Detection of forest fire using Wireless sensor network

Author 1. Ms. Sneha Patil

Author 2. Mr. Sumedh Deshpande
Author 3. Mr. Rushikesh Shingade
- Dept. of Computer and IT Engg., VJTI, Mumbai.
Abstract- Indias forest cover of 19.5 percentage of geographical area is deteriorating at an alarming rate. Among other factors, forest fires
are a major cause of degradation of Indian forests. It results in massive destruction and impact the ecology, society and economy of the
region. Estimated average tangible annual loss due to forest fires in country is Rs.440 crore (US dollar 100 million approximately) [2].
According to a Forest Survey of India report, about 50 per cent of forest areas in the country are prone to fire.[1] In this document, we
propose a new real-time forest fire detection method by using wireless sensor networks. Our goal is to detect and predict forest fire
promptly and accurately in order to minimize the loss of forests, wild animals, and people in the forest fire. Wireless sensor networks are
increasingly applied in the field of environmental and ecological monitoring. Especially in difficult and harsh environments, it has
advantages that traditional monitor systems lack. To monitor temperature in the forest in a more timely and precise way, we pointed out
unique advantages of safety in data transmission, flexibility in building the network, and low cost and energy requirements for a forest
fire monitoring system based on a wireless sensor technology that we designed. We propose this system as a first attempt and complement
to existing forest fire monitoring and prevention methods. It provides a solid basis in terms of hardware for the application of advanced
wireless sensor network technology.
Keywords: Wireless sensor network, DYMO, TinyOS, Simulation

I.INTRODUCTION
The use of this technology throughout society "could well dwarf
previous milestones in the information revolution": U.S.
National Research Council Report, 2001. In 2003 it was
heralded as one of "10 emerging technologies that will change
the world". This revolutionary technology is known as
Wireless Sensor Networks (WSNs). A sensor network is an
infrastructure comprised of sensing (measuring), computing,
and communication elements that gives an administrator the
ability to instrument, observe, and react to events and
phenomena in a specified environment. A wireless sensor
network (WSN) is a wireless net- work consisting of spatially
distributed autonomous devices using sensors to cooperatively
monitor physical or environmental conditions, such as
temperature, sound, vibration, pressure, motion or pollutants, at
different locations. There are four basic components in a sensor
network: (1) an assembly of distributed or localized sensors; (2)
an interconnecting network (usually, but not always, wirelessbased); (3) a central point of information clustering; and (4) a
set of computing resources at the central point (or beyond) to
handle data correlation, event trending, status querying, and data
mining. The communication between sensor nodes starts with
data being forwarded via multiple hops, to a sink (sometimes
denoted as controller or monitor) that can use it locally or is
connected to other networks (e.g., the Internet) through a
gateway. The design issues for network set up and
communication include cost, energy efficiency, energy
sufficiency, ability to cope with node failures, nature of node i.e.
homogenous or heterogeneous, communication failures,
scalability to large scale of deployment, environment
accessibility, ease of use and many more. A Crossbow

Technology, Inc., system was selected to provide information on

how to use the TinyOS technology. Our goal is to introduce
implementation of a potentially user-friendly WSN system with
its use for detection of forest fire.
This paper provides description accompanied by figures to help
the reader get a better understanding of what we have
accomplished as our final year project.
Contributions: (objective, dymo)This proposed method,
present a probabilistic duplicate detection method for
hierarchical multimedia data called XMLMultiDup. This
method considers all parameters and aspects for comparison of
XML datasets which contain multimedia database like images,
audio and videos. The algorithm presented here extends work in
[1] significantly improving level of detecting duplication and
efficiency and this paper also considers relational databases for
finding duplications by converting it into hierarchical datasets.

Figure 1: Two XML elements that represent the same Film. Nodes are labelled
by their XML tag name.

