基于语义认知网络的数字孪生网络架构研究

Research on Semantic-aware Networks for Digital Twin Network

数字孪生网络是一种新兴的技术架构,通过对物理网络中的实体进行数字化表示,有望实现物理网络与孪生网络的闭环系统。然而,在传统通信网络框架下仍然难以满足数字孪生网络高保真度、高时效性和高资源利用率等需求。语义认知网络作为一种智能的信息交互范式,侧重对数据含义的理解和处理,可显著提高传输效率和准确性,从而满足数字孪生网络高保真度和高时效性的需求。为此,提出了一种基于语义认知网络的数字孪生网络架构。该架构通过融合语义认知网络,可实现高保真度的孪生体模型构建、网络决策的智能优化以及高效的数据传输。在该架构的基础上,以基于语义认知网络的网络数据增强为案例,验证了语义认知网络可以在模型聚合过程中同时考虑样本数据量和分布相似性,训练得到具备生成高精度合成数据的孪生模型。

Digital twin network is an emerging technological architecture that aims to create a closed-loop system between physical and twin networks via digitally representing the entities in the physical network.However,it is still difficult for traditional communication architecture to meet the high-fidelity,real-time responsiveness,and efficient resource utilization requirements of digital twin networks.As an intelligent paradigm for information interaction,semantic-aware network focuses on understanding and processing the data meaning,significantly improving transmission efficiency and accuracy to meet the high-fidelity and high-timeliness requirements of digital twin networks.Therefore,a digital twin network architecture based on semantic-aware networks is proposed.By integrating semantic-aware networks,the proposed architecture can achieve high-fidelity twin model construction,intelligent optimization of network decisions,and efficient data transmission.Using semantic-aware networks in the context of network data augmentation as a case study,it is demonstrated that semantic-aware networks can simultaneously consider the data quantity and distribution similarity during the model aggregation process,resulting in a twin model capable of generating high-precision synthetic data.