A GIS DATABASE FOR TIME-EVOLVING SPATIAL OBJECTS 
Dae-Soo Cho, In-Sung Jang, Kyoung-Wook Min, Jong-Hyun Park 
LBS Research Team, Telematics Research Division, ETRI, 161 Gajeong-dong, Yuseong-gu, Daejeon, 305-600 Korea - (junest, 
e4dol2, kwmin92, jhp)@etri.re.kr 
KEY WORDS: Spatial Information Sciences, Databases, Modelling, Query, GIS , 
ABSTRACT: 
In this paper, we have designed a data model for moving objects and implemented it. The moving objects are time-evolving spatial 
objects, that is, their geometries are dynamically changed as time varies. Generally, a GIS database stores and manages the spatial 
objects, of which geometries are rarely changed. The traditional G 
objects, due to the fact of geometries being frequently changed and all of the history information of moving objects being managed. 
To manage moving objects efficiently, we have added new data types, such as moving point and moving polygon, to the traditional 
GIS data type. We have also defined the semantic of underlying operators for those data types. It is expected that the GIS database 
we have developed makes it feasible to developing a wide range of database applications managing moving objects, such as cars, 
aircraft, ships, mobile phone user, hurricanes, oil spills in the sea, forest fires, armies, and tribes of people in history. 
1. INTRODUCTION 
Recently, various types of location-based services have 
obtained increasingly high attention according to the extensive 
spread of mobile handset, which is capable of accessing 
wireless internet, and the development of location 
determination technology, that is represented by GPS (Global 
Positioning System). Location-based services are related the 
moving objects which change their locations through time. 
Therefore, to provide location-based services efficiently, it is 
required that an efficient system which could acquire, store, and 
query the large number of locations. The time-evolving 
locations of moving objects are not efficiently managed by 
existing commercial Database Management System (DBMS) as 
well as Geographic Information System (GIS). The reason is 
that there is a critical set of capabilities that are needed by 
moving objects database applications (Wolfson et al. 1993), 
such as location-based services, and are lacking in existing 
DBMS and GIS. The needed capabilities are location data 
model for moving objects, query language for moving objects, 
location index for moving objects, and so on. 
Previous works for moving objects can be classified into two 
main groups; works related to location data models and query 
languages (Sistla et al., 1997; Forlizzi et al., 2000; Wolfson et 
al. 1998; Güting et al, 2000) and works related to indexing 
locations (Pfoser et al., 2000; Kollios et al., 1999; Nascimento 
and Silva, 1998; Vazirgiannis et al, 1998; Song and 
Roussopoulos, 1987). These works, also, can be classified by 
works for current and future. locations (Sistla et al, 1997; 
Kollios et al, 1999; Wolfson et al. 1998; Song and 
Roussopoulos, 1987) and works for trajectories (past locations) 
of moving objects(Pfoser et al., 2000; Forlizzi et al., 2000; 
Nascimento and Silva, 1998; Vazirgiannis et al., 1998; Güting 
et al, 2000). Other type of previous works to is related to 
generate synthetic data (Pfoser and Theodoridis, 2000; 
Theodoridis et al., 1999; IBM). Location data generator, which 
is capable of simulating real-world moving objects, are needed 
because it is not possible to obtain real datasets, either they do 
not exist or they are not accessible. 
IS database, therefore, has a difficulty to manage the moving 
The purpose of this paper is to design and implement the overall 
112 
architecture of a Moving Objects Database (MODB) which is 
applicable to the real-world applications. We have integrated 
various kinds of works related to moving objects into the 
MODB. The rest of the paper is organized as follows: Firstly, 
we will discuss the overall architecture of MODB. Then, we 
will explain each of six modules of which the system consists. 
Finally, we will conclude by giving directions for future work. 
2. OVERALL ARCHITECUTRE OF MODB 
The Moving Objects Database (MODB) devised in this paper is 
depicted by Figure 1. It is composed of six modules, Intelligent 
(location) Acquisition Module (IAM), a Query Processing 
Module (QPM), a Buffer Management Module (BMM), a 
Location Indexing Module (LIM), a Location Storage Module 
(LSM), and an Attribute Storage Module (ASM). 
Intelligent Acquisition Module (IAM): According to the 
location acquisition policies we are proposed such as static 
acquisition policy, distance-based acquisition policy, region- 
based acquisition policy, and predict-based acquisition policy, 
[AM acquires the current location of moving objects and reports 
it into the QPM. The policies determine when IAM acquires the 
location of a moving object and how many threads IAM uses to 
acquire the locations of all objects. The objective of IAM is as 
follows. When location based services prevail into the wireless 
internet applications; we can easily predict that transmission 
overhead is so heavy to acquire the locations of large 
subscribers and vehicles between MODB and location server. 
To solve this problem, MODB must support IAM that lessens 
transmission overhead as much as possible and guarantees 
stable system state. So, we have proposed the techniques of 
minimizing overheads of transmission in acquiring locations of 
so many moving objects. 
Query Processing Module (QPM): First of all, we defined 
query interfaces to issue user’s request. Also, we defined the 
moving objects model, which is composed of data structures 
and operations to represent the moving objects. User’s request 
Lc 
nx 
sh 
me 
me 
Lo 
rer 
ma 
Sto 
At 
the 
We 
We 
anc 
The 
mo 
loc 
mo 
and