40120130406001

International Journal of Electronics and JOURNALEngineering & Technology (IJECET), ISSN 0976 –
INTERNATIONAL Communication OF ELECTRONICS AND
6464(Print), ISSN 0976 – 6472(Online) Volume 4, Issue 6, November-December (2013), © IAEME

COMMUNICATION ENGINEERING & TECHNOLOGY (IJECET)

ISSN 0976 – 6464(Print)
ISSN 0976 – 6472(Online)
Volume 4, Issue 6, November-December, 2013, pp. 01-13
© IAEME: www.iaeme.com/ijecet.asp
Journal Impact Factor (2013): 5.8896 (Calculated by GISI)
www.jifactor.com

IJECET
©IAEME

AN EFFICIENT CLUSTER HEAD SELECTION SCHEME FOR DYNAMIC
SLEEP TIME IN WIRELESS SENSOR NETWORK
Rashid M. Awadi1, Rawya Y. Rizk2, Mohamed I. Habib2, and Amira A. M. Elsonbaty3
1

Dept. of Communication, Faculty of Engineering, Mansoura University, Egypt.
2
Electrical Engineering Department, Port Said, Egypt.
3
Higher Institute of Computer and Business Administration, El zarka, Damietta, Egypt.

ABSTRACT
Energy consumption is the key challenge in the Wireless Sensor Networks (WSNs). Recent
advances in WSNs enable us to develop minimum energy consumption clustering algorithms.
Clustering is an effective way for reducing energy consumption of sensor nodes as well as the cost of
transmission. One of the other main sources of energy waste in a WSN is idle listening, i.e., nodes
consuming energy to sample an idle channel. The sleep time is an approach utilizes known traffic
statistics and optimally controls the sleep interval between consecutive wake-ups of the receiver so
that the expected total energy spent during each transmission is minimized. This paper presents a
complete solution of the energy consumption problem in WSNs. It combines a dynamic sleep time
approach, and an efficient cluster head selection scheme to save energy in WSNs. Simulation results
are included to compare fixed sleep times to dynamic control policy time. This study classifies the
lifetime into different types and gives the corresponding cluster head selection method to achieve the
life-time extension objective. Simulation results show that the energy is significantly reduced
compared with the previous clustering based routing algorithm for the sensor networks.
Keywords: Data aggregation, Clustering based algorithms, dynamic sleep, Energy efficiency,
Wireless Sensor Networks, WSNs.
1.

INTRODUCTION

Due to existing and emerging applications in various situations, wireless sensor networks
(WSNs) have recently emerged as a premier research topic [1, 2]. A WSN consists of a number of
small-sized sensor nodes spreading over a geographical area and a sink node where the end user can
access data. All nodes are equipped with capabilities of sensing, data processing, and communicating
with each other by means of a wireless ad hoc network. A wide range of tasks can be performed by
these tiny devices, such as condition-based maintenance and the monitoring of a large area with
respect to some given physical quantity [3].
1

International Journal of Electronics and Communication Engineering & Technology (IJECET), ISSN 0976 –

One of the limitations of wireless sensor nodes is their inherent limited energy resource.
Besides maximizing the lifetime of the sensor node, it is preferable to distribute the energy dissipated
throughout the WSN in order to minimize maintenance and maximize the overall system
performance.
To support scalability, nodes are often grouped into disjoint clusters. Each cluster would have
a leader, often referred as cluster head (CH). A CH is responsible for not only the general request but
also assisting the general nodes to route the sensed data to the target nodes. The power-consumption
of a CH is higher than of a general (non-CH) node. Therefore, the CH selection will affect the
lifetime of a WSN.
Low Energy Adaptive Clustering Hierarchy (LEACH) is a well known clustering based
protocol that utilizes randomized rotation of local cluster base stations to evenly distribute the energy
load among the sensors in the network. LEACH uses localized coordination to enable scalability and
robustness for dynamic networks, and incorporates data aggregation into the routing protocol to
reduce the amount of information that must be transmitted to the base station. The cluster heads are
randomly chosen in order to randomize the distribution of the energy consumption and load among
the sensors, and therefore taking the first step towards evenly distributing the energy consumption
through the system’s lifetime. It has the disadvantage of irrationality of choosing cluster head and of
unbalanced energy consumption of nodes [4].
The sleep time is an approach utilizes known traffic statistics and optimally controls the sleep
interval between consecutive wake-ups of the receiver so that the expected total energy spent during
each transmission is minimized.
This paper proposes a data aggregation operation based on a scalable cluster architecture
whose basic idea is as follows: First, position information is used to select some sensors as CHs.
Then further it organizes those nodes into clusters. As in LEACH approach, which is broken up into
rounds, where each round begins with a set-up phase when the clusters are organized, followed by a
steady-state phase when data transfers to the base station occur. In order to minimize overhead, the
steady-state phase is long compared to the set-up phase. In each cluster, dynamic sleep time is
applied with data aggregation.
In this paper, a complete solution of the energy consumption problem in WSNs is presented.
An efficient cluster head selection scheme to select the cluster-heads based on the required energy to
do the transmission to the sink is presented. A dynamic sleep time approach is also proposed to save
energy in WSNs.
The rest of this paper is organized as follows. Section 2 presents the proposed approach.
Section 3 presents the simulation results of the proposed protocol and comparisons with the closely
related protocols. Finally, Section 4 introduces the main conclusions and the future work.
2.

