New Approach for Selecting Cluster Head based on LEACH Protocol
for Wireless Sensor Networks
Wided Abidi and Tahar Ezzedine
University of Tunis El Manar, Engineering School of Tunis, Communications Systems Laboratory, Tunis, Tunisia
Keywords: Wireless Sensors Networks, LEACH Protocol, Cluster Head Election, Energy Efficiency, Network
Lifetime.
Abstract: From hierarchical routing protocols, Low-Energy Adaptive Clustering Hierarchy (LEACH) has been
considered as one of the effective algorithms that optimize energy and prolong the lifetime of network. In
this paper, we propose a new approach of electing Cluster Head (CH) based on LEACH protocol. The
selection of Cluster Head (CH) in LEACH is carried out randomly. In our proposed approach, we consider
three fundamental criteria: the remaining energy, the number of neighbours within cluster range and the
distance between node and CH. In fact, in our algorithm, we include these factors in calculation of
threshold. Simulation results shows that our proposed approach beats LEACH protocol in regards of
prolonging the lifetime of network and saving residual energy.
1 INTRODUCTION
Wireless Sensor Network (WSN) consists of large
number of tiny devices called sensor nodes (Anastasi
et al., 2008). These Nodes are deployed randomly in
a geographical area. Their roles are to sense, collect,
aggregate and send data between each other or to a
Base Station (BS) located outside of the sensor area.
This communication costs important energy
consumption. On the other hand, sensor nodes use
batteries as power source that are limited resources.
In addition, this power source is usually not
replaceable or rechargeable. Hence, the need to
extend the lifetime of nodes and minimize the
energy consumption is necessary.
Due to the energy constraints of the large number of
deployed sensors, routing in WSN becomes very
challenging and many routing protocols have been
developed (Al-Karaki and Kamal, 2004). In
hierarchical routing protocols, network is divided in
a number of clusters. In each cluster, there is only
one node that communicates with the BS called
Cluster Head (CH). By selecting a CH, the routing
overhead of non CH nodes is reduced since these
nodes have only to send data to CH. These protocols
use data aggregation and fusion in order to reduce
the number of transmitted messages to the BS. And
furthermore, all nodes have a chance to be a CH (Al-
Karaki and Kamal, 2004), (Katiyar, 2011). From
hierarchical routing protocols, Low Energy Adaptive
Clustering Hierarchy (LEACH) (Heinzelman et al.,
2000), (Heinzelman, 2000) is one of the most
famous protocols that use dynamic clustering. We
will give an overview of this protocol and its
shortcomings in the following section. In this paper,
we propose a new approach for selecting CH to
avoid some deficiencies of LEACH protocol. Since
LEACH does not take into account remaining
energy of the node and the distance between node
and BS in choosing the CH. In our new algorithm,
the selection of CH is based on three factors:
residual energy, distance between the CH and sink
and the number of neighbor nodes within the cluster
range. Thus, elected CH must have at the same time
a high residual energy, maximum number of
neighbor and finally a low distance to sink. By
considering these factors, we can save energy
consumption and prolong the lifetime of the network
and good results will be shown by simulations later
in the paper.
The rest of the research work is organized as
follows. Related work is presented in section 2
Section 3 details the proposed algorithm to select
CH. Simulation results are shown and discussed in
section 4. We conclude in section 5.
114
Abidi, W. and Ezzedine, T.
New Approach for Selecting Cluster Head based on LEACH Protocol for Wireless Sensor Networks.
DOI: 10.5220/0006336101140120
In Proceedings of the 12th International Conference on Evaluation of Novel Approaches to Software Engineering (ENASE 2017), pages 114-120
ISBN: 978-989-758-250-9
Copyright © 2017 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
2 RELATED WORK
Over recent years, various hierarchical protocols and
algorithms are developed to enhance the energy
efficiency in WSN. In this section, we give an
overview of some of them.
Low Energy Adaptive Clustering Hierarchy
(LEACH) has been introduced by Heinzelman, et al.
(2000) to reduce power consumption. LEACH
divides network to clusters and only one node (CH)
in each cluster is the leader and it changes each
round. CH communicates directly with the BS to
send data and uses data aggregation technique what
reduce energy consumption and prolong the lifetime
of the WSN.
LEACH centralized (LEACH-C) has been
proposed also by Heinzelman (2000), (Geetha et al.
2012). This is a centralized clustering algorithm. It
uses the BS to elect CHs. In fact, the BS receives
information about the position and energy level of
each sensor node in the WSN. Then, BS elects a
number of nodes as CH for each round and finally
based on minimal power for transmitting, clusters
are formed.
