Localization of wireless sensor network

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395 -0056
Volume: 04 Issue: 02 | Feb -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 5.181 | ISO 9001:2008 Certified Journal | Page 1338
LOCALIZATION OF WIRELESS SENSOR NETWORK
Range free anchor-based algorithm using Monte Carlo Localization
Prof. Usha Neekeleetan1, Princess Mariam Zawu2.
Head of Department, of EC L D Engineering College, Ahmedabad, Gujarat, India1
PG Student [ECS], Dept. of EC, L.D College of Engineering, Ahmedabad, Gujarat, India2
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Abstract— recent proceedings in radio and
embedded systems have enabled the increase of wireless
sensor networks. Wireless sensor networksaretremendously
being used in different environment to perform various
monitoring tasks such as search, rescue, disaster relief, target
tracking and a number of tasks in smart environments. In
many of those tasks, node localization is inherently one of the
system parameters. Node localization isrequiredtoreportthe
origin of events, assist in group querying of sensors, routing
and to and also to know the answer of network coverage. So,
one of the fundamental challenges in wireless sensor is node
localization. This paper presents an accurate range-free
localization scheme for nodes in mobile wireless sensor
networks. As it is already known that the sequential Monte
Carlo localization methodworkswellforlocalizationinmobile
WSNs. Based on the sequential Monte Carlo method, the
TSBMCL algorithm utilizes the nodes for localization.
Keywords—wireless sensor networks; mobile
WSNs; localization; TSBMCL; Mobile Location;
Monte Carlo
1. INTRODUCTION
The massive advances of microelectromechanical
systems (MEMS), computing and communication
technology have fomented the emergenceofmassively
distributed, wireless sensor networks consisting of
hundreds and thousands of nodes. Each node is able to
sense the environment, perform simple computations
and communicate with its other sensors or to the
central unit. One way of deployingthesensornetworks
is to scatter the nodes throughout some region of
interest. This makes the network topology random.
Since there is no a priori communication protocol, the
network is ad hoc. These networks are tremendously
being implemented to perform a number of tasks,
ranging from environmental and natural habitat
monitoring to home networking, medical applications
and smart battlefields. Sensor network can signal a
machine malfunction to the control center in a factory
or it can warn about smoke on a remote forest hill
indicating that a forest fire is about to start. On the
other hand wireless sensor nodes can be designed to
detect the ground vibrations generated by silent
footsteps of a burglar and trigger an alarm. Since most
applications depend on a successful localization, i.e. to
compute their positions in some fixed coordinate
system, it is of great importance to design efficient
localization algorithms. In large scale ad hocnetworks,
node localization can assist in routing. In the smart
kindergarten node localization can be used to monitor
the progress of the children by tracking their
interaction with toys and also with each other. It can
also be used in hospital environments to keep track of
equipment, patients, doctors and nurses. For these
advantages precise knowledge of node localization in
ad hoc sensor networks is an active field of research in
wireless networking.Unfortunately,foralargenumber
of sensor nodes, straightforward solution of adding
GPS to all nodes in the network is not feasible because:
In the presence of dense forests, mountains or other
obstacles that block the line-of-sight from GPS
consumption of GPS will reduce the battery life of the
sensor nodes and also reduce the effective lifetime of
h large number
of nodes, the production cost factor of GPS is an
small. But the size of GPS and its antenna increases the
sensor node form factor.Forthesereasonsanalternate
solution of GPS is required which is cost effective,
rapidly deployable and can operate in diverse
environments. The two environments were one that
allows preplace anchor nodes when distribution is
ready and has high accuracy and the other is one that
those not allow preplace and requires accuracy of
about 30%[5]. At present, most of the localization

