NL2036715B1 - Adaptive cube indexing method and system - Google Patents

Abstract

An adaptive cube indexing method and system is provided. It includes: for a topological relationship between adaptive cube tiles and geometries, setting topology type encoding rules to segment and calibrate a geometry index generated herein; segmenting geometry identif1ers of different geometric types of multi-point, line, surface, body, or grid geometries expressing earth phenomena based on a multi-level tree structure of a tiled adaptive cube and the topology type encoding rules, to generate and store multi-level topology indexes, defining a request tile region based on an access request to obtain access topology type encoding rules, obtaining geometry indexes within the tile region based on the topology type encoding rules, filtering away redundant geometry indexes within the tile region, accessing corresponding target geometry data based on the filtered geometry indexes. Efficiency of accessing a target geometry of the adaptive cube and performance of an adaptive cube-based earth phenomenon analysis-ready application is improved. 1/1 Analysis-ready application client Access request Request agent module R t . . . gefmugtnry Geometry Adaptlvecube application data index servrce platform Geomet ry access module Indexes Indexes Indexes of Index of level 1 of level 2 level n databa _ &“sz Multi-leveltopology indexes se ii aml>:is- Indexing module ‘ application J handle Database cluster Adaptive cube storage system FIG. l

Description

ADAPTIVE CUBE INDEXING METHOD AND SYSTEM

FIELD OF TECHNOLOGY

[0001] The present invention belongs to the technical field of big data, and particularly relates to an adaptive cube indexing method and system.

BACKGROUND

[0002] Now we have entered the era of big data, especially with the continuous development of geographic information applications, corresponding earth online analysis- ready applications put forward new requirements for indexing and access of multi-dimensional geometric cubes, and the data volume of the multi-dimensional geometric cubes is geometrically growing. A data source of a single geometric type is gradually unable to meet the increasing real-time or quasi-real-time demand of online real-time analysis of the geographic information applications. Data of different geometric types is mostly distributed in databases of well-known data collection and management institutions at home and abroad, each data center has a set of relatively independent data indexing and access methods, and data provision methods are not the same. Standardized GIS cloud computing technology and

WebGIS data services may connect multiple distributed resources through standard protocols to support collaborative operations, for example, ISO TC211, the open geospatial consortium (OGC) and other mainstream standardization organizations provide a series of data directories or data interface service specifications for collaborative discovery of geographic information data through standardization activities. For example, the OGC CSW specification is a Web- based directory service specification that lacks geometry data indexing details; although OGC

WCS and WFS may provide geometry details, they lack an efficient tiled indexing mechanism.

Itis difficult for standardized Web protocols to directly penetrate into the internal data index level. In particular, with regard to tiled indexes for big data, geometry data indexing and access methods for very large earth scenario data need to be further explored to support real-time analysis and computation among database clusters.

[0003] In addition, in the scenario that the analysis-ready application needs only part of the data, a traditional full data access method wastes bandwidth and may not satisfy the i performance requirements of the analysis-ready application. With the continuous development and maturity of NoSQL databases, the application requirements for mass storage and real-time analysis may be well satisfied. The NoSQL databases may support storage of data of different geometry types. For geometry data, there may be various indexing methods to improve the speed and efficiency of data query, for example, on the basis of various properties of geometries, such as shape, size, and dimension. However, when earth scenario data is very large, it is also necessary to segment indexes effectively.

[0004] BeiDou grid code is a discrete, multi-scale regional location identification system developed on the basis of the geospatial subdivision theory of geographical coordinates subdividing grid with one dimension integral coding on 2n-Tree (GeoSOT), and may easily establish an intrinsic interconnection with any entity object and various different data in the same region range, which may satisty different levels of data index management demands from macro to fine, so the code may be considered to be used for indexing of geometry identifiers of different geometry types in the earth phenomenon scenarios.