THE PROPOSED SOLUTION TO ENERGY EFFICIENT PROBLEM

One of the most important things that affect the performance of the system and to reduce the
energy is the optimal choice for the CH, according to a number of factors (residual energy, distance,
etc,). Another energy saving solution is to use a sleep time. This paper suggest two algorithms. First
algorithm for head selection is the Associated Cluster Head Array (ACHR). The second algorithm is
the Dynamic Sleep time for Aggregationed Data (DSDA). It uses dynamic sleep time rather than
fixed sleep time to reduce the time wasted and thus wasted energy.
2-1. The Proposed ACHR Algorithm
This paper applies head selection algorithm where the whole network is separated into a
number of clusters. Each cluster has a CH. Fig (1) shows the pseudo code of the head selection
algorithm.
2


In set-up phase, the sensor nodes form the constant cluster and the cluster-head elected based
on the residual energy of the individual node calculation with constant clustering and the node
scheduling scheme is adopted in each cluster. In this paper, static clusters are formed using virtual
points whose locations are determined by the sink. First of all, all sensor nodes check their locations
using the low-power GPS device. Then, they send their coordinates to the sink. The CH
communicates with the local base station, then the local base station feed data to the entire network
of base stations, and terminal user can access useful information.
The cluster formation of the proposed approach aims at balancing the energy load of CHs.
Each CH is responsible only for received data from the cluster members, performs aggregation
process over the received data and then to the BS. BS chooses the cluster heads according the current
information of nodes at the beginning of each round.
In each cluster, termed as Ci, one of sensor nodes in Ci is elected as the cluster head CHi. A
CH is responsible for receiving the sensed data of the other sensor nodes in cluster and routing to the
sink. The CH is selected from the sensor nodes in the same cluster, where the selection is performed
round-by-round. Therefore, the energy consumption of the CH is higher than of the other nodes. In
order to balance the energy consumption for elongating the lifetime of this WSN, the CH in a cluster
is alternate among sensor nodes. Therefore, the CH selection manner will affect the lifetime of this
network.
When a residual energy of CH node falling below the threshold value, it triggers a new CH
candidacy event by informing the BS that it is unable to perform its duties as a CH any more.
Subsequently the BS will inform this to all other CHs thus initiating a CH rotation phase. The next
round cluster head is selected in advance to avoid the deadlock when an old cluster head dies.
The proposed algorithm suggested associated CH array which contains Node- ID and
Location-ID. It orders decreasing by residual energy of sensors. Cluster nodes gathering data from
environment and send it to the CH. While processing of Cluster head node the energy is reduced. So
if the energy of CH becomes below to the non-cluster head nodes energies means next round should
be processed. ACHR calls the next CH for each cluster to start the new round. In this algorithm the
formation cluster and members of each cluster is fixed. It doesn't need re_clustering.
Pseudo code 1: The ACHR algorithm
1 Setup ( )
2 Sensor nodes send information about its current location to the base station
3 Geographical group the data
4 Determining good clusters
5 In each cluster, all sensor nodes sent energy level to BS
6 If sensor node with highest energy level then
7 Choose CH and unicast the information to the member node
8 Else
9 Choose nodes as member node
10 End if
11 Nodes send data to CH (transmission uses a minimal energy)
12 The radio of each non CH can be turned off until the nodes allocated transmission time
13 CH performs data aggregation
14 Aggregated data is sent to BS (transmission takes high energy)
15 Decreasing ordered array according the energy level
16 If (CH's residual energy < the threshold value)
17 New round without selection the CH using ACHR
18 New rounds without re_clustering
19 With each new round choose the next ID.
20 End.
Fig (1): Pseudo code 1: Head selection algorithm.
3