Power-Efficient Gathering in Sensor Information
Systems (PEGASIS) has been elaborated by Lindsey
& Raghavendra (2002). This greedy algorithm is
based on forming a chain structure from sensor
nodes. In fact, each node in the network transmits
and receives data only from a neighbor. Only one
node is selected from the chain to send data to the
sink. It uses data aggregation like LEACH protocol
but don’t use clustering. The use of chain and the
absence of clusters train several threats and attacks
and furthermore, communication overhead is
increased.
The Hybrid Energy Efficient Distributed
(HEED) has been developed by Younis & Fahmy
(2004). In this is clustering protocol, probability to
elect a CH take into account three factors residual
energy, communication cost and average minimum
reachability power (AMRP). It uses the same
communication method as LEACH protocol but
HEED protocol has a well balanced energy and
longer network lifetime than LEACH.
Reference (Taheri et al., 2010) presents the
protocol HEED Non-Probabilistic approach and
Fuzzy Logic (HEED-NPF). In this protocol, election
of cluster head selection is based on Fuzzy Logic
which uses node degree and node centrality as input
parameters. The output is the Fuzzy cost. Each node
in network chooses the CH with least cost and joins
it. This technique is more effective to prolong the
lifetime of network than HEED.
Reference (Taruna et al., 2012) proposes a new
approach based on LEACH and covers the CH
selection phase. In fact, the proposed algorithm
calculates the center point between the sensor node
and the Base Station. Then, the node chooses the
closest CH to the center point and gets bind to it to
form clusters.
In the reference (Singh et al., 2013), authors
focus on selecting CH to save energy consumption
and lifetime of the network. They consider the
remaining energy of nodes and give analysis and
simulations when the BS is inside or outside the
network area.
Optical-LEACH (O-LEACH) is an improved of
LEACH. It was introduced in (El Khediri et al.,
2014) as a clustering hierarchy, an optical and
adaptive protocol that minimizes energy
consumption. In this reference, the node should have
a current energy greater than ten percent to become a
CH.
The reference (Sharma et al., 2015) calculates a
new threshold which is based on node energy,
distance between sensor node and BS, distance
between CH and BS. The analysis of simulation
results proves that this new algorithm is better in
term of balancing the node energy and prolonging
the network lifetime.
In the reference (Li and Huo, 2016), authors
propose a new algorithm that firstly calculate the
optimal cluster number by considering location
adaptability and data aggregation rate. Secondly,
they present a new threshold based on remaining
energy, initial energy, average energy consumption,
and node degree to select CH. Thirdly, a self-
adaptive uneven clustering algorithm is proposed
that takes node degree into consideration and solve
the “hot spot” problem. And finally, they propose a
solution to solve “isolated nodes problem”.
3 LEACH PROTOCOL
LEACH (Heinzelman et al., 2000), (Heinzelman,
2000) is a hierarchical protocol that is based on data
aggregation, dynamic allocation of CH and local
control on data transmission. It operates by round to
round and each round comprises three phases:
Advertisement phase, cluster set-up phase and
steady-state phase.
Advertisement phase: Firstly, each node in the
network decides if it will be a Cluster Head (CH) or
not for present round. This decision depends on the
desired percentage of CHs in the network and the
number of times the node is served as CH so far. In
New Approach for Selecting Cluster Head based on LEACH Protocol for Wireless Sensor Networks
115
fact each node i choose a random number between 0
and 1. If this number is less than a threshold T(i), i
becomes a CH.
T
(
i
)
=
∗(
)
i
f
i∈G
0otherwise
(1)
Where P is the desired percentage of cluster
heads, r is the current round, and G is the set of
nodes that have not been cluster-heads in the last
rounds. After CH election phase, each CH
broadcasts advertising messages to the remaining
nodes inviting it to choose which of the CHs they
will join and finally, clusters are created for the
current round. The choice of remaining sensor nodes
will depends on the signal strength of the received
broadcasting messages.
Cluster setup phase: Each remaining node
communicates its decision to the chosen CH node
that it will be belong to the cluster. To receive this
information, all CHs keep their receivers on during
this phase. Based on the number of nodes in the
cluster, the CH creates a time division multiple
access (TDMA) schedule and informs other sensor
nodes when it can transmit.
Steady-state phase: In this phase, transmission
data starts. Sensor nodes send their data in their own
time slot and their radio can be turned off. CH must
keep their radio on to receive all data from nodes.
