Scaling out moving objects for geo-fence proximity determination
An event processing system for distributing geospatial computations with geo-fences is disclosed. The system can partition and distribute geo-fences efficiently for parallel computation, and can track a potentially unlimited number of moving objects. A computing device determines a minimum bounding rectangle (MBR) covering a geographic area indexed by a geo-grid, and partitions the MBR and a set of geo-fences to a plurality of processing nodes. The computing device receives input events of a stream comprising locations within the geographic area. The device distributes the events to partitions on processing nodes, based at least on the events' locations. The device can determine a partition identifier corresponding to a respective location in constant time. The computing device then combines partition results to obtain a single geospatial computation result.
1. A method, comprising:
determining, by a computing device, a geo grid minimum bounding rectangle (MBR) covering a geographic area indexed by a geo grid;
partitioning the geo grid minimum bounding rectangle to a respective processing node of a plurality of processing nodes;
replicating, by the computing device, a set of geo fences to the plurality of processing nodes, wherein the geo fences bound regions within the geographic area;
receiving input events of a stream, a respective input event comprising respective location information for a location within the geographic area;
distributing the respective input event to a respective partition on the respective processing node based at least in part on the respective location information;
receiving, from the respective processing node, an output result corresponding to a respective partition result;
combining, by the computing device, the respective partition result with a second partition result to obtain a single result corresponding to a geospatial computation; and
outputting the single result corresponding to the geospatial computation for a user.
2. The method of claim 1 , wherein distributing the respective input event to the respective partition further comprises determining a partition identifier corresponding to the respective location information in constant time.
3. The method of claim 2 , wherein determining the partition identifier corresponding to the respective location information in constant time is based at least in part on a grid cell partitioner partition representation.
4. The method of claim 2 , wherein determining the partition identifier corresponding to the respective location information in constant time is based at least in part on computing the partition identifier.
5. The method of claim 4 , wherein computing the partition identifier is based at least in part on a modulus operation and/or a matrix operation.
6. The method of claim 2 , further comprising mapping the partition identifier to the respective processing node based at least in part on a modulus operation, a division operation, and/or a custom function.
7. The method of claim 1 , further comprising adjusting a cell size of the geo grid.
8. The method of claim 1 , wherein, responsive to the respective location information of the respective input event matching a respective geo fence, the respective partition result is computed, by the respective processing node, as a spatial function applied to at least the respective location information.
9. A system, comprising:
a memory configured to store computer-executable instructions; and
a processor configured to access the memory and execute the computer-executable instructions to perform a method for distributing a geospatial computation, the method comprising:
determining a geo grid minimum bounding rectangle (MBR) covering a geographic area indexed by a geo grid;
partitioning the geo grid minimum bounding rectangle to a respective processing node of a plurality of processing nodes;
replicating a set of geo fences to the plurality of processing nodes, wherein the geo fences bound regions within the geographic area;
receiving input events of a stream, a respective input event comprising respective location information for a location within the geographic area;
distributing the respective input event to a respective partition on the respective processing node based at least in part on the respective location information;
receiving, from the respective processing node, an output result corresponding to a respective partition result;
combining the respective partition result with a second partition result to obtain a single result corresponding to a geospatial computation; and
outputting the single result corresponding to the geospatial computation for a user.
10. The system of claim 9 , wherein while distributing the respective input event to the respective partition, the processor further performs determining, from a grid cell partitioner partition representation, a partition identifier corresponding to the respective location information in constant time.
11. The system of claim 10 , wherein determining the partition identifier corresponding to the respective location information in constant time is based at least in part on computing the partition identifier.
12. The system of claim 11 , wherein computing the partition identifier is based at least in part on a modulus operation and/or a matrix operation.
13. The system of claim 10 , wherein the processor further performs mapping the partition identifier to the respective processing node based at least in part on a modulus operation, a division operation, and/or a custom function.
14. The system of claim 9 , wherein the processor further performs adjusting a cell size of the geo grid.
15. The system of claim 9 , wherein, responsive to the respective location information of the respective input event matching a respective geo fence, the respective partition result is computed, by the respective processing node, as a spatial function applied to at least the respective location information.
16. A non-transitory computer-readable medium storing computer-executable code that, when executed by a processor, cause the processor to perform a method for distributing a geospatial computation, the method comprising:
determining a geo grid minimum bounding rectangle (MBR) covering a geographic area indexed by a geo grid;
partitioning the geo grid minimum bounding rectangle to a respective processing node of a plurality of processing nodes;
replicating a set of geo fences to the plurality of processing nodes, wherein the geo fences bound regions within the geographic area;
receiving input events of a stream, a respective input event comprising respective location information for a location within the geographic area;
distributing the respective input event to a respective partition on the respective processing node based at least in part on the respective location information;
receiving, from the respective processing node, an output result corresponding to a respective partition result;
combining the respective partition result with a second partition result to obtain a single result corresponding to a geospatial computation; and
outputting the single result corresponding to the geospatial computation for a user.
17. The computer-readable medium of claim 16 , wherein while distributing the respective input event to the respective partition, the processor further performs determining a partition identifier corresponding to the respective location information based at least in part on computing the partition identifier.
18. The computer-readable medium of claim 17 , wherein computing the partition identifier is based at least in part on a modulus operation and/or a matrix operation.
19. The computer-readable medium of claim 17 , wherein the processor further performs mapping the partition identifier to the respective processing node based at least in part on a modulus operation, a division operation, and/or a custom function.
20. The computer-readable medium of claim 16 , wherein, responsive to the respective location information of the respective input event matching a respective geo fence, the respective partition result is computed, by the respective processing node, as a spatial function applied to at least the respective location information.