A Massive Spatial Object Storage Method Fused with Encoding and Multi-Version Data

Abstract

The invention provides a mass space object storing method fusing codes and multi-version data. Multi-dimension space data are converted to one-dimensional space via low-precision grid coding and high-precision space object coding and then redundant data trimming involved in a query process is achieved via a strategy of lexicographical order location of data blocks and low-redundancy data query, so that the range query performance is effectively improved. When the coding precision is 8 Base-32 characters, the covering area of a grid unit is about 725m<2>; when a query radius is 1000m, query response time is not longer than 1s. The method can effectively meet data query requirements such as tour resource recommendation based on positions of tourists in smart tourism.

Description

technical field [0001] The invention relates to the technical field of distributed storage and access of spatial object data, in particular to a method for storing massive spatial objects that combines encoding and multi-version data. Background technique [0002] Location-based services need to frequently query massive spatial objects and obtain real-time responses, which requires efficient storage models and access optimization technology support. Using the existing relational model database to manage massive spatial objects, first of all, due to the need to adopt a fixed pattern to organize spatial objects, it cannot well meet the storage requirements of different attributes of different spatial objects, which will lead to a large amount of data redundancy; secondly, when dealing with such as range When querying, since spatial objects with close geographical locations do not need to meet the proximity of storage locations, there is no advantage in data I / O; at the same ti...

AI Technical Summary

Problems solved by technology

Using the existing relational model database to manage massive spatial objects, first of all, due to the need to adopt a fixed pattern to organize spatial objects, it cannot well meet the storage requirements of different attributes of different spatial objects, which will lead to a large amount of data redundancy; secondly, when dealing with such as range When querying, since spatial objects with close geographical locations do not need to meet the proximity of storage locations, there is no advantage in data I / O; at the same time, there is also a lack of scalability in the face of large-scale data

The existing distributed model can provide the scalability of massive spatial object management, but lacks effective access optimization technology for spatial object data; at the same time, there is no effective theoretical relationship between storage accuracy and access performance

Images