Energy Consumption
This paper assumes that nodes operate in a periodic sleep schedule where each sleep period
consists of a sleep interval and a wake-up interval. The sleep interval is the time duration when the
node’s radio is off in each sleep period. The wake-up interval is the time duration that a node has its
radio on to transmit packet.
This paper applied a simple model of [5] for the radio hardware energy dissipation where the
transmitter dissipates energy to run the radio electronics and the power amplifier, and the receiver
dissipates energy to run the radio electronics. Both the free space and the multipath fading channel
models were used, depending on the distance between the transmitter and receiver. Power control
can be used to invert this loss by appropriately setting the power amplifier, if the distance is less than
a distance threshold, the free space (fs) model is used; otherwise, the multipath (mp) model is used.
Fig. (20 show the four major components and associated energy cost parameters of a typical sensor
node.
To compute the energy consumption, the equations (1-10) are considered. Table (10 presents
the parameters used in the head selection algorithm. The following assumptions are considered in the
network model.
• All micro-sensor nodes deployed within a square area are homogeneous.
• All micro-sensors and the BS are stationary after deployment.
• All micro-sensor nodes use only the initial battery power and are not recharged.
• All micro-sensor nodes are location-aware (e.g., sensor nodes are equipped with GPS receivers or
location detection devices).
Table (1): The Parameters of the head selection algorithm.
L
‫ܧ‬௘௟௘௖
‫ܧ‬௔௠௣
‫ܧ‬௙௦

‫ܧ‬஽஺
‫ܧ‬௜௡௧௘௥
‫்ܧ‬௫௜,௝

‫ܧ‬ோ௫௜
ECH
Emem
Ecluster
N
N
D
d0
ଶ
݀௧௢஼ு

The number of bits in each data message
The per bit energy dissipation for running the transceiver circuitry
The amplifier parameter for the multipath propagation
The amplifier parameter for the free space propagation
The energy for data aggregation
The energy dissipation of data transmission from the cluster head to
the base station
The energy dissipation for transmitting an l bit message from the
transmitter to receiver
Energy required for receiving an l bits message
The energy dissipation of the cluster head ECH
The energy consumed by one member node
The average energy dissipated in cluster
Number of nodes
Number of clusters
Distance from the transmitter to receiver
Threshold distance
Distance from the member node to the cluster head

4


Fig (2): Four major components and associated energy cost parameters of a typical sensor node.
‫א‬௙௦

(1)

d0=ට‫א‬௔௠௣

ସ
‫்ܧ‬௫௜,௝ ൌ ‫ܧ‬௘௟௘௖ . ݈ ൅ ‫ܧ‬௔௠௣ . ݀௜.௝ . ݈ (Multipath d<d0)

(2)

ଶ
‫்ܧ‬௫௜,௝ ൌ ‫ܧ‬௘௟௘௖ . ݈ ൅ ‫ܧ‬௙௦ . ݀௜.௝ . ݈ (Free space d≥d0)

(3)

ECH=݈‫ܧ‬௘௟௘௖ ቀ ௡ െ 1ቁ ൅ ݈‫ܧ‬஽஺ ௡ ൅ ‫ܧ‬௜௡௧௘௥

(5)

‫ܧ‬ோ௫௜ ൌ ݈. ‫ܧ‬௘௟௘௖
ே

(4)

ே

ଶ
‫ܧ‬௠௘௠ ൌ ‫ܧ‬௘௟௘௖ . ݈ ൅ ‫ܧ‬௙௦ . ݀௧௢஼ு . ݈

Ecluster=ECH+(

ே

௡

Etotal =n Ecluster

(6)

െ 1ሻ‫ܧ‬௠௘௠

(7)

The average area of the cluster is approximately
The area is Square=m2….m=ට ௡

௅మ

௅మ
௡

(8)
(9)
(10)