Here the main contribution compared to previous work is and

objectives of this system are 1) to detect duplicates in
hierarchical data which contain multimedia data e.g. images,
audio and video using XMLMultiDup method. 2) To convert
relational database into XML and then detect duplicates as
above. 3) To compare datasets according to user choice and
display results e.g. limited no. of levels to be compared. 4) To
increase efficiency and effectiveness of duplicate detection in
comparison of multimedia databases. 5) To eliminate or update
duplicates, to reduce size of databases in data warehouses. 6) To
consider all probabilities of XML trees for comparison for
example part of tree, structure of trees, levels of tree, values and
contents within trees
and complete subtrees to find
duplications.
Structure: This paper is organized as follows: Section 2
presents related work. Section 3 summarizes methodology of
the proposed system. Our strategies to this algorithm are then
presented in Section 4.Working environment of these techniques
over artificial and real world data in Section 5. Finally, in
Section 6 we conclude and present suggestions for future work.
II.

RELATED WORK

In [2]
Ananthakrishna has
exploited
dimensional
hierarchies typically associated with dimensional tables in
data warehouses to develop duplicate elimination algorithm
called Delphi, which significantly reduces the number of
false positives without missing out on detecting duplicates. He
rely on hierarchies to detect an important class of
equivalence errors in each relation, and to efficiently reduce
the number of false positives.
Carvalho and Silva proposed a similarity-based approach in [3]
to identifying similar identities among objects from multiple
Web sources. This approach works like the join operation in
relational databases. In the traditional join operation, the
equality condition identies tuples that can be joined together.
In this approach, a similarity function that is based on
information retrieval techniques takes the place of the equality
condition. In this paper, we present four different strategies to
dene the similarity function using the vector space model and
describe experimental results that show, for Web sources of
three different application domains, that our approach is quite
effective in nding objects with similar identities, achieving
precision levels above 75%.
DogmatiX is a generalized framework for duplicate detection
[4], dividing the problem into three components: candidate
definition defining which objects has to be compared, duplicate
definition defining when two duplicate candidates are actually
duplicates, and duplicate detection means how to efficiently find
those duplicates. The algorithm is very effective in the first
scenario: Edit distance should compensate typos, and our
similarity measure is specifically designed to identify duplicates
despite missing data. On the other hand, synonyms, although
having the same meaning, are recognized as contradictory data
and the similarity decreases. They are more difficult to detect
without additional knowledge, such as a thesaurus or a
dictionary. Thus, we expect the second scenario to yield poorer
results.
Milano Propose a novel distance measure for XML data, the
structure aware XML distance [5] that copes with the flexibility

which is usual for XML files, but takes into proper account the
semantics implicit in structural schema information. The
structure aware XML distance treats XML data as unordered.
The edit distance between tokens t1 and t2 is the minimum
number of edit operations (delete, insert, transpose, and replace)
required to change t1 to t2; we normalize this value with the
sum of their lengths
In [6] author has proposed a novel method for detecting
duplicates in XML which has structural diversity. This method
uses a Bayesian network to compute the probability of any two
XML objects being duplicates. Here author has considered not
only children elements but also complete subtrees. Computing
all probabilities, this method performs accurately on various
datasets. Following figure shows two XML trees which contain
duplicate data although value represented differently.
Base for proposed system presented in [1], has extended work
done in [6] by adding pruning algorithm to improve the
efficiency of the network evaluation. That is to reduce the no. of
comparisons where the pairs which are incapable of reaching a
given duplicate probability threshold are decreased. It requires
user to give input, since the user only needs to provide the
attributes to be compared, their respective default probability
values, and a similarity value. However, the system worked in
good manner that it allows to use different similarity measures
and different combinations of probabilities.
III.

