OPTIMIZED CLUSTER ESTABLISHMENT AND CLUSTER-HEAD SELECTION APPROACH IN WSN

International Journal of Computer Networks & Communications (IJCNC) Vol.13, No.4, July 2021
DOI: 10.5121/ijcnc.2021.13404 53
OPTIMIZED CLUSTER ESTABLISHMENT AND
CLUSTER-HEAD SELECTION APPROACH IN WSN
Shivshanker. P. Biradar1
and Dr. T. S. Vishwanath2
1
ECE department, LAEC, Bidar, Karnataka, India
2
ECE department, BKIT, Bhalki, Karnataka, India
ABSTRACT
In recent years, limited resources of user products and energy-saving are recognized as the major
challenges of Wireless Sensor Networks (WSNs). Clustering is a practical technique that can reduce all
energy consumption and provide stability of workload that causes a larger difference in energy depletion
among other nodes and cluster heads (CHs). In addition, clustering is the solution of energy-efficient for
maximizing the network longevity and improvising energy efficiency. In this paper, a novel OCE-CHS
(Optimized Cluster Establishment and Cluster-Head Selection) approach for sensor nodes is represented to
improvise the packet success ratio and reduce the average energy-dissipation. The main contribution of
this paper is categorized into two processes, first, the clustering algorithm is improvised that periodically
chooses the optimal set of the CHs according to the speed of the average node and average-node energy.
This is considerably distinguished from node-based clustering that utilizes a distributed clustering
algorithm to choose CHs based on the speed of the current node and remaining node energy. Second, more
than one factor is assumed for the detached node to join the optimal cluster. In the result section, we
discuss our clustering protocols implementation of optimal CH-selection to evade the death of SNs,
maximizing throughput, and further improvise the network lifetime by minimizing energy consumption.
KEYWORDS
Wireless Sensor Networks (WSNs), Cluster Heads (CHs), Clustering Approach, Energy Efficient (EE),
LEACH.
1. INTRODUCTION
The improvement in communication technology and the increase in the demand for smart devices
generated novel dimensions for WSNs. WSNs contain different/same types of nodes and BS
(Base Station). The BS transmits commands to the nodes in the region of sense. It gathers data
from SNs (sensor-nodes), executes very simple processing, and transmits the data through the
internet to end-users. WSNs are self-managed network systems that contain several distributed
SNs. The main purpose of WSN is to monitor the conditions like environment, temperature, and
humidity or to recognize monitoring environment, disaster monitoring, traffic-management, and
warfare monitoring [1-2]. Anyways, WSNs have resource constraints. This constraint adds less
memory, limited battery power, and limited bandwidth. Among these, the power of a limited
battery is one of the main constraints. When the energy efficiency, it is pretty challenging to
replace or recharge the battery. There exist many mechanisms like radio optimization, EE
(energy-efficient) routing, duty routing, clustering [3-5] in order to minimize energy
consumption. The diverse applications and resource constraints of WSNs pose higher challenges.
WSNs nodes generally have essential limitations in terms of available energy as well as
communication and computing capabilities.

