News > Inventing a method for rapid detection of Android malware by means of graphic patterns


  Print        Send to Friend

Inventing a method for rapid detection of Android malware by means of graphic patterns

Using graphic patterns, the researchers of TMU's Computers Faculty managed to develop a method for rapid detection of Android malwares. The results of this study showed that the new method can detect malwares more rapidly and is able to identify the relevant malware in 93.36% of the cases.

Recent studies show that Android operating system is the dominant system in more than 75% of the mobile market. High utilization of Android devices makes them the main target for security and privacy breaching. Therefore, the study of Android malwares is of great importance. In this study which has done by Mahsa Husseini, as Master’s thesis in Computer engineering, a new method was developed based on graphic patterns to detect the malwares in Android devices. The new method consists of two steps, identifying the graphic patterns and detecting them. In the first stage, at first one graph which is sensitive to security is produced for each detected malware and then this graph is refined to reduce the graphic pattern production time. Then a graphic pattern will be produced for each malware family by repeatable pattern detection. In detection stage, by comparing adopting the retrieved graph which is sensitive to Android security, graphic pattern is used for Android malwares detection.

Two data sets were used to evaluate the new method. The first data set includes Android malwares classified into different categories and the second data set includes some ordinary Android applications. The results of the study showed that the suggested method can detect Android malwares with a detection rate of 93.36%. This research was supervised by Dr. Mahdi Abadi, faculty member of Computer Engineering Department.

14:26 - 2018/02/05    /    number : 6595    /    Show Count : 875