2-2. the proposed DSDA algorithm
This work dynamically adjusts the sleep interval of each node based on the delay constraint
and the network condition changes. The dynamic sleep time between following channel polling uses
available statistical network traffic information and prior observations. It allows the sleep time to
dynamically change in each sleep interval according to a message sent to the receiver.
In [6], a dynamic sleep time control method is presented. However, the sleep time obtained is
a conservative one since it guarantees the expected delay conditioning on the arrival occurs during
the sleep time, which is a strict constraint. Also this approach only focuses on the next message. In
this paper, a dynamic sleep time control is proposed where it can batch a few messages together and
share the same preamble, and overcome expected delay constraint. The original ﬁxed sleep time is
divided into two parameters: a fixed “expected delay” and the corresponding dynamic sleep time.
Table (2) presents the list of parameters used in the dynamic sleep algorithm.
5


Table (2): The list of parameters used in the dynamic sleep algorithm.
Variables

Definition

F(x)

Uniform distribution function

Zt

Sleep time

D

Random preamble delay denoted

[a, b]

Inter_arrival time distribution is uniform

V

Max (t, a)

data Time (i)

Interarrival time

time2listen(i),
time2sleep (i)

The time remaining for a node to be in the listen or sleep state

Tagg

The actual aggregation period Taggr

Τ

Deferred period for aggregation

cycle Time

The duration of the frame (cycle)

Sleep _ state

Nodes have turned off their radio for communication;
Can still sense and process data;
Nodes cannot transmit and receive packets from other nodes in the
network.

Listen _state

Nodes send and receive data packets.
During that time aggregation can be performed.

The following equations are used in the dynamic time sleep control. A uniform distribution U
[a, b], where 0 ≤ a < b is considered [7].
z୲ =ቐ

ഥ
2D ൅ v െ t,

ഥ ൑ ଵ ሺb െ vሻ
D ଶ

ଵ
ഥ ଵ
D ൅ ଶ ሺb െ tሻ ൅ ଶ ሺv െ tሻ,

0,
‫ۓ‬
ۖx െ a
Fሺxሻ ൌ
,
‫۔‬b െ a
ۖ
‫ە‬
1,

x൏ܽ

(11)

ഥ ൐ ଵ ሺb െ vሻ
D ଶ
(12)

a൑x൏ܾ

x൒b

Data Aggregation
Data aggregation aims to combine responses from multiple sensors into a single message.
Reducing the number of messages transmitted in a network can greatly reduce the amount of energy
consumed [8, 9]. Fig. (2) shows the Pseudo code that determines the values of the data aggregation
period, Taggr, for the different states and cases.

6


Pseudo code 2: Data aggregation time calculation.
1 BEGIN
2 data Time (i)
3 time2listenTime (i)
4 time2sleep (i)
5 Tagg = 0
6 If τ > listen Period AND status = sleep
7 Tagg = time2listenTime + cycle Time
8 else if τ > listen Period AND status = listen
9 Tagg = time2listenTime
10 else if τ < listen Period AND status = sleep
11 Tagg = time2listenTime + τ
12 else if τ < listen Period AND status = listen
(if sufficient time to be sent in the same cycle)
13 Tagg = τ
14 Else Tagg = time2listenTime
15 END
Fig (3): Pseudo code 2: The data aggregation time calculation.
3.

SIMULATION RESULTS

3-1. ACHR Algorithm
This section presents the performance of the proposed head selection algorithm. The
simulation is done with different numbers of nodes randomly deployed between (x=0, y=0) and
(x=100, y=100) by meter. Simulation carries out a comparison between the proposed algorithm, and
the efficient cluster head selection scheme for data aggregation (ECHSSDA) that was presented in
[10].
Table (3) shows the required time needed for completing both ACHR algorithm and
ECHSSDA algorithm. It is clear that the proposed ACHR greatly saves the time with respect to the
ECHSSDA algorithm.
Table (3): The required time needed for completing ACHR algorithm and ECHSSDA algorithm.
N_nodes

Algorithm

100

500

100

500

100

ECHSSDA

1000

2400

3800

5200

1000

100

1000

100

ECHSSDA

2000

2400

6400

8600

1600

100

1600

100

ECHSSDA

2000

4200

6400

8600

1800

100

1800

100

ECHSSDA

3600

7200

1080

14400

ACHR

80

cluster4

ACHR

60

cluster3

ACHR

40

cluster2

ACHR

20

cluster1

ACHR

2500

100

2500

100

ECHSSDA

14400

10800

7200

3600

7


Figures (4-6) show that the total required time spent in ECHSSDA algorithm is larger than
the total required time spent in ACHR algorithm. Only one cluster is created in ACHR algorithm and
remains constant throughout the time so it does not consume extra time with each cycle for the
establishment of a group as it happens with ESSCHDA with more required time.

