A big data-based information system service method
By constructing a directed acyclic graph, masking physical arrival timestamps, calculating uncertainty indexes, and skipping database lock decision instructions, the system response performance avalanche problem caused by traditional discrete state machine logic in high-concurrency out-of-order data stream processing of big data information systems is solved, thereby improving the stability and accuracy of the system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN YUNHENG NETWORK TECHNOLOGY CO LTD
- Filing Date
- 2026-04-23
- Publication Date
- 2026-07-21
AI Technical Summary
When processing high-concurrency out-of-order data streams, existing big data information systems suffer from a significant drop in system response performance due to the large amount of invalid idle time in the underlying memory bandwidth and clock cycles caused by traditional discrete state machine logic and database transaction locking mechanisms.
By constructing a directed acyclic graph, masking the physical arrival timestamp, building an adjacency matrix in the memory buffer based on the business state preceding pointer, performing feature space projection operation, calculating the uncertainty index, compiling the difference increment matrix in the memory buffer ring, skipping the database lock determination instruction, directly performing matrix addition operation, and generating the system state matrix.
It reduces the probability of logical errors caused by network jitter, improves the speed of central processing unit correlation extraction, suppresses state machine logic drift and timing inversion, reduces system response latency and computing power idle queuing, and improves system stability and accuracy.
Smart Images

Figure CN122432218A_ABST