i2ct_submission_105

Abstract - Wireless Sensor Networks (WSN) is rapidly
increasing area of research. In WSN, many sensor nodes are
spatially placed in particular region. These sensor nodes are
gather the information about surrounding environment like
temperature, pressure, humidity and transmit these data
packets to sink node. Reliability is one of the important
qualities of service parameter of WSN. It is consistency in
calculating results. A system will give uniform results then
we can say that it is reliable system. In this paper, effect of
reporting rate and packet size is estimated on the reliability
of a given system. Reliability is observed by different
parameters such as packet delivery ratio, packet loss ratio
against the packet size and reporting rate.
Keywords – Reliability, Packet Delivery Ratio,
Congestion, Packet Size, Reporting rate.
I. INTRODUCTION
Wireless sensor networks (WSN) consists of
thousands of sensor nodes located in given environment
in order to good judgment of humidity, pressure,
temperature and so on. These collected data is sending to
the base station via gateways [1]. Sensor nodes are having
inadequate energy while sink node is with limitless
energy. These sensor nodes are able to intellect, course
data and communicate these data to the sink node. During
the data transmission process, source node will send data
packets to sink node expected that it must be all received
at destination sink node [2]. The data packets which get
received at sink node send an acknowledgement so that
source node gets informed about successful data transfer.
But due to congestion along the path there may be
possibility of packet loss. In that case, retransmission of
lost packet took placed [3].
There are various qualities of service parameters
WSN such as reliability, congestion, throughput and lots
of. Reliability is evenness in measuring the results.
Reliability is achieved if a system will give uniform
results. Reliability is said to be high if number of packets
send by source node completely received at sink node. As
the number of packets reached at sink node increases,
reliability is also increases. Congestion is traffic in front
of sensor nodes while their data transmission process [4].
Congestion are of two types that is node-level and link-
level congestion. Throughput is the delivery of packets
from source node to sink node during certain time
duration. WSN is very beneficial than wired network
because it minimizes implementation cost and time, nodes
with low energy easily get replaced by new sensor nodes.
There are various applications of WSN as in healthcare
sector, environmental applications, home usages, military
applications, and so on [5].
Source Node
Sink Node
Sensor Nodes
Sensor field
Fig.1 Wireless sensor network
Fig.1 shows that wireless sensor networks in which
various sensor nodes are distributed in sensor field.
Source node sends data packets to sink node. The path
traced by data transfer process is shown by arrows in
fig.1. These data packets are forwarded to sink node by
particular path. Fig.1 gives clear idea regarding WSN.
II. RELATED WORK
Hamed Yousefi et. al proposed that data
aggregation is efficient method to minimize number of
packet transmission as well as it also helps to remove the
redundancy of packet data in a given environment [6].
Despite of forwarding all data packets to the sink node,
we hold the packets efficiently and allow more time for
data aggregation process. It is ultimately beneficial for
maximizing efficiency of data aggregation. In this paper
author proposed a structure free, real time data
aggregation protocol (RAG). It is used for energy saving
during data aggregation process. Our RAG protocol used
two different policies for data aggregation such as firstly
judiciously waiting policy and secondly Real time data
aware any cast policy. In judiciously waiting policy
delaying packets along the data transmission path without
missing their deadlines so that sufficient time is available
for data aggregation process. In real time data aware any
cast policy, building proper routing paths for transmission
process at MAC layer level takes placed [7]. Data
Evaluation of Reliability in Wireless Sensor Networks
Praful P. Maktedar Vivek S. Deshpande
Department of Information Technology, Department of Information Technology,
MIT College of Engineering, MIT College of Engineering,
Pune, India. Pune, India
prafulmaktedar8@gmail.com vsd.deshpande@gmail.com
International Conference on Convergence of Technology - 2014
978-1-4799-3759-2/14/$31.00©2014 IEEE 1

