ISSN : 0976-8491 (Online) | ISSN : 2229-4333 (Print) IJCST Vol.

6, Issue 2, April - June 2015

Data Aggregation in Wireless Sensor Networks (WSN)

1
Shrijana Pradhan, 2Komal Prasad, 3Kritie Chandra Gupta, 4Tannushree Bhujel
1,2,3,4
Sikkim Manipal University, INDIA

Abstract II. Data Aggregation

Wireless sensor network consists of small, light weighted sensor Data aggregation uses various approaches to aggregate the data
nodes that are positioned in distant and dense areas to monitor and sensed by the sensor nodes. Aggregation function is performed
sense various conditions in the areas of interest. The sensor nodes on the sensed data from the sensors using some approach such
arrange themselves into a network with the help of routing protocol. as LEACH (Low Energy Adaptive Clustering Hierarchy). Data
The use of sensors helps in increasing the energy efficiency but aggregation is the technique of collecting and processing the
at the same time it also increases the data redundancy. Energy data in such a way so as to reduce the energy consumption by
efficient data aggregation protocols are required for effective removal of the redundant data and hence enhances the network life
energy utilization resulting in the enhancement of the network cycle [4]. There are many WSN applications that do not require
lifetime. The protocols should be designed in an efficient way to data from individual sensor nodes but instead put more focus on
make the best use of its limited available resources. The absence of statistical quantities such as mean, variance,etc. Without the use
data aggregation protocols may lead to transmission of redundant of data aggregation technique, redundant data may increase the
data which leads to the extra energy consumption. Various data performance overhead and the collision frequency [3]. Heuristic
aggregation techniques have been proposed in wireless sensor based data aggregation protocol increases the operation of sensor
networks. This paper proposes data aggregation using the network in various traffic conditions in the network [1].
technique of mean, variance and standard deviation to decrease
the data redundancy and reduction in the data packet size sensed III. Overview of Leach Protocol
by the sensor nodes being transmitted to the Base Station and also LEACH is abbreviated for Low Energy Adaptive Clustering
compares the simulation with the LEACH protocol. Hierarchy and is one of the first cluster based hierarchical protocol.
LEACH uses multi-hop communication in which the sensor
Keywords nodes sends the data to the respective cluster head rather than
Wireless Sensor Network, Data Aggregation, Mean, Variance, directly sending the data to the base station [9]. In this protocol,
Standard Deviation, LEACH, Multi-Hop, TDMA, Direct a local cluster is formed by combining the sensor nodes. A cluster
Transmission Protocol head(CH) is chosen from the sensor nodes in a local cluster. The
main aim of the Cluster Head (CH) is to collect the data from the
I. Introduction sensor node and send it to the base station [5]. In this approach,
Wireless sensor network comprises sensor nodes deployed a node becomes a CH with a probability ‘p’ and broadcasts its
randomly or deterministically in the network to sense the events[2]. decision to all other nodes [3]. Then, each non cluster head node
The limitation of battery lifetime is one of the main constraint of determines its cluster by choosing the CH with which it has
wireless sensor networks. The main objective of the sensor nodes least communication energy. The role of cluster head (CH) is
is to sense the data and send it to the base station. The direct passed to the other nodes periodically so as to balance the load.
transmission of the data to the base station will not be efficient Hence, lifetime of the network is prolonged. This is performed
because of the distance between them. Thus, a better technique by choosing a random number by each node or threshold value
needs to be employed in order to communicate data across the T (n) between 0 and 1. If the threshold value T(n) is more than
network to the base station and prolong network lifetime which random number, the node will then become the cluster-head for
can be achieved by various routing protocols and data gathering the current round r.
mechanisms. This is done by various data aggregation techniques Where,
that are deployed to decrease the energy intake. Data aggregation is
the technique of collecting and processing the data in such a way so
as to reduce the energy consumption by removal of the redundant
data and hence enhances the network life cycle [7].Without the
use of data aggregation technique redundant data will increase the
performance overhead and the collision frequency [6]. Normally,
grouping of sensor nodes into clusters is widely adopted in large p = desired percentage of CH nodes in the sensor population
scale WSN environments. The hierarchical routing protocols such r = current round number
as LEACH (Low Energy Adaptive Clustering Hierarchy) imply G= set of nodes which have not been CH in the last 1/p round
cluster based organization of the sensor nodes on which data [5, 3, 9].
aggregation techniques can be performed. Such mechanism leads
to higher efficiency and better utilization of the wireless sensor IV. Literature Survey
network. The cluster formation process leads to a hierarchical Currently a lot of researches are being done about the wireless
network structure, where each cluster has a cluster head which sensor network protocols. New methods are being proposed in
usually performs the task of aggregation on the data that is sensed the clustering algorithm and data aggregation algorithms for
by its members (sensor nodes). Data transmission is a very energy efficient utilization of the energy of the sensor nodes. One of
consuming process therefore, it is essential to minimize number of the fundamental procedures for saving energy consumption is
packets that propagates along the network. Moreover, clustering data aggregation [4]. The energy of the sensor nodes is saved
also helps to stabilize the network topology and enhance network by eliminating redundancy with the help of data aggregation.
operation. Since most of the energy is consumed in transmission thus

w w w. i j c s t. c o m International Journal of Computer Science And Technology   233

