Modeling Internet-Scale Policies for Cleaning Up Malware
The results of the authors' experiments using ASIM indicate that when filtering wicked traffic, the best targets for intervention are a small group of the largest ASes. Specifically, they find that intervention by the top 0.2% of ASes (in terms of size) is more effective than intervention by a randomly chosen subset of 30% of all ASes. However, they show that this efficacy rapidly drops off if less than three quarters of that top 0.2% intervene. This is an issue of importance if not all the largest ASes fall within the same regulatory domain, such as a nation-state.