[0005] Since the earth phenomena tend to have a certain continuity, geometries may often cross subdivided regions, and the index redundancy often differs under different topology conditions, so repeated access to geometry data may be avoided by filtering away redundant indexes. Compared to a single index, an index organization of a multi-level tree structure may have a plurality of indexing solutions, and large tile retrieval regions tend to have higher redundant indexes. By selecting a suitable index hierarchy solution, redundant indexes may be minimized, and on this basis, related geometries of desired cube tiles are accessed, which may reduce the consumption of the bandwidth and the index retrieval hash rate in the real-time analysis process, and may significantly improve the performance for large-scale geometry data online analysis applications.

SUMMARY

[0006] In order to solve the above problems, the present invention provides an adaptive cube indexing method and system. By means of the method, the efficiency of accessing a target geometry of an adaptive cube is improved, and the performance of an adaptive cube-based earth phenomenon analysis-ready application is enhanced. 2

[0007] In order to achieve the above purpose, the technical solutions of the present invention are as follows. An adaptive cube indexing method includes: for a topological relationship between adaptive cube tiles and geometries, setting topology type encoding rules to segment and calibrate a geometry index generated herein; segmenting geometry identifiers of different geometric types of multi-point, line, surface, body, or grid geometries expressing earth phenomena on the basis of a multi-level tree structure of a tiled adaptive cube and the topology type encoding rules, to generate multi-level topology indexes and store same; and defining a request tile region on the basis of an access request to obtain access topology type encoding rules, obtaining geometry indexes in a tile region range on the basis of the topology type encoding rules, filtering away redundant geometry indexes in the tile region range, and accessing corresponding target geometry data on the basis of the filtered geometry indexes, so as to improve access performance of the adaptive cube.

[0008] In an embodiment of the present invention, the topology type encoding rules are: each topology type code consists of a 3-digit binary number, and each digit represents whether an intersection between a geometry and an interior, boundary and exterior of an adaptive cube tile is null, and is O when the intersection is null and is 1 when the intersection is not null; the topology type codes are used as basic units t to form a topology type code set, denoted as {t;}; in a case where the topology type code set is used for subdivision, geometries in earth phenomenon subject data are subdivided according to the topology type code set, and geometries that satisfy any {t;} element are indexed by corresponding tiles to calibrate geometries indexes generated according to rules herein; and in a case where the topology type code set is used for access, geometry indexes that satisfy any {t;} element in retrieved data are selected.

[0009] In an embodiment of the present invention, the multi-level tree structure of the tiled adaptive cube is that an adaptive cube including a spatial dimension X, a spatial dimension ¥, and an adaptive dimension } is tiled and divided into tiles hierarchically, each parent node tile being a union set of all child node tiles, each tile being an adaptive cube tile, each tile including a plurality of data pieces, and each data piece including one type of geometries; a root node tile is R and has # child node tiles, respectively being Ti, Ta, ..., Ts, and each child 3 node tile including own child node tiles, denoted as parent node tile = U (child node tile), where U denotes a union set; a hierarchical tree structure is constituted, the hierarchical tree structure is represented recursively, and the next hierarchy of T; (0<i<#u+1) has m child nodes, respectively being Ta, Ta, ..., Tm; a tile is made to be T;, including & data pieces, respectively being Py, Pa, ..., Pr; and each data piece includes a collection of geometry sets of different geometric types, respectively being Gi, Go, ..., Gx, and geometries in each geometry set have the same geometric type, denoted as {g1}, {g2}, … {ge}, where i, j, k, m, and # are all positive integers.

[0010] In an embodiment of the present invention, the adaptive cube including the spatial dimension X, the spatial dimension Y, and the adaptive dimension }”is a cube in which variable information of data, including elevation, time and variables, is arranged in order in the adaptive dimension.

[0011] In an embodiment of the present invention, the geometry identifiers of different geometric types of the multi-point, line, surface, body, or grid geometries expressing the earth phenomena are segmented on the basis of the multi-level tree structure of the tiled adaptive cube and the given topology type encoding rules, and topology type codes satisfied are calibrated in the geometric indexes, to generate earth phenomenon subject data indexes of the multi-level tree structure.

