Demo: Immersive Digital Twin Networks for Industrial Internet of Things

Le résumé fourni par la source

With the rapid advancement of the Industrial Internet of Things (IIoT), Digital Twin Networks (DTNs) have emerged as critical enablers for industrial network optimization. However, current full-space industrial DTNs face persistent challenges: neglect of propagation-impacting industrial factors, computational inefficiency, and poor generalization in Artificial Intelligence (AI)-based prediction method. This demo establishes an immersive IIoT-DTNs framework for real logistics scenarios, where cargo stacking rates are considered as the primary industrial dynamic factor. Hence, the 3D scenes reflecting varying stacking rates are generated for Ray Tracing (RT)-based channel modeling. The proposed Adaptive Sampling-based Sparse RT Calculation (AdaSpRT) method strategically selects critical propagation areas for RT calculation, replacing random sampling. Then, AI-based channel prediction can be trained and fine-tuned with the RT data for remaining sample prediction. Evaluations across two distinct environments, high-density shelving vs. open layout, demonstrate 22.0% and 12.4% efficiency improvements over conventional RT while maintaining Mean Absolute Errors (MAE) of 3.44 dB and 1.56 dB respectively.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.