Efficient local strategies for vaccination and network attack

Department of Physics, Umeå University - 901 87 Umeå, Sweden, and Department of Physics, University of Michigan - Ann Arbor, MI 48109, USA

Received: 8 September 2004
Accepted: 22 October 2004

Abstract

We study how a fraction of a population should be vaccinated to most efficiently stop epidemics. Our starting point is that only local information —about the neighborhood of specific vertices— is usable in practice. For this reason we consider only local vaccination strategies. The efficiency of the vaccination strategies is investigated with both static and dynamic measures. We test the strategies on real-world networks and model networks with real-world characteristics. Among other things, we find that the most efficient strategy for many real-world situations is to iteratively vaccinate the neighbor of the previous vaccinee that has the most links out of the neighborhood.