Information Technology Reference
In-Depth Information
approach for communication could be extended to sensor networks. This research
program resulted in many WSN systems like acoustic tracking of low-flying aircraft
[ 7 ] or Remote Battlefield Sensor System (REMBASS). These solutions were very
expensive and could be used only for a dedicated military purpose. The main focus of
the WSN research remained in the military area until the end of 1990s when the first
motes for environmental monitoring were developed [ 8 ].The availability of low-cost
sensor nodes has resulted in the emergence of many other potential applications,
from industrial sensing to infrastructure security and health care. Since the end of
1990s many companies have created vertical WSN systems which can only be used
for a single application. To stimulate WSN market development by increasing the
interoperability between these dedicated systems, the companies developed commu-
nication standards for WSNs (Zigbee [ 9 ], WirelessHART [ 10 ] and ISA-SP100 [ 11 ]).
Although these standards are already mature and many companies sell products
complaint with them, WSN market has not expanded significantly since their
publication. Therefore currently, most WSN systems are deployed as vertical
solutions and the users do not tend to apply them for multiple applications. As a
result the price of WSN nodes is still high and the WSN suppliers try to respond to
the market demand by developing systems only for dedicated application.
The WSN research community realized the weaknesses of the vertical solutions
(described in Section 5.1.1) and one of the first solutions that addresses these problems
was IrisNet project from Intel Research started around 2000 [ 12 ]. The project aims to
develop a scalable software infrastructure that employs data mining to let Internet
users query real-time and historical video information produced by Web cams and
other sensors. ÌrisNet takes database-centric approach in its design and users can
query for measured data using XPATH query language (see Section 5.2.2.1).
Since the emergence of IrisNet system many solutions have been proposed
which facilitate deployment of horizontal applications. They share many features
of IrisNet, e.g., database-like design approach (Hourglass [ 13 ], SenseWeb [ 14 ]) or
multi-tier architecture (JWebDust [ 15 ], Janus [ 6 ]). Existing sensor networks inte-
gration frameworks differ mainly by the number of provided features and also by
the maturity of the system implementation. Some of them have only been
implemented as a prototype (e.g., Janus) and others have been widely used for
various applications and are constantly being improved by the research community
(e.g., GSN). But none of the proposed solutions have been used in a commercial
application and only the future can show if the sensor networks integration frame-
works would contribute to the expansion of the WSN market.
5.2.1.2
Summary of Features of the Existing Integration Frameworks
Table 5.1 summarizes nonfunctional properties of the surveyed existing SNIFs. The
detailed analysis of each presented SNIF is provided below. Most of the proposals
enable the interactions of applications with different heterogeneous SANs and
support many necessary types of interactions such as query- and event-based inter-
actions or streaming. Some approaches such as IrisNet and JWebDust provide
Search WWH ::




Custom Search