融合多特征的Android恶意软件检测方法

doi:10.3969/j.issn.1671-1122.2020.01.010

摘要/Abstract

摘要：

针对当前基于机器学习的Android恶意软件检测方法特征构建维度单一,难以全方位表征Android恶意软件行为特点的问题,文章提出一种融合软件行为特征、AndroidManifest.xml文件结构特征和Android恶意软件分析经验特征的恶意软件检测方法。该方法提取Android应用的Dalvik操作码N-gram语义信息、系统敏感API、系统Intent、系统Category、敏感权限和相关经验特征,多方位表征Android恶意软件的行为并构建特征向量,采用基于XGBoost的集成学习算法构建分类模型,实现对恶意软件的准确分类。在公开数据集DREBIN和AMD上进行实验,实验结果表明,该方法能够达到高于97%的检测准确率,有效提升了Android恶意软件的检测效果。

关键词: Android, 恶意软件, 融合多特征, XGBoost

Abstract:

Aiming at the current problem that the feature construction of Android malware detection method based on machine learning has a single dimension and it is difficult to comprehensively characterize the behavior characteristics of Android malware, this paper proposes a malicious software detection method that integrates the behavior characteristics of software, the structural characteristics of AndroidManifest.xml file and the characteristics of Android malware analysis experience. This method extracts the N-gram semantic information, system sensitive API, system Intent, system Category, sensitive authority and relevant experience characteristics of the Dalvik operand code of Android application, characterizes the behavior of Android malware in multiple directions, and constructs the feature vector. Then, the integrated learning algorithm based on XGBoost is used to construct the classification model, so as to realize the accurate classification of malware. Experiments were conducted on DREBIN and AMD in the open data set, and the experimental results showed that this method could achieve a detection accuracy of over 97%, which effectively improved the detection effect of Android malware.

Key words: Android, malware, multi-feature, XGBoost

中图分类号:

TP309

侯留洋, 罗森林, 潘丽敏, 张笈. 融合多特征的Android恶意软件检测方法[J]. 信息网络安全, 2020, 20(1): 67-74.

HOU Liuyang, LUO Senlin, PAN Limin, ZHANG Ji. Multi-feature Android Malware Detection Method[J]. Netinfo Security, 2020, 20(1): 67-74.

图/表 12

图1

图2

表1

表2

表3

表4

表5

表6

表7

表8

表9

表10

参考文献 20

[1]	FORNI A A. Gartner Says Worldwide Sales of Smartphones Grew 9 Percent in First Quarter of 2017[EB/OL]. , 2019-6-5.
[2]	NICOLAS L. Number of Available Applications in the Google Play Store from December 2009 to December 2017[EB/OL]. , 2019-6-5.
[3]	RALF B. Malware Trends 2017[EB/OL]. , 2019-6-5.
[4]	WU Dongjie, MAO C H, WEI T E, et al.DroidMat: Android Malware Detection through Manifest and API Calls Tracing[C]//IEEE. 2012 Seventh Asia Joint Conference on IEEE, August 9-10, 2012, Tokyo Japan. New Jersey: IEEE, 2012: 62-70.
[5]	AU K W Y, ZHOU Yifan, ZHANG Huang, et al. PScout: Analyzing the Android Permission Specification[C]//ACM. ACM Conference on Computer & Communications Security ACM, October 16-18, 2012, Raleigh, NC, USA. New York: ACM, 2012: 217-228.
[6]	PANDITA R, XIAO Xusheng, YANG Wei, et al.WHYPER: Towards Automating Risk Assessment of Mobile Applications[C]//ACM. 22nd USENIX Conference on Security, August 14-16, 2013, Berkeley, CA, USA. New York: ACM, 2013: 571-584.
[7]	IDRESS F, RAJARAJAN M.Investigating the Android Intents and Permissions for Malware Detection[C]//IEEE. 10th International Conference on Wireless and Mobile Computing, Networking and Communications(WiMob), October 8-10, 2014, Larnaca, Cyprus. New Jersey: IEEE, 2014: 354-358.
[8]	LI Li, BARTEL A, BISSYANDE T F, et al.IccTA: Detecting Inter-Component Privacy Leaks in Android Apps[C]//IEEE. 37th International Conference on Software Engineering(ICSE 2015), May 16-24, 2015, Florence/Firenze, Italy. New Jersey: IEEE, 2015: 280-291.
[9]	ARP D, SPREITZENBARTH M, HUBNER M, et al.Drebin: Effective and Explainable Detection of Android Malware in Your Pocket[C]//IEEE. 2014 Network and Distributed System Security Symposium, February 23-26, 2014, San Diego, California, USA. New Jersey: IEEE, 2014: 23-38.
[10]	CHIN E, FELT A P, GREENWOOD K, et al.Analyzing Interapplication Communication in Android[J]. Signo Pensam, 2006, 27(49): 36-45.
[11]	ZHOU Wu, ZHOU Yajin, GRACE M, et al.Fast, Scalable Detection of Piggybacked Mobile Applications[C]//ACM.3rd ACM Conference on Data and Application Security and Privacy, February 18-20, 2013, San Antonio, TX, USA. New York: ACM, 2013: 185-196.
[12]	HOFFMANN J, USSATH M, HOLZ T, et al.Slicing Droids: Program Slicing for Smali Code[C]//ACM. 28th Annual ACM Symposium on Applied Computing, March 18-22, 2013, Coimbra, Portugal. New York: ACM, 2013: 1844-1851.
[13]	YU Wei, GE Linqiang, XU Guobin, et al. Towards Neural Network Based Malware Detection on Android Mobile Devices[EB/OL]. , 2019-6-5.
[14]	ZHOU Wu, ZHOU Yajin, JIANG Xuxian, et al.Detecting Repackaged Smartphone Applications in Third-party Android Marketplaces[C]//ACM. 2nd ACM Conference on Data and Application Security and Privacy, February 7-9, 2012, San Antonio, Texas, USA. New York: ACM, 2012: 317-326.
[15]	WOGNSEN E R, KARLSEN H S, OLESEN M C, et al. Formalisation and Analysis of Dalvik Bytecode[EB/OL]. , 2019-7-10.
[16]	AAFER Y, DU Wenliang, YIN Heng.DroidAPIMiner: Mining API-level Features for Robust Malware Detection in ANDROID[C]//Springer. International Conference on Security and Privacy in Communication Systems, September 25-28, Sydney, NSW, Australia. Heidelberg: Springer, 2013: 86-103.
[17]	WEI Fengguo, LI Yuping, ROY S, et al.Deep Ground Truth Analysis of Current Android Malware[C]//Springger. International Conference on Detection of Intrusions and Malware, and Vulnerability Assessment, July 6-7, 2017, Bonn, Germany. Heidelberg: Springer, 2017: 252-276.
[18]	LI Jin, SUN Lichao, YAN Qiben, et al.Significant Permission Identification for Machine-learning-based Android Malware Detection[J]. IEEE Transactions on Industrial Informatics, 2018, 14(7): 3216-3225.
[19]	XIE Niannian, ZENG Fanping, QIN Xiaoxia, et al.RepassDroid: Automatic Detection of Android Malware Based on Essential Permissions and Semantic Features of Sensitive APIs[C]//IEEE. 2018 International Symposium on Theoretical Aspects of Software Engineering(TASE), August 14-16, 2018, Guangzhou, China. New Jersey: IEEE, 2018: 52-59.
[20]	ZHANG Hanqing, LUO Senlin, ZHANG Yifei, et al. An Efficient Android Malware Detection System Based on Method-level Behavioral Semantic Analysis[EB/OL]. , 2019-8-12.

