Empowering disaster response: Advanced Network Slicing solutions for reliable Wi-Fi and 5G communications

Abstract

In the context of increasing global challenges, such as climate change, geopolitical unrest, and a range of natural disasters, the need for robust Public Protection and Disaster Relief (PPDR) management strategies has emerged to face phases such as risk mitigation, preparations, and recovery mechanisms. Within such a scenario, communication technologies are an essential part of emergency-response strategies. Both cellular and non-cellular networks, such as 5G and Wi-Fi, are crucial in supporting PPDR operations. 5G networks are essential for covering extensive areas, for instance, enabling video surveillance through drones, whereas Wi-Fi networks are better suited for localized applications, such as in temporary shelters or field hospitals. During environmental disasters, these networks often experience substantial traffic loads due to the high demand for diverse services, each with varying network requirements, such as enhanced Mobile Broadband (eMBB) and Ultra-Reliable Low-Latency Communications (URLLC). The increased network load poses a risk of impairing PPDR services, as the network may be unable to meet the stringent requirements necessary for first responders. Consequently, the implementation of Network Slicing techniques becomes critical to ensure flexibility, isolation, and dynamic prioritization of disaster management services, thereby guaranteeing the necessary network performance for effective emergency response. In this paper, we deploy real-life Network Slicing mechanism for Wi-Fi and 5G networks, to guarantee the network requirements for PPDR services. We evaluate the performance of Wi-Fi and 5G networks (throughput, latency, and packet loss), separately, for different scenarios. First, we introduce a dynamic Network Slicing mechanism for Wi-Fi networks. This mechanism, based on Software-Defined Networking (SDN) and In-band Network Telemetry (INT), incorporates an algorithm that mimics human-like reasoning to dynamically allocate network resources, such as airtime, using physical Wi-Fi equipment in real-world environments. Subsequently, we present a Network Slicing configuration for a real-world 5G network, deployed using actual hardware, utilizing a modular Open Radio Access Network (O-RAN) architecture.