HealthEdge: A Proactive DRL-Driven Service Reconfiguration Mechanism for Pre-Hospital Emergency Response in Edge Networks

Digital health initiatives are increasingly reliant on Internet of Things (IoT) applications to provide timely and efficient care, with pre-hospital emergency services being a critical component. These services are inherently delay-sensitive, necessitating robust resource management to ensure rapid response times and high-quality care. However, IoT devices in edge networks, such as emergency response units, are often resource-constrained, posing challenges in meeting the dynamic demands of emergency medical services. The imperative of pre-hospital emergency response demands an urgent reevaluation of current resource allocation strategies in edge networks. Traditional static dispatch protocols often fall short in dynamically optimizing ambulance deployment, leading to life-threatening delays and suboptimal patient outcomes. This paper introduces an innovative service reconfiguration mechanism that proactively addresses the pressing need for real-time, intelligent service allocation in emergency health services. This mechanism aims to re-schedule the configuration of resources provided by IoT devices through service migration technique. Extensive experiments are conducted upon datasets of pre-hospital emergency in Shenhzen, and experimental results show that our mechanism is more efficient than baseline techniques in satisfying delay constraints of IoT applications, decreasing energy consumption of IoT devices, and improving the service efficiency of pre-hospital emergency.