[0012] In an embodiment of the present invention, the access request includes the access topology type encoding rules given according to a request range given by a coordinate system spatially compatible with the adaptive cube; the request range is subdivided according to the multi-level tree structure of the tiled adaptive cube, a tile hierarchy corresponding to a smallest tile region covering the request range is selected, and the corresponding tile region is the defined request tile region.

[0013] In an embodiment of the present invention, the geometry indexes in the defined tile region range are obtained on the basis of the topology type encoding rules in the access request, and according to a rule that geometry index of tile region = U (geometry indexes of tiles within the tile region), where U denotes a union set, the redundant geometry indexes in the tile region range are filtered away, and the corresponding target geometry data is accessed on the basis of the filtered geometry indexes. 4

[0014] The present invention further provides an adaptive cube indexing system. The system includes: an indexing module, mounted on an adaptive cube storage system, and configured to generate multi-level topology indexes according to a multi-level tree structure of a given tiled adaptive cube; a request agent module, mounted on an adaptive cube application service platform, and configured to access the multi-level topology indexes in an index database, and define a request tile region according to a cube request range in an analysis-ready application of a client; and a geometry access module, mounted on an online adaptive cube application service platform, and configured to obtain geometry indexes within the tile region according to an index range, obtain corresponding geometries from the adaptive cube storage system, and return the geometries to the client, so as to complete the access.

[0015] In an embodiment of the present invention, the steps of the method as described above are performed.

[0016] Compared to the prior art, the present invention has the following beneficial effects: the present invention is an earth phenomenon-oriented adaptive cube indexing and access technique, which provides an efficient solution for geometry data indexing and access of a NoSQL database. Storage and indexing strategies may be flexibly adjusted according to the actual demand, the transmission of redundant geometry data and the consumption of a fusion hash rate of a corresponding tile region are reduced, and the performance of an earth phenomenon-oriented geographic information analysis-ready application is improved.

BRIEF DESCRIPTION OF THE DRAWINGS

[0017] FIG. 1 is a schematic structural diagram of modules according to the present invention.

DESCRIPTION OF THE EMBODIMENTS

[0018] The technical solutions of the present invention are specifically described below in conjunction with the accompanying drawings.

[0019] As shown in FIG. 1, an embodiment of the present invention provides an earth phenomenon-oriented adaptive cube indexing method. The method mainly includes the following processing steps: for a topological relationship between adaptive cube tiles and 5 geometries, set topology type encoding rules to segment and calibrate a geometry index generated herein; segment geometry identifiers of different geometric types of multi-point, line, surface, body, or grid geometries expressing earth phenomena on the basis of a multi-level tree structure of a tiled adaptive cube and the topology type encoding rules, to generate multi-level topology indexes and store same; and define a request tile region on the basis of an access request to obtain access topology type encoding rules, obtain geometry indexes in a tile region range on the basis of the topology type encoding rules, filter away redundant geometry indexes in the tile region range, and access corresponding target geometry data on the basis of the filtered geometry indexes, so as to improve the access efficiency of the adaptive cube.

[0020] This embodiment also gives examples of the topology type encoding rules, such as {001, 010, 011, 110, 111}, which represent a geometry being inside an adaptive cube tile, a geometry being at the boundary of an adaptive cube tile, a geometry being inside and at the boundary of an adaptive cube tile, a geometry being at the boundary of and outside an adaptive cube tile, and a geometry being inside, at the boundary of and outside an adaptive cube tile, respectively.