aggregation is valuable approach for saving energy.
Temporal and spatial convergences are the two essential
situations for data aggregation method. Therefore, it is
needed that data packets must be meet up at same node at
a same time during their transmission. To obtain high
aggregation ratio, we must spotlight on best time control
mechanisms and organization of correct routes for data
transmission [8].
Dawei Gong et al planned that in lossy wireless
sensor networks with a mobile data collector where most
of the links are unreliable. So for reliable data delivery we
must have to perform retransmission of packets. But it
results in more energy consumption. Hence it introduced
the finest method for collection of different sensor nodes
in a given environment that is clustering method [9]. In a
clustering method, nodes are grouped into clusters. Sensor
nodes are transferred data to their cluster head and from
cluster head these data packets further forwarded to the
sink node with mobile data collector. Main goal of
clustering methods are to improve energy effectiveness
which is eventually accountable for rising network
lifetime [10]. The cluster head selection is performed in
either of two ways such as, the node with minimum
selection weight among all the nodes present in network
will declare itself as a cluster head or by randomly singled
out the ID number between all the nodes presents in
network of same minimum weight [11]. All nodes are
associated with cluster head by minimum energy cost
paths. A minimum energy cost path is a sum of weights of
all links along the path.
Sudip Misra et al proposed that a simple, least
time, energy efficient routing protocol with one level data
aggregation protocol (LEO). It ensures two main criteria
for congestion control and reliability that is absolute time
and node energy [12]. In order to minimize time for data
transmission process either it took shortest path or reduces
number of retransmission by each node. The whole
transmission path may not essential to identified rather
than each node knows its neighboring nodes only. So
routing table contains the information of its neighboring
node due to which less memory is required [13]. Each
routing table contains two types of information as first,
absolute time taken by packet from node to the base
station. Second the residual energy of node. A packet is
sent from node when received at base station then an
acknowledgement is send which is further used in route
maintenance process [14]. The node will broadcast the
message that it is not capable to take part in the data
transmission process when energy of a node is less than
its threshold level. The routing tables are updated by the
nodes coming in the range of dying node and discard the
information of that dying node. So that it ensures
reliability of data transfer process [15].
III. PERFORMANCE ANALYSIS
We used Network Simulator tool (NS2) for
implementing our scenario. We used 30 different sensor
nodes with the packet size 50 bytes. Area of sensor field
is 1000*1000 sq. m. We used MAC protocol as IEEE
802.11 and Ad-hoc on demand routing protocol (AODV).
On the basis of above scenario, we plotted four graphs.
Fig.2 Number of packets as a function of reporting rate in
packets per seconds
The graph of number of packets as a function of
reporting rate in packets per seconds. It shows Packet
sends and received graph by keeping packet size same as
50 bytes. Firstly when reporting rate is 10, number of
packets send by source node is low, but as reporting rate
increases further there is continuous increase in number of
packets send. Secondly, when reporting rate is 10, number
of packets receive by sink node is low, but as reporting
rate increases further there is slowly increase in number of
packets receive refer Fig.2.
Fig.3 Packet percentage as a function of Reporting rate in
packets per seconds
Fig.3 shows the graph of Packet Delivery Ratio in
percentage as a function of Reporting rate in packets per
seconds. It shows Packet delivery ratio (PDR) graph. At
first, when reporting rate is 10 then PDR is high. But as
soon as reporting rate is increasing, there is decrease in
PDR. This is happen because due to increase in
978-1-4799-3759-2/14/$31.00©2014 IEEE 2

congestion in the network so packets do not get specific
path to reach to sink node. Hence PDR is decreases with
respect to increase in reporting rate.
Fig.4 Packet percentage as a function of Reporting rate in
packets per seconds
The graph of Packet Loss Ratio in percentage as a
function of Reporting rate in packets per seconds as
shown in Fig.4. Initially, when reporting rate is 10, PLR is
low. But as reporting rate increases there is respective
increase in PLR because due to increase in reporting rate
congestion is occurred in network and chances of packets
loss increases. Hence PLR is decrease with respect to
increase in reporting rate.
Fig.5 Number of packets as a function of Packet size in
bytes
The graph of Number of packets as a function of
Packet size in bytes. Fig.5 shows Packet sends and
received graph by keeping reporting rate same as 10
packets per seconds. Firstly when Packet size is 10,
number of packets sends by source node is almost
constant. Secondly, when Packet size is 10, number of
packets receive by sink node is low, but as Packet size
increases further there is slowly increase in number of
packets receive. But it is up to threshold point 30, after
that as packet size increases packets receive slowly
decreases.
Fig.6 Packet percentage as a function of Packet size in
bytes
The graph of Packet Delivery Ratio in percentage as
a function of Packet size in bytes. At primary state, when
packet size is 10 then PDR is low. But as soon as packet
size is increasing, there is increase in PDR. This is happen
because due to increase in packet size large amount of
information is received at sink node. But it is up to
threshold point 30, after that as packet size increases PDR
slowly decreases refer Fig.6.
Fig.7 Packet percentage as a function of Packet size in
bytes
The graph of Packet Loss Ratio in percentage as a
function of Packet size in bytes. At first, when packet size
is 10 then PLR is high. But as soon as packet size is 20,
PLR suddenly decreases and after that up to point 30 PLR
rapidly increasing as shown in Fig.7. This is happen
because due to increase in packet size huge amount of
information is collective in single packet. If packet is lost
then more amount of information is missing. But it is up
978-1-4799-3759-2/14/$31.00©2014 IEEE 3

