IEEE INTERNET OF THINGS JOURNAL, VOL. …, NO.

…, MONTH 2019

Residual Energy Based Cluster-head Selection

in WSNs for IoT Application
Trupti Mayee Behera, Sushanta Kumar Mohapatra, Umesh Chandra Samal, Mohammad. S. Khan,
Mahmoud Daneshmand, and Amir H. Gandomi

Abstract—Wireless sensor networks (WSN) groups specialized Till date, about 5 billion smart devices are already
transducers that provide sensing services to Internet of Things connected and by 2020 about 50 billion devices to be
(IoT) devices with limited energy and storage resources. Since connected [1]. The number of people actually communicating
replacement or recharging of batteries in sensor nodes is almost
may exceed the number of devices/machines connected to
impossible, power consumption becomes one of the crucial design
issues in WSN. Clustering algorithm plays an important role in them virtually. This will generate huge traffic where humans
power conservation for the energy constrained network. Choosing may become the minority of generators and receivers of traffic
a cluster head can appropriately balance the load in the network [3]. This gives the reason for exploring IoT for various
thereby reducing energy consumption and enhancing lifetime. The research areas owing to its challenges and opportunities
paper focuses on an efficient cluster head election scheme that WSN acts as a medium that bridges the virtual digital world
rotates the cluster head position among the nodes with higher
to the real world. Tiny sensors or actuators connected with
energy level as compared to other. The algorithm considers initial
energy, residual energy and an optimum value of cluster heads to each other are responsible for sensing and transferring the
elect the next group of cluster heads for the network that suits for values to the Internet. WSN comprises sensor nodes deployed
IoT applications such as environmental monitoring, smart cities, in a network field to monitor various physical and
and systems. Simulation analysis shows the modified version environmental parameters. The routing path of data from the
performs better than the LEACH protocol by enhancing the sensing node to the sink node or base station (BS) should be
throughput by 60%, lifetime by 66%, and residual energy by 64%.
designed in an energy efficient manner since recharging the
Index Terms— WSN, IoT, CH selection, Residual energy,
sensor battery is practically impossible [4].
Lifetime, Energy efficient. Different from the ad-hoc network, WSN meant for IoT
application faces numerous challenges in terms of a number of
I. INTRODUCTION sensor nodes, hardware, mode of communication, battery
power and computational cost to name a few. Apart from
I INTERNET of Things (IoT) is an ecosystem of
interconnected devices and objects via the internet enabling
them to send and receive data. It is an invisible but intelligent
sensing, the sensors used in the IoT paradigm are assigned
with additional functionalities and has to face new challenges
in terms of QoS (quality of service), security and power
network that senses, controls and can be programmed [1], by
management [5]. Some of these issues are addressed by
employing embedded technology to communicate with each
adopting various technological changes in primitive protocols
other. The IoT provides immediate access to information
and schemes used for WSN. QoS requirements in IoT based
related to any device with high productivity and efficiency [2].
WSN faces significant challenges like extreme resource
Manuscript received XXX XX, XXXX; accepted XXX XX, XXXX. Date of content, redundancy in data, dynamic size of the network, less
publication XXXX XX, XXXX; date of current version XXXX XX, XXXX. reliable medium, heterogeneous network, and multiple BS or
Trupti Mayee Behera, Sushanta Kumar Mohapatra, and Umesh Chandra sink nodes [6]. The key security issues in WSN includes data
Samal are with School of Electronics Engineering, KIIT University, authenticity and confidentiality, data integrity and freshness in
Bhubaneswar, Odisha 751024, India. (e-mail: [email protected] ;
[email protected] ; [email protected] ). data [7].
M. S. Khan is with Department of Computer & Information Sciences, East Reduction of power consumption has always been a crux
Tennessee State University, Johnson City, USA (e-mail: issue in designing WSNs. Recent research outcome has come
[email protected] ). up with different ideas to reduce energy and extend network
M. Daneshmand is with the School of Business, Stevens Institute of
longevity for proper utilization of resources. Routing
Technology, Hoboken, NJ 07030, USA and he is Steering Committee Chair of
IEEE IoT J (e-mail: [email protected]).
algorithm plays a crucial role in the process. Clustering builds
A.H. Gandomi (corresponding author) is with the School of Business, a hierarchy of clusters or groups of sensing nodes that collects
Stevens Institute of Technology, Hoboken, NJ 07030, USA, (e-mail: and transfers the data to its respective cluster head (CH). The
[email protected]). CH then groups the data and sends the fused to sink node or
Color versions of one or more of the figures in this paper are available base station (BS) which acts as middleware between the end
online at http://ieeexplore.ieee.org.
user and the network. Among the clustering algorithm,
Digital Object Identifier XXXXXXXXXXXXXXXX
LEACH (Low Energy Adaptive Clustering Hierarchy) is a
classical protocol that considers energy for hierarchical
routing of data [8]. The network is grouped into clusters, and
"Copyright (c) 2012 IEEE. Personal use of this material is permitted. However, permission to use this material
for any other purposes must be obtained from the IEEE by sending a request to [email protected] "
 IEEE INTERNET OF THINGS JOURNAL, VOL. …, NO. …, MONTH 2019

