Management Apparatus Device Assignment Similarity Clustering

The patent merges devices with similar attribute information into the same subtree of the management tree. By calculating similarity between device attributes and organizing devices into clustered groups within the tree structure, the system reduces the number of distinct leaf nodes that need to be assigned. This clustering approach allows multiple devices to share common node keys along their paths to the root, thereby reducing overall data transmission volume while maintaining efficient assignment processes.

The patent performs preliminary organization of devices into similarity-based clusters before actual key distribution occurs. By pre-calculating attribute similarities and establishing the optimized tree structure in advance, the system prepares assignment paths that minimize future key distribution overhead. When new devices join, they can be quickly integrated into existing clusters rather than requiring complete reassignment, thereby reducing the time required for device onboarding while maintaining an operationally simple key management process.

The patent applies local quality by organizing devices into distinct subtrees based on their attribute characteristics. Each subtree represents a local group with homogeneous properties, allowing the system to apply optimized key distribution strategies specific to each cluster while maintaining the overall security structure of the management tree. This localized organization enables efficient grouping for transmission without compromising the security guarantees provided by the hierarchical key structure, as each subtree still maintains proper encryption relationships with its ancestors.