A recent paper called “Worm Propagation Strategies in an IPv6 Internet”, written by Steven M. Bellovin, Angelos Keromytis, and Bill Cheswick, examines whether or not the deployment of IPv6 will in fact provide a substantial level of barrier against worms. Shared below are the introductory paragraphs from this paper.

In recent years, the internet has been plagued by a number of worms. One popular mechanism that worms use to detect vulnerable targets is random IP address-space probing. This is feasible in the current Internet due to the use of 32-bit addresses, which allow fast-operating worms to scan the entire address space in a matter of a few hours. The question has arisen whether or not their spread will be affected by the deployment of IPv6. In particular, it has been suggested that the 128-bit IPv6 address space (relative to the current 32-bit IPv4 address space) will make life harder for the worm writers: assuming that the total number of hosts on the Internet does not suddenly increase by a similar factor, the work factor for finding a target in an IPv6 Internet will increase by approximately 2⁹⁶, rendering random scanning seemingly prohibitively expensive.

Some worms, such as Melissa, spread by email. These worms will not be affected by the adoption of IPv6; though the space of possible email addresses is vast, these worms typically consult databases such as Microsoft Outlook’s address book.

On the other hand, life will indeed be harder for address-space scanners, such as Code Red and Slammer. Clever heuristics can cut the search space dramatically. More specifically, multi-level searching and spreading techniques can negate the defender’s advantage. However, the code size required for worms will increase, which may help prevent Slammer-like attacks. This has created the impression that an IPv6 Internet would be impervious to similar kinds of worms.

In the past, there have been two forms of address space scans. Some worms use a uniformly distributed random number generator to select new target addresses. This strategy is indeed unlikely to succeed in an IPv6 world. Other worms preferentially spread locally, by biasing the search space toward addresses within the same network or subnet. This will be a more successful strategy, though at first glance the 80-bit local space (nearly twice Avogadro’s number!) would seem to be a formidable obstacle. We observe that certain strategies can improve the attacker’s odds. In particular, by taking advantage of local knowledge and patterns in address-space assignment, the attack program can cut the search space considerably.

We discuss a number of strategies worms could use in an IPv6-based Internet to find new targets. We separate these into two categories, wide area and local-area searches, somewhat mirroring the IPv6 address architecture. We argue that worms will use different types of information sources to first determine existing networks and establish a presence there, and then spread locally inside an organization. We hope to illustrate that simple reliance on the IPv6 address space for protection against scanning worms is not a wise defensive strategy, and we suggest areas where research could assist in detecting and limiting future worm propagation.