Fig (4): The required time (ARCH.vs.ECHSSDA) for 20 nodes.


8


3-1. DSDA Algorithm
This section presents the performance of the proposed DSDA algorithm with uniform
distribution function. In this simulation, our experiment model performed on 100 nodes which were
randomly deployed and distributed in a 100×100 m2 area. Sensor nodes contain two kinds of nodes:
sink nodes (no energy restriction) and common nodes (with energy restriction). All nodes have no
mobility since the nodes are fixed in most of the applications of WSNs. Aggregation time were
applied in dynamic sleep time equations. Long preamble time is double the event's arrival times, but
short preamble is the event's arrival time.
Tables (4) and (5) present the sleep time control obtained under uniform distribution (short long) variable preamble. Figures (7-10) show how applied the principle of dynamic sleep time is
much better than fixed sleep time. As well as, the use of long or short variable preamble is better
than fixed preamble periods because it helps to reduce power consumption and also idle times. It is
also shown that the principle of data collection helps to enlarge sleep time so reduce the energy
consumed in the data transfer.
Table (4): The sleep time control obtained under uniform distribution (short - long) variable preamble.
Zt(sleep)
AGE (Ti)
4.5
46
22
38
36
51
16
19
59
30
54
12
39

Long preamble
µs(2×y)
Downlink
9
92
44
76
72
102
32
38
118
60
108
24
78

Short preamble
µs(y)
Uplink
4.5
46
22
38
36
51
16
19
59
30
54
12
39

9

short preamble

Long preamble

59.5
59.5
59.5
59.5
59.5
59.5
59.5
59.5
59.5
59.5
59.5
59.5
59.5

64
105.5
81.5
97.5
95.5
110.5
75.5
78.5
118.5
89.5
113.5
71.5
98.5


Fig (7): The sleep time control obtained under uniform distribution.

Fig (8): The sleep time control obtained under uniform distribution (short - long) variable preamble.

10


Table (5): The sleep time control obtained under uniform distribution (aggregation messages) for
(short-long) variable preamble.

AGE
(Ti)
0.19
9.5
4.5
46
22
38
36
7.55
51
16
19
59
30
54
12
39
31

Long
preamble
µs(2×y)
Downlink
0.38
19
9
92
44
76
72
15.1
102
32
38
118
60
108
24
78
62

Short
preamble
µs(y)
Uplink
0.19
9.5
4.5
46
22
38
36
7.55
51
16
19
59
30
54
12
39
31

Tagg

long preamble

short preamble

30.19
39.5
34.5
210
52
210
210
37.55
210
46
49
210
60
210
42
210
210

89.88
108.5
98.5
315.5
133.5
307.5
305.5
104.6
320.5
121.5
127.5
328.5
149.5
323.5
113.5
308.5
300.5

89.69
99
94
269.5
111.5
269.5
269.5
97.05
269.5
105.5
108.5
269.5
119.5
269.5
101.5
269.5
269.5

Fig. (9): The sleep time control obtained under uniform distribution (aggregation messages) for
(short-long) variable preamble.

11


Fig. (10) shows different cases of sleep time control obtained under uniform distribution.
4.

CONCLUSION

This paper provides a complete solution of energy consumption in WSNs. Two algorithms
are proposed. First one for head selection is the Associated Cluster Head Array (ACHR) that helps to
reduce the energy due to the optimal choice for the CH. The second algorithm is the Dynamic Sleep
time for Aggregationed Data (DSDA). It uses dynamic sleep time rather than fixed sleep time to
reduce the time wasted and thus wasted energy.
Simulation results conclude that the total required time spent in the proposed ACHR
algorithm is much smaller than it in the case of ECHSSDA algorithm. Simulation results also show
that the proposed dynamic sleep time with data aggregation model consumes less power less than a
fixed sleep time in WSNs, either with single message, or aggregation message. Use long preamble
period gives a 10% longer than short preamble period, but the use of long periods of preamble
without purpose gives more consumption energy.
Future versions of this work will include an energy-based threshold to account for nonuniform energy nodes. The duration of the steady-state phase, and the complexity of the environment
will be researched in the future.

