Self-maintained Network Digital Twin for Human-centric Wireless Metaverse

Jiaxi Wang,Yixue Hao,Long Hu,Tong Zhang,Xiaoqiang Ma,Min Chen

DOI: https://doi.org/10.1109/mnet.2023.3337644

IF: 10.294

2024-01-01

IEEE Network

Abstract:Network digital twin is a key enabler for the human-centric wireless metaverse, requiring fine-grained replication, high-fidelity screen rendering, and the integration of emerging intelligent technologies. However, existing frameworks for the network digital twin overlook the integration of human-centric features within the metaverse and the importance of asynchronous data collection, both of which are indispensable for actualizing the metaverse and attaining self-maintenance capabilities. To address this issue, in this article, we propose a human-centric framework and a self-maintained mechanism for the network digital twin, utilizing continuous prediction and error tolerance to enhance the performance of DT decision-making. Specifically, by considering the interplay among different components, we present a human-centric framework for the network digital twin, comprising the device twin layer, network twin layer, artificial intelligence service layer, and user intent layer. In consideration of asynchronous information, we propose a self-maintenance mechanism facilitated by two key capabilities: error tolerance and continuous prediction. Furthermore, we conduct a resource allocation experiment to validate the efficiency of the proposed framework and methods. The results demonstrate that the framework reduce latency by employing request prediction and robust optimization.