IJCST Vol. 6, Issue 2, April - June 2015 ISSN : 0976-8491 (Online) | ISSN : 2229-4333 (Print)

data aggregation is useful for reducing battery consumption • Calculates mean (µ)
of the sensor nodes. Thesetechniques not only save the battery
consumption but also enhances the network life. To achieve high
energy efficiency clusters are formed by deploying sensor nodes • Calculates variance(v)
deterministically [1]. Each cluster has one cluster head whose aim
is to collect the data from the sensor nodes and forward it to the
base station. Even size clusters are more efficient because energy • Perform standard deviation (σ) on the received data
consumption in even size clusters is 28% and 34% for uneven
clusters [1]. It has been observed that 70% of energy is consumed
in transmission [3]. Since most of the sensed data is redundant due
to geographically assembled sensors hence energy can be saved N=no of nodes, xi=sensed value, µ=mean
by data aggregation. For organizing the network into a connected • Calculates the upper limit by performing
hierarchy an effective approach is considered to be clustering[10]. UL=mean(µ)+standard deviation(σ)
The communication capacity and the energy of the nodes in the • Calculates the lower limit by performing
network are limited but there are some factors that are necessary LL=mean(µ)-standard deviation(σ)
for the transmission of data like the energy efficiency, efficient and
real-time throughput [4]. An efficient mechanism of WSN is the Table 1: Simulation Parameter Table
Data Aggregation, which can reduce the probability of collision
Sl No. Simulation Parameters Values
and also efficiently the energy for transmission [4]. Cluster based
data aggregation protocol have some side effects one of which 1. Environment size 100*100
is latency. In data aggregation protocols latency occurs because
to save energy these protocols use the TDMA(Time Division 2. Number of nodes 100
Multiple Access) schedule. Each member of the cluster has to send
3. Number of rounds 5 to 200
the data to its cluster head only in the assigned time slot and the
cluster head waits until it receives all the data only after which it 4. Initial energy per node (Eo) 0.1J
can perform the aggregation on the data which becomes the main
cause for the latency in the network. Direct Transmission protocol Total energy of the network
5. 10J
is also one of the techniques proposed for data aggregation. Using system
this protocol each sensor nodes forwards the data directly to the
EDA(Data Aggregation
sink. This is a simple technique which requires a lot of energy 6. 5nJ
energy per bit)
in each node if the base station is far away from the nodes. Thus
the battery of the nodes get drained out quickly and the delay 7. Base station location(x,y) 50,50
also increases [11].

V. Proposed Scheme VI. Results and Discussions

In this approach, the sensors nodes are organized in an order The proposed protocol is simulated and compared with the
such that the sensor nodes form a group of clusters on which LEACH protocol.
the data aggregation technique will be performed. On receiving
the value from sensor node, the cluster head will store it in an
array. After receiving the data from the second sensor node the
cluster head will check for redundancy by comparing it with the
previously stored values so that the duplicate ones are discarded.
This process takes place for a fixed time period. When the timeout
occurs, the cluster heads calculate mean (µ) and variance (v) to
perform standard deviation (σ) on the collected data. The fused
data packet is sent over to the base station so the base station is
aware of the values being generated by the sensor nodes. This
approach ensures decreasing data redundancy at the base station
such that only critical information is forwarded along the network
which further reduces processing time, energy consumption and
efficiency of the network. Fig. 1: Energy Consumption Comparison Graph of Direct
For the proposed algorithm: Transmission Protocol, LEACH and Our Proposed Protocol
At the time of initialization, energy of each sensor node in the
network is same. Fig. 1 indicates the comparison graph of energy consumption (in
1. Cluster head selection using clustering protocol(LEACH) joules) between Direct Transmission protocol, LEACH, and our
2. Cluster Head Data Collection proposed protocol. The initial energy considered for each protocol
• The CH creates a TDMA schedule and assigns each node a is 0.1J. At round 6 the energy consumed by Direct Transmission
time slot in which it can transmit the sensed data. Protocol, LEACH and the proposed protocol are 0.002393J,
• The CH collects sensed data and performs following 0.001611J, 0.001213J respectively. From fig. 1 it is clearly shown
operations. that the proposed protocol is more energy efficient as compared
• Checks the sensed data for redundancy until timeout. to the others as it consumes less energy.

234 International Journal of Computer Science And Technology w w w. i j c s t. c o m

ISSN : 0976-8491 (Online) | ISSN : 2229-4333 (Print) IJCST Vol. 6, Issue 2, April - June 2015

VIII. Acknowledgements
We would like to extend our gratitude to our Project Guide for
giving us the right suggestion in the right direction. We would
also like to thank all the authors of the papers whom we have
mentioned in the reference section. Last but not the least, we
would like to thank our HOD, Department of CSE and SMIT for
giving us the opportunity to work in this project.

