Please use this identifier to cite or link to this item:
Title: Wireless sensor networks for forest fire detection
Authors: Le, Thu Nga.
Keywords: DRNTU::Engineering
Issue Date: 2009
Abstract: For the last few decades, increasing incidents of forest fire have caused tremendous human and property loss and raised significant global concern. This project aims to contribute a novel method for forest fire detection by localizing and tracking certain groups of animals using Wireless Sensor Network (WSN) technology. To do this, a number of sensor nodes are attached to selected animals. Whenever the temperature sensed at these animals’ proximity rises beyond a predefined threshold, the localization module would be activated and the subjects’ motion paths are analyzed. The region of forest fire is estimated based on 2 indicators: • A group of animals is observed to run away from a certain area, and • The temperature sensed at the animals’ surrounding environment is high. With this initiative, we figure out that a central technical issue of our project is localization technique for mobile sensor nodes. A large number of localization techniques in WSN have been previously proposed. However, with the author’s knowledge, none of them is exclusively targeted for the purpose of forest fire detection; therefore inevitably have certain disadvantages such as high complexity, low efficiency and large power consumption when being adopted in forest fire detecting systems. The student and her FYP supervisor have proposed a new simple and efficient sensor-based localization algorithm which is particularly tailored for forest fire detecting systems. This new algorithm overcomes the inherent drawbacks of existing techniques by the following features: • Grid-based technique: Unlike almost exiting algorithms which determine a point-based location, our algorithm estimates the region-based or grid-based location of the sensor node. On one hand, we believe a region-based location is good enough for forest fire detection, because forest fires are normally spread in a region instead of occurring in a single point, and they are often rescued by helicopters which can quickly locate the actual fire location given an estimated region from certain height. On the other hand, estimating a region-based location requires a much simpler process with less iteration than determining a point-based position, thus significantly reduces complexity, enhances efficiency as well as saves sensors’ power consumption. • Event-triggering mechanism: the localization module is not activated all the time; instead it is only triggered when temperature sensed at any sensor node rises beyond limit. As the temperature decreases into the allowable range again, localization function would automatically stop. By this way, the sensors draw less battery power, which considerably lengthens the system lifetime. Despite of its simplicity, the algorithm can achieve a good performance in terms of accuracy and area of outputted region. Its key performance indicators include: • Probability of accurate detection is 100%, which means the actual node location is always within the estimated region-based location. • Probability of returning a small-enough region-based location is maintained at 60% and above for monitored area with dimension up to 6 times of sensors’ transmission range. In addition to the theoretical development and evaluation in simulation environment, the newly-proposed algorithm was also implemented in Crossbow’s set of sensor hardware. The experimental results have shown that the algorithm is highly implementable, and can work reasonably well in practice. Nevertheless, together with a good localization technique, a suitable multi-hop routing protocol designed for communication between mobile nodes and base station, which is beyond the scope of this project, is also required to obtain high accuracy and stability for localization systems.
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
  Restricted Access
1.72 MBAdobe PDFView/Open

Page view(s)

checked on Sep 26, 2020


checked on Sep 26, 2020

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.