Ann. Telecommun. (2018) 73:353–365
https://doi.org/10.1007/s12243-017-0611-6
A dynamic harmony search-based fuzzy clustering protocol
for energy-efficient wireless sensor networks
Osama Moh’d Alia
1
Received: 15 June 2017 / Accepted: 1 October 2017 / Published online: 10 October 2017
© Institut Mines-T
´
el
´
ecom and Springer-Verlag France SAS 2017
Abstract In the development of cluster-based energy-
efficient protocols for wireless sensor networks (WSNs),
a particularly challenging problem is the dynamic organi-
zation of sensors into a wireless communication network
and the routing of sensed information from the field sen-
sors to a remote base station (BS) in a manner that prolongs
the lifetime of WSNs. This paper presents a new energy-
efficient clustering protocol for WSNs, which can minimize
total network energy dissipation while maximizing network
lifetime. The protocol is divided into two parts. The first
deals with constructing an infrastructure for the given WSN.
A newly developed algorithm, based on a harmony search
(HS), automatically determines the optimal number of clus-
ters and allocates sensors into these clusters. This algorithm
also eliminates the need to set the number of clusters a
priori. The second part is concerned with the process of
sending sensed data from nodes to their cluster head and
then to the BS. A decentralized fuzzy clustering algorithm
is proposed, where the selection of cluster heads in each
round is locally made in each cluster during the network
lifetime. Simulation results demonstrate that the proposed
protocol can achieve an optimal number of clusters, prolong
the network lifetime and increase the data delivery at the
BS, when compared to other well-known clustering-based
routing protocols.
Osama Moh’d Alia
sm
alia@yahoo.com; oalia@ut.edu.sa
1
Department of Computer Science, Faculty of Computers and
Information Technology, University of Tabuk, PO box 741,
Tabuk 71491, Saudi Arabia
Keywords Wireless sensor networks · Energy-efficient
routing protocols · Harmony search algorithm · Fuzzy
clustering
1 Introduction
Wireless sensor networks (WSNs) have potential appli-
cations in civilian and military domains, including envi-
ronmental monitoring, surveillance, healthcare, intelligent
building control, traffic control and object tracking [1–3].
WSNs consist of a large number of autonomous sensor
nodes equipped with sensing capabilities, wireless com-
munication interfaces and limited processing and energy
resources.
One or more powerful base stations (BSs) act as the final
destination of the data. WSNs are used for distributed and
cooperative sensing of physical phenomena and events of
interest [4].
Energy consumption needed to pass sensory data to the
final destination for each node is a major concern when
designing WSN routing protocols [5–7]. Economical usage
of sensor nodes is paramount in order to prolong the oper-
ational lifetime of the network. Moreover, the sensor nodes
are equipped with limited energy sources (lightweight bat-
tery); therefore, replacing these batteries may be inappli-
cable or impossible [8–10]. Thus, optimal battery energy
usage can prolong use and improve performance efficiency
of WSNs. Designing routing algorithms, which can con-
sume less energy, while maintaining the efficiency and
robustness of WSNs, is hence highly desirable.
During the last decade, WSN researchers have focused on
solving the above-mentioned problem by proposing power
management protocols alongside the evolution of hardware-
based solutions [11–13]. Cluster-based techniques, which
@Phil_Robichaud
@deepthiw
@JoseServera