17-11-2012, 03:36 PM
Research in wireless sensor networks (WSNs)
Research in wireless sensor networks (WSNs).doc (Size: 143 KB / Downloads: 55)
Abstract
Research in wireless sensor networks (WSNs) has experienced a significant growth in recent years. One topic of special
interest is the use of WSNs in the detection of forest fires. In this paper. The proposed algorithm based on information fusion techniques. This algorithm uses a threshold
method and nodes equipped with temperature, humidity and light sensors.
Introduction
Mobile ad hoc networks (MANETs) are complex distributed systems comprised of wireless devices .
In a MANET, no infrastructure is required to enable information exchange among the nodes, since they are
able to self-organize using diverse network topologies and multi-hop communication techniques. Wireless
sensor networks (WSNs) are a special case of MANETs. In a WSN, the nodes that conform the network
are low power devices equipped with one or more sensors, a processor, memory, a power supply, a radio,
and an actuator . The nodes or motes of a WSN are able to sense the environment conditions, such
as temperature, humidity, barometric pressure, biological or chemical conditions, among others. Wireless
sensor networks can be used in many applications.
Due to their memory and processing limitations, the nodes of a WSN are equipped with a radio to
transfer wirelessly the important and partially processed data to a base station. The base station receives
the information from the nodes, process them and take the appropriate actions; for instance, inform a user
that an event of interest has taken place. Commonly the main power source in a sensor node is a battery.
However, a secondary power supply can be used, such as solar panels.
Among theWSNs applications, forest fire detection can be very helpful in avoiding human and material
losses
Since forest fire detection has been a main concern in many countries, diverse techniques have been
devised to monitor forest fires. One of the most traditional methods is the use lookout towers located at
high points. Unfortunately, this method has limitations due to the unreliability of human observations. For
this reason, some techniques based on the use of technology have been proposed, as the Automatic Video
Surveillance System or the use of satellite images. However, the low spatial and temporal resolution
of the satellite images may cause a delay in the fire detection process.
The use of multiple sensor sources, and the deployment of the sensor nodes in areas not
visible to the satellite, increase the probability of a more accurate and early fire detection.
WSNs event detection applications are in essence fusion information processes Fusion information
can be defined as “the use of the theory, techniques and tools created and applied to exploit the synergy in
the information acquired from multiple sources (sensor, databases, information gathered by humans, etc.),
in such a way that the resulting decision or action is in some sense better (qualitatively or quantitatively, in
terms of accuracy, robustness, etc.) than would be possible if any of these sources were used individually
without such synergy exploitation” .
RelatedWork
One of the first proposals of the use of wireless communication for fire detection was the SIGMASPACE
system, that used smoke detectors .In Chen et al. proposed a method based on data fusion.
The algorithm used temperature, smoke density and CO density values, and neural networks and fuzzy
inference to determine if a fire has occurred. However, none of these proposals are based on the use ofWSN
technology.
Doolin and Sitar described in a system to monitor forest fires, based on the use ofWSNs. The nodes
of the system were able to sense temperature, humidity and barometric pressure, and were equipped with
GPS units.
Yu et al. proposed in a paradigm for forest fires detection that use a WSN. It considered that the
motes included temperature, relative humidity, smoke and wind speed sensors. The nodes of the WSN are
organized in clusters, and assign a node as a cluster leader. Processing the data sensed by the nodes, the
cluster leader calculates a weather index using a neural network. This index is sent by each cluster leader to
the manager node, which determine if there is a risk of a fire.
FireWxNet was proposed by Hartung et al. in a multi-tiered portable wireless system for monitoring
weather conditions in rugged wildland fire environments. The motes of the WSN sensed temperature,
relative humidity, and wind speed and direction. FireWxNet was designed to provide a better comprehension
of the weather conditions related with the presence of fire, and did not considered its detection.
Sha et al. proposed the FireNet architecture in to support fire rescue operations. In their proposal,
each firefighter in the sensor field carries a sensor node, which can sense parameters of interest. The sensor
board records all the information expected by the incident commander and the fire department, that is used
for real-time decisions and post-event analysis.
The Forest-Fires Surveillance System (FFSS) was proposed by Son et al. in which consisted of
a WSN, a middleware and a web application. The nodes of the WSN measure temperature and humidity,
and detect smoke. The middleware program and the web application analyze the collected data. An index,
defined by the South Korea government, was used to trigger an alarm notification indicating that a fire may
occur.
Finally, da Penha et al. proposed and evaluated two algorithms for forest fire detection, based
on information fusion techniques. They used a WSN to sense light and temperature values. The proposed
algorithms were based in the threshold and Dampster-Shafer methods.
Algorithm Based on the Threshold Method
In order to devise an algorithm to detect forest fires using a WSN, it was important to understand the
prevailing environmental conditions when a fire occurs. To achieve this, The collected measurements of
temperature, light and relative humidity during the months of July and August, since these are the months
with the highest incidence of fires Here temperature, light and relative humidity sensors used since they
are available in many WSN platforms. The decision is made to devise a proposal based on the use of economic
and highly available sensors. The use of more sensor will add reliability to the detection process, but will
increase the energy consumption and the deployment costs. Thus, the goal was to make a compromise
between efficiency and cost.
The measurements were taken without the occurrence of fires. The environmental conditions were measured when some fires are artificially generated. The experiments were conducted in order to simulate the existing conditions in the early stages of a fire. Consider the sensor nodes were placed on the top of trees of the
grounds of the university campus. To simulate the fires torches are used. The nodes were exposed to the
torch flames for periods of ten minutes.