the sensor node transmits its data to the corresponding CH. The primary goal of any sensor network is to route the data
The protocol randomly selects CHs in a stochastic manner for assembled by sensors and forward it towards the BS. The
each round. The CH communicates with each node of the simplest method to communicate data is direct transmission
cluster called member nodes to collect the sensed data. The where the nodes have to direct its data to the base station or
CH assigns TDMA (Time Division Multiple Access) sink node. However, if the distance between sink and network
schedules to its corresponding cluster member. The member is large, the node will die out quickly due to unnecessary
node can transmit data during the allotted time slot. The data energy consumption [14]. Clustering algorithm reduces the
is then checked for redundancy and compressed before unwanted power consumption in delivering data to BS by
communicating with the sink node grouping the network into clusters. Each cluster is assigned a
The CHs directly communicate with BS in LEACH CH that sends data to BS. An important phase in the clustering
protocol; hence the power consumption in sending data from algorithm is the CH election process that should guarantee
CH to BS will be more as compared to the communication uniform energy distribution among the sensor nodes [15].
between the CHs. As a result, the CHs will exhaust its energy LEACH protocol has intensively been modified by
within a short period of time. Multi-hop communication, on researchers to improve the network performance. Technical
the other hand, can be helpful to overcome this problem, but researchers are contributing vigorously in enhancing existing
still not effective in cases of small networks. algorithms for better performance of the IoT system [16]. An
Electing a CH is a sophisticated job as various factors have energy-efficient trust derivation method was discussed in [17]
to be considered for selection of best node in the cluster [9]. for WSN-based IoT networks. The scheme uses risk strategy
The factors include the distance between nodes, residual analysis to reduce network overhead by deriving an optimal
energy, mobility and throughput of each node. number of recommendations. The energy-aware scheme
LEACH algorithm enhances the lifetime of the network in maintains adequate security and also reduces the latency of the
comparison to direct or multi-hop transmission but still has network. A time-based CH selection is proposed in [18] called
many limitations. The election of cluster heads is done TB-LEACH that sets well-distributed clusters and enhances
randomly which does not ensure proper distribution and the lifetime by 20 to 30%. The distance between nodes and BS
optimal solution. The nodes with lesser energy have equal are considered for threshold based CH selection in [15] that
priority as that of those with a higher energy level to be improves lifetime by 10%.
elected as CH. So, when a node of low residual energy gets Thein et al. in [14] have modified the probability for the
selected to serve as CH, it dies out quickly resulting in shorter selection of CH based on the residual energy of each node.
network span [10]. The paper also considers the optimal value of CH but for fixed
The paper aims to select the CH considering important values like 1 and 6. The network lifetime enhances by 40-
parameters like the initial energy, remaining energy of the 50%. Another CH selection method for aggregation of data is
individual node and the optimal number of CHs in the discussed in [19] that eliminates redundancy and enhances the
network. The modification is done in the classical LEACH network lifetime. The threshold value is modified by
algorithm. With the completion of each round, the residual considering a hotness factor that defines the relative hotness of
energy of the non-CH nodes are checked, and the one with the a particular sensor node to that of the network.
higher energy level in comparison to others has a higher CH is selected using particle swarm optimization (PSO) in
probability for CH selection for the current round. This would [20]. The criteria for selection have an objective function in
prevent the network to die out too early thereby enhancing the terms of node degree, intra-cluster distance, residual energy,
network lifetime. and a number of optimal CHs. The model performs better in
The paper is structured as follows: Section II provides the terms of various network metrics in comparison to various
required background and related work. Our proposed system routing protocols. PSO-ECHS is discussed in [21], where PSO
model and algorithm are discussed in Section III. Section IV based CH selection is made using parameters like node-to-
discusses the simulation results and analysis. Finally, the node distance, distance to BS and residual energy. Another
article is concluded in Section V. optimization technique called Grouped Grey Wolf Search
Optimization is used in [22] for security-aware CH selection
II. RELATED WORK to improve the network lifetime.
One of the major issues of IoT is to handle a large number An improvement of LEACH was proposed in [23] where
of sensors that will be deployed, in terms of the cost of residual energy plays an important role in CH election. A
servicing and maintenance [11]. Further replacing sensor simple Multi-hop approach to LEACH was also studied, and it
batteries which are already located in the network field can be is found that both protocols perform better than LEACH by
a tedious job [12]. For instance, in case a sensor is to be extending lifetime after a certain period of time. A non-
deployed on a certain animal or species, it requires the battery probabilistic multi-criteria based CH selection was presented
of the sensor to outlive the animal which is far more in [24] where the network is divided into separate zones. The
achievable. This leads to another important challenge which is CH or zone head is selected using the ANP (Analytical
power management. Reliable end-to-end data transmission Network Process) decision tool. A set of parameters have been
with proper congestion control and low packet loss ratio are collected from where the best parameters have been selected
some of the other major concerns in WSN [13]. for zone head selection.
Behera et al.: Residual Energy Based Cluster-head Selection in WSNs for IoT Application

