Browse Prior Art Database

An apparatus and approach to archive massive historical spatiotemporal data in IoT environment

IP.com Disclosure Number: IPCOM000237719D
Publication Date: 2014-Jul-07
Document File: 5 page(s) / 297K

Publishing Venue

The IP.com Prior Art Database

Abstract

The disclosure discloses an apparatus and approach to archive massive historical spatiotemporal data in real time with efficient storage and also convenient OLTP-style spatial temporal query support. The proposed apparatus can support real time spatiotemporal historian via combining timeseries storage and trajectory summary storage based on road network, and achieve high insert throughput with low spatial storage and index cost. The proposed data achieve process can archive spatiotemporal data and trajectory summary data with high insert throughput by map-matching and trajectory cleaning. The proposed spatiotemporal query process based on two-layer spatiotemporal index and VTI framework can support spatiotemporal native SQL query effectively.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 50% of the total text.

Page 01 of 5

An apparatus and approach to archive massive historical spatiotemporal data in IoT environment

Requirements of IOT data management in Connected Vehicles/devices, Telecom M2M, Fleet management include real time processing high throughput spatiotemporal events with persistency requirements and quick spatial temporal query, especially for trajectory query.

Current data management for spatiotemporal solution can not meet these requirements. For example, real-time databases stores data in flat files and organized by time but have no spatial support. Rational databases have spatial extender but the write throughput is poor due to online spatial index building.

The disclosure discloses an apparatus and approach to archive massive historical spatiotemporal data in real time with efficient storage and also convenient OLTP-style spatial temporal query support. The system architecture is shown below:

There are mainly four components in the system to archive massive historical spatiotemporal data in real time with efficient storage and support convenient OLTP-style spatial temporal query:


Map Loading Component: This component generates the static road network with edges and junctions. Edge represents a city block with the following attributes: (1) Polyline(ST_LineString) with spatial index: the block polyline given by a sequence of points (x1,y1),(x2,y2),...,(xn,yn). Usually the block is a straight line, i.e. given by two points. (2) Linkid (varchar): The block id number. Junction represents a intersection with the following attributes: (1) Point(ST_Point): the location of the junction.
(2) ID: the identity of junction.


Summary Data Write Component: Since a moving objectis moving in the exist network, its trajectory is a list of edges (called trajectory summary) in the network. This

1



Page 02 of 5

component matches the trajectories of moving objects into the road network and represents the trajectory as a list of linkids (trajectory summary), whichare stored in the relational table with B-tree index so as to offload the heavy directly update of sub-trajectory with spatial index.


Raw Data Write Component: This component cleanses the raw GPS data based on road network, such as outlier filter and trajectory compression, and stores the cleansed GPS data into TimeSeries table, which provides compact, efficient storage for an unlimited number of time-indexed observations for each moving object (vehicle).


Query Engine with SQL extension Component: This component provides spatiotemporal native SQL (OLTP-style) to query spatiotemporal data. Two-layer spatiotemporal index is used to achieve fast query response.

Two-layer spatiotemporal index is shown in the following figure.

The first layer index is using linkid to partition trajectory. The linkid can link a trajectory to spatial index built in the static road network. R-tree index of road network can be used to find the proper linkid related to spatial condition. The second layer index is using 2-dim...