Non-cooperative satellite network handover behavior inference method and device based on timing model

CN121887258BActive Publication Date: 2026-06-12NAT UNIV OF DEFENSE TECH

Patent Information

Authority / Receiving Office: CN · China
Patent Type: Patents(China)
Current Assignee / Owner: NAT UNIV OF DEFENSE TECH
Filing Date: 2025-08-19
Publication Date: 2026-06-12

AI Technical Summary

⚠Technical Problem

Existing technologies cannot effectively predict future connection paths and critical handover events in non-cooperative satellite networks, resulting in low prediction accuracy and an inability to anticipate handover timing or simulate dynamic handover rules.

⚗Method used

A time-series model-based approach is adopted to construct time-series samples by acquiring satellite orbit data, calculate basic physical characteristics, expected connection duration and handover cost, and use an LSTM network for model training to predict future satellite connection probabilities and identify handover behavior.

🎯Benefits of technology

It improves the accuracy of satellite connectivity prediction, enables early prediction of handover events, enhances situational awareness, provides judgments that are more in line with engineering practice, and makes prediction results more accurate and interpretable.

✦ Generated by Eureka AI based on patent content.

Smart Images

Figure CN121887258B_ABST

Patent Text Reader

Abstract

The application relates to a non-cooperative satellite network switching behavior inference method and device based on a time sequence model. The method comprises the following steps: generating a dynamic physical parameter sequence of all visible satellites within a time window for a specified ground observation point through public orbit data. Two key core features, namely an expected connection duration used to represent potential benefits of connection and a switching cost used to represent a cost of switching from one satellite to another, which can indirectly reflect the switching decision logic of the non-cooperative satellite network, are introduced and quantified. Then, the dynamic data containing the innovative features and the time sequence samples of the basic physical features are input into a time sequence model. The model can understand the trade-off between maximizing the expected connection duration of connection and minimizing the switching cost through learning a large amount of historical or simulated data, so that the connection satellite and the possible switching event at a future time step can be inferred with high precision and high reliability.

Need to check novelty before this filing date? Find Prior Art