The IoT, being a ubiquitous network, connects smart • Nodes are deployed randomly and transmit its data
devices and objects to the cloud. WSN provides a platform for periodically
the collection and communication of data to monitor and • Each room has a CH that communicates with the BS either
control the physical world for the betterment of the society in the single hop or multi-hop communication.
[25]. Communicating wireless technologies drains more power • The BS receives the data from each CH and spread it to the
as compared to the devices meant to receive or sit idle. The cloud.
rising number of smart devices connecting to the internet has The environment-monitoring applications [27] requires the
made energy conservation a surplus parameter in IoT proper routing of data so that the network energy can be used
designing. Developing energy-efficient techniques for effectively. In case a node with lesser residual energy is
deployment of sensor networks have always been a elected as CH in one of the room, it will lead to the end of
challenging task for researchers. When incorporated with IoT, transmission of data from that room. As a result, the end user
power becomes a more crucial issue owing to the number of will not receive complete information for monitoring of
devices being connected in large scale. environmental conditions. The communication model used in
To maintain IoT standards, researchers have focussed on [28] shown in Fig. 2 has been considered to study the behavior
device energy-conserving techniques such as clustering where of the proposed model.
the choice of CH should be done judiciously. Various methods
for efficient CH selection was studied from the above
literature that enhances the network performance. However,
important parameters like residual energy, initial energy and
an optimum number of clusters in the network, have not been
considered to the best of our knowledge for modification in
the threshold value for CH selection.

III. SYSTEM MODEL

The rapid increase in population density in urban areas
requires modern infrastructures with suitable services to meet Fig. 2. Radio energy model
the requirements of the city inhabitants. Hence latest advances
in communication technologies such as IoT has been in
demand to provide a framework for the development of smart
cities [26].

Fig. 1. Environment monitoring using IoT

Fig. 3. Flowchart of LEACH protocol

This section presents an environmental monitoring scenario
that uses WSN as an integral part of IoT. The nodes are For short distance communication as to that of between
grouped in four different rooms to form clusters as shown in nodes and CHs within the room considers the free-space
Fig.1. Let there be eight sensor nodes in each room where only model and for longer distance transmission like between CHs
one node can become the CH (marked red) for each instant of and sink follows multi-path fading model [29]. For
time. The sink node collects data from the CHs of each room symmetrical propagation channel, the power consumption in
and sends the fused information to the end user. transmitting ‘k’ bits of data in a packet to a sensor placed ‘d’
For the system model, some reasonable assumptions have meters away can be written as [30]:
been adopted as follows:
• Nodes are static and homogeneous with initial energy 0.5J
ETx (k , d ) = ETx _ ele (k ) + ETx _ mp (k , d ) (1)
and are distributed in rooms to monitor variables such as
humidity, temperature, sound, and luminosity.
• BS/sink is fixed and installed in the middle of the network
 IEEE INTERNET OF THINGS JOURNAL, VOL. …, NO. …, MONTH 2019

