shown that it is not expensive to be effective.
