一种物联网设备数据的动态存储管理方法及系统
By using hierarchical balanced clustering and dynamic data allocation based on temperature thresholds, combined with a knapsack problem solver to optimize migration strategies, the problems of coarse data grouping and static migration strategies in IoT device data management are solved, achieving efficient utilization of storage resources and improved query performance.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SUZHOU GUANWEN STORAGE TECH CO LTD
- Filing Date
- 2025-10-24
- Publication Date
- 2026-07-17
AI Technical Summary
In existing IoT device data management, the data grouping granularity is coarse, which cannot effectively capture the data locality under complex query patterns, resulting in an increase in the number of cross-block queries and I/O overhead. Static migration strategies lead to wasted network bandwidth and unbalanced storage allocation, making it difficult to achieve a performance-cost balance under the limited resources of edge devices.
By configuring monitoring indicator metadata information, a structured metadata table and workload set are generated. A hierarchical balanced clustering algorithm is used to encode data tuples into feature vectors and group them into blocks of fixed capacity. The temperature values of the blocks are initialized using a hierarchical clustering tree and dynamically allocated to cloud, edge and cold storage layers. The migration strategy is optimized by a knapsack problem solver and combined with a device layer cache replacement strategy to achieve adaptive data management.
It enhances the locality of data within blocks, reduces the number of cross-layer queries, improves query response speed, reduces access latency, optimizes storage resource utilization, reduces network overhead, and improves cache hit rate.
Smart Images
Figure CN121434207B_ABST