Distributed Estimation and Selective Control of Epidemics Spreading Over Complex Networks

In this work we propose a novel control scheme for the containment of infectious epidemics spreading over networks of individuals. The proposed control scheme consists of a distributed estimator, which provides the estimate of the state of each individual, and a resource allocation policy, which selects potential individuals to test and quarantine, and to vaccinate, taking into account possibly limited numbers of tests and vaccine doses available. The proposed strategy provides interventions on an individual basis rather than average actions at the population level, as commonly done in most of the existing literature. Moreover it exploits the available observations on the individual states in a closed-loop fashion, contrary to the open-loop approach used on the part of the literature focusing on individual-based control strategies. Simulation results show how the proposed approach produces a clear improvement in the containment of epidemics compared to traditional strategies and that our approach is robust with respect to uncertainties on the knowledge of the network.