LEACH Protocol has several advantages
(Heinzelman, 2000), (Haneef and Deng, 2011):
Comparing with direct communication, LEACH
protocol achieves more than 7 reductions in
dissipated energy. In addition, lifetime of the
network is raised due to dynamic clustering. By
using aggregation technique, LEACH reduces data
message sent to the BS. During setup phase, it uses
TDMA mechanism to minimize the conflict between
clusters. Finally, since LEACH is a distributed
protocol; it doesn’t need global knowledge of
network. But also it has a certain number of
shortcomings (Heinzelman et al., 2000), (Haneef and
Deng, 2011), (Yan, and Liu, 2011): CHs are elected
randomly and residual energy of the node is not
taken into account for cluster formation. CHs in the
network have not a uniform distribution. It happens
that sometimes these nodes are concentrated in one
part of network which trains loss of energy. After
aggregation, CHs send data to the sink in single hop
for that LEACH is not applicable to large networks.
In each round, all sensor nodes participate in
forming new clusters which dissipates energy. Data
aggregation is applied each round if there is a
change in data packages or not which cost some
unnecessary energy of cluster-heads.
4 PROPOSED WORK
4.1 Cluster Head Selection Approach
The main shortcoming of LEACH is the random
selection of CH that is applied to all sensor nodes
without taking into account any factor. In reality, to
increase the lifetime of network and energy
efficiency, we need to change the threshold of
electing CH. In other words, we must consider three
essential factors: the distance between the node and
the BS, the residual energy and the number of
neighbor nodes within the cluster range, to calculate
the threshold. Therefore, by including distance
between the node and the BS, data transmission
overhead is minimized. The flowchart shown in
Figure 1 explains as well our proposed approach.
When considering the remaining energy of node
each round and alive neighbors, we can optimize the
election of CHs. Thus, nodes having at the same
time high residual energy, short distance to the sink
and several neighbors are chosen as CHs. By
incorporating above criteria, we can use a cost
function which is expressed as:
cost
(
i
)
=α
E
(
i
)
E
+β
N
(
i
)
N
+
γ
D
(
i
)
−D
D
−D
(2)
Where E
rem
(i) is the remaining energy of node i,
E
init
is the initial energy, N
nb
(i) is the number of
neighbors of node i, N
alive
is the number of alive
nodes, D
toBS
(i) is the distance between the node i and
the BS, D
toBSmin
is the distance between the closest
node to the BS and the BS and D
toBSmax
is the
maximum distance to the BS. Then the threshold can
be written as follows:
T
(
i
)
=
P
1−P∗(rmod
1
P
)
∗cost(i)i
f
i∈G
0 otherwise
(3)
After selecting CHs, the remaining nodes have to
choose its cluster for each round. The choice of
nodes is based on the distance between the node and
the CH. Nodes opt to the closest one and gets bind to
it to form clusters.
ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering
116
Figure 1 : Flowchart of the proposed approach.
α, β and γ are weights parameters between 0 and 1
that are determined through Analytic Hierarchy
Process (AHP) method.
First, the energy weight parameter should be
greater than β and γ. In fact, the node to become CH
should have the maximum energy level (Li et al.,
2014) that influences results in terms of the first
node dead (FND). Then, considering the reference
(Li and Huo, 2016), the node neighbors factor is
more effective than the distance between the node
and the base station so β should be greater than γ.
Based on this information and by applying the
AHP method, we can elaborate the paired
comparison matrix as shown in Table 1 where A is
the remaining energy criterion, B is alive neighbor
criterion and C is the distance to BS criterion.
Table 1: Paired Comparison Matrix.
Criteria A B C
Priority
vector
A 1 3 7 64,34%
B 1/3 1 5 28,28%
C 1/7 1/5 1 7,38%
Sum 31/21 21/5 13 100%
λ
max
=3.0967, CI=0.0484, CR=8.34% < 10%
(acceptable)
Note that the priority vector is obtained from
normalized Eigen vector of the matrix and presents
the α, β and γ weights values respectively. The
diagonal of the matrix is always 1 and the lower
triangular matrix is filled using formula a
ji
=1/a
ij
.
Where a
ij
denotes the ratio of the i
th
criterion weight
to the j
th
criterion weight. As shown in Table 2, the
fundamental 1 to 9 scale can be used to rank the
judgments.
λ
max
is the Eigen value and is obtained from the
summation of products between each element of
Eigen vector and the sum of columns of the
reciprocal matrix.
