基于同一家族恶意软件在行为上的相似性特征,提出了一种基于行为的Android恶意软件家族聚类方法.该方法构建了软件行为刻画特征集合,通过定制ROM的方式来构建行为捕获机制并采集恶意软件的行为日志,基于行为日志提炼恶意软件特征集,使用DBSCAN(density-based spatial clustering of applications with noise)聚类算法进行家族聚类.通过对大量已经人工分类的恶意软件进行评估,实验结果表明,在最优情况下,本方法在聚类准确率上达到了91.3%,在测试样本识别预测上正确率达到了82.3%.
In this paper,we propose a behavior-based malware cluster method inspired by the characteristic that Android malwares in a same family behave similarly to speed up the procedure of malware detection and analysis.With agood understanding of Android malware,it defines features for characterizing malware behavior,and extracts them from a customized Android ROM while malware running.With these collected feature data,DBSCAN algorithm is employed to cluster malwares into various families.The evaluation results shows that the accuracy rate of malware clustering can reach 91.3% at best,and the correct prediction rate is 82.3%for evaluated samples.