algorithms are designed for the static wireless sensor
network. If the static sensor network localization
algorithm is applied to the mobile sensor network, a
series of problems will be caused, for instance, the
accuracy of localization will be reduced because of the
mobility of nodes, the node energy consumption is
accelerated, etc. Some proposals are emerged in terms
of the node localization of the mobile wireless sensor
network. In the literature [6], this paper verifies the
correction and reliability of ROA (Rectangle
Overlapping Approach).They alsomadethesimulation
of the ROA algorithm by using the Borland c++ builder
6.0 program languages.Theresearchersalsodiscussed
the traverse strategy of a moving beacon, which is
moving in horizontal and vertical directions and
moving in random direction. The paper also discussed
the three important values,thesensingrangeofacircle
and a beacon moving each direction and the angle by
the directional antenna. The researchers also explored
the effects among each of the three factors through
simulations. In the literature [2], the author puts
forward the TSBMCL algorithm according to the MCB
algorithm. On account of the insufficient anchor nodes
near the unknown nodes, the ordinary nodes with the
good locations are screened as the assistantpositionof
the temporary anchor nodes. In the algorithm,itisable
to satisfy the location requirementofmobileWSNwith
the low cost and high accuracy. The sampling
procedure is complicated for the classic Monte Carlo
localization algorithm while the sampling efficiency is
low. This paper proposes a Monte Carlo localization
algorithm in the combination of hop count/distance
transformation model. In this algorithm, it can avoid
the direct use of node communication range to
determine the sampling area and there is no
requirement on the additional ranging hardware. A
better location performance is achieved.
I. Localization using range-free anchor
based algorithm using Monte Carlo
localization
A range-free localization algorithm for mobile
sensor networks based on the Sequential Monte
Carlo method. The Monte Carlo method has been
extensively used in robotics where a robot
estimates its localization based on its motion,
perception and possibly a relearned map of its
environment. Monte Carlo method as used in
robotics to support the localization of sensors in a
free, unmapped terrain. The authors assume a
sensor has little control and knowledge over its
movement, in contrast to a robot. Apart from the
experiments with MCL, there are at the moment
few localization protocols specifically designed
with mobile wireless sensors in mind. Most of the
papers presenting localization algorithms suggest
that supporting mobility can be achieved by
rerunning the localization algorithm after some
time interval, either static or adaptable. While this
is not optimal but feasible in some cases, the
whole class of algorithms using information from
distant nodes or iterative approaches will suffer
from severe information decay. At the time the
information reaches a distant node that wants to
use it, it is very likely that the whole network
configuration has changed. A node will therefore
always calculate an inaccurate location, not due to
the lack of information or to the intrinsic
inaccuracy of the calculations it uses, but due to
the way its localization algorithm gathers this
information.
MONTE CARLO LOCALIZATION
1)Bayes filtering
The MCL represents the robot's positional
certainty at an arbitrary location in a given grid
map. A robot calculates the posterior probability
by using the Bayes filter [9] based on the
odometry and range data as follows:
Bel(x,) = xt ZO:t UO:t) (l)
Where x, denotes the robot pose (x, y, 0) at time
t, zo:= {Zo, Zl,*z. } denotes the measurements of
the range sensor (e.g. Laser scanner, sonar, IR
sensor, etc.) up to t, and uo:t = {U0, ul,..., utf is the

odometric data from the wheel encoder.
Reliability of measurements varies with the
accuracy of the range sensor. In order to cope
with various uncertainties, probability models
are used to reflect the errors: sensor model(or
perception model) and motion model (or action
model).The Bayes filter is conducted in two
steps (i.e., prediction and update), which can be
represented by the following equation:Bel(xt) =
7 p(zt xt ) JP(xt xt1, ut ) Bel(xt-I )dxt-1 where q
is the normalizing constant, p(xt xt,-1 u,) is the
motion model, and p(zt xt) is the perception
model. We assume that both the motion and
sensor models are described by a Gaussian
distribution and the noise is modelled by the
Gaussian noise with zero mean.
1) Particle filters
The particle filter used in MCL represents a
posterior distribution p(xt zo t, uo ) by a set of
random samples drawn from this distribution.
Each particle corresponds to the robot pose (x, y,
0). Among several variants of the particle filter,
the SIR (Sampling Importance Resampling)
algorithm is adopted in this research [11]. The
approach is composed of the next three
steps;sampling,importanceweightingand
resampling. In sampling, the new sample set X',
is generated according to the motion model p(xt
x, l, uj) from the past sample set X,_l distributed
by Bel(x, 1). In importance weighting, the
importance factor ot(l) is evaluated using the
sensor model,where q is a normalization
constant. p(zt xt'(0) is calculated using the
similarity measure function is. The positional
probability is computed by the evaluation
function, which calculates range differences
between the laser scan data Zt and the expected
reference range data xt(i) computed from a grid
map. In resampling, the new sample set Xt is
randomly chosen from X't according to the
distribution defined by importance factor (o)
The importance factor o,(i) of the sample set X,
at time t is initialized to 1/N. Through the
recursive computation of three steps, the
samples converge to the pose with highest
probability.
B). TOPOLOGICAL INFORMATION
Topological Map building based on Thinning
Algorithm. Topological information is abstraction
of the environment in terms of the nodes
representing discrete places and the edges
connecting them together. The topological
information can be generated either by the GVG
method or the thinning method. The GVG method
is robust to various environments and can be
extended to the higher-dimensional space.
However, the map createsthe boundaryedgesand
weak meet points which are unnecessary in
navigation. On the other hand, the thinning
method does not create such information and is
robust to sensor noise and various environments
because it isbasedon the probabilisticframework.
A thinning method is one of the image processing
algorithms which are used to detect the skeleton
of images. Fig. 1 illustrates the concept ofthinning.
The objects on the left can be described
satisfactorily by the structure composed of
connected lines (i.e., 'T' shape drawn with thin
lines on the right). Note that connectivity of the
structure is still preserved even with thin lines. In
the case of mobile robots, the connected lines are
used as paths on which a robot travels without
colliding with other objects.
II. Model
Temporary-seed based Monte Carlo
localization (TSBMCL), including two main
parts. The first is the process of voting
temporary seeds, and the second one is the
process that the temporary seeds aid other
nodes for localization. The whole TSBMCL

