A Survey on Measuring Anonymity in Anonymous Communication Systems

The popularity of Internet applications has made communication privacy an increasingly important security requirement. As an important aspect of privacy, anonymity ensures that a subject may use resources or services without disclosing user identity or corresponding relationships. Since the seminal work by Chaum for anonymous communication, many different anonymous communication systems, and anonymous protocols have been developed and investigated extensively. In recent years, anonymous communication systems have evolved from academic tools used by specialists to mass-market software used by millions of ordinary people. How to evaluate and quantify the anonymity that different anonymous communication systems can offer has been a new focus. Though some efforts have been made on anonymity metrics and measuring techniques, systematic research on measuring anonymity of anonymous communication systems is still needed. In this paper, we give a comprehensive overview of state-of-the-art research on measuring anonymity. We first summarize the anonymous mechanisms applied by different anonymous systems. We then introduce the formalization of the notion of anonymity for measuring the anonymity degree. Further, metrics based on various theories and approaches are reviewed. We particularly elaborate on the metrics based on information theory as a separate section because of its popularity and multiple branches. The metrics based on information entropy is probably of the greatest interest.