混合编码数据流解码方法及相关装置
By introducing a large language model and an intelligent agent process execution engine, the structure of the hybrid encoded data stream is automatically parsed and decoding logic rules are generated, solving the problems of rigidity and high maintenance costs in traditional methods, and realizing adaptive and efficient decoding of hybrid encoded data streams.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- HEFEI IFLY DIGITAL TECH CO LTD
- Filing Date
- 2026-02-13
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional hybrid encoded data stream decoding methods cannot adaptively and intelligently parse and process unknown or dynamically changing hybrid encoded data, leading to communication link interruptions and high maintenance costs.
A large language model is introduced as the decoding logic rule analysis model. Through pattern recognition and semantic understanding capabilities, the structure of the hybrid encoded data stream is automatically parsed to generate decoding logic rules. The intelligent agent process execution engine then performs data packet segmentation, grouping, parallel decoding, and reassembly.
It enables adaptive parsing and processing of unknown or dynamically changing mixed encoded data at the receiving end, reducing maintenance costs and improving system flexibility and decoding efficiency.
Smart Images
Figure CN121708946B_ABST