54
EE is a most eminent feature of the routing protocols for the WSNs, required to prolong the
network-lifetime and to improvise the performance of communication and also receives much
attention. Different communication nodes and cluster formation of sending data have been the
most popular methods. In general, compared to the cluster-based routing-protocols and non-
clustering-protocols can proficiently utilize SNs in a network [6]. A CH, called cluster leader, is
in removing charge as correlated data that can reduce the volume of final data. Afterward, CH
will send aggregated data to BS [7-8]. In the protocols of cluster-based routing, SNs are parted
into several clusters to minimize the consumption of energy for longer communication distances.
The clustering can reduce all energy consumption and also balance workload that causes a larger
difference in energy depletion among other nodes and CHs. Furthermore, clustering is the
solution of EE for maximizing the network longevity and improvising energy-efficiency.
Anyways, clustering protocols implement optimal CH-selection to evade the death of SNs and
further improvise the network lifetime [9-10].
According to [11], SNs consume more energy through communication during the process of
computation. Few protocols assume more than one hop communication and nodes nearer to BS
that has excessive transmission overhead leading to the energy gaps in the field of the sensor
[12]. In order to address the problem of energy and improvise the lifetime of the network, several
clustering protocols have been introduced for the WSNs. The LEACH is an eminent hierarchical
routing protocol with respect to saving energy with routing protocols [13]. In this, all networks
are parted into many clusters and every single cluster chooses the node in a probabilistic manner
as the CH, which is in sending the data, receiving a charge, compressing from non-CH-nodes to
BS. In order to reduce the WSNs-energy consumption, the protocol of LEACH routing defines
only one CH-node in every single cluster and chooses on a rotation basis. Although the efficiency
of the LEACH-protocol has been studied, it has some disadvantages that need improvement due
to choosing the CH based on round-robin, the number of CHs is not reasonable in every single
round, and nodes at the network will be selected as CH. On the other side, there is a consideration
of threshold condition, distribution of CHs, and remaining energy of every single node after the
end of every single round.
The routing protocol of WSNs is categorized into hierarchical, plane, location-based routing [14].
The routing protocol is utilizing as the LEACH protocol. The LEACH protocol is a representative
protocol between the protocols of hierarchical routing [15]. The protocol of LEACH improvised
EE of the network with the help of the clustering method but it doesn’t assume the movement of
situation nodes. Therefore, if the nodes are less as the rate of DT (data-transmission). The
protocol of LEACH-mobile is defined as a protocol, which develops the issue of node
movements [16]. In the protocol of LEACH-mobile, the given nodes are failed and moved to
transmit the data, the issue was resolved by generating a failed node to re-transmitting and re-
cluster the data. Anyways, the energy is consumed for the identification of moved nodes and the
consumption of energy is higher than the LEACH protocol.
In this paper, a novel OCE-CHS (Optimized Cluster Establishment and Cluster-Head Selection)
approach for SNs is represented to improvise the PDR (packet delivery-ratio) and reduce the
average energy-dissipation. The main contribution of this paper is 2-fold, first, the CA (clustering
algorithm) is improvised that periodically chooses the optimal set of the CHs according to the
speed of the average node and average-node energy. This is considerably distinguished from
node-based clustering that utilizes a distributed-clustering algorithm to choose CHs based on the
speed of the current node and remaining node energy. Second, more than one factor is assumed
for the detached node to join the optimal cluster.
The complete paper is organized in such a way that second section demonstrate the complete
literature survey of considered topic, 3rd
section demonstrates the important equation and their

