A bandwidth allocation method and device in a digital twin scene, equipment and medium

By using reinforcement learning to predict the distribution of users' access in a digital twin system and optimizing bandwidth allocation strategies, the problem of limited gateway bandwidth was solved, thereby improving the real-time performance and user experience of digital twin applications.

CN120499006BActive Publication Date: 2026-06-30SOUTH CHINA UNIV OF TECH
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
SOUTH CHINA UNIV OF TECH
Filing Date
2025-04-21
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

In a digital twin system, the limited bandwidth of the regional gateway cannot provide unlimited bandwidth for the sensors corresponding to each physical object, resulting in data transmission delays. This affects the real-time performance and data freshness of the virtual twin, and fails to effectively support the real-time performance and user experience of digital twin applications.

Method used

The system receives access data from the monitored area through a smart gateway, uses reinforcement learning to predict the distribution of visitors, and iterates and optimizes the bandwidth allocation strategy within a preset time to ensure that hotspot areas receive reasonable bandwidth and non-hotspot areas receive reasonable compensation bandwidth, thereby achieving timely data transmission and optimizing user experience.

Benefits of technology

It improves the real-time performance and user experience of digital twin applications under limited bandwidth conditions, reduces information age, ensures the real-time performance and reasonable allocation of data sources, and optimizes the user experience.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120499006B_ABST
    Figure CN120499006B_ABST
Patent Text Reader

Abstract

This invention discloses a bandwidth allocation method, apparatus, device, and medium in a digital twin scenario. The method includes: a smart gateway in the digital twin application system receiving access data from a monitored area and viewing the data information; storing the current access data of the monitored area in a cache and predicting the population access distribution within the monitored area in the next scheduling cycle; iteratively optimizing the predicted population access distribution using a reinforcement learning-based method within a preset time; selecting the optimal strategy during the iteration process as the bandwidth allocation strategy for the next scheduling cycle, and distributing the strategy to sensors and / or smart terminals in the digital twin application system. This invention predicts the population access distribution, enabling targeted identification of grids containing hot data and reasonable differentiated bandwidth allocation, thereby ensuring the real-time performance of the digital twin application. This invention can be widely applied in the field of communication technology.
Need to check novelty before this filing date? Find Prior Art