METHODOLOGY

A method described in [1], the author has extended his

previous work by increasing efficiency and effectiveness for
duplicate detection in hierarchical data, but proposed system
will be useful for both simple and multimedia data. Here the
input will be two XML trees or datasets; for this we can use real
world data or standard dataset. The first phase of this
XMLMulitDup method is to input either XML data or simple
relational data which will be converted into XML for
comparison and duplicate detection. The choice of user will be
taken for comparison i. e. whether to compare structures of tree,
values of tree or contents of the tree. The second contribution of
proposed system is to input dataset which contain any type of
multimedia data which contain images, audio or videos. We will
first compute prior, computational and final probabilities using
Bayesian Network. The algorithm and formulae given below
will be used for this whole recursive process. But while doing
this, if we use Bayesian network there is an issue of complexity
of O(n X n). Hence a pruning technique will be used which
reduces no. of comparisons using pruning algorithm. Here the
contents of multimedia data will be later compared by MD5
hash Key algorithm, if structure and values found duplicate.
Following figure shows the architecture of proposed system,
which includes combination of three algorithms.
1. Bayesian network
2. Pruning Algorithm
3. MD5 Hashkey Algorithm

Third Layer
This layer is also called Client Layer (Visualization, Analysis
tool) It provides the user visualization software and graphical
interface for managing the network. Moteview is used for this
purpose that bundles software from all three layers to provide an
end to end solution.
Algorithm : Detection of forest fire
Input: Sensed temperature
Output: Forest fire is detected or not

Figure 2: Architecture of Proposed System

Here the proposed system uses all these algorithms but needs
small user intervention. User has to provide the parameter by
which comparison will perform. And second the action to be
performed after duplicate detection which may be elimination,
updation, or any other operation. Next section will describe all
algorithms and example which show how the original trees
shown in figure 1 will be converted to Bayesian network.
IV. DESIGN AND SPECIFICATION
In this system, there are three design layers. They are as
follows: [31]
First Layer
This layer is also called Mote Layer(TinyOS firmware). XMesh
resides here which is the software that runs on the cloud of
sensor nodes forming a mesh network. The nodes N1 to N8
form the mesh network as shown in diagram below. The XMesh
provides the networking algorithms required to form a reliable
communication backbone connecting all the nodes within the
mesh cloud to server.

Analyze the average temperature of the region according to

Season (Winter/Summer)
Sensor nodes (MDA 100 & IRIS) are deployed in the forest
3. Each sensor node is assigned a unique identification
number (uid)
4. Radio board (MIB520) is used as the gateway for
receiving information
5. FOR all sensor nodes belonging to a Network DO
5.1 Nodes sense data (e.g., temperature) continuously
5.2 This data is passed to the Reception Centre via
multiple hops
5.3 The reception centre then examines the temperature
difference of current and average temperature
5.4 For a particular region,
[(IF season is summer AND if this temperature
difference is greater than 10 degrees) OR (IF season is winter
AND if this temperature difference is greater than 10 degrees )]
THEN the fire brigade is informed about the
forest fire.
ELSE No action is performed
END FOR
Sequence diagram for Forwarding Data :
A Sensor node is a part of group called cluster, which reads data
from surrounding. If sink is not in its range, it passes data to
other sensor node and so on till it is passed to a node in range
which is called super node. The data at each step is aggregated
and passed forward.
The super node then forwards it to the sink from where it
reaches the reception centre. The data is examined here and if
conditions of forest fire are met, a message is sent to fire
brigade which takes further course of action by reaching that
sensor node.

Figure 4.1: System Design

Second Layer
This layer is also called Server Layer(Database, Logger). It is an
always-on facility that handles translation and buffering of data
from the wireless network and provides the bridge between the
wireless motes and clients. Xserve, which is installed with
MoteWorks, is used for the purpose of displaying sensor
message packet contents as they arrive on the PC over serial
port. This is shown with arrows in the diagram.