REFERENCES
[1]
[2]
[3]

A. Willig, K. Matheus, and A.Wolisz, "Wireless technology in industrial networks",
Proceedings of the IEEE, vol. 93, no. 6, pp. 1130–1151, 2005.
A. Wheeler, "Commercial applications of wireless sensor networks using zigbee", IEEE
Communications Magazine, vol. 45, no. 4, pp.70 –77, 2007.
X. Jia, J. Wu, and Y. He," Minimum Data Aggregation Time Problem in Wireless Sensor
Networks",MSN 2005, LNCS 3794, pp. 133–142, 2005.

12


[4]

[5]
[6]
[7]
[8]

[9]

[10]

[11]

[12]

[13]

[14]

[15]

Stella.k, Symeon .p, "Energy efficient framework for data gathering in wireless sensor
network via the combination of sleeping mac and data aggregation strategies", Int. J. Sensor
Networks, vol. 10, nos. 1/2, 2011.
Z. Wei-hua, L. La-yuan, Z.Liu-min, W. Xuan-zheng, "Energy Consumption Balance
Improvement of LEACH of WSN", Chinese Journal of Sensors and Actuators, 2008.
X. Ning, Christos, and G. Cassandras "Dynamic sleep time control in wireless sensor
networks", ACM Transactions on Sensor Networks, vol. 6, no. 3, June 2010.
J. S. Chen, Z. W. Hong, N. Chung, "Efficient Cluster Head Selection Methods for Wireless
Sensor Networks", journal of networks ,vol. 5, no.8, 2010.
I. Solis, K. Obraczka. “In-network aggregation trade-offs for data collection in wireless
sensor networks,” Computer engineering department, University of California, Santa Cruz.
August 2003.
K. Maraiya, K. Kant, N. Gupta I "Efficient Cluster Head Selection Scheme for Data
Aggregation in Wireless Sensor Network", International Journal of Computer Applications,
vol. 23, no.9, June 2011.
R. Saravanakumar, S. G. Susila, and J. Raja, "Energy efficient constant cluster node
scheduling protocol for wireless sensor networks", in. Proc. of Wseas Transactions on
Communication, vol. 10, no. 4, April 2011.
Syed Abdul Sattar, Mohamed Mubarak.T, Vidya Pv And Appa Rao, “Corona Based Energy
Efficient Clustering In WSN” International Journal of Advanced Research in Engineering &
Technology (IJARET), Volume 4, Issue 3, 2013, pp. 233 - 242, ISSN Print: 0976-6480, ISSN
Online: 0976-6499, Published by IAEME.
Mr.Yogesh V Patil, Mr. Pratik Gite and Mr.Sanjay Thakur, “Automatic Cluster Formation
and Assigning Address For Wireless Sensor Network”, International Journal of Computer
Engineering & Technology (IJCET), Volume 4, Issue 4, 2013, pp. 116 - 121, ISSN Print:
0976 – 6367, ISSN Online: 0976 – 6375, Published by IAEME.
Wategaonkar D.N and Deshpande V.S., “On Improvement of Performance For Transport
Protocol Using Sectoring Scheme In WSN”, International Journal of Computer Engineering
& Technology (IJCET), Volume 4, Issue 4, 2013, pp. 275 - 281, ISSN Print: 0976 – 6367,
ISSN Online: 0976 – 6375, Published by IAEME.
Bharathi M A, Vijaya Kumar B P , Manjaiah D.H, “Power Efficient Data Aggregation Based
on Swarm Intelligence and Game Theoretic Approach In Wireless Sensor Network”,
International Journal of Computer Engineering & Technology (IJCET), Volume 3, Issue 3,
2012, pp. 184 - 199, ISSN Print: 0976 – 6367, ISSN Online: 0976 – 6375, Published by
IAEME
Mohanaradhya, Sumithra Devi K A and Andhe Dharani, “Distance Based Cluster Head
Section in Sensor Networks For Efficient Energy Utilization” International Journal of
Advanced Research in Engineering & Technology (IJARET), Volume 4, Issue 1, 2013, pp.
50 - 58, ISSN Print: 0976-6480, ISSN Online: 0976-6499, Published by IAEME

13

40120130406001

More Related Content

40120130406001