CI is the Consistency Index and is calculated by:
CI=
λ
−n
n−1
(4)
CR is the Consistency ratio and is calculated by:
CR=
CI
RI
(5)
Where RI is the Random Index whose values are
cited in table 3.
New Approach for Selecting Cluster Head based on LEACH Protocol for Wireless Sensor Networks
117
Table 2: A fundamental 1 to 9 scale.
Number Rating
Verbal Judgment of
Preferences
1 Equally preferred
3 Moderately preferred
5 Strongly preferred
7 Very strongly preferred
9 Extremely preferred
Table 3: Random Consistency Index (RI).
Dimension RI
1 0
2 0
3 0,58
4 0,9
5 1,12
6 1,24
7 1,32
8 1,41
9 1,45
10 1,49
4.2 Energy Model
In our research, we have used the same energy
model as the traditional LEACH (
Heinzelman et al.,
2000), as shown in Figure 2.
Figure 2 : The radio energy consumption model.
Note that E
elec
is the energy consumption per bit
for running transmitter or receiver circuitry, k is the
number of bits, ε
fs
and ε
mp
are proportional constant
of the energy consumption for the transmit amplifier
in free space channel model (ε
. k .d
2
power loss)
and multipath fading channel model (ε
. k .d
4
power loss), respectively and d is the distance
between transmitter and receiver. Thus we can
deduce the energy consumed to transmit k bits along
a distance d through a free space channel model is:
E
(
k,d
)
=E
∗k+ε
∗k∗d
(6)
Or multipath fading channel is:
E
(
k,d
)
=E
∗k+ε
∗k∗d
(7)
And the energy to receive these bits is:
E
(
k
)
=E
∗k
(8)
5 SIMULATIONS AND
NUMERICAL RESULTS
In this section, simulations are performed via Matlab
software in the same conditions. We have compared
between our proposed approach and LEACH
protocol using parameters listed in Table 4.
As shown in Figure 3, we consider a WSN with
randomly distributed sensor nodes in 100×100
network field. Initially, all nodes are normal nodes
and have the same energy value. Normal nodes
appear in blue point and the BS sets outside of the
sensor area and appears in green point.
Table 4 : Parameters System.
Simulation area 100x100m
2
Number of Round 1000
Number of nodes 200
desired percentage of CH 0.1
Initial energy of node 0.5 J
Transmission/ Reception
energy per bit E
elec
50 nJ/bit
Transmitter Amplifier
energy dissipation free Space
10 pJ/bit/m
2
Transmitter Amplifier
energy dissipation multiPath
0.0013
pJ/bit/m
4
Base Station location Located at
50x175
Figure 3 : Initial Wireless Sensor Network.
ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering
118
In order to evaluate the performance and
efficiency of our proposed approach, we focus on
the number of alive nodes, dead nodes and total
residual energy of the network respectively in
figures 4, 5 and 6.
Figure 4 shows a comparison between LEACH
protocol and our proposed work of the total number
of nodes that are still alive in each round. Our
proposed approach prolongs the lifetime of the
network practically between 30% and 40%
comparing to LEACH protocol.
Figure 5 gives the number of dead nodes per
round in the network. We can see clearly that our
proposed algorithm beats LEACH protocol in term
of the First Dead Node (FDN). In fact, for LEACH,
the FND is after 119 rounds and for our approach is
after 274 rounds. This proves that network lifetime
is well prolonged by the new cost of selection CHs.
Figure 4 : The number of alive nodes per round.
Figure 5 : The number of dead nodes per round.
Figure 6 : Total residual energy per round.
Finally, we focus on the total remaining energy
of nodes. Figure 6 shows the decreasing of this
energy per round. When comparing our approach to
the LEACH, it’s visible that in our proposed work,
residual energy decreases slower than LEACH. In
other words, our approach is effective to save energy
consumption better than LEACH. This proves that
we have chosen an efficient way for selecting CH
based on residual energy of the node, the number of
neighbours of the node and the distance between the
node and the BS.
6 CONCLUSIONS
The main purpose is to increase the lifetime of WSN
and saving energy consumption. One of the major
shortcomings of LEACH protocol is the probability
of selecting CH. We relied on this shortcoming and
we have proposed a new strategy to select CH by
including residual energy, distance between the node
and the BS and the number of neighbor of node
within the cluster range. The simulation results show
the round that the first node dies is delayed by about
57% than that in LEACH. The lifetime of the
network is prolonged about 40%. The remaining
energy in our approach decreases more slowly than
that in LEACH algorithm. All these results prove
that our proposed strategy is effective in reducing
the energy consumption and prolonging the network
lifetime.
