Subject Training

Context: Digital Twins (DTs) rely on continuous, bidirectional communication between a physical system and its virtual counterpart. This communication often depends on wireless networks (Wi-Fi, 5G, LoRaWAN, etc.), which must handle heterogeneous traffic, variable channel conditions, and resource constraints [2, 3]. Ensuring real-time synchronization between the physical system and the DT is crucial for accuracy, responsiveness, and reliability, particularly in industrial and IoT contexts where latency, jitter, packet loss, and bandwidth limitations can degrade performance. Traditional static resource allocation methods in wireless networks cannot guarantee optimal performance under dynamic and unpredictable conditions [3, 1]. Adaptive algorithms, leveraging concepts from AI, reinforcement learning, or heuristic optimization, offer promising approaches to dynamically manage network resources such as bandwidth, transmission power, and scheduling policies. By optimizing how these resources are allocated, it is possible to improve both network performance and the fidelity of the DT, while also reducing energy consumption and unnecessary overhead.

Problem Statement: Ensuring reliable synchronization between physical systems and digital twins over wireless networks raises several unresolved challenges : 1. Which network parameters (latency, jitter, packet loss, throughput, energy consumption) most critically affect the fidelity and responsiveness of digital twins? 2. How to design adaptive algorithms that can dynamically reallocate wireless resources in real time under varying and unpredictable network conditions? 3. How to balance conflicting objectives, such as maximizing DT accuracy and responsiveness while minimizing energy consumption and avoiding network congestion? 4. How to validate and compare the effectiveness of adaptive allocation strategies across different wireless technologies (e.g., Wi-Fi, 5G, LoRaWAN) and digital twin use cases? The work aims to address these questions by proposing and evaluating adaptive resource allocation methods capable of optimizing wireless network behavior to sustain accurate and efficient digital twin synchronization.

Références [1] Nikolaos Apostolakis et al. « Digital twins for next-generation mobile networks : Applications and solutions ». In : IEEE Communications Magazine 61.11 (2023), p. 80-86. [2] Latif U Khan et al. « Digital twin of wireless systems : Overview, taxonomy, challenges, and opportunities ». In : IEEE Communications Surveys & Tutorials 24.4 (2022), p. 2230-2254. [3] Shah Zeb et al. « Industrial digital twins at the nexus of NextG wireless networks and computational intelligence : A survey ». In : Journal of Network and Computer Applications 200 (2022), p. 103309.