E NERGY D EGREE D ISTANCE C LUSTERING A LGORITHM FOR W Sns

International Journal of Wireless & Mobile Networks (IJWMN) Vol. 7, No. 2, April 2015
DOI : 10.5121/ijwmn.2015.7205 59
ENERGY DEGREE DISTANCE CLUSTERING ALGORITHM
FOR WSNS
Awatef Ben Fradj Guiloufi, Salim El Khediri, Nejah Nasri
and Abdennaceur Kachouri
National Engineering School of Sfax, Sfax, Tunisia
ABSTRACT
Energy consumption is an important parameter in the context of the wireless sensor networks (WSNs).
Several factors can cause energy over consumption such as mobility, node position (relay or gateway),
retransmissions... In this paper, we described a new Energy-Degree Distance(EDD) Clustering Algorithm
for the WSNs. A node with higher residual energy, higher degree and closer to the base station is more
likely elected as a clusterhead (CH). The inter cluster and intra cluster communications are realized on one
hop. The principal goal of our algorithm is to minimize the energy consumption and balance energy
consumption among all nodes. By comparing EDD clustering algorithm with LEACH algorithm,
simulation results have shown its effectiveness in saving energy.
KEYWORDS
Wireless Sensor networks (WSNs), clustering, clusterhead, energy, LEACH.
1. INTRODUCTION
A WSNs is a set of nodes called sensors which are able to detect a particular information and to
send it to the Base Station (BS). The sensors are inter-connected using a wireless radio
communication ([1] ,[2]). The WSNs are characterized by an absence of infrastructure, a
resources constraints, an heterogeneity and a dynamics structure. For that, it is important to
design an auto-organized virtual topology ([3] ,[4]). For these virtual topology, several methods
were defined in the literature such as clustering algorithm and the dorsal structure [5].
The clustering procedure is to cut the structure in small zones (called cluster) which are managed
by a leader called Cluster Head (CH) ([6], [7]).
For clustering method, the WSNs architecture can be presented into three layers: a sensors nodes
which are the receivers of the data, the CHs, and the BS [3].
For WSNs, the important metric tool is generally energy consumption (network lifetime) [8]. In
fact, the lifetime is a fundamental parameter in the context of availability in the WSNs ([9] ,[10]).
In this paper, we proposed an energy efficient clustering algorithm called EDD. The objective of
this algorithm is to minimize the energy consumption among all nodes. We propose also to
compare between the performances of dynamic clustering and static clustering.
This paper is organized as follows. Section 2 describes the model of the network. Section 3
presents the proposed algorithm. The simulation results are described in the last section.

60
2. EDD CLUSTERING ALGORITHM
In this section, we define a new clustering algorithm called EDD (Energy-Degree Distance)
Clustering Algorithm for WSNs of which the goal is the minimization of the energy consumption
and the maximization of network lifetime.
2.1. Network Model
We will model the WSN by a graph G = (V, E) where V represents the set of sensors and E =
{(u, v) ⊂V2
│ d (u, v) ≤ R} represents the wireless connections between nodes. R is the
transmission range and D(u, v) represents the Euclidean distance between the node u and the
node v. Table I defines all the parameters and their definitions.
Table 1 Parameters and definitions
Parametres Definitions
N Number of nodes
Ec(u) Energy consumed of u
DisBS(u) Distance to the base station
Deg(u) Degree of u
Weight(u) Weight of u
2.2. Energy Consumption Model
The energy consumption is generally the most important parameter for WSNs evaluation phase. It
depends in fact on the nodes’ characteristics (nature of data processing, transmitted power,
standby mode, …), and nodes role during the communication [11].
The consumed energy is defined by this equation [12]:
Εχχχχ = Ec/capture+Ec/treatment+Ec/communication (1)
Where:
• Ec/capture: is the consumed energy by a sensor during the capture unit activation.
• Ec/treatment: is the consumed energy by the sensor during the activation of its treatment unit.
• Ec/communication is the consumed energy by the sensor during the activation of its
communication unit.
The consumed power by the communication unit is generally very high than the consumed
energy by the treatment unit and the capture unit. In fact, the transfert of one bit can consume as
much as the execution of a thousands instructions [13]. For this, we can neglect the consumed
energy of the capture unit, and the treatment unit compared to the energy consumed by the
communication unit. In this case, the equation (1) will be thus:
Εχχχχ = Ec/communication (2)
The communication energy is equal to the sum of the emission energy and the reception energy:
Ec/communication = ETX+ERX (3)
Referring to [14], the transmission the energy and the reception energy can be defined as follows:

