ZTE places third in the ITU Graph Neural Networking Challenge

PRESS RELEASE: ZTE Corporation announced that it was ranked third in the Graph Neural Networking Challenge 2021 for the data model it proposed in the ITU AI / ML (Artificial Intelligence / Machine Learning) in the 5G Challenge.

With the data model, ZTE obtained the MAPE (Mean Absolute Percentage Error) of 1.85 in a real large-scale network, which is significantly better than the test result (MAPE> 300) without the optimization of the ‘algorithm. The Graph Neural Networking Challenge is part of the AI ​​/ ML in 5G Challenge series and is organized by the Barcelona Neural Networking Center and the Universitat Politecnica de Catalunya BarcelonaTech (BNN-UPC).

Graph Neural Networking Challenge aims to build a large-scale digital twin with Graph Neural Networks (GNN) and simulate network performance using the GNN algorithm. Traditional networked digital twin technology cannot achieve high simulation precision with low computation cost. Whereas GNN is the only machine learning technology to maintain a balance between them. In addition, GNN can provide higher simulation accuracy with low computational resource requirements and facilitate the deployment of lab network training results into real networks.

The difficulty of applying GNN in a digital twin network is that the performance indicators of a small-scale network used for machine learning training are vastly different from those of a true large-scale network. Therefore, the learning model obtained in the small network cannot be applied in real networks.

To address this challenge, ZTE changed the predictor variable in the port usage latency model, assuming that the port usage ranges at different network scales are close to each other. Next, ZTE calculated the latency based on the relationship between port usage and latency, to solve the problem of different ranges and greatly improve the prediction accuracy of small-scale network formation models.

To date, ZTE has integrated this technology into research and planning of network self-repair and provision of rapid services to improve their reliability. Based on the principle of this algorithm, ZTE has developed an automatic network optimization algorithm, so that the intelligent management and control system can recommend the most desirable network configuration solution according to the current state. network and user needs. Going forward, ZTE will continue to use its AI / ML technologies to deliver smarter, more efficient and more convenient communication networks to all customers.

Subscribe to our daily newsletter