1009 
A NEW PEER-TO-PEER-BASED INTEROPERABLE SPATIAL SENSOR WEB 
ARCHITECTURE 
S.H.L. Liang 
Department of Geomatics Engineering, University of Calgary, Calgary, Alberta, CANADA T2N 1N4 
steve.liang@ucalgary.ca 
Commission I, ThS-1 
KEY WORDS: GIS, IntemetAVeb, Interoperability, Multisensor, Distributed 
ABSTRACT: 
With the rapid advances in sensor network, and information and communication technologies, the vision of a World-Wide Sensor 
Web (WSW) is becoming a reality. However, there is a lack of a spatial information infrastructure that could aggregate the 
independent geo-sensor networks into a coherent Spatial Sensor Web (SSW). The SSW vision brings two architectural challenges to 
today’s GIService systems: (1) scalability and (2) interoperability. This paper introduces GeoSWIFT 2.0, a new scalable and 
interoperable SSW architecture. GeoSWIFT 2.0 is scalable - it removes all single points of failure and system performance 
bottlenecks by using a fully decentralized P2P spatial query framework. GeoSWIFT 2.0 is also interoperable - it integrates 
interoperable sensor web standards with the scalable P2P framework. 
1. BACKGROUND 
Distributed sensor networks are attracting more and more 
interest in applications for large-scale monitoring of the 
environment, civil structures, roadways, animal habitats, etc. 
With the rapidly increasing number of large-scale sensor 
network deployments, the vision of a World-Wide Sensor Web 
(WSW) is becoming a reality. Similar to the World-Wide Web 
(WWW), which acts essentially as a “World-Wide Computer”, 
the Sensor Web can be considered as a “World-Wide Sensor” or 
a “cyberinfrastructure” that instruments and monitors the 
physical world at temporal and spatial scales that are currently 
impossible. Ranging from video camera networks that monitor 
real-time traffic to matchbox-sized wireless sensor networks 
embedded in the environment to monitor habitats, the WSW 
will generate tremendous volumes of priceless data, enabling 
scientists to observe previously unobservable phenomena. 
One major reason that the development of the WSW has been 
greatly limited is the lack of an infrastructure that connects 
many heterogeneous sensor networks to the applications that 
desire sensor network data. Data is the raison d'être of any 
sensing exercise. However, today’s WSW researchers have 
focused on distributed sensor networking rather than on sensor 
data management (Balazinska et al., 2007). 
Moreover when it comes to sensor data, the phrase “spatial is 
special” is particularly relevant. Sensing is essentially a 
spatially based sampling process in which each sensor data can 
generally be associated with location information. Within the 
context of the WSW, the phrase means that handling spatial 
properties of sensor data requires special algorithms, data 
models, databases, data presentations, system architectures, etc. 
There is a desire for a spatial information infrastructure 
designed specifically for the WSW. The spatial information 
infrastructure would aggregate the independent geo-sensor 
networks into a coherent Spatial Sensor Web (SSW). The main 
goal of this paper is to propose a new scalable and interoperable 
GIService architecture for the Spatial Sensor Web. 
2. ARCHITECTURAL DESIGN CHALLENGES AND 
REQUIREMENTS 
The SSW vision brings exciting and innovative applications. 
However, it also brings its architectural design unique 
challenges. Below, we outline two major challenges. 
2.1 Scaling to accommodate an enormous amount of 
mobile and transient sensors 
The number of sensors in the WSW could be enormous. These 
geographically distributed sensors would generate a massive 
amount of spatially referenced data streams. Many of the 
sensors may be mobile because they have to update locations 
often, while many wireless and battery-powered sensors may be 
transient because they frequently have to connect/reconnect. 
The above factors are challenging today’s Internet GIService’s 
scalability. The existing GIService architectures’ centralized 
topologies are not designed for such large-scale and highly 
dynamic data sources. When existing architectures scale to 
accommodate users and sensors, its centralized components 
make the system vulnerable in that they are single points of 
failure. The centralized components are also system 
performance bottlenecks in that all additional system loads are 
added to them. A solution to making the system scalable by 
removing the architectures’ centralized components is critical. 
2.2 Allowing heterogeneous sensor networks to 
interoperate 
Today’s sensor networks are not interoperable. In other words, 
they cannot transparently allow each other to access, 
interchange, understand, and use the sensing resources. One of 
the major reasons for this inability to interoperate is that sensor 
networks are computers deployed in the fields. In order to 
accommodate the severely constrained environments, these 
sensor networks are built vertically with specialized hardware,