项目名称: 基于API的静态插桩技术与Android平台恶意代码检测
项目编号: No.61272078
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 茅兵
作者单位: 南京大学
项目金额: 80万元
中文摘要: 随着智能移动终端的普及,用户正在手机与平板电脑上越来越多地使用或存储个人隐私信息。然而,在这个新兴的技术平台上,大量使用了混淆以及隐蔽技术的变种恶意代码正不断地涌现,侵害着用户的隐私。本项目以Android这个广为使用的平台作为对象,针对当前恶意代码往往是变种的、混淆的、隐蔽的,三个新特征,提出基于静态插桩的恶意代码检测技术。具体来说,我们的研究将包括三个方面。首先,API粒度程序插桩技术为我们提供高效的程序运行监控手段。其次,恶意代码行为建模技术为我们捕获恶意行为,确认变种的、代码混淆的恶意程序。最后,基于可疑API接口的识别技术为我们识破职能混淆与合谋攻击两种隐蔽攻击。本项目的研究一方面希望能够推进基于API粒度的静态插桩技术与Android恶意代码检测理论;另一方面希望能够构造通用的静态插桩工具集,为Android恶意代码分析与检测提供支持,为用户提供实用的全方位保护。
中文关键词: Android;API;敏感;恶意软件;检测
英文摘要: In these days, smart mobile phones or pads have stored numerous vital privacy information. However, on these novel platforms, a great deal of mobile malwares, which are obfuscated, concealed and varietal, are emerging and purloining users' privacy information. In order to handle malwares with these three features, we propose a malware detection approach using API-grained program instrumentation technique and choose Google Android as our target OS. Our system is supposed to solve three problems. First, we plan to implement program instrumentation technique as an efficient runtime monitor mechanism. Second, we study malware behavior modeling utilized to capture obfuscated and varietal malicious programs. Third, we study recognition technique to defend two concealed permission attacks, confused deputy attack and collusion attack. As a summary, we propose the theory of API-grained program instrumentation and Android malware detection. Meanwhile, we plan to implement a general program instrumentation tool set, which can give support to analyze and detect Android malicious programs and roundly protect users.
英文关键词: Android;API;sensitive;malware;detect