Site network state identification method, device, electronic equipment and system

By building a device network diagram and introducing an attention mechanism, the problem of low prediction accuracy and network status recognition accuracy in the prior art is solved, and higher prediction accuracy and recognition accuracy are achieved, which is suitable for existing mobile communication networks.

CN119676735APending Publication Date: 2025-03-21CHINA MOBILE GROUP DESIGN INST +1
0 Cites 0 Cited by

Patent Information

Application Number
CN202411792195.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-12-06
Publication Date
2025-03-21

AI Technical Summary

Technical Problem

The prior art considers factors in predicting mobile data traffic, which is difficult to conform to the actual distribution of existing mobile communication networks, resulting in low prediction accuracy and network state recognition accuracy.

Method used

By constructing a device network diagram that embeds isomorphic and heterogeneous information, combining multiple spatial correlation impact differences measurements, and introducing attention mechanisms in the time dimension and spatial dimension respectively, we capture the differences in the degree of influence of different access network devices in the time sequence and space in mobile data traffic prediction.

Benefits of technology

The prediction accuracy of mobile data traffic is improved, thereby improving the recognition accuracy of the target network status of the access network equipment, and more in line with the actual situation of the existing communication network.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention provides a site network state identification method, device, electronic equipment and system, and the method comprises the steps: enabling access network equipment data to be embedded into an access network equipment graph, and obtaining an equipment network graph; based on an aggregation space attention matrix and a time attention matrix corresponding to the equipment network diagram, aggregating initial adjacent time sequence feature sequences extracted from the equipment network diagram to obtain a first adjacent time sequence feature sequence corresponding to the equipment network diagram, and aggregating initial period time sequence feature sequences extracted from the equipment network diagram to obtain a second adjacent time sequence feature sequence corresponding to the equipment network diagram; obtaining a first period time sequence feature sequence corresponding to the equipment network diagram; and determining a target traffic prediction value corresponding to each access network device in the device network diagram based on the first adjacent time sequence feature sequence and the first period time sequence feature sequence. According to the invention, the prediction precision of the mobile data traffic can be improved, so that the recognition precision of the station network state is improved.
Need to check novelty before this filing date? Find Prior Art