Circumventing network filtering with polymorphic protocol shapeshifting

Abstract

As use of the Internet has expanded to become ubiquitous, so has the use of filtering technology to selectively block access to content. In order to restore access to blocked content, filtering circumvention technologies have also been developed. This dissertation addresses the technical methods for developing a circumvention technology that is both useful and pragmatic. This goal was achieved by using modeling of filters as the basis for constructing filtering-resistant encodings that provide adaptability to filtering methods. The result was an engine that can generate encodings customized to work with different circumvention tools and to circumvent different filters. Through a value-sensitive design process a circumvention tool was developed that utilizes this engine to provide access to credible and relevant information for users on networks that implement filtering. This research has implications for the practical effectiveness of filtering technologies and the methodologies for the creation of circumvention tools.