Tree-Order Scan Spatial Query Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Evaluating spatial predicates for large sets of spatial objects in two or more dimensions is resource-intensive and slows down query responsiveness due to the need for numerous operations, many of which are unnecessary.
Innovation Solution
A tree-order scan technique is used to identify a minimal bounding rectangle for querying spatial data, employing a columnar data store with space-filling curves like Z-order and Hilbert curves for efficient indexing and dictionary compression, allowing for faster filtering and refinement of spatial queries.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional spatial predicate evaluation methods are used on large sets of spatial objects, then comprehensive spatial query coverage is achieved, but query runtime performance deteriorates and resource consumption increases
Solution Approach 1:
The patent segments spatial data into hierarchical levels using a space-subdivision tree structure (e.g., quadtree, octree). Spatial objects are organized into coarse-grained parent nodes and fine-grained child nodes, allowing queries to be processed at multiple resolution levels. This segmentation enables the system to quickly eliminate large portions of irrelevant data at higher levels before examining detailed objects at lower levels, dramatically reducing the number of spatial predicate operations required.
Solution Approach 2:
The patent performs preliminary spatial indexing and bounding volume hierarchy (BVH) construction before query execution. Spatial objects are pre-organized into tree structures with pre-computed bounding volumes at each node. During query processing, these pre-established structures enable rapid filtering of non-candidate objects through bounding volume intersection tests, avoiding the need to perform detailed spatial predicate evaluations on all objects in the dataset.
2Measurement precision
If detailed spatial predicate evaluation is performed on all spatial objects, then query accuracy is maintained, but memory consumption increases
Solution Approach 1:
The patent divides spatial data into hierarchical segments organized in a space-subdivision tree. Only the spatial bounds and aggregation statistics for each tree node are stored in memory, rather than complete geometric representations of all individual objects. This segmentation allows the system to maintain accurate spatial query capabilities while using memory proportional to the number of nodes in the hierarchical structure, which is significantly smaller than storing all detailed object data.
Solution Approach 2:
The patent introduces a hierarchical dimension to spatial data organization by constructing space-subdivision trees with multiple levels of abstraction. This additional hierarchical dimension allows the system to represent extensive spatial datasets compactly by storing aggregated bounding volume information at parent nodes, reducing the memory footprint while preserving the ability to perform accurate spatial queries through selective descent into child nodes.
3Reliability
If comprehensive spatial operations are performed to ensure accurate query results, then query reliability is improved, but query responsiveness deteriorates
Solution Approach 1:
The patent segments the query processing task into multiple hierarchical levels. At each level, the system performs simplified bounding volume intersection tests that are computationally inexpensive. Only when bounding volumes intersect does the system proceed to evaluate detailed spatial predicates on actual objects. This multi-level segmentation ensures query reliability by systematically checking all potential candidates while dramatically reducing the time spent on actual spatial predicate evaluations by eliminating non-candidates early in the process.
Solution Approach 2:
The patent performs preliminary filtering using bounding volume hierarchies before executing detailed spatial predicate evaluations. The BVH structure pre-organizes objects and their bounding volumes, allowing the system to quickly identify candidate objects that could potentially satisfy the query condition. This preliminary action maintains query reliability by ensuring all valid candidates are considered, while reducing query execution time by avoiding detailed evaluations on objects that cannot possibly satisfy the query.
Data Source
AI summary
A query of spatial data is received by a database comprising a columnar data store storing data in a column-oriented structure. Thereafter, a minimal bounding rectangle associated with the query is identified using a tree-order scanning technique. A spatial data set that corresponds to the received query is then mapped to the physical storage in the database using the identified minimal bounding rectangle. Next, the spatial data set is then retrieved. Related apparatus, systems, techniques and articles are also described.