⎧⎪ Eele * k + E fs * k*d 2 , d ≤ d 0 removing any redundant bits. Actual data routing occurs
E Tx (k , d ) = ⎨ (2) during the steady-state stage, where the collected data is
4
⎪⎩ Eele * k + Emp * k*d , d > d 0 forwarded to the BS by the CHs of the network.
ERx (k ) = ERx _ ele (k ) + kEele (3)
Where Eele is the per bit energy consumption by receiver or Set-up stage:
For the first round, the clusters and CHs are formed using
transmitter. Emp is a transmission parameter for multi-path normal LEACH algorithm, where CHs are selected using
fading and E fs is for the free-space model [31]. equation (4). After data transfer, each node in the network
expends some amount of energy which is different for every
LEACH is a primitive single-hop clustering protocol that node. The expenditure of power depends on the distance
saves an enormous amount of energy as compared to non- separating the sending and receiving nodes represented as ‘d’.
Hence for the next round, the CH is elected using a modified
clustering algorithms [32]. Once the nodes are deployed,
equation given as
sensors group together to form clusters with one CH in each
cluster for data aggregation. The protocol is implemented in ⎧ P E
⎪⎪ × residual k opt ; for all n Ú G
rounds. Clusters are formed dynamically and the cluster heads 1 Einitial
T (n) = ⎨1 − P(r mod ) (5)
are elected randomly. Each node in the cluster has equal ⎪ P
probability to be elected as CH which intends to balance the ⎪⎩0; Otherwise
energy dissipation. The residual energy is checked constantly Where Eresidual is the remaining energy level of the node and
by the sink until the lifetime ends, i.e. all nodes die out their Einitial is the initial assigned energy level. The optimal number
battery power. The steps involved for each round in LEACH is of cluster kopt can be written as in [33].
depicted in the flowchart given in Figure 3. E fs
With ‘n’ sensor nodes in the field, let there be ‘m’ clusters. kopt = n M (6)
2π E d 4 (2m − 1) E − mE
In LEACH algorithm, P represents the probability for a node amp 0 DA
to opt for CH. Before the start of the first round, each node ‘i’ ‘M’ represents the network diameter and E0 is the initial
generates a random number between the interval [0,1]. If the energy supplied to each node.
number is found to be less than a threshold value T(n) given Once the CHs for the current round are selected, they send
by the equation (4), then that node will become CH for that their CH announcement information to member nodes in the
round. respective clusters. The sensing nodes check the signal
P strength of the request message and decide the CHs it wants to
⎧
⎪⎪ ; for all n Ú G join. The CH then broadcasts TDMA (Time Division Multiple
1 Access) schedules for the member nodes to transmit data in
T (n) = ⎨1 − P(r mod ) (4)
⎪ P different time slots to avoid data collision. The process then
⎪⎩0; Otherwise continues for the rest of the rounds till all the nodes in the
network exhaust all its energy.
With each round, the CH changes based on the electing
probability which indicates that all the nodes in the cluster Steady-state stage:
have the same chance to be elected as CH irrespective of its
residual energy. Equi-probable CH election process gives rise
to the possibility of electing a CH with minimal residual
energy which will die out quickly as compared to the one with
comparative higher energy level. Therefore, the residual
energy of each node is included in the equation of election
probability of CH such that the nodes with higher energy level
have a greater chance to be elected as CH. This in return
ensures equal distribution of power in the network thus
enhancing network lifetime.
In order to combat this problem, an advanced algorithm is
proposed called R-LEACH. The algorithm is divided into
rounds with each round consisting of cluster formation and
steady-state stages.

IV. PROPOSED WORK

The proposed protocol represents a hierarchical clustering
algorithm that involves two stages: set-up and steady-state
stages. In the initial set-up phase, the sensor nodes are
deployed in the network and are sub-divided into clusters
headed by a CH responsible for the collection of data from
sensing nodes. The data is fused to reduce the volume by Fig. 4. Flowchart of R-LEACH
Behera et al.: Residual Energy Based Cluster-head Selection in WSNs for IoT Application

During the time slot assigned to each node, transmission of 100

data to CHs occurs. Only the transmitting node remains active LEACH
and all other nodes in the cluster will turn-off its radio to save R-LEACH
80

Number of Dead Nodes

energy. After all the nodes in the cluster have finished
transferring data, the CH will start processing the data. The
CH receives and then aggregates the data to remove any 60
redundancy and compress the information as much as possible
for fair utilization of bandwidth. The CHs then forwards the 40
data to the sink or BS in either single-hop or multi-hop
communication. The entire process is depicted in a flowchart
20
as shown in Fig. 4.