[0021] This embodiment also gives an example of performing segmentation on the basis of the multi-level tree structure of the tiled adaptive cube and the topology type encoding rules to generate the multi-level topology indexes. A multi-level tree model of the tiled adaptive cube is built in the spatial plane according to the "BeiDou Grid Location Code" (GB/T 39409- 2020) and in the variable axes according to different themes (e.g., time, elevation, and time- elevation combinations), different geometric types of geometries such as multi-point, line, surface, body or grid geometries of earth phenomena are stored with MongoDB, and geometry identifiers thereof are segmented according to the constructed multi-level tree structure of the tiled adaptive cube and the set topology type encoding rules, so that the multi-level topology indexes are generated and stored in a MySQL index database.

[0022] This embodiment also gives an example of an earth phenomenon-oriented adaptive cube request. The request includes a request range (a lower limit of longitude, an upper limit of longitude; a lower limit of dimension, an upper limit of dimension; a lower limit of time, and an upper limit of time) in a space-time coordinate system of WGS84, as well as access topology type encoding rules {001, 010, 011, 110, 111}. 6

[0023] A request agent module mounted on an adaptive cube application service platform receives a cube request initiated by a client, reads a corresponding scenario index, calculates, according to the multi-level tree structure of the tiled adaptive cube, to determine a tile hierarchy and a tile region range corresponding to the smallest tile region covering the request range, obtains a geometry index of the defined tile region range on the basis of the topology type encoding rules in the cube request, then filters away indexes of redundant geometry identifiers in the tile region range through MySQL DISTINCT according to a rule of geometry index of tile region = U (geometry indexes of tiles in the tile region), and transmits filtered geometry indexes and an analysis-ready application handle to a geometry access module of an adaptive cube storage system.

[0024] The geometry access module of the adaptive cube storage system reads corresponding target geometry data from MongoDB according to the received geometry indexes, and transmits the data to the analysis-ready application client according to the analysis-ready application handle.

[0025] An embodiment of the present invention further provides an earth phenomenon- oriented adaptive cube indexing system. The system includes: an indexing module, mounted on a MongoDB-based adaptive cube storage system, and configured to generate multi-level topology indexes according to a multi-level tree structure of a given tiled adaptive cube; a request agent module, mounted on an adaptive cube application service platform, and configured to access the multi-level topology indexes in an index database, and define a request tile region according to a cube request range in an analysis-ready application of a client; and a geometry access module, mounted on an online adaptive cube application service platform, and configured to obtain geometry indexes within the tile region according to an index range, obtain corresponding geometries from the MongoDB-based adaptive cube storage system, and return the geometries to the client, so as to complete the access.

[0026] In summary, the earth phenomenon-oriented adaptive cube indexing method and system according to the present invention segment the earth phenomenon-oriented geographic information geometry indexes through the multi-level tree structure of the tiled adaptive cube according to the topology type encoding rules and store same, determine the suitable index hierarchy and the smallest tile region covering the request range, filter away the redundant geometry indexes in the tile region range, and on this basis, access related geometries of desired cube tiles, which may reduce the consumption of the bandwidth and the index retrieval hash rate in the real-time analysis process, and may significantly improve the performance for large- scale geometry data online analysis applications.

[0027] The above are preferred embodiments of the present invention, and all changes made in accordance with the technical solutions of the present invention, insofar as the resulting function does not exceed the scope of the technical solutions of the present invention, fall within the scope of protection of the present invention. 8

Claims (9)