New Approach for Selecting Cluster Head based on LEACH Protocol for Wireless Sensor Networks
119
REFERENCES
Al-Karaki, J.N., Kamal, A.E., 2004. Routing techniques in
WSN: A survey. In IEEE Wireless communication,
vol. 11, no. 6, pp. 6.
Anastasi, G., Conti, M., Di Francesco, M., Passarella, A.,
2008. Energy conservation in wireless sensor
networks: A survey. In Ad Hoc Networks Journal, vol.
7, issue 3, pp.537-568
Geetha, V., Kallapur, P.V., Tellajeera, S., 2012. Clustering
in Wireless Sensor Networks: Performance
Comparison of Leach & Leach-c protocols using Ns2.
Procedia Technology, vol. 4, pp. 163–170
Haneef, M,. Deng, Z., 2011. Comparative analysis of
classical routing protocol LEACH and its updated
variants that improved network life time by addressing
shortcomings in Wireless Sensor Network. In 2011
Seventh International Conference on Mobile Ad-hoc
and Sensor Networks (MSN), pp. 361-363.
Heinzelman, W. B., 2000. Application Specific Protocol
Architectures for Wireless Networks, Ph.D.thesis,
Dept. Electrical Engineering and Computer Science,
Massachusetts Institute of Technology, Cambridge,
MA.
Heinzelman, W. B., Chandrakasan, A., Balakrishnan, H.,
2000. Energy-efficient communication protocol for
wireless microsensor networks. In Proceedings of the
33rd Annual Hawaii International Conference on
System Sciences (HICSS ’00), vol. 8, pp. 1–10, Maui,
Hawaii, USA.
Katiyar, N.V., Chand, Soni, S., 2011. A Survey on
Clustering Algorithms for Heterogeneous Wireless
Sensor Networks. International Journal on Advanced
Networking and Applications, vol. 2, no. 4, pp.745-
754.
Li, J., Huo, J., 2016. Uneven clustering routing algorithm
based on optimal clustering for wireless sensor
networks. In Journal of Communications, vol. 11, no.
2, pp. 132-142.
Li, J., Jiang, X., Lu, I.T., 2014. Energy Balance Routing
Algorithm Based on Virtual MIMO Scheme for
Wireless Sensor Networks. In Hindawi Publishing
Corporation Journal of Sensors, vol. 2014, pp. 1-7.
Lindsey, S., Raghavendra, C.S., 2002. PEGASIS: Power-
efficient Gathering in Sensor Information System. In
Proceedings IEEE Aerospace Conference, vol. 3, Big
Sky, MT, Mar., pp. 1125-1130.
Singh, A.P., Sharma, N., Roy, N.R., 2013. Residual
Energy and Distance based Energy-Efficient
Communication Protocol for Wireless Sensor
Network. International Journal of Computer
Applications, vol. 74, no. 12, pp. 11-16.
Taheri, H., Neamatollahi, P., Naghibzadeh, M.,
Yaghmaee, M.H., 2010. Improving on HEED protocol
of wireless sensor networks using non probabilistic
approach and fuzzy logic (HEED-NPF). In
Proceedings 5th International Symposium on
Telecommunications, Tehran, Iran, pp. 193-198.
Taruna, S., Kohli, S., Purohit, G.N., 2012. Distance Based
Energy Efficient Selection Of Nodes To Cluster Head
In Homogeneous Wireless Sensor Networks. In
International Journal of Wireless & Mobile Networks
(IJWMN), vol. 4, No. 4.
Yan, J.F., Liu, Y.L., 2011. Improved LEACH routing
protocol for large scale wireless sensor networks
routing. In Proceedings of International Conference
on Electronics, Communications and Control
(ICECC), pp. 3754-3757.
Younis, O., Fahmy, S., 2004. Heed: A hybrid, Energy-
efficient, Distributed Clustering Approach for Ad-hoc
Networks. In IEEE Transactions on Mobile
Computing, vol. 3, no. 4, pp.366-369.
El Khediri, S., Nasri, N., Wei, A., Kachourid, A., 2014. A
New Approach for Clustering in Wireless Sensors
Networks Based on LEACH. Procedia Computer
Science, vol 32, pp. 1180–1185.
Sharma, R., Mishra, N., Srivastava, S., 2015. A proposed
energy efficient distance based cluster head (DBCH)
Algorithm: An Improvement over LEACH. Procedia
Computer Science, vol. 57, pp. 807-814.
ENASE 2017 - 12th International Conference on Evaluation of Novel Approaches to Software Engineering
120