V. SIMULATIONS 0
0 500 1000 1500 2000 2500 3000
The network parameters considered for MATLAB Rounds
simulation for the network model are described in Table 1. Fig. 6. Network Lifetime
The packet size is considered to be 4000 bits. 100 nodes are
deployed randomly with BS placed in the center of the The number of actual data packets sent to sink is shown in
network area as shown in Fig. 5. Fig.7. Since the CHs are selected based on the remaining
TABLE I energy of each node, it effectively reduces the energy
SIMULATION PARAMETERS dissipation in transferring data. As a result, data transmission
Parameters Value frequency increases and more packets are successfully
Network diameter 100 meters2 transmitted to the BS as compared to that in LEACH protocol.
Total number of nodes (n) 100 nodes The average energy expenditure of the network is shown in
Total network energy (E0) 0.5 J
Fig. 8. The residual energy depletes faster in LEACH than that
Energy dissipation: receiving (Eamp) 0.0013 pJ/bit/m4
Energy dissipation: free space model (Efs) 10 pJ/bit/m2
of R-LEACH. Since the energy for modified LEACH depletes
Energy dissipation: power amplifier (Eamp) 100 pJ/bit/m2 at a slow rate, the network lifetime also extends to more
Energy dissipation: aggregation (EDA) 5 nJ/bit number of rounds.
5
x 10
2

1.5 LEACH
Paackets send to BS

R-LEACH

0.5

0
0 500 1000 1500 2000 2500 3000
Rounds
Fig. 7. Packets to BS
50
Fig. 5. Node Deployment LEACH
R-LEACH
A. Network Analysis 40
Average Residual Energy

The simulation result in Fig.6 shows the network life for

both the LEACH and R-LEACH protocols.
30
The first node for LEACH dies out at 1092 round whereas
for R-LEACH it is at 1382 rounds. Similarly, the last node for
LEACH dies out at 1510 rounds whereas for R-LEACH it is at 20
2474 rounds. LEACH protocol assumes CHs dissipates the
same energy for each round that leads to inefficient CH 10
selection and affects the network lifespan. R-LEACH selects
CHs considering the residual energy of nodes and an optimal
number of clusters together thereby enhancing the network 0
0 500 1000 1500 2000 2500 3000
lifetime to more number of rounds. Rounds
Fig. 8. Average Residual energy
 IEEE INTERNET OF THINGS JOURNAL, VOL. …, NO. …, MONTH 2019

2
x 10 network for all the energy values than other routing protocols
for an IoT based sensor network. LEACH does not consider
LEACH
R-LEACH energy level for the threshold value of CH selection; hence the
1.5 first node dies out after less number of rounds in all cases.
CBDAS and GHND are grid-based WSNs that consider few
Throughput

parameters for CH selection and perform better than LEACH

1 protocol. IGHND divides the network into zones and selects a
zone head considering parameters like residual energy,
average distance, and priority of a node. None of the methods
0.5
considers the optimum value of clusters in the network
resulting in inefficient CH selection. R-LEACH takes into
account both the important parameters which result in an
0
0 500 1000 1500 2000 2500 3000 improved stable network.
Rounds
Fig. 9. Throughput

Throughput represents the ratio between actually

transmitted data packets to the successfully received data at
BS or sink. The higher the ratio, the better is the performance.
Fig. 9 show the graph of the throughput of the two protocols.
It is clear that due to the modification in the threshold value of
CH selection, the throughput is increased by 60% for R-
LEACH. Hence, it can be concluded that the modified
protocol performs better than LEACH protocol and can be
used extensively for homogeneous networks.
B. Variation of energy and its effect on Network metrics
Network Stability and Network Lifetime are two important
parameters to deliver good network performance in an IoT- Fig. 11. Half Node Dead
based environment consistently. Network Stability is the time
from the beginning of the network till the death of the first
node (FND) and Network Lifetime is determined by the time
elapsed between FND and LND (death of the last node) in the
network. To study the behavior of proposed R-LEACH
protocol with respect to these metrics, we analyze the
parameters like FND for network stability, LND for network
lifetime and HND in the network. For a fair comparison with
recent methods proposed for IoT based WSN like IGHND
[24], GHND [34], CBDAS [35], we consider the packet size to
be 2000. The initial energy E0 of 0.25, 0.5 and 1 J are
considered for network analysis.

Fig.12. Network Lifetime

The proposed R-LEACH model maximizes the network

lifespan by covering more rounds of operation and
transmitting more packets to the BS. R-LEACH outperforms
LEACH, CBDAS, GHND, and IGHND in terms of HND and
FND for all energy values because it selects stable nodes as
CH. In LEACH, the CH selection is done in a randomized
manner leading to a short span of the network. CBDAS
consumes extra energy in chain formation and data
transmission from header to rest of the nodes adding extra
load to the battery life. GHND and IGHND consider multiple
Fig. 10. Network Stability parameters for zone head selection but does not consider the
number of zones or cluster in the network which effects the
The result in Fig.10 shows R-LEACH provides a more stable network lifetime.
Behera et al.: Residual Energy Based Cluster-head Selection in WSNs for IoT Application

