Sensor Localization in Wireless Sensor Networks via Base Station

1
Sambit Shreeram Sahoo, 2 Samir Kumar Sahoo, 3
Sandeep Kumar Pradhan, 4 Sandhyarani Paramanik, 5 Sanjibani Gachhayat
1234
Students of MCA, Raajdhani Engineering College, Bhubaneswar
5
Asst.Prof., Department of CSE & MCA, Raajdhani Engineering College, Bhubaneswar
[email protected], [email protected], [email protected], [email protected], [email protected]

ABSTRACT:- Sensors in a Wireless Sensor Network are typically deployed at random. In order
to establish the network, the position of the sensors is required to be localized. Once this activity is
complete, a routing algorithm is used to establish the network. Whilst many methods have been
contemplated in the past, this paper presents a method to find the exact location of sensors using
Received Signal Strength Indicator (RSSI) values coupled with triangulation techniques. Further,
these computations can be undertaken by the base station thereby reducing considerable energy and
computational overheads for an otherwise resource hungry wireless sensor network.

Keywords: WSN, Sensor Localization, Triangulation Technique, Energy Consumption Reduction

Introduction Use of Global Positioning System (GPS) is not a very

Wireless Sensor Networks are one of the emerging useful approach in WSNs where the number of
computing technologies which have displayed sensors are large and rather than absolute position, the
maximum research potential and are now gradually relative position with respect to a reference within or
inching from the concept stage and lab trials into to a outside the network is of interest. Research in
practical and commercially viable reality. An localization can be broadly categorized in two
application field of inhospitable environs like techniques namely the range based localization and
underground mines1 is what has been chosen as the range-free localization. The range based techniques
forerunner in the Wireless Sensor Networks use radio signal strength, triangulation and
applications. Wireless Sensor platforms are now trilateration to estimate the position of the node. In
equipped with a variety of environmental monitoring range free system, the nodes are aware if the nodes in
sensors for acoustic sensing, temperature sensing, the vicinity, their approximate physical range and
pressure sensing, humidity measurements and also location estimate. This system typically does not rely
lately video capture. Newer applications include long on radio signal strengths and do not require the
term monitoring of undersea coral reefs2, Cardiac hardware for the same. This paper concentrates on the
applications3 and wireless filter monitoring system for range free localization. In range free schemes some of
compressed air4. As the data increases, the capturing the nodes are aware of their positional data. Such
and transmitting of it becomes complex. So the nodes are termed as anchors while the rest are called
sensors must be capable of retaining the energy, as normal nodes. Anchors are static while the normal
transmitting the voluminous and complex data, and nodes may be static or dynamic. The normal nodes
must logically decide the optimal path or route for acquire the positional information of the anchor nodes
data exchange. Optimal deployment of sensors, their and then compute their own positions with the
routing, data fidelity and integrity with energy anchors. This process involves no exchange of
conservation to maximize network life thus become ranging information and hence relatively low cost
the key research areas. sensors can be utilized to establish the network. The
Methods of localization environment also is not able to disturb the
Localization is the process of identifying the connectivity information between the nodes which
position of the deployed sensors in the Region of make such a network more robust compared to the
Interest (ROI) so as to establish a routing algorithm. range based scheme. In5, a heterogeneous network is
————— considered which contains the nodes with the well-
*Author for Correspondence known location information. Here, the anchors beacon
E-mail: [email protected] their position to the neighbors. The neighbors keep

1
track of the received beacons. A simple centroid the location of the sensor nodes is crucial. In the
model is used to approximate the listening nodes present study have been undertaken using simulations.
location with the help of the proximity information. The sensors were deployed randomly in the ROI. The
This protocol is referred as centroid algorithm 6,7 make sensors have been assumed to have fully charged
the anchors flood their location information to the battery and all the sensors are assumed to have the
network rather than utilizing single hop broadcasts. A same battery capacity. The rate of change of Received
running hop count is maintained at each node along Signal Strength Indicator (RSSI) vs distance is also
the way of the broadcast. Nodes compute their assumed to be identical for all the sensors. The
location based on hop count and average distance per changes in the RSSI values with distance is measured
hop from a particular anchor. The authors term this by the real mote Figure 1. The base station is
method as DV-Hop. A similar algorithm, Amorphous connected to the laptop which receives the data from
Positioning algorithm uses offline hop-distance the other mote. For the experiment, the position of the
estimates, improving location estimations through the sender mote is changed in a range from one to fifty
neighbor information exchange. meters. This experiment is done in outdoor and the
The author has also proposed the coordinate system nesC programming in TOSSIM with the real sensors
with the available local information for ad hoc sensor Figure 2.The change in the RSSI with distance is
networks8. Mobile Anchor Positioning uses a scheme shown in Figure 3.The sensors as soon as they are
where anchor nodes are used for localization of sensor deployed transmit their identity as beacons. All the
nodes without using additional hardware9. Annealing nodes are programmed to aggregate the RSSI value
–Differential Evolution with Mobile Anchor for the nodes from which they are receiving
Positioning approach proposed by9 is also quite transmission in a Table 1. The Table 1 has node name
promising towards improving localization accuracy. and the corresponding RSSI value. This activity will
Range free localization using fuzzy logic 10 is another cease within 10 seconds of deployment and all the
approach to optimize the localization problem. The sensors will then become passive listeners. The data
anchor positioning (APIT)11 scheme is proposed, aggregated in the 10 seconds duration will be stored
which is optimal in an environment having irregular in the memory of the sensors. Once the base station
radio patterns and random node placement. The receives information on all the nodes, localization
method also has a low communication overhead.
Variety of application scenarios have been checked
also with focus on error handling and trapping. The
results illustrate that the range-free schemes are
suitable to support the various applications in field of
wireless sensor networks and they are accurate with a
slight degradation in performance.
Proposed methodology
Once the sensors are deployed, to establish a
network the initial action is to find out the positions of
these sensors. This activity is called localization. In Fig. 1 — Schematic of deployed sensors and their range with
order to make the measured data significant, finding respect to neighboring sensors

