CloudIntellMal: An advanced cloud based intelligent malware detection framework to analyze android applications

作     者：Mishra, Preeti Jain, Tanmay Aggarwal, Palak Paul, Gunjan Gupta, Brij B. Attar, Razaz Waheeb Gaurav, Akshat 

作者机构：Doon Univ Sch Technol Dept Comp Sci Dehra Dun India Reverie Language Technol Ltd Bangalore India Cognizant Ghaziabad India Amazon Bangalore India Asia Univ Dept Comp Sci & Informat Engn Taichung Taiwan Kyung Hee Univ 26 Kyungheedae Ro Seoul South Korea Symbiosis Int Univ Symbiosis Ctr Informat Technol SCIT Pune India Univ Petr & Energy Studies UPES Ctr Interdisciplinary Res Dehra Dun India Princess Nourah bint Abdulrahman Univ Coll Business Adm Management Dept POB 84428 Riyadh 11671 Saudi Arabia Ronin Inst Montclair NJ USA 

出 版 物：《COMPUTERS & ELECTRICAL ENGINEERING》 (计算机与电工)

年 卷 期：2024年第119卷第PartA期

核心收录：

学科分类：0808[工学-电气工程] 08[工学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：SERB-POWER Grant under Science and Engineering Research Board, Department of Science and Technology (SERB-DST) , Govt. of India [SPG/2021/002003] Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia [PNURSP2024R 343] 

主　　题：Android malware detection Static analysis Amazon EC2 Amazon S3 Malware detection Artificial Intelligence 

摘      要：Serverless computing has become very popular in recent times which facilitates a greater abstraction in virtualization technology. With the rapid development in serverless computing, a new paradigm has been evolved to design and develop android applications. However, the android platform has become one of the most vulnerable targets due to the increase in the number of users and hence raises a strong security concern to develop an advanced security framework. In this paper, a serverless computing based intelligent malware detection framework, called CloudIntellMal is proposed to detect malware targeting android-based applications. The framework incorporates some popular key cloud services such as Amazon Web Services (AWS) Elastic Cloud Compute (EC2) and Simple Storage Service (S3) to store and pre-process the logs taken from the end user devices. An efficient feature extraction algorithm is designed that employs Bag of n-grams to develop feature vectors. The framework runs the machine learning algorithm in the cloud infrastructure to learn malicious patterns. The decision model is loaded in EC2 at the time of detection phase to classify the monitored apps. A prototype of the approach has been developed and validated using Drebin dataset and results seem to be promising.