VI. CONCLUSION [14] M. C. M. Thein and T. Thein, “An energy efficient cluster-head
selection for wireless sensor networks,” in Intelligent systems, modelling
Since energy and lifetime are two major constraints in and simulation (ISMS), 2010 international conference on, 2010, pp.
designing any routing protocol for WSN, much research has 287–291.
been done to achieve the goal. Choosing an energy-efficient [15] S. H. Kang and T. Nguyen, “Distance based thresholds for cluster head
selection in wireless sensor networks,” IEEE Commun. Lett., vol. 16, no.
routing algorithm that distributes the load in the network evenly 9, pp. 1396–1399, 2012.
is a challenging process. LEACH protocol ensures an adaptive [16] J. A. Stankovic, “Research directions for the internet of things,” IEEE
algorithm but still has some limitations. A modified CH Internet Things J., vol. 1, no. 1, pp. 3–9, 2014.
selection algorithm has been suggested in this paper that aims to [17] J. Duan, D. Gao, D. Yang, C. H. Foh, and H.-H. Chen, “An energy-
aware trust derivation scheme with game theoretic approach in wireless
extend the network lifetime by controlling the energy sensor networks for IoT applications,” IEEE Internet Things J., vol. 1,
dissipation in the network. The enhanced routing process can be no. 1, pp. 58–69, 2014.
used effectively in scenarios like environmental monitoring [18] H. Junping, J. Yuhui, and D. Liang, “A time-based cluster-head
selection algorithm for LEACH,” in Computers and Communications,
using IoT as the protocol delivers a better result for 2008. ISCC 2008. IEEE Symposium on, 2008, pp. 1172–1176.
homogeneous networks in comparison to LEACH. Simulation [19] K. Maraiya, K. Kant, and N. Gupta, “Efficient cluster head selection
result shows improved network performance for metrics such as scheme for data aggregation in wireless sensor network,” Int. J. Comput.
residual energy, packets sent to BS, throughput and lifetime. Appl., vol. 23, no. 9, pp. 10–18, 2011.
The current work can be extended by considering more [20] B. Singh and D. K. Lobiyal, “A novel energy-aware cluster head
selection based on particle swarm optimization for wireless sensor
parameters for CH selection in a network with mobile nodes networks,” Human-Centric Comput. Inf. Sci., vol. 2, no. 1, p. 13, 2012.
that changes its position frequently. The proposed model can [21] P. C. S. Rao, P. K. Jana, and H. Banka, “A particle swarm optimization
also be tested on different realistic scenarios for a WSN based based energy efficient cluster head selection algorithm for wireless
sensor networks,” Wirel. networks, vol. 23, no. 7, pp. 2005–2020, 2017.
IoT system.
[22] A. Shankar, N. Jaisankar, M. S. Khan, R. Patan, and B. Balamurugan,
“Hybrid model for security-aware cluster head selection in wireless
REFERENCES sensor networks,” IET Wirel. Sens. Syst., 2018.
[23] F. Xiangning and S. Yulin, “Improvement on LEACH protocol of
wireless sensor network,” in Sensor Technologies and Applications,
[1] J. Chase, “The evolution of the internet of things,” Texas Instruments,
2007. SensorComm 2007. International Conference on, 2007, pp. 260–
2013.
264.
[2] D. Bandyopadhyay and J. Sen, “Internet of things: Applications and
[24] H. Farman et al., “Multi-criteria based zone head selection in Internet of
challenges in technology and standardization,” Wirel. Pers. Commun.,
vol. 58, no. 1, pp. 49–69, 2011. Things based wireless sensor networks,” Futur. Gener. Comput. Syst.,
2018.
[3] L. Tan and N. Wang, “Future internet: The internet of things,” in 2010
[25] S. Kallam, R. B. Madda, C.-Y. Chen, R. Patan, and D. Cheelu, “Low
3rd International Conference on Advanced Computer Theory and
Engineering (ICACTE), 2010, vol. 5, pp. V5--376. energy-aware communication process in IoT using the green computing
approach,” IET Networks, vol. 7, no. 4, pp. 258–264, 2017.
[4] W. B. Heinzelman, A. P. Chandrakasan, and H. Balakrishnan, “An
[26] J. Jin, J. Gubbi, S. Marusic, and M. Palaniswami, “An information
