Mitigating Biological Epidemic on Heterogeneous Social Networks

TL;DR

This paper develops an optimal control framework for epidemic mitigation on heterogeneous social networks, analyzing vaccination and treatment strategies across different network topologies to improve outbreak management.

Contribution

It introduces a degree-based compartmental model with separate control signals for network groups, optimizing strategies for various network types including real-world data.

Findings

Optimal strategies significantly outperform non-optimal heuristics.

Degree classes critical for epidemic control vary with network topology.

Cost of controls influences resource allocation and strategy effectiveness.

Abstract

Recent Covid-19 pandemic has demonstrated the need of efficient epidemic outbreak management. We study the optimal control problem of minimizing the fraction of infected population by applying vaccination and treatment control strategies, while at the same time minimizing the cost of implementing them. We model the epidemic using the degree based Susceptible-Infected-Recovered (SIR) compartmental model. We study the impact of varying network topologies on the optimal epidemic management strategies and present results for the Erdos-Renyi, scale free, and real world networks. For efficient computational modeling we form groups of groups of degree classes, and apply separate vaccination and treatment control signals to each group. This allows us to identify the degree classes that play a significant role in mitigating the epidemic for a given network topology. We compare the optimal control strategy with non optimal strategies (constant control and no control) and study the effect of various model parameters on the system. We identify which strategy (vaccination/treatment) plays a significant role in controlling the epidemic on different network topologies. We also study the effect of the cost of vaccination and treatment controls on the resource allocation. We find that the optimal strategy achieves significant improvements over the non optimal heuristics for all networks studied in this paper. Our results may be of interest to governments and healthcare authorities for devising effective vaccination and treatment campaigns during an epidemic outbreak.