algorithm is given as follows: The procedure
of voting temporary seeds is depicted below.
(a). All the seeds who are always aware of
their locations flood information in the whole
network, the data frames include
(xseed,yseed),ID Seed, TTL, denotes the
coordinate value of the seed, ID seed denotes
the seed’s ID number, and TTL denotes the
maximum flooding hops initialized by 2.
(b). All the nodes who need localization
record the coordinate value and the ID
number of the seeds whose flooding
information they can receive. If the TTL value
is 2, then the node set it 1 and forward this
data frame to other nodes. If the TTL is 1, the
node sets it 0 and never forwards it. Thus all
the seeds’ data frames could be forwarded
once at most.
After the phase of seeds’ information flooding,
every node gets an anchor box under the
algorithm of MCB, then it could further get its
weight by which the node could judge whether it
could become a temporary seed or not.
Localization would be realized.
The formula for weight computation is given as
follows:
In formula (1),TA denotes the weight
value,Sanchorbox denotes the area of the node’s
anchor box, Nnos denotes the number of seeds
within two hops, α is a factor for adjusting the
weight’s scaling. Meanwhile, two thresholds
should be set. The upper limit is TAup, and the
lower limit is TAdown. If the TA Value of one node
exceeds TAdown, and then the node is localized
under the algorithm of MCB. If the TA value
Exceeds TAup, then thisnode becomesa temporary
seed. If the TA value is less than TAdown, then this
node is not localized and it waits for the flooding
information of the temporary seeds.
III. Performance
In real world applications, we often have to
improve the localization accuracy in some partsof
the whole monitoring region according to the
position where the emergency happens. For
example, when fire takes place in the forest
demonstrates that the TSBMCL scheme could
effectively optimize the localization situation in
the mobile WSNs.
The localization failure rate of MCL/TI was
significantly lower than that of standard MCL in
Fig. 5 when the number of samples was identical.
For example, when 1000 samples were used, the
failure rate of 6% for MCL/TI is much lower than
that of 25% for the standard MCL. The reason for
the superior performance of the proposed
scheme is that the sample density of MCL/TI is
much higher than that of the standard MCL. Also,
the localization failure rate of MCL using 3000
samples was similar to that of MCL/TI using 900
samples. This result shows that the required
number of samples can be reduced to 30% of the
MCL samples for the typical office environment.

IV. Conclusion
This paper presented an accurate, range-free
localization schemes for nodes in mobile wireless
sensor networks.
As is already known, the sequential Monte Carlo
localization method works well for localization in
mobile WSNs. Basedon the sequentialMonte Carlo
method, the TSBMCL algorithm utilizes the nodes
which are localized well to aid other nodes for
localization. Our simulation results show that the
TSBMCL algorithm outperforms the MCB
algorithm by improving the localization accuracy
of the network.
V.References
[1] Ju Mei,Jinyan Gao,and Di Chen “Range-Free
Monte Carlo Localization for Mobile Wireless
Sensor Networks “Journal of IEEE international
conference on wireless sensor.
[2] Tae-Bum Kwon, Ju-Ho Yang, Jae-Bok Song,
Woojin Chung” Efficiency Improvement in Monte
Carlo Localization
Through Topological Information” Journal of
IEEE International conference of mobile wireless
sensor.
[3] Xi-Rong Bao, Shi Zhang, Ding-Yu Xue,
2008”Research on the Self-localization of
Wireless Sensor Networks”, Journal of IEEE
International conference on embedded software
and system
[4] YAO-HUNG WU AND WEI-MEI CHEN, 2009
“Localization Of Wireless Sensor Networks Using
A Moving Beacon With Directional Antenna”.,
Journal of IEEE on high performance computing
and computation
[5] Younes Ahmadi, Naser Neda and Reza
Ghazizade”Range Free Localization in Wireless
Sensor Networks for Homogenous and Non-
Homogenous Environment “

Localization of wireless sensor network

More Related Content

What's hot (16)

Viewers also liked (20)

Similar to Localization of wireless sensor network (20)

More from IRJET Journal (20)

Recently uploaded (20)

Localization of wireless sensor network