Toward Autonomous Trusted Networks-From Digital Twin Perspective

With the rapid growth of online applications and the increasing complexity of network architecture, network security faces significant challenges. As a promising approach, the network digital twin brings fresh opportunities for building more trustworthy networks. In this article, we analyze the challenges faced when designing cybersecurity twins and the methodology to address them. Then, the autonomous trusted network (ATN) framework is proposed, which leverages network digital twins, to enhance network security. The ATN framework incorporates flexible data aggregation, a security twin model, and an intelligent configuration model. It collects critical network data, trains security twin models using graph neural networks (GNNs), and generates intelligent configuration policies for proactive defense. By integrating network digital twins, advanced AI technologies, and programmable data planes, the ATN framework enables proactive threat detection, real-time response, and optimization, thereby enhancing the stability and reliability of networks in the face of evolving security challenges. The effectiveness of the ATN framework is demonstrated through a case study on Link-Flooding Attacks (LFA). The ATN framework effectively predicts and mitigates LFA by limiting attack traffic while ensuring the uninterrupted flow of legitimate traffic. The results show that the ATN framework outperforms traditional mitigation methods, such as traffic filtering and rerouting, in terms of promptness and efficiency.