符号	语义	数量	代表的Dalvik指令
C	Compare	5	Cmp-float\|cmpg-float\|cmpl_double\|cmpg-double \|cmp-long
D	Definition	11	Const\|const/4\|const/16\|const-wide\|const/high\|const-string
M	manipulation	13	Move\|move-wide\|move-object\|move-result\|move-exception
R	Reture	4	Reture\|return-void\|return-wide\|return-object
L	Monitor	2	Monitor-enter\|monitor-exit
G	Jump	3	Goto\|goto/16\|goto/32
I	Judgemnet	12	If-eq\|if-ne\|if-lt\|if-ge\|if-gt\|if-le\|if-eqz\|if-nez\|if-gez\|if-gtz\|if-lez
T	Reading	21	Aget\|iget\|sget\|aget-wide\|aget-object\|aget-boolea\|aget-byte\|aget-char
P	Writing	21	Ainput\|iput\|sput\|aput-wide\|aput-object\|aput-boolea\| aput-byte\|aput-char
V	Method call	15	Invoke-virtual\|invoke-super\|invike-direct\|invoke-static

特征类别	数量	特征部分实例
敏感API特征	629	Android/net/ConnectivityManager;startUsingNetworkFeature Android/net/wifi/p2p/WifiP2pManager;initialize Android/net/ConnectivityManager;stopUsingNetworkFeature Android/provider/Browser;getVisitedHistory Android/location/LocationManager;getProvider
敏感权限特征	21	Android.permission.CAMERA Android.permission.READ_CONTACTS Android.permission.WRITE_CONTACTS Android.permission.READ_PHONE_STATE Android.permission.READ_CALL_LOG
系统Action特征	41	Android.intent.action.MAIN Android.intent.action.VIEW Android.intent.action.PICK_ACTIVITY Android.intent.action.MEDIA_MOUNTED Android.intent.action.MEDIA_UNMOUNTED
系统Category	11	Android.intent.category.DEFAULT Android.intent.category.BROWSABLE Android.intent.category.TAB Android.intent.category.CAR_DOCK Android.intent.category.DESK_DOCK
经验特征	17	权限数量资源文件中包含图像的文件个数资源文件中包含可执行文件个数 DEX文件中所有类的个数超过10个 DEX文件中所有方法的个数超过1000个

特征	特征详细信息
恶意行为特征	3-gram信息,629个敏感API
文件结构特征	41个系统Action,11个系统Category,7个敏感权限
经验特征	Service数量等17个

数据名称	特征数	特征说明
DREBIN	1697	敏感权限、敏感API、系统Action、系统Category、3-gram和经验特征
AMD	1697	敏感权限、敏感API、系统Action、系统Category、3-gram和经验特征

软件名称	来源	说明
Xshell6	Netsarang	终端模拟管理系统
Winscp	SourceForge	远程文件传输工具
androguard	Python库	静态解析Android应用文件
NLTK	Python库	提取操作码N-gram信息