55
respective explanation of our proposed approach. In addition, result and simulation analysis using
proposed model has shown in section 4 and at last we conclude our proposed approach in section
5.
2. LITERATURE SURVEY
Since the SNs are powered by batteries, once a few nodes are exhausted, the network may not
work properly. Therefore, energy constraint has been the main key to operate an efficient WSN.
Few traditional techniques minimize energy consumption and transmission distance of path, thus
extending all network lifetime [17,18]. Authors in [17] choose neighbor-node with fewer hops
from a source node to SN (sink-node) is utilized as a determinant. In paper [18], introduced
ladder-algorithm based ACO (ant-colony-optimization) to solve the issue of the energy
consumption in the routing. The algorithm utilized the ACO method to define the transmit paths
that can efficiently minimize the consumption of energy. In another way, this algorithm basically
based on min-hop count is similar to min-energy routing [19]. Although this type of method can
minimize the consumption of energy, it has some drawbacks that undertake the DT (data
transmission) in a period of time while the other nodes are idle. They select a forwarding node
that doesn’t assume residual energy that is very easy to generate few nodes and an outcome is not
even energy distribution among the nodes. Thus, the lifetime of a network is at less level.
The routing protocols like GBR (gradient-based-routing) [20], LEACH [21], PEGASIS [22], DD
(directed diffusion) [23], SPIN (Sensor-Protocol for Information through Negotiation) [24] were
proposed for the efficient routing of multi-hop in the WSNs [25] [26]. SPIN can’t ensure a
hundred percent packet delivery from the source node to BS. Moreover, the SPIN requires
complete topology knowledge. DD is known as the protocol of data-centric and it is required for
sink-node to generate, re-route, and transfer intermittent-updates packets. The main object of
designing DD was efficiency in the consumption of energy, resulting in maximizing the
expectancy of network life. In order to minimize the consumption of energy, DD utilized data
processing and compression within the network.
A number of various protocols have been introduced for WSN-node location and localization-
based routing protocol. These contain GEDIR [27], GAF [28], SPAN [29], GEAR, GOAFR [30]
and MFR. These papers referred to fact that the main benefit of these given protocols is the
capability to recognize the correct location of SN within sensor-network. The localization of the
node is connected to the efficiency of energy-WSN. It saves the resources of energy-WSN [31].
Anyways, these protocols outcome in the loss of power because of its node-distribution and
geographical-topology in WSN. Whereas in paper [32], the author has exploited a novel load-
balancing protocol for handling the consumed energy by SNs in WSN in 2019. The simulation
was done by utilizing MATLAB, the performance comparison was estimated utilizing 2 protocols
like SEP and LEACH.
The authors [33], introduced GWO (grey-wolf-optimization) to solve the problem of CH
selection. The function of suitable fitness was used to confirm WSN coverage and it is fed in
GWO to discover the optimum. The introduced model outcomes are associated with the routing
protocol of LEACH. Four different models such as performance, a lifetime of residual energy,
network throughput are analyzed. Lastly, the proposed model outperforms LEACH in entire
topologies by various indicators. Whereas in [34], the author introduced an EE-CH-selection
algorithm that is based on WOA-C and WOA (whale-optimization-algorithm) was introduced.
Therefore, it supports the selection of an energy-aware CH-based fitness function that assumes
the remaining energy node and the sum of the adjacent energy nodes. Lastly, an implemented
algorithm was estimated based on throughput, overall stability, EE, a lifetime of the network.

56
Moreover, the WOA-C performance was evaluated against protocols of another standard
contemporary routing to represent its superiority over another model.
The cluster-based routing has been proposed with various contributions. They are classified into
constraints for CH-selection and algorithms for CH-selection. The constraints for the selection
process of CH for routing protocol have been introduced in [35] whereas [36] introduces the
algorithm of metaheuristic optimization for the selection process of CH. Since this research aims
to improvise the optimization algorithm for the selection process of CH, they assume [35-36] for
further analysis and study. Moreover, their paper focused on FA and DA that take the update
based on random solutions, distribution, and encircling. It leads an algorithm to coverage prior to
traditional algorithms.
This paper [37] introduces an improvised function of heuristic in an ACO and assumes residual
energy and distances of nodes to discover an optimal path of DT. In [38], the mechanism for the
WSN-routing that can be more effective regarding node energy, criteria of the route length, end-
to-end delay is represented. This technique utilized ant-colony based on the routing algorithm and
the local inquiries is to discover an optimal route. The above algorithm [37-38] plays much
attention to residual path and path-length while ignoring the effects of energy-statistics like
minimum energy and average energy that could lead to an imbalance in an energy-dissipation of
all networks. Whereas, paper [39] represent ACO based routing-protocol for the WSNs
containing the stationary nodes. It gives an effective method of multi-path-DT to obtain reliable
communication in node faults while assuming the levels of energy level. Anyways, an algorithm
considers the given network, which is static and can’t be implemented in more than one sink
node.
3. OPTIMIZED CLUSTER ESTABLISHMENT AND CLUSTER-HEAD
SELECTION
The cluster-based WSN-SN (sensor-node) with single hop-communication and gateways inside
the cluster is represented in below figure-1. The advantages of cluster-based WSN are given as
follows: per cluster one representative node requires to be involved in the routing process and DA
(data-aggregation), it significantly minimizes the consumption of energy. Additionally, it
conserves bandwidth of communication as the SNs requires to interact with CHs and evades
redundant messages between them. The cluster can be managed in terms of energy; it improvises
the network-scalability. However, it is most crucial if SNs are not-properly allocated to CHs for
the formation of the cluster, then a few CHs may have overloaded with a higher number of SNs.
Such overloaded may maximize the latency in the communication and degrade the performance
of the WSN.