to threshold point 30, after that as packet size increases
PLR slowly decreases
.
IV. CONCLUSION AND FUTURE WORK
In this paper, after the study of six different graphs in
performance analysis we identified some notable changes
occurred during the data transmission process. By
observing the graph of number of packets send and
receive as a function of reporting rate in packets per
seconds, number of packets send and receive increases
with the increase in reporting rate. The graph of PDR and
PLR in percentage as a function of reporting rate shows
that increase in reporting rate causes particular decrease in
PDR but particular increase in PLR. The graph of Number
of packets as a function of Packet size in bytes shows that
number of packets send is constant but at the start number
of packets received is low, as the packet size increases,
packets received increases. However it is up to threshold
point 30 after that as packet size increases packet received
decreases. The graph of Packet Delivery Ratio in
percentage as a function of Packet size in bytes shows that
increase in packet size causes increase in PDR but after
threshold point 30 as packet size increases PDR
decreases.
By Observing the graph of Packet Loss Ratio in
percentage as a function of Packet size in bytes knows
initially PLR is high but when packet size is increase then
PLR suddenly decreases up to point 20 after that PLR
suddenly increases up to point 30. A point 30 is threshold
point, after that as packet size increases PLR slowly
decreases.. These graphs show that packet size and
reporting rate are affecting on all the packets which are
available in given networks. Also congestion, energy,
throughput are other parameters which affects on PDR
and PLR. So to increase the reliability of a given system
we have to increase PDR also minimize the energy
consumed and congestion in the network. These are issues
for future works.
REFERANCES
[1] Praful P. Maktedar, Vivek S.
Deshpande,”Interpretation of Reliability in Wireless
Sensor Networks”, 2013 IEEE CUBE International
Conference on Cloud and Ubiquitous Computing and
Emerging Technologies (CUBE), pp 104-107,Nov- 2013.
[2] Mohammadreza Balouchestani, Kaamran Raahemifar
and Sridhar Krishnan,” Increasing The Reliability of
Wireless Sensor network With new Testing Approach
Based on Compressed Sensing Theory”, pp 1-4, 2011
[3] Praful P. Maktedar, Vivek S. Deshpande,J. B.
Helonde, V. M. Wadhai,” Performance Analysis of
Reliability in Wireless Sensor Networks”, International
Journal of Innovative Technology and Exploring
Engineering (IJITEE), vol 2, issues 4, pp 299-302, 2013.
[4] Hung Ta Pai, “Reliability Based Adaptive Distributed
Classification in Wireless Sensor Network”, vol 59, issues
9, pp 4543-45, 2010.
[5] Shoubhik Mukhopadhyay, Debashis Panigrahi,”Model
Based Techniques for Data Reliability in Wireless Sensor
Network”, vol 8, issues 4, April 2009.
[6] Hamed Youse_, Ali Movaghar, Mohammad Hossein
Yeganeh, Naser Ali naghipour,”Structure Free Real Time
Data Aggregation in wireless sensor networks”, Elsevier
Computer Communication, pp. 1132-1140, Nov. 2011.
[7] Gajendra S. Vyas, Vivek S. Deshpande,”Performance
Analysis of Congestion in Wireless Sensor Networks”,
2013 IEEE 3rd International Advanced Computing
Conference (IACC) ,pp 254-257, Feb.- 2013.
[8] Giuseppe Campobello, Alessandro Lenardi and Sergio
Palazzo,” Improving Energy Saving and Reliability in
Wireless Sensor Network Using Simple CRT Based
Packet Forwarding Solution”, vol 20, issues 1, pp 191-
205, 2012.
[9]Dawie Gong, Yuanyuan yang, Zhexi Pan,”Energy
Efficient Clustering in Lossy Wireless Sensor Networks”,
Elsevier Computer Communication, pp. 1323-1336, June
2013.
[10] Gajendra Sanjay Vyas, Vivek S.
Deshpande,”Performance of Congestion in Wireless
Sensor Networks Using Redundant Nodes”, 2013 IEEE
CUBE International Conference on Cloud and Ubiquitous
Computing and Emerging Technologies (CUBE), pp 73-
76, Nov- 2013.
[11]Hong Luo, Huadong Ma and Sajal K. Das,”Data
Fusion with Desired Reliability in Wireless Sensor
networks”,vol. 22, March 2011.
[12] Sudip Misra, P. Dias Thomasinous, “A Simple, least
Time, and energy efficient routing protocol with one level
Data Aggregation in wireless sensor networks”, Journal of
System and Software, Elsevier Computer
Communication, pp.852-860, 2010.
[13] Swati D. Kadu, Vivek S.
Deshpande,”Characterization of Throuput in Wireless
Sensor Networks For MAC and Routing Protocol”, 2013
IEEE CUBE International Conference on Cloud and
Ubiquitous Computing and Emerging Technologies
(CUBE), pp 108-111, Nov- 2013.
[14]R.R.Rout, S.K. Ghosh, S. Chakrabarti,”Cooperative
routing for Wireless Sensor Network using Network
Coding”, vol. 2, issues 2, pp. 75-85,2012.
[15] Swati D. Kadu, Vivek S. Deshpande,”Handling
Throuput in Wireless Sensor Networks”, 2012 IEEE
International Conference on Computational Intelligence
and Computing Research (ICCIC), pp 1-4, dec- 2012.
978-1-4799-3759-2/14/$31.00©2014 IEEE 4

i2ct_submission_105

More Related Content

What's hot (18)

Similar to i2ct_submission_105 (20)

i2ct_submission_105