application-specific protocol architecture for wireless microsensor
framework for creating a smart city through internet of things,” IEEE
networks,” IEEE Trans. Wirel. Commun., vol. 1, no. 4, pp. 660–670,
Internet Things J., vol. 1, no. 2, pp. 112–121, 2014.
2002.
[5] J. Gubbi, R. Buyya, S. Marusic, and M. Palaniswami, “Internet of [27] E. Souto, G. Guimarães, G. Vasconcelos, M. Vieira, N. Rosa, and C.
Ferraz, “A message-oriented middleware for sensor networks,” in
Things (IoT): A vision, architectural elements, and future directions,”
Futur. Gener. Comput. Syst., vol. 29, no. 7, pp. 1645–1660, 2013. Proceedings of the 2nd workshop on Middleware for pervasive and ad-
hoc computing, 2004, pp. 127–134.
[6] B. Bhuyan, H. K. D. Sarma, N. Sarma, A. Kar, and R. Mall, “Quality of
[28] J. Li and P. Mohapatra, “An analytical model for the energy hole
service (QoS) provisions in wireless sensor networks and related
challenges,” Wirel. Sens. Netw., vol. 2, no. 11, p. 861, 2010. problem in many-to-one sensor networks,” in IEEE vehicular technology
conference, 2005, vol. 62, no. 4, p. 2721.
[7] Q. Jing, A. V Vasilakos, J. Wan, J. Lu, and D. Qiu, “Security of the
[29] B. Elbhiri, R. Saadane, D. Aboutajdine, and others, “Developed
Internet of Things: perspectives and challenges,” Wirel. Networks, vol.
Distributed Energy-Efficient Clustering (DDEEC) for heterogeneous
20, no. 8, pp. 2481–2501, 2014.
wireless sensor networks,” in I/V Communications and Mobile Network
[8] W. R. Heinzelman, A. Chandrakasan, and H. Balakrishnan, “Energy- (ISVC), 2010 5th International Symposium on, 2010, pp. 1–4.
efficient communication protocol for wireless microsensor networks,” in
[30] D. S. Kim and Y. J. Chung, “Self-Organization Routing Protocol
Proceedings of the 33rd Annual Hawaii International Conference on
System Sciences, 2000, p. 10 pp. vol.2-. Supporting Mobile Nodes for Wireless Sensor Network,” in Computer
and Computational Sciences, 2006. IMSCCS ’06. First International
[9] M. Chatterjee, S. K. Das, and D. Turgut, “An on-demand weighted Multi-Symposiums on, 2006, vol. 2, pp. 622–626.
clustering algorithm (WCA) for ad hoc networks,” in Global
Telecommunications Conference, 2000. GLOBECOM’00. IEEE, 2000, [31] T. M. Behera, U. C. Samal, and S. K. Mohapatra, “Energy Efficient
Modified LEACH Protocol for IoT Application,” IET Wirel. Sens. Syst.,
vol. 3, pp. 1697–1701.
vol. 8, no. 5, pp. 223 – 228, 2018.
[10] J. Xu, N. Jin, X. Lou, T. Peng, Q. Zhou, and Y. Chen, “Improvement of
[32] Y. Li, N. Yu, W. Zhang, W. Zhao, X. You, and M. Daneshmand,
LEACH protocol for WSN,” in Fuzzy Systems and Knowledge
“Enhancing the performance of LEACH protocol in wireless sensor
Discovery (FSKD), 2012 9th International Conference on, 2012, pp.
networks,” in Computer Communications Workshops (INFOCOM
2174–2177.
WKSHPS), 2011 IEEE Conference on, 2011, pp. 223–228.
[11] Y.-K. Chen, “Challenges and opportunities of the internet of things,” in
17th Asia and South Pacific Design Automation Conference, 2012, pp. [33] S. Hussain and A. W. Matin, “Energy efficient hierarchical cluster-based
routing for wireless sensor networks,” Jodrey Sch. Comput. Sci. Acadia
383–388.
Univ. Wolfville, Nov. Scotia, Canada, Tech. Rep., pp. 1–33, 2005.
[12] A. Ali, Y. Ming, T. Si, S. Iram, and S. Chakraborty, “Enhancement of
RWSN Lifetime via Firework Clustering Algorithm Validated by [34] H. Farman, H. Javed, J. Ahmad, B. Jan, and M. Zeeshan, “Grid-based
hybrid network deployment approach for energy efficient wireless
ANN,” Information, vol. 9, no. 3, p. 60, 2018.
sensor networks,” J. Sensors, vol. 2016, 2016.
[13] C. Wang, K. Sohraby, B. Li, M. Daneshmand, and Y. Hu, “A survey of
[35] Y.-K. Chiang, N.-C. Wang, and C.-H. Hsieh, “A cycle-based data
transport protocols for wireless sensor networks,” IEEE Netw., vol. 20,
aggregation scheme for grid-based wireless sensor networks,” Sensors,
no. 3, pp. 34–40, 2006.
vol. 14, no. 5, pp. 8447–8464, 2014.
 IEEE INTERNET OF THINGS JOURNAL, VOL. …, NO. …, MONTH 2019

