Optimal discrete resource allocation on metapopulation networks for suppressing spatial spread of epidemic

Suppressing an epidemic among regions within limited medical resources has attracted widespread attention but remains challenges. Thereby, within a fixed budget, a reasonable strategy for resource deployment inevitably plays a significant role on constraining the epidemic. In this paper, a metapopulation network model is constructed to simulate the spatial evolution of the epidemic coupled with the discrete resource descriptions, which are mapped to the parameters with binary scopes. Besides, to obtain the optimal resource deployment, we adopt the Binary Particle Swarm Optimization (BPSO) algorithm to search the optimal solution. Experimental results show that BPSO can achieve effective discrete resource deployment, and its solution is significantly better than random allocation of resource. When the budget is low, the vaccine resources are allocated preferentially, with few curative resources allocated. With the budget increasing, all the patches are almost filled with the vaccine resources, and the curative resources are gradually allocated. In addition, under limited budget, the resources are allocated into patches by a scattered manner, and the network with intensive community structure or not has few impacts on resource allocation, which further implies that resources cannot be over concentrated in contiguous areas.

Automated summary

Suppressing an epidemic in regions with limited medical resources is challenging but crucial. This study constructs a metapopulation network model to simulate the spatial evolution of the epidemic and adopts the Binary Particle Swarm Optimization (BPSO) algorithm to find the optimal resource deployment. Experimental results show that BPSO can effectively allocate resources and outperforms random allocation. In cases of low budget, vaccine resources are prioritized while curative resources are allocated sparingly. As the budget increases, more areas are filled with vaccine resources and curative resources are gradually allocated. Additionally, under a limited budget, resource allocation is scattered and the network's community structure has minimal impact on allocation, indicating the need to avoid over-concentration of resources.