Figure 4.4: Sequence Diagram(Forwarding Data)

V.EXPERIMENTAL SETUP
The sensor nodes are configured using mote software. The
codes are burned on nodes and base station. Sensor nodes are
deployed in the environment whose surrounding temperature
has to be measured. The radio board is connected to the usb port
on computing device which acts as the gateway. Simulation of
DYMO routing protocol is done using TOSSIM (Tiny OS
Simulator) [28]
The data received from Terminal:

VI. CONCLUSION
In this project report, we succeeded in understanding the basics
of Wireless Sensor Network (WSN), the software and hardware
required for it along with their specifications required to build a
wireless sensor network system. By optimally routing the data
with the reference of dymo routing protocol and extracting the
information from it effectively, Wireless sensor network provide
insights about the early detection of forest fire efficiently thus
making an environment friendly system by reducing the impact
caused by forest fire.
VII. FUTURE WORK

Application User Interface

In this section, the GUI (graphical user interface) of the
application is shownMonitoring: The sensed data (temperature in this case) from
node 01 is displayed in the figure.

This work can be further extended to include the following

features:
Scalability: We have implemented forest _re detection for a
particular season and region. This can be further extended to
include different seasons and regions.
Accurate detection: Along with temperature other parameters
can also be considered like humidity to accurately detect the
forest fire. Thus, wireless sensor networks can be a very
promising technology to be used in future.
.
REFERENCES

Figure 5.3: Snapshot of Monitoring

Detection: The sensed data is examined continuously for a
drastic change in temperature. Once found, a message is
displayed "Fire detected at node 01" as shown in figure.
Figure 5.4: Alert Of Detection

[1] Luis Leitao, Pavel Calado, and Melanie Herschel, Efficient and Effective
Duplicate Detection in Hierarchical Data IEEE Trans. on Knowledge and Data
Engineering, Vol. 25, No. 5, May 2013.
[2] R. Ananthakrishna, S. Chaudhuri, and V. Ganti, Eliminating Fuzzy
Duplicates in Data Warehouses, Proc. Conf. Very Large Databases (VLDB),
pp. 586-597, 2002.
[3] J.C.P. Carvalho and A.S. da Silva, Finding Similar Identities among Objects
from Multiple Web Sources, Proc. CIKM Workshop Web Information and Data
Management (WIDM), pp. 90-93, 2003.
[4] M. Weis and F. Naumann, Dogmatix Tracks Down Duplicates in XML
Proc. ACM SIGMOD Conf. Management of Data, pp. 431-442, 2005.
[5] D. Milano, M. Scannapieco, and T. Catarci, Structure Aware XML Object
Identification, Proc. VLDB Workshop Clean Databases (CleanDB), 2006.
[6] L. Leitao, P. Calado, and M. Weis, Structure-Based Inference of XML
Similarity for Fuzzy Duplicate Detection, Proc. 16th ACM Intl Conf.
Information and Knowledge
Management, pp. 293-302, 2007.
[7] K.-H. Lee, Y.-C. Choy, and S.-B. Cho, An efficient algorithm to compute
differences between structured documents IEEE Transactions on Knowledge
and Data Engineering (TKDE), vol. 16, no. 8, pp. 965979, Aug. 2004.
[8] E. Rahm and H.H. Do, Data Cleaning: Problems and Current Approaches
IEEE Data Eng. Bull., vol. 23, no. 4, pp. 3-13, Dec. 2000.

[9] L. Leitao and P. Calado, Duplicate Detection through Structure

Optimization, Proc. 20th ACM Intl Conf. Information and Knowledge
Management, pp. 443-452, 2011.
[10] M.A. Hernandez and S.J. Stolfo, The Merge/Purge Problem for Large
Databases, Proc. ACM SIGMOD Conf. Management of Data, pp. 127-138,
1995.

[11] S. Guha, H.V. Jagadish, N. Koudas, D. Srivastava, and T. Yu, Approximate

XML Joins, Proc. ACM SIGMOD Conf. Management of Data, 2002.
[12] Joe Tekli, Richard Chbeir, Kokou Yetongnon An overview on XML
similarity: Background, current trends and future directions,Computer Science
Review, Volume 3,Issue 3, August aphy. Personal hobbies will be deleted from
the biography.