According to the US university new technique is based on the use of motion sensors, position and mobile environment to meet user gestures.
The malware on mobile is a growing problem, but researchers at the University of Alabama, USA, discovered a new way to detect when doubtful mobile applications begin to act strangely.
The technique relies on the use of motion sensors, position and mobile environment to learn the gestures that users often do when a phone call is started, take pictures or use NFC reader to scan your credit card.
Some mobile malware programs and abuse of these services and security researchers predict that number will only grow.
The technology developed by researchers at UAB can control these three services and check that attempts to access them are associated with natural gestures expected by users. If not, you probably were initiated by malware. The study, which involved collecting data from real-life situations to train technology, showed that detect different gestures and use them to differentiate between the actions initiated by users and other automated to do something with a high degree of accuracy. Therefore, the technique can be a viable defense against malware.
The technology does not require root access on the device and is better than that based on the signature used by most mobile antivirus services, according to the director of the UAB systems security and privacy, Nitesh Saxena approach.
“The current antivirus does not detect new forms of malware, due to the lack of signatures for these malware” Saxena, an author of the study. “By contrast, our approach does not apply to any firm.”