Spatial Index Range Union Computation Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing relational database systems are inefficient when computing the union of bounding boxes filtered by entity criteria, as they often rely on a brute-force approach that processes all entities, wasting resources and causing latency, especially on devices with limited computing power or large databases.
Innovation Solution
A technique that creates a backchannel object to maintain a current range union result and the most recent bounding box tested, excluding targets whose bounding boxes are completely contained within the current result and only testing entities that meet the filtering criteria, thereby avoiding unnecessary comparisons.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a brute-force approach is used to compute the union of bounding boxes filtered by entity criteria, then all entities are processed to ensure completeness, but processing time and computational resources are wasted on entities that cannot contribute to the final result
Solution Approach 1:
The patent applies preliminary action by performing a bounding box containment test before evaluating entity criteria. The system checks whether a target's bounding box is completely contained within the current range union result, and if so, excludes it from further processing. This preliminary spatial filter prevents wasted computation on entities that cannot contribute to the final range union, while maintaining completeness by only excluding entities that are spatially redundant.
2Measurement precision
If entity filtering criteria are applied to all entities in the spatial index, then accurate filtered results are obtained, but unnecessary comparisons increase processing overhead
Solution Approach 1:
The patent segments the filtering process into two distinct stages: (1) a spatial pre-filtering stage that uses bounding box containment tests to identify potentially contributing entities, and (2) a criteria evaluation stage that applies entity filtering criteria only to those pre-selected entities. This segmentation dramatically reduces the number of criteria evaluations needed while ensuring accuracy, as the spatial filter eliminates only entities that are spatially redundant.
3Measurement precision
If the spatial index structure (R-tree) is fully traversed to ensure all entities are considered, then complete results are achieved, but the complexity and time consumption increase significantly
Solution Approach 1:
The patent integrates a preliminary bounding box containment check into the R-tree traversal process. As the query descends through the R-tree levels, each target node and entity is first tested against the current range union result using bounding box containment. Only targets that pass this preliminary spatial test proceed to criteria evaluation and potential inclusion in the result. This maintains complete entity coverage while simplifying the processing logic by early-excluding redundant targets.
Data Source
AI summary
In one embodiment, a union of the bounding boxes is computed of entities in a relational database indexed by a spatial index, filtered according to some criteria. For each of a set of targets to be considered, a bounding box of the target is tested against a current range union result. If the bounding box is completely contained in the current range union result, the target is excluded from further consideration. Otherwise, the target is processed. When the target is an entity, it is determined whether the entity meets the entity filtering criteria. If so a union is computed and the current range union result updated. When each of the targets to be considered in the spatial index has been subject to test or excluded, the current range union result is returned as a final range union result.