Fig. 2 — Experiment with real sensors

2
 Fig. 4 — Comparison of Average Location Error

Results
Fig. 3 — RSSI Vs Distance
The approach used for localizing the sensor mote
Table 1 — RSSI values Between Nodes uses a simple calculation for solving the equations.
A B C D E The memory requirement of this approach is
A -30 -20 -13 -40 comparatively less as it does not include complex
B -31 -23 -34 -15 calculations or requirement of the previous readings.
C -19 -18 -40 -43 The Average Location Error (ALE) is calculated as
D -14 -30 -35 -25 shown in equation one.
E -38 -17 -44 -27

∑√(K𝑐𝑎𝑙−Kr)2+
(ALE = r)2
F𝑐𝑎𝑙−F
algorithm is run to finalize the positions of all nodes. ... (1)
�
�
The positions are estimated based on mapping the computation capability is available.
sensors based on their relative RSSI values from other
sensors which can be solved as simple algebraic
equations. The Algorithm 1 used for the location of
the sensors. The base station then identifies the
clusters and cluster heads based on predefined
algorithm and transmits this information back to all
the nodes through the clusters.
Algorithm 1: Get Location - GETLOC( )
 Input : Distance based on RSSI
 Output: location of a node
1. Get the distance between the nodes based on RSSI
values
2. Form the equations based on the distance
3. Define origin
4. Solving the equations to find the location
5. Calculate the values of the location of nodes
6. Return
The energy consumption would be minimum for
localization and clustering as the sensors will be
active only for the initial phase of 10 seconds and
then only when the beacon message is received from
the nodes, the sensors again go into active
transmission. The process also enables the clustering /
selection of clusterhead activity to be undertaken by
the base station thereby reducing this overload from
the sensor nodes. Complex algorithms can also be
employed at the base station where abundant

3
 The ALE is calculated as difference of calculated
values of X and Y location as X cal& Y cal with the
real values Xr and Yr divided by the total number of
sensors nodes as N. Figure 4 shows the graph of the
ALE (m) Vs the number of sensors nodes of the
proposed method vis a vis other methods. The
method proposed in the paper removes a
considerable computational overhead from the
sensors. The entire localization and subsequent
actions like routing, identification of cluster heads
etc. is now undertaken by the base station which is
assumed to have sufficient energy and
computational resources. Further as can be seen
from Figure 4 the ALE has reduced for the proposed
method.
Conclusions
The paper presents a simplistic method of
localization of sensors using a novel method in
which the base station aggregates the RSSI values of
all sensors as soon as they are deployed. The base
station maps the sensors on the geographical area and
also undertakes clustering and cluster head selection.
This would considerably reduce the energy as well as
computation overhead of the individual sensors. A
real time and a considerably advanced and complex
control on routing and fault tolerance can be
executed through the base station. Extremely
complex algorithms also can be executed at the base
station. Real time connectivity to a remote location
through a cloud backbone also can be established.

4
86 J SCI IND RES VOL 77 FEBRUARY 2018

References 6 Osornio-Rios R A, Identification of Positioning System for

1 Singh M Marjit & Mandal J K, Logistic Regression Based Industrial Applications using Neural Network, J Sci Ind Res,
Reliability Analysis for Mobile Ad Hoc Network with Fixed 03 (2017) 141-144.
Maximum Speed and Varying Pause Times, J Sci Ind Res, 2 7 Singh M Marjit & Mandal J K, Identification of Positioning
(2017) 81-84. System for Reliability of MANET under the Influence of
2 Noonsang P, Jaroensutasinee K & Jaroensutasinee M, Black Hole Attack in Adhoc on Demand Distance Vector
Wireless Sensor Network for Ocean Observatory, J Geo Routing Protocol, J Sci Ind Res, 07 (2017) 423-426.
Mar Sci, 7 (2015) 0975-1033. 8 Roselin J & Latha P, Energy Efficient Coverage Using
3 Gowri T & Kumar P R, An Innovative Adaptive Noise Artificial Bee Colony Optimization in Wireless Sensor
Canceler Family for Cardiac Signal Filtering: Application to Networks, J Sci Ind Res, 01 (2016) 19-27.
Wireless Body Sensor Network, J Sci Ind Res, 11 (2016) 9 Sivakuma S & Venkatesan R,Performance Evaluation of
671-675. Hybrid Evolutionary Algorithms in minimising localisation
4 Ignjatovic I, Šešlija D, Tarjan L & Dudic S, Wireless sensor error for wireless sensor networks, J Sci Ind Res, (2016)
system for monitoring of compressed air filters, J Sci Ind 289-295.
Res, 5 (2012) 334-340. 10 Parwekar P, Reddy R, An efficient fuzzy localization
approach in Wireless Sensor Network, Proc Int Conf on
5 Dhanalakshmi R, Vadivel A & Parthiban P, Shortest Path
Fuzzy Systems (Hyderabad) (2013) 1098-7584.
Routing in Solar Powered WSNs Using Soft Computing
11 Liao W, Lee Y & Kedia S, Mobile Anchor Positioning for
Techniques, J Sci Ind Res, 01(2017) 23-27.
Wireless Sensor Networks, J IET Comm,7 (2011) 914-921.