61
ETX(K,d) = Eelec * K + εamp * K * dλ
. (4)
ERX (K) = Eelec * K. (5)
Where:
• K: message length (bits).
• D: distance between transmitting node and receiving node (m).
• λ: of way loss exhibitor, λ>=2.
• Eelec: emission /reception energy, Eelec = 50 nJ/bit.
• εamp: transmission amplification coefficient, εamp = 100 pJ/bit/m2
2.3. ClusterHeads election procedure
Step 1: each sensor transmits a “hello” message to its neighbors, for the discovery of 1-hop
neighborhood.
Step 2: each node u ϵ V found its degree. The Deg(u) is the cardinality of the set E1={(u,
v) ⊂V2
│ d (u, v) ≤ R}
(6)
Step 3: each node u ϵ V compute its energy consumed Ec(u).
Step 4: each node u ϵ V compute its distance to the BS.
La distance à la station de base est calculée en se basant sur l'équation suivante:
(7)
Step 5: Sensors calculate their weights, the weight is defined as follows:
Weight(u) = Ec/com (u) + 1/ D(u) +DisBS(u) (8)
Step 6: The sensors distribute their weights to their neighbors.
Step 7: The sensor which has the weakest weight is declared as a CH, it puts its state = “CH” and
transmits a message “clusterhead_elected” (containing its ID) to its neighbors.
Step 8: These laters, after receiving this message, declare themselves like “Nm”, transmit to the
CH a message “clusterhead_accepted”, and note the identity of their CHs in their databases.
The EDD clustering algotithm can be described by this organigram:
Figure 1. Organigram

62
2.4. Dynamic clustering VS static clustering
There exist two method of clustering: dynamic clustering and static clustering.
For the dynamic clustering, the update of the cluster and CHs is done each round, in the other
hands, the change relates to the totality of the structure.
For the static cluster, the cluster are static and for each cluster, the change only relates to CH, in
this case, the number of clusters is always the same and CH of each cluster changes according
to its properties.
3. SIMULATION RESULTS
The simulation results are implemented using Matlab 7.0.1 tool. The WSNs is composed of a
number of nodes which varies between 10 and 200 nodes.
The simulation of the proposed algorithm was carried out during 10 deactivation intervals T
(standby mode) in a space of 150 m × 150 m and the range of the nodes (Tx-Arranges) is 40 m.
The size of a measured data package for sensors and send towards their clusterheads is 4000 bits.
Fig. 2 shows the communication structure of network with 30 nodes.
In Figure. 2, red o represent the CH, yellow triangle represent the ordinary sensor node, blue lines
represent the communication between CH and its members, and black * represent the BS.
Figure2. Communication structure of network
Then, we suppose to make a comparison between the dynamic clustering method (DC) and the
static clustering method (SC) by applying the proposed clustering algorithm EDD.
Figure.3 represents the control traffic for both method: EDD- DC indicates that the update is
done in a dynamic way and EDD- SC indicates that the method of clustering is static.
We notice that the control traffic is higher for the dynamic method, more control packages are
transmitted during the cluster construction phase.
For the static method, control packages are tarnsmitted only if there is a CH change
procedure. A higher control traffic can represent a cause of energy overconsumption, but, the
energy consumption for the control traffic is generally negligible than that consumed during the
transmission of information.

63
Figure 3 . Control traffic
Figure.4 represents the average energy consumption for both methods. We notice that the
average energy consumption of the dynamic method is higher, this method consumes more
energy during the clusterhead election phase and the clusters update phase.
Figure 4. Energy consumption
Figure. 5 represents the first node die for two methods. The best results are given by EDD- DC,
this method tends to balance the load of energy consumption between nodes, in this case the
network lifetime increases. EDD- SC set up of the static clusters and the update concerne only
the CHs. This method can involves the choice of CHs having a weak energy reserve, which
results the crushing of its battery.

64
Figure 5. First node die
Among the most known clustering algorithms in literature, we distinguish, LEACH
algorithm [8]. LEACH is a famous algorithm which the goal is the minimization of the
energy consumption in the WSNs.
We wish in this part to compare our algorithm with the LEACH clustering algorithm.
Figure. 6 represents the energy consumption VS the network size. We can note that the
values obtained by our algorithm are rather low compared to those obtained by LEACH.
These results indicate that our algorithm is more effective and can prolong the network
lifetime and ensure its good performance.
Figure 6. EDD-DC VS LEACH
4. CONCLUSIONS
We presented in this paper a new clustering algorithm for the WSNs called EDD. The principal
goal of our algorithm is the prolongation of the network lifetime, two parameters were taken into
account for the choice of CHs: the nodes energy consumption and their degree.
The simulation results show that our algorithm is more effective in energy and increase the
network lifetime.

