Abstract
Disasters severely damage cellular network infrastructures, weakening network communication service capability (CSC) and impeding post-disaster efforts. Therefore, evaluating network ability to resist disasters and recovering CSC are crucial. In this paper, we introduce a novel metric, network capacity scalability (NCS), defined by spatial throughput (ST) and its standard deviation to characterize CSC in disasters. By revealing the impact of disasters on NCS, network resistance to disasters can be reflected. Specifically, a critical disaster intensity (CDI) is derived, below which the effect of disasters on NCS is negligible and networks are disaster-resistant. However, NCS rapidly deteriorates once CDI is exceeded, necessitating recovery strategies. In response, we design a CSC compensation strategy where uncrewed aerial vehicle access points (UAPs) are supplemented, and a critical UAP density maximizing NCS is provided. Notably, compared to ST, NCS can more promptly reflect CSC deterioration, enabling rapider strategy implementation. Moreover, we also demonstrate that fluctuating burst service superlinearly worsens NCS, revealing the network's poor tolerance to burst service. In this light, we propose a coverage adjustment strategy for terrestrial base stations and UAPs. Simulation results show that the negative effect of burst service can be significantly mitigated, which verifies the effectiveness of the proposed strategy.