9 
A MULTI-SCALE LINE STRUCTURE FOR PROGRESSIVE TRANSMISSION OF 
VECTOR DATA ACROSS THE WEB 
REN Yingchao 3 ’ *, ZHU Lin b , YANG Chongjun 3 
a The State Key Laboratory of Remote Sensing Information Sciences, IRSA, CAS, Beijing, 100101 
b College of Resources Environment and Tourism, Capital Normal University, Beijing, 100037, China - 
yingchaoren@gmail.com 
Youth Forum 
KEY WORDS: Multi-scale Line, Simplification, Multi-way Tree, Progressive Transmission 
ABSTRACT: 
This paper presents a multi-way tree based multi-scale line structure for progressive transmission of vector data. By topological 
restraints, the structure maintains the consistent topological relations among the lines of different scales. By integrating the data of 
the same scale into one certain level node, the structure provides more efficient hierarchical data management. And the order of the 
points in the original line is described by the relative order of points in the multi-scale line tree, so that we can restore the line in any 
scale, which also supports the modification of multi-scale line without rebuilding of the multi-scale structure. Meanwhile, the data in 
different scale are independent of each other, so that we can store the data in different scale into different tables in database, which 
will improve the first access efficiency. 
1. INTRODUCTION 
Geographic space is a complicated system with geographic 
elements with self-similarity in different scales interacting with 
each other. Geographic space in different scales represents 
different geographic phenomena and geographic laws. In the era 
of paper maps, people produced a series of maps with different 
map scales to reflect different geographic phenomena in 
different scales. 
Geographic Information System (GIS) is an electronic tool to 
manage maps. The concept of map scale in traditional map has 
also changed. The functions of zoom in and zoom out make 
maps be able to be displayed in any map scales. However, 
traditional GIS can only manage maps with a single scale. So 
zoom in operation doesn’t bring into the increasing of map 
information, while zoom out operation makes map information 
be too dense to be discerned so as to make it hard for users to 
get main information in the map. Secondly, operations of zoom 
in and zoom out are just to make maps with a single map scale 
to be displayed with different view scales which cannot reflect 
geographic phenomena in different scales in nature. 
The data volume of geographical data is usually massive. 
Although the bandwidth of the network has got remarkable 
improved, the rapid growth of spatial data and the users’ 
demands to the high resolution spatial data soon exhausts the 
limitedly improved bandwidth, which situation is more serious 
in wireless network. Nowadays, the limited bandwidth of the 
network has become an important bottleneck for the web 
issuing and access of the massive, high resolution spatial data 
across the web, which appears more serious to mobile devices 
based on the wireless network, such as cell phones, PDAs, on 
which a great deal of detailed data in the high resolution spatial 
data cannot be displayed, while still occupy bandwidth to 
download. 
Progressive transmission of vector data provides a promising 
solution to that problem. The skeleton data with low resolution 
are transmitted to the client at first, then by users’ request the 
detailed increment data are transmitted to the client and 
integrate with the existing data on the client to restore the high 
resolution data, so as to reduce the redundant data transmission, 
improve the utilization of the bandwidth while preserving the 
precision of the original data. 
The key to the progressive transmission of vector data is to 
generate the multi-scale representation of spatial data. The ideal 
way is just to maintain a single base map with very large map 
scale. By user’s demands, maps with small scales are generated 
from that base map on the fly. The abstract of map is a process 
of map generalization. Map generalization should preserve 
topological and geometric consistency among maps with 
different scales. So map generalization has to deal with 
complicated and time-consuming geometric computation, and 
meanwhile there exist a great deal of subjective factors which 
are hard to be regularized. Till now, the efficiency and the 
results of map generalization still cannot satisfy the demands of 
on-the-fly map generalization. Anther way is to maintain a 
multi-version map database. Similar to traditional paper maps, 
the multi-version technology integrates a series of maps with 
different map scales together. By users’ requests for different 
map scales, the system accesses maps with corresponding scale 
and returns them to the users, so that users can get maps with 
corresponding map scales in different view scales. The multi 
version technology satisfies the demands of users for maps with 
different scales. However, it also brings some new problems. 
Firstly, the system has to maintain maps with different scales, 
which heavies the burdens of system. Secondly, data update 
Supported by the 863 Program of China (Grant No. 2006AA12Z208) and the CAS Innovation Program (Grant No. KZCX2-YW- 
304-02).