Protocol Vulnerability Analysis




Network protocols continue to suffer from well documented vulnerabilities. Despite this, a practical methodology for classifying these vulnerabilities does not exist. In this paper, we present such a methodology.

We have developed a grammar for expressing network protocol exploits in terms of vulnerabilities and symptoms. Vulnerabilities are defined by characteristics, conditions which must hold for a vulnerability to exist. Symptoms are the violations of policy enabled by vulnerabilities. Exploits, then, are the pairing of vulnerabilities with their corresponding symptoms.

Using our grammar, we analyzed many protocols and present our classifications visually using syntax trees. We detail the classification process, and discuss future applications of this work towards a secure protocol design framework.