References
[1] Amrita Ruperee, ShikhaNema, Sanjay Pawar”, Achieving
Energy Efficiency and Increasing Network Life in Wireless
Sensor Network”, IEEE,2014
[2] Bala Krishna, Noble Vashishta”, Energy Efficient Data
Fig. 2: First Node Dead Comparison Graph Aggregation Techniques in Wireless Sensor Networks",
International Conference on Computational Intelligence and
Fig. 2 shows the first node dead graph between LEACH and the Communication Networks, 2013.
proposed protocol. The graph shows that for 100 nodes the first [3] Divya Prabha, Vishal Kumar Arora,"A Survey on LEACH
node dies at round number 183 in case of LEACH and at 186 in and its Descendant Protocols in Wireless Sensor Network”,
case of the proposed protocol. International Conference on Communication, Computing &
Systems, 2014.
[4] JiPeng, Li Yupeng, Jiang Jingqi, Wang Tianbao,"A Clustering
Protocol for Data Aggregation in Wireless Sensor Network”,
International Conference on Control Engineering and
Communication Technology, 2012.
[5] Meenakshi Sharma, Kalpana Sharma,"An Energy Efficient
Extended LEACH (EEE LEACH)”, International Conference
on Communication Systems and Network Technologies,
2012.
[6] Miloud Bagaa, NoureddineLasla, AbdelraoufOuadjaout,
YacineChallal,"SEDAN: Secure and Efficient protocol
for Data Aggregation in wireless sensor Networks”, IEEE
Conference on Local Computer Networks, 2007.
[7] MousamDagar, Shilpa Mahajan,"Data Aggregation in
Fig. 3: Transmission Time Comparison Graph Wireless Sensor Network: A Survey”, International Journal
of Information and Computation Technology, 2013.
Fig. 3 shows that the Transmission time in case of the proposed [8] Nandini. S. Patil, Prof P.R.Patil,"Data Aggregation in Wireless
protocol is greater than that of the LEACH protocol because the Sensor Network”, International Conference on Computational
proposed protocol performs aggregation using mean and standard Intelligence and Computing Research, 2010.
deviation to decrease the data packet size. Hence, in case of the [9] N. Thangadurai, R.Dhanasekaran,"Energy Efficient Cluster
proposed technique the cluster head instead of sending all the data based Routing Protocol for Wireless Sensor Networks”,
sensed by the sensor nodes , it will just send the upper range and International Journal of Computer Applications, 2013.
lower range of the values sensed by the sensor nodes. [10] Ossama Younis, Marwan Krunz, Srinivasan
Ramasubramanian,“Node Clustering in Wireless Sensor
VII. Conclusion Networks: Recent Developments and Deployment
In this paper, we considered LEACH protocol which is a wireless Challenges", IEEE, 2006.
sensor network routing protocol and proposed a new data [11] S Nithyakalyani, Dr. S. Suresh Kumar,"Data Aggregation in
aggregation technique by using LEACH protocol for clustering. Wireless Sensor Network using node Clustering Algorithms-
Then our data aggregation protocol is successfully simulated and A Comparative Study”, IEEE Conference on Information
compared with LEACH protocol. From the simulation results we and Communication Technologies, 2013.
can draw the conclusion that this protocol is more energy-efficient [12] SuatOzdemir, Yang Xiao,"Polynimial Regression Based
than LEACH with greater network lifetime. We discuss regarding Secure Data Aggregation For Wireless Sensor Network”,
the energy constraint in wireless sensor network. Replenishment IEEE Communication Society, 2010.
of energy for sensor node is a big challenge in WSN as these
nodes once deployed may not be physically accessible. Large
data packet transmissions increase overhead and consumes more
energy. This motivates for design of efficient data aggregation
technique. To employ efficient data aggregation technique that
can collect and model data into a construct with reduced packet
size and data redundancy is the aim of the project. This data
aggregation technique is expected to produce satisfactory results
for the optimal use of the battery power of the sensor nodes
deployed in an area.

w w w. i j c s t. c o m International Journal of Computer Science And Technology   235

IJCST Vol. 6, Issue 2, April - June 2015 ISSN : 0976-8491 (Online) | ISSN : 2229-4333 (Print)

Shrijana Pradhan received her B.E

degree in Computer Science and
Engineering from Government
Engineering College, Bilaspur (CG)
in 2011.She is currently an Assistant
Professor II in the Department of
Computer Science and Engineering
at SMIT, SMU. Her research interests
include wireless sensor networks and
graph theory.

Komal Prasad is pursuing her B.Tech

degree in Computer Science and
Engineering from Sikkim Manipal
Institute of Technology, Majitar,
Sikkim, India from 2011. Her research
interests include wireless sensor
networks.

Kritie Chandra Gupta is pursuing her

B.Tech degree in Computer Science
and Engineering from Sikkim Manipal
Institute of Technology, Majitar, Sikkim,
India from 2011. Her research interests
include wireless sensor networks.

Tannushree Bhujel is pursuing her

B.Tech degree in Computer Science
and Engineering from Sikkim Manipal
Institute of Technology, Majitar,
Sikkim, India from 2011. Her research
interests include wireless sensor
networks.

236 International Journal of Computer Science And Technology w w w. i j c s t. c o m