57
Figure 1. Sensor-node and Cluster-based WSN
The WSN model is examined with the SNs. It is considered as a network model that contains
SNs scattered in an area of meters. An assumption of the network model is represented
below:
 The BS (Base station) is fixed and it is located in the field of the sensor.
 The BS and FNs are neither energy-constraint while the SNs are energy-constraint.
 Every single node knows its energy level, velocity, and current position.
 Every single node can use the power-setting in order to change transmitted power through
RSSI (received signal-strength-location).
We believe that given assumptions are reasonable for real networks. The FNs are powered by
energy-harvesting and thus have the continuous power of supply. Every single node identifies its
current positions and speed that can be achieved by the algorithms of GPS and location.
Additionally, every single node can estimate its energy level based on specific hardware of the
energy model. GPS is to estimate the position. Anyways, it is very expensive to train with all
nodes and that can’t function in the indoor-applications while there is no way to for LOS (line-of-
sight) to a satellite.
The consumption of SNs is measured by the same model as energy dissipation of radio-hardware
where transmitter dissipates of energy is to run power amplifier and radio electronics, and also
receiver disperses the energy to rum radio electronics. Where is defined as the distance between
transmitter and receiver. If distance is less to the assumed threshold the open model of
channel space is assumed with the loss of energy ( or else multipath channel model ( ) is
assumed. In order to receive and transmit a number of the bit message signal at the distance ,
the cost of receiving and transmitting is given by: .
Where distance is less then
(1)

58
Where distance is higher or similar to
(2)
Here, defines the energy bit/dissipation at the receiving and transmitting process, where
defines for an amplified energy parameter and also corresponds to the open model of space
channel. Furthermore, is defined as the amplified energy parameter and corresponds to the
model of the multipath channel.
From the given network, all SNs is given as , defines the scale of time
polling ( ). Begin stage, the round for every single message towards the BS, where
range value of from 1 to . Anyways, a transferred message consists of node-location,
node level of energy and present the node speed. According to the transmitted message,
firstly BS calculates the average energy node followed by
.
Also, we calculate the average speed of the node as:
(3)
Where . The BS selects the number of nodes from set ( ) and assumed as:
(4)
The given symbol defines for the optimal CHs number, represents node sets, and
represents cardinality. Furthermore, the level of energy nodes is given above,
then the value of average/threshold energy and calculated current speeds are given below, then
the speed of the average value can be given as and
.
While assuming a condition, where , this implies that the number of set nodes in
have higher energy levels and lower speeds are trained to become the CHs during the particular
round. Whereas assuming a condition , this condition implies that the set of nodes
in having higher energy levels and lower speeds that are probable to become the CHs at the
time of a particular round.
The number of set nodes is assumed to carry the process of optimization to obtain optimal CHs
from that becomes the CHs at a specific round. Here, we have many processing levels in order
to obtain the optimal CHs like completion/final level, initialization level, and neighborhood level.
Firstly, the set of stochastically CHs is created and defined by from the given
and to confirm . is assumed to discover out a new set of the nodes
which represents the random perturbation of the nodes at the finding level of the

59
neighborhood. Furthermore, novel coordinates of nodes in are calculated via the
coordinates of in .
(5)
(6)
Where the function of the rand is assumed to create a random number and defines the
maximal change at present random perturbation.
Afterward, it found neighborhood-nodes in that placed around , so node chooses a new
CHs set and defined with the help of (this can be written as ). The current
state is represented via CHs set with the help of cost-function whereas the neighborhood
state is represented via CHs set with the cost-function .
Let assume iteration, the current state will alter to the state of the neighborhood and
probability is calculated as follows:
(7)
The given eq. (11) is true when , otherwise the the probability will be only
one. Where, defined as the parameter of the control factor and given by:
(8)
Whereas, is defined as the value of hyperparameter that ranges from 100-1000. The
calculated value help to decide better clