During the recent HUAWEI CONNECT 2025 event held in Shanghai on September 20, 2025, Huawei introduced its upgraded Xinghe Intelligent WAN Solution. This new offering is designed to enhance network monetization and improve industry intelligence through a robust IP network foundation that ensures a deterministic experience for users.
At the event’s roundtable discussion titled “AI-Powered WAN, Digital-Intelligent Future,” John Cai, President of the Router Domain within Huawei’s Data Communication Product Line, identified four significant challenges currently confronting Wide Area Networks (WANs). First, the global push for digitalization necessitates the removal of data silos to facilitate network convergence. Second, the emergence of advanced services requires higher standards for bandwidth and latency. Third, with the integration of AI, even a minimal packet loss of 0.1% can lead to a 50% decrease in computing efficiency on traditional networks, highlighting the need for greater transmission efficiency. Finally, advancements in quantum computing pose new risks to conventional network security.
To combat these issues, the Xinghe Intelligent WAN Solution offers four essential capabilities that aim to redefine network performance.
Key Features of the Xinghe Intelligent WAN Solution
The first feature is all-service convergence, achieved through technologies like slicing and SRv6, which ensure service isolation and reduce provisioning time to mere minutes. The second capability is a deterministic network; AI-powered application identification boasts over 95% accuracy in recognizing encrypted flows, enabling the automatic identification of video conference freezes and the acceleration of critical services.
Additionally, the solution includes a high-computing-efficiency WAN, where a flow awareness engine can identify 99% of large data flows, enhancing throughput by up to 90%. The Starnet lossless algorithm guarantees long-distance transmission with zero packet loss and keeps computing efficiency loss below 5%. Moreover, the Starnet vector engine is designed to protect sensitive data effectively.
The final capability focuses on intrinsic security, utilizing quantum-resistant encryption alongside Xsec to facilitate the secure and simplified deployment of encrypted networks for confidential data transmission.
At the roundtable, various customers and partners highlighted their successes in collaboration with Huawei. Mohammad Mustaghfirin, Chief Technology Officer of Surge Indonesia, shared that the company set a new benchmark for self-built broadband operations in transportation using IPv6 Enhanced networking, which has significantly improved connectivity for users. Kenneth Bagarukayo, Commissioner of Data Networks at Uganda’s Ministry of ICT & NG, noted a remarkable reduction in deployment time for each MDA from two hours to just one minute, thanks to SRv6. This innovation has enabled a versatile, multi-purpose network.
Furthermore, Zhang Xianguo, General Manager of Network R&D at DYXnet, revealed that their collaboration with Huawei resulted in the co-development of an AI-native ultra-connectivity architecture. This architecture allows for elastic scaling and RDMA-based long-distance file transfers that are more than twenty times faster than traditional TCP methods.
As Huawei looks to the future, the company remains committed to advancing AI-network integration. The aim is to propel various industries toward intelligent transformation with the Xinghe Intelligent WAN Solution. By establishing intelligent, efficient, and reliable deterministic network foundations, Huawei seeks to serve customers globally and meet the evolving demands of the digital landscape.
