Efficient Browser Identification with JavaScript Engine Fingerprinting

For security and privacy on the Internet the webbrowser has become an important piece of software. Nowadays even mobil devices include a powerful browser, and in combination with cloud computing the webbrowser mimics more and more operating system functionality. For many use cases, legit server operators as well as attackers need to find out the exact version of a particular user’s webbrowser. While the user agent string (which is currently used for that purpose) can be easily manipulated, certain aspects of the underlying software cannot. In our approach we will use the deviance of Javascript engine implementations to identify the browser accurately. This approach is efficient and has low computational overhead, and might be used to identify the version of the webbrowser regardless of the possibly manipulated user agent string.