Geospatial Correlation Indexing with Spacetime Hashes
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing geospatial correlation techniques based on clustering algorithms and spatial data search algorithms are data-intensive and slow, becoming increasingly inefficient as the number of records grows to billions or trillions, making it difficult to automatically detect significant events and fuse data across sources based on spatial and temporal correlations.
Innovation Solution
The use of spacetime hashes to generate an index that encodes spatial-temporal data into a searchable format, allowing for the creation of a search data structure that groups subsets of spacetime hashes, enabling fast and scalable geospatial correlation by using either trie or bitmap data structures to improve search speed.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional clustering algorithms and spatial data search algorithms are used for geospatial correlation, then the system can process geospatial data, but the processing speed becomes slow and data-intensive when dealing with billions or trillions of records
Solution Approach 1:
The patent segments the geospatial data processing by dividing it into two distinct phases: an offline indexing phase that pre-processes and organizes data into spacetime hash structures, and an online query phase that performs rapid correlation searches. This segmentation allows the system to handle large volumes of data efficiently by moving the computationally intensive work to the offline phase, thereby improving online processing speed without sacrificing comprehensive data analysis
Solution Approach 2:
The patent applies preliminary action by pre-computing and storing spacetime hashes for all geospatial records during an offline indexing phase. This pre-processing creates a specialized data structure that enables extremely fast correlation queries online. The spacetime hash computation, which combines spatial coordinates and temporal information, is performed in advance so that during query operations, the system only needs to perform simple hash comparisons rather than complex spatial algorithms, dramatically improving processing speed for large datasets
2Quantity of substance
If the number of records increases to billions or trillions, then more geospatial information is available for analysis, but existing techniques become slower and more data-intensive
Solution Approach 1:
The patent replaces traditional mechanical spatial search algorithms with a hash-based computational approach. Instead of using complex spatial indexing structures like R-trees or grid-based methods that require iterative searching and comparison, the system computes spacetime hashes that encode both spatial and temporal information into fixed-size values. This substitution transforms the search problem from a mechanical, iterative process into a direct computational lookup, reducing time complexity from O(n) or O(log n) to O(1) for exact matches and enabling efficient handling of billions or trillions of records without increasing processing time
3Reliability
If traditional spatial data search algorithms are used, then geospatial correlations can be detected, but the methods are relatively slow and become increasingly inefficient at scale
Solution Approach 1:
The patent changes the fundamental parameters of the search problem by transforming continuous geospatial coordinates and temporal information into discrete spacetime hash values. This parameter transformation allows the system to maintain correlation detection accuracy by preserving the essential spatial-temporal relationships in the hash structure, while simultaneously dramatically improving processing efficiency. The spacetime hash function is designed to maintain locality and temporal proximity properties, ensuring that correlated events produce similar or matching hash values, thus preserving detection reliability while enabling scalable processing of large datasets
Data Source
AI summary
Disclosed are methods, systems and non-transitory computer readable memory for geospatial correlation. For instance, a method may include: obtaining a plurality of records, wherein each record has a key value that points to the record in a database, and each of the plurality of records includes at least data for an observation of an entity; and generating an index based on the plurality of records. The generating the index may include: determining a plurality of observations based on the plurality of records, an observation including a device identifier and time-and-geolocation data for a record of the plurality of records; encoding the plurality of observations into a plurality of spacetime hashes; and forming a search data structure of the index for a plurality of key values based on the plurality of spacetime hashes, wherein the search data structure groups subsets of the plurality of spacetime hashes.