T. M. Behera got her B.Tech degree editor of IEEE Access. He has been on technical program
in Electronics & Comm. Engineering committees of various international conferences and
from Biju Pattnaik University of technical reviewer of various international journals in his
Technology in 2007. She received her field. He is a member of IEEE.
M.Tech in Communication System
from KIIT University in 2012. She Dr. Mahmoud Daneshmand is Co-
has 10 years of teaching experience Founder and Professor of Department
and is currently an Assistant Professor of Business Intelligence & Analytics;
in SOEE at KIIT University, and Professor of Department of
Bhubaneswar. Her research area broadly includes Wireless Computer Science at Stevens Institute
Sensor Networks and its application in IoT. of Technology. He has more than 40
years of Industry & University
Dr. S. K. Mohapatra received his experience as Professor, Researcher,
Ph.D. from National Institute of Assistant Chief Scientist, Executive
Technology, Rourkela, in the year Director, Distinguished Member of Technical Staff,
2016. He is currently an Assistant Technology Leader, Chairman of Department, and Dean of
Professor in SOEE, KIIT University, School at: Bell Laboratories; AT&T Shannon Labs–
Bhubaneswar and having 18 years of Research; University of California, Berkeley; University of
teaching experience. His research Texas, Austin; Sharif University of Technology; University
interests include Modeling and of Tehran; New York University; and Stevens Institute of
Simulation of Nanoscale Devices and its application in IoT. Technology. He received his Ph.D. and M.S. degrees in
Energy-efficient Wireless Sensor Networking, Adhoc Statistics from the University of California, Berkeley; M.S.
Networks, Metamaterial absorbers in THz application, UWB- and B.S. degrees in Mathematics from the University of
MIMO and Reconfigurable Antenna. He has been a part of Tehran.
committee member of various international conferences,
Editorial Board Member and Reviewer of international Dr. Amir H. Gandomi is an
journals. He is a life member of ISTE, IETE, CSI, OITS, and Assistant Professor of Analytics &
member of IEEE. Information Systems at School of
Business, Stevens Institute of
Dr. U. C. Samal, received his B.E. Technology. Prior to joining Stevens,
degree in Electrical Engineering from Dr. Gandomi was a distinguished
VSSUT, Burla, Sambalpur, India in research fellow in headquarter of
2003 and M. Tech. degree in BEACON NSF center located at
Electronic System and Comm. Michigan State University. He received his Ph.D. in
Engineering from National Institute of engineering and used to be a lecturer in several universities.
Technology, Rourkela, in 2006. He Dr. Gandomi has published over one hundred and thirty
obtained his Ph. D. degree from the journal papers and four books. Some of those publications
Department of Electrical Engineering, Indian Institute of are now among the hottest papers in the field and
Technology (IIT), Kanpur, India in 2015. His area of collectively have been cited more than 11,000 times (h-
specialization lies in wireless communication systems, signal index = 52). He has been named as Highly Cited Researcher
processing techniques for communication, 5G wireless (top 1%) for two consecutive years, 2017 and 2018, and one
communication technologies cognitive radio and wireless of the world’s most influential scientific minds. Dr.
sensor networks. He has 3 years of industry and 4 years of Gandomi is currently ranked 20th in GP bibliography
teaching experience. Currently, he is working as Assistant among more than 11,000 researchers. He has also served as
Professor (II) at KIIT University, Bhubaneswar, Odisha. associate editor, editor and guest editor in several
prestigious journals and has delivered several
Dr. Mohammad S. Khan is currently keynote/invited talks. Dr. Gandomi is part of a NASA
an Assistant Professor of Computing technology cluster on Big Data, Artificial Intelligence, and
at East Tennessee State University Machine Learning. His research interests are global
and the director of Network Science optimization and (big) data mining using machine learning
and Analysis Lab (NSAL). He and evolutionary computations in particular.
received his M.Sc. and Ph.D. in
Computer Science and Computer
Engineering from the University of
Louisville, Kentucky, USA, in 2011
and 2013, respectively. His primary area of research is in
ad-hoc networks, network tomography, connected vehicles,
and vehicular social networks. He currently serves as Co-
Editor-in-Chief of International Journal of Grid and High-
Performance Computing (IJGHPC) and as an associate