MinerGuard: A Solution to Detect Browser-Based Cryptocurrency Mining through Machine Learning

• Main Authors: Yen-Jung Lai, 賴彥蓉
• Other Authors: Fu-Hau Hsu
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/9k9m6u

碩士 === 國立中央大學 === 資訊工程學系 === 106 === Since Coinhive released its browser-based cryptocurrency mining code in September 2017, many websites embed mining JavaScript to mine cryptocurrency by using CPU resources without the consent of the device owner, it’s called Cryptojacking. And Cryptojacking has become the latest attack trend in computer security field. Many security specialists provide some methods to block the mining scripts, such as filtering mining scripts by blacklist. However, due to the significant increase in the Cryptojacking attacks, the static blacklist mechanism has become useless to protect users in time. In this paper, we design and implement the mining identification mechanism which based on the observation of users’ computer resources. Our mechanism observes the changes of CPU usages in time to identify whether or not a website uses the mining scripts and notify the users. The experiment results show that our system is more accurate than the blacklist mechanism and our system does not need to update system regularly. But the blacklist mechanism has to update blacklist constantly. Abuse of web mining scripts and illegal acts of Cryptojacking are becoming more and more serious. The way to prevent Cryptojacking effectively will become a new issue for security. And the goal of our study is to protect people from becoming miners.