Enhancing flood resilience in critical communication infrastructures: empirical analysis and software-defined architecture design

Mobile communication systems serve as a lifeline for disaster response, yet catastrophic floods—exemplified by the Beijing 23·7 and Zhengzhou 7·20 events—have exposed profound vulnerabilities that trigger cascading failures in emergency command and rescue operations. While existing interventions predominantly focus on temporary hardware deployments (e.g., UAVs), the potential of software-defined optimization within existing 5G terrestrial communication infrastructure remains critically underexplored. This study addresses this gap through a dual-layered research design that bridges technology and disaster management practice. First, we establish an empirical foundation by applying an LLM-assisted mixed-methods analysis to a comprehensive multi-source corpus. This corpus integrates academic literature, official disaster reports, and original interviews with personnel from emergency management departments and technical experts from telecommunications operators. This analysis identifies and characterizes five core communication failure modes in real-world flood scenarios, clarifying the critical infrastructure vulnerabilities that precipitate decision-making paralysis and systemic governance failures. Second, grounded in these empirical findings, we develop an integrated theoretical framework and a dynamic operational architecture for 5G base station (gNB) resilience. This architecture systematically maps specific 5G software-defined capabilities to the identified disaster challenges. The results provide a cost-effective, scalable strategy for enhancing critical infrastructure resilience without requiring extensive hardware modifications. This research offers actionable guidelines for policy makers and emergency planners to maintain information pathways during extreme climate-related risks, directly contributing to the modernization of urban safety governance.