Methods for intelligent mapping of the IPv6 address space

Abstract

Due to the rapid growth of the Internet, the available pool of unique addresses in version four of the Internet Protocol (IPv4) is nearly depleted. As a result, the next generation protocol, IPv6, is now widely implemented and rapidly being adopted. This thesis examines new methods for active mapping of the IPv6 topology, i.e., router and link discovery. Better characterization of the IPv6 topology can provide the Department of Defense and other federal agencies the ability to defend networks and more effectively respond to cyber attacks. However, given that the IPv6 address space is roughly 79 billion billion billion times larger than the IPv4 space, mapping in IPv6 introduces new network measurement challenges. As a first step toward intelligent IPv6 topology discovery, this thesis takes lessons learned in efficient IPv4 mapping algorithms and attempts to apply them to IPv6. We develop several novel IPv6 mapping techniques and evaluate their probing time and topological coverage as compared to current state-of-the-art systems. Finally, we uncover previously unrecognized properties of several IPv6 deployments, infer network topologies, and characterize common IPv6 subnetting structure.