Surviving Disaster Events Via Dynamic in-Network Processing Assisted by Network Digital Twins

• Published in: 2023 International Conference on Information and Communication Technologies for Disaster Management (ICT-DM) pp. 1 - 6
• Main Authors: De Trizio, Federica, Sciddurlo, Giancarlo, Cianci, Ilaria, Striccoli, Domenico, Piro, Giuseppe, Boggia, Gennaro
• Format: Conference Proceeding
• Language: English
• Published: IEEE 13.09.2023
• Subjects: Autonomous aerial vehicles; Computer simulation; Decision making; Digital twins; Quality of service; Security; Task analysis
• ISSN: 2643-6868
• DOI: 10.1109/ICT-DM58371.2023.10286937

In post-disaster situations, search and rescue operations are actively supported by Unmanned Aerial Vehicles (UAVs) technology. As it is well known, UAV-based monitoring services require computational effort for in-network processing, which drones cannot easily support. This gives rise to the need to offload tasks to available and nearby network segments. However, the disruption caused by natural disasters often renders networks and services unavailable and unreachable. While some recent scientific contributions have already explored the idea of offloading computing tasks to support network elements in post-disaster situations, there is still a need to formulate a lightweight approach aimed at dynamically selecting suitable, trusted, and available network segments for in-network processing. In line with these premises, a novel service orchestration scheme for selecting the most suitable domain to offload processing tasks generated by drones is presented. Precisely, the conceived scheme addresses the domains affected by natural disasters and replaces them with available network segments for in-network processing, leveraging the Network Digital Twin representation. The adopted strategy employs a lightweight multi-criteria decision-making methodology to jointly consider the quality of service and trustworthiness parameters of the domains crossed by a drone. The effectiveness of the proposed solution is investigated through computer simulations. The obtained results demonstrate that the conceived approach offers up to 75.61% service availability for in-network processing compared to baseline techniques.