摘要
作为世界上最流行的移动操作系统,Android正面临着快速增长的恶意软件的威胁。如何快速高效地检测出Android恶意软件对保证用户手机安全具有十分重大的意义。从Android软件的权限出发,统计了4000多个恶意应用和2000多个正常应用的权限分布情况,依据特征权限在恶意应用和正常应用中的分布规律,设计了一种轻量级的快速检测方法 LWD(Light Weight Detection)。LWD根据特征权限在恶意应用中的使用频率和在正常应用中的使用频率的不同来定量分析特征权限恶意程度值,并以此计算每个样本的恶意程度值是否超过规定阈值来判断该样本是否属于恶意应用。实验结果表明,与市场上主流的杀毒软件相比,LWD方法具有较好的检测率。而且LWD是基于单一的权限特征对恶意软件进行检测,因此具有较高的时间效率。作为一种轻量级检测方法,LWD可以为更进一步深入检测恶意应用提供参考依据。
As the most popular mobile operating system in the world,Android platform has been under the threat of the quickly growth of malware. It makes a significant sense to find solutions to detect the malware quickly. After analyzing the character permissions usage in more than 4000 malware and more than 2000 benign applications,a solution named LWD（ Light Wight Detection） to detect malware is proposed. LWD defines the malicious value of character permissions on basis of the permissions usage frequency in malware and benign applications. Then,the application will be judged whether it is a malware by calculating the malicious value of the application. The results of experiment show that LWD has a better detection compared with other popular anti-virus software. As a single permission characterization,LWD has good time efficiency and it is able to provide foundation of further malware detection.
出处
《计算机应用与软件》
2017年第1期306-310,320,共6页
Computer Applications and Software
关键词
智能手机
安卓系统
特征权限
LWD
频率
Smart phone
Android system
Character permissions
LWD
Frequency