65
REFERENCES
[1] C. Tidjane Kone, (2011) ‘’Conception de l’architecture d’un réseau de capteurs sans fil de grande
dimension,’’PHD, University of Henri Poincaré Nancy I.
[2] G. Chalhoub, (2010) ‘’Les réseaux de capteurs sans fil’’, Ph.D, University of Clermont, Auvergne.
[3] L. Samper, (2008) ‘‘Modélisations et analyses de Réseaux de capteurs’’, PHD, VERIMAG
laboratory.
[4] R. Kacimi, (2009)’’Techniques de conservation d’énergie pour les réseaux de capteurs sans fil,’’
PHD, TOULOUSE UNIVERSITY.
[5] M. Khan & J. Misic. (2009) ‘’On the lifetime of wireless sensor networks,’’ ACM Transactions on
Sensor Networks (TOSN), Vol, 5, No. 5.
[6] R. Kuntz, (2010) ‘‘Medium access control facing the dynamics of wireless sensor networks, ‘‘PHD,
University of Strasbourg.
[7] D. Kumar, D. Trilok, C. Aseri & R. B. Patel, (2009) ‘’EEHC : energy efficient heterogeneous
clustered scheme for wireless sensor networks,’’ Computer Communication., Vol. 32, No. 4, pp 662–
667.
[8] W. B. Heinzelman, (2000) ‘’Application-specific protocol architectures for wireless networks,’’ PHD,
B.S.Cornell University.
[9] Y. Ossama & S. Fahmy, (2004) ‘’HEED : A hybrid energy-efficient distributed clustering
approach for ad hoc sensor networks,’’ IEEE Transactions on Mobile Computing, Vol. 3, No. 4,
pp 366 – 379.
[10] Duan, Changmin & Hong Fan, (2007) ‘' A distributed energy balance clustering protocol for
heterogeneous wireless sensor neworks''. In : Proc. Int. Conf. Wireless Communications, Networking
and Mobile Computing WiCom. pp. 2469–2473.
[11] V. Raghunathan, C. Schurgers, S. Park & M.B. Srivastava, (2002) ‘’Energy- aware wireless micro-
sensor networks,’’ IEEE Signal Processing Magazine, Vol. 19, No. 2, pp 40 –50.
[12] Mitton, Nathalie, Busson, Anthony & Fleury, Eric. (2004) ‘’Self-organization in large scale adhoc
networks‘’. In : Mediterranean ad hoc Networking Workshop (Med-Hoc-Net’04). Bodrum, Turquie.
[13] Wendi Rabiner Heinzelman, Anantha Chandrakasan, & Hari Balakrishnan, (2000) “Energy-efficient
communication protocol for wireless microsensor networks,” In : Proc IEEE 33rd Hawaii
International Conference on System Sciences.
[14] G. J. Pottie & W. J. Kaiser, (2000) ‘’Wireless integrated network sensors,’’ Commun. ACM, Vol.
43, No. 5, pp. 51-58.
Authors
Awatef BENFRADJ was born in Gabes, Tunisia in 1983. She received the engineering
degree in Telecommunications and Networks from the University of Gabes, in 2007. In
2009, she obtained the master degree in Networks from Higher School of Communication
of Tunisia. Since 2011, she has been a researcher within the laboratory of electronics and
information technology, Sfax University.
Salim El Khediri, a Ph.D student at the National School of Engineers of Sfax, Tunisia
(Laboratory of Electrical and Information Technology) and in partnership collaboration
education with CEDRIC laboratory of CNAM (National Conservatory of Arts and Crafts
, Paris). Since 2009 working in the capacity of assistant teacher of computer sciences at
the Faculty of Sciences Gafsa – Tunisia.
Nejah NASRI was born in Menzel Bouzaienne Village, in 1981. He received the B.S. and
M.S. degrees in electronic engineering from the University of Sfax, in 2006 and the Ph.D.
degree in electronic and telecommunication engineering from Toulouse University, France,
in 2010. From 2006 to 2009, he was a Research Assistant with Higher Institute of
Computer and Communication Techniques (ISITCom), Hammam Sousse, Tunisia. Since 2010, he has been
an Assistant Professor with the Informatics Engineering Department, Gafsa University. He is the author of

66
more than 50 articles. His research interests include engineering of wireless sensors networks and wireless
underwater communication.
Abdennaceur KACHOURI was born in Sfax, Tunisia, in 1954. He received the
engineering diploma from National school of Engineering of Sfax in 1981, a Master
degree in Measurement and Instrumentation from National school of Bordeaux
(ENSERB) of France in 1981, a Doctorate in Measurement and Instrumentation from
ENSERB, in 1983. He “works” on several cooperation with communication research
groups in Tunisia and France. Currently, he is Permanent Professor at ENIS School of
Engineering and member in the “LETI” Laboratory ENIS Sfax

E NERGY D EGREE D ISTANCE C LUSTERING A LGORITHM FOR W Sns

More Related Content

E NERGY D EGREE D ISTANCE C LUSTERING A LGORITHM FOR W Sns