CONCLUSIONS 1. Adaptive cube indexing method, characterized in that, it includes: - in view of a topological relationship between an adaptive cube tile and the geometry, coding rules for the topology type are set to segment and calibrate a geometry index generated by a sample file; - the geometry of various geometric types representing multi-point, line, surface, volume or grid geometry expressing earth phenomena is identified and segmented according to the coding rules for multi-level tree structure and topology types based on tile-based adaptive cubes, and multi-level topology indexes are generated and stored; - based on an access request, a request tile area is defined, encoding rules for the access topology type are obtained, the geometry index of the tile area range is obtained based on the encoding rules for the topology type, the redundant geometry index of the tile area is filtered, and the corresponding target geometry data is accessed based on the filtered geometry index, so as to improve the performance of adaptive cube access. 2. The adaptive cube indexing method according to claim 1, characterized in that: the topology type coding rules are as follows: Each topology type code consists of 3 binary digits, each digit indicating whether the intersection of the geometry and the interior of the adaptive cube tile, the boundary and the exterior is empty, empty O and non-empty 1; the topology type code set is composed with the topology type code as the basic unit t, which is represented as {ti}; when the topology type code set is set for partitioning, the geometry in the earth phenomenon subject data is partitioned according to the topology type code set, and the geometry satisfying a {ti} element is indexed by the corresponding tile, which is used to calibrate the geometry index generated according to the rules of this file; when setting the topology type coding set for access, it is used to select the geometry index satisfying a {ti} element in the retrieved data. 9 3. An adaptive cube indexing method according to claim 1, characterized in that: the multi-level tree-like structure of the tiled adaptive cube is as follows: the tiling is an adaptive cube consisting of spatial dimension X, spatial dimension Y, adaptive dimension V, and divides the tiles hierarchically, each parent node tile 1s the union of all the child node tiles, each tile is an adaptive cube tile, each tile contains a plurality of data segments, and each data tile contains a class of geometric bodies; the node of the parent tile is R and has n child node tiles, which are Ti, Tz, … , Ty, each child tile contains its own child tile, which is represented as the parent tile = U (child tile), where U stands for union, form a hierarchical tree structure, which is represented recursively, and the next level of Tioci<n+n has m child nodes, namely Ti, Ti, … Tin; Let the tile T; are composed of k data slices, denoted Pi, Pa, …, Py; each data slice contains a collection of geometry sets of different geometry types, Gl, G2, … , Gk, the geometry in each geometric set has the same geometric type, denoted {gi}, {2}, … » tg}; ij, k m, n are all positive integers. 4. The adaptive cube indexing method according to claim 3, characterized in that: the adaptive cube consisting of spatial dimension X, spatial dimension Y and adaptive dimension V is a cube that arranges the variable information of the data in an orderly manner on the adaptive dimension, including height, time and variables. 5. Adaptive cube indexing method according to any one of claims 1 to 4, characterized in that: the geometry identifiers of different geometric types of multi-point, line, surface, volume or grid geometry expressing terrestrial phenomena are segmented according to the multi-level tree structure based on tile-based adaptive cubes and the given topology type coding rules, and the topology type codes satisfying the requirements are calibrated into the geometry index and the multi-level tree structure of the terrestrial phenomenon index is generated. 6. An adaptive cube indexing method according to any one of claims 1 to 4, characterized in that: the access request contains the request range specified according to the coordinate system compatible with the adaptive cube space, and the coding rules 10 for the access topology type are given; the request range is divided based on the multi-level tree structure of the tile-based adaptive cube, and the tile level corresponding to the smallest tile area covering the request range is selected and the corresponding tile area is the defined request tile area. 7. Adaptive cube indexing method according to any one of claims 1 to 4, characterized in that: according to the rule of tile area geometry index = U (tile area geometry index) in the access request, the redundant tile area geometry index is filtered based on the union of the U representative set and the corresponding target geometry data is accessed based on the filtered geometry index. 8. An adaptive cube indexing system, characterized in that it comprises: an index module mounted on an adaptive cube storage system, for generating a multi-level topological index according to a multi-level tree structure of a given tiled adaptive cube; a request proxy module on an adaptive cube application service platform accessing a multi-level topology index in an index database for defining a request tile region based on the client's immediate analysis of a request range of the cube in a directly analyzable application of a client; and a geometry access module equipped on an online adaptive cube application service platform, for obtaining geometry indices within the tile region obtained according to an index range, for obtaining corresponding geometries from the adaptive cube storage system and returning them to the client, thus completing access. 9. An adaptive cube indexing system according to claim 8, for performing the method steps described in any of claims 1 to 7. tl