"A New and Non-Intrusive Method for Preventing Cyber Attacks on Android Devices"

Android is the most targeted mobile operating system by malware. Researchers at the Singapore Management University (SMU) have discovered a new way to prevent cyberattacks on Android devices. The method is described as dynamic, intelligent, and non-intrusive in detecting malware on Android devices. The researchers are leveraging a side-channel for detecting sensitive and unusual behaviors on mobile apps. Their method is convenient as it does not require rooting or gaining privilege control from Android users. Android operating system upgrades do not affect the detection method. The method of detection also does not breach the Personal Data Protection Act of 2012 since it does not extract data in its performance. The research team designed the side-channel monitoring system by taking input from side-channel readings and using artificial intelligence and deep machine learning (ML) to train a deep neural network model to determine if sensitive or uncharacteristic behavior has been exhibited on mobile apps. This approach to monitoring and detection offers researchers a way to dynamically monitor apps' behaviors instead of statically analyzing each app's code. Using this method, stealthy attacks can be detected. Testing of the technique showed that it could detect sensitive behavior, with a 98.5 percent accuracy rate. This article continues to discuss the growing sophistication of cyberattacks, the heavy targeting of the Android operating system by hackers, challenges associated with designing a malware detection system for Android, and the side-channel monitoring solution designed by SMU researchers to protect Android devices from cyberattacks.

SMU reports "A New and Non-Intrusive Method for Preventing Cyber Attacks on Android Devices"