一种基于启发式特征的语义知识库识别方法
By calculating the Euclidean distance between the feature vectors of the sender and receiver knowledge bases in a semantic communication system, and employing a transformer architecture and heuristic feature methods, the problem of semantic knowledge base mismatch is solved, achieving accurate recognition and matching, and improving communication performance and data transmission efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING INST OF TECH
- Filing Date
- 2025-07-07
- Publication Date
- 2026-07-17
AI Technical Summary
In existing semantic communication systems, the mismatch between the semantic knowledge bases of the sender and receiver leads to performance degradation, and there is a lack of effective identification and matching methods.
A semantic knowledge base recognition method based on heuristic features is adopted. The Euclidean distance between the feature vectors of the transceiver knowledge base is calculated by the base station. The feature vector with the smallest Euclidean distance is selected for matching. A semantic encoding and decoding model with a transformer architecture is used, and the model is optimized by nested or non-nested update methods.
It achieves accurate recognition and matching of semantic knowledge base, reduces the amount of data transmitted for recognition, and improves communication reliability and bandwidth utilization.
Smart Images
Figure CN120897229B_ABST