We aim to better understand the features of network protocols that facilitate denial of service attacks, in order to design more robust protocols and architectures in the future and evaluate existing designs more accurately.

Denial of service attacks pose a serious threat to the current Internet architecture. The most common attack methods used today leverage just a small handful of vulnerabilities in common network protocols such as TCP and DNS. To avoid repeating this problem, great care must be taken to identify potential vulnerabilities when evaluating proposals for new protocols and future Internet architectures. Our goal is to develop a model for designing more robust protocols and identifying flaws in existing ones. This model will enable us to understand the vulnerabilities inherent in features like mobility, to determine which layer of the protocol stack is best suited to handle such features securely, and to more accurately compare seemingly disparate protocols.