The use of Digital Twins of the IoT infrastructure resilience

The development of the IoT has enabled increased automation of industrial processes and information systems, which has made systems more interconnected and more complex. However, this increased complexity has also brought new security and resilience challenges. Disturbances in the environment can lead to failure of the IoT system, and these disturbances can be caused by factors and can lead to malfunctions in the IoT system, degradation of its performance, or even complete failure. The principle of digital twin consists in creating a digital replica of a physical system or a process, which can be used to simulate and test operational scenarios. In the case of IoT systems, this means creating a digital replica of the entire system, including sensors, networks, servers and applications. By using digital twins, it is possible to simulate disturbances in the IoT system and test the system's ability to cope with them. Digital twins can also be used to monitor the system in real time, comparing data from the real system with data from the digital twin to detect anomalies. If an anomaly is detected, the digital twin can be used to simulate remediation scenarios and assess the effectiveness of different remediation strategies. The use of a digital twin can play a key role in enhancing the resilience of the IoT system by using simulation and modeling of the behavior of the IoT system under normal and disturbed conditions, taking into account the different sources of disturbance. The digital twin can thus be used to test different system configurations and to predict possible failures. The issue related to the use of a digital twin for an IoT system could therefore focus on how to use this technology to detect and prevent potential disturbances, how to realistically model the IoT system, how to integrate the digital twin in the daily management of the system, etc. In summary, the principle of digital twins can be used to strengthen the resilience of IoT systems by guaranteeing reliable and real-time interactions. Digital twins can simulate disturbances and test the system's ability to cope with them, as well as monitor the system in real time and detect anomalies. Objectives and perspectives of the work: IoT infrastructures are complex systems that must transmit, process, analyze, merge and synthesize data from multiple sensors to create situational awareness and understanding. However, these systems are subject to various disturbances that can alter the normal functioning of this infrastructure and make it vulnerable. In order to meet these challenges, the thesis proposes to use the concept of digital twins to improve the management, maintenance and performance of an IoT system, taking into account potential disturbances. Digital twins are virtual replicas of a physical system that allow real-time monitoring, control and simulation of IoT infrastructure behavior. The thesis aims to qualify the operation of the IoT infrastructure in an uncertain, evolving and non-cooperative environment, in degraded mode if necessary. It seeks to study the reconfiguration agility necessary for a restoration of the operational capacities of the infrastructure (resilience) and an increase in its robustness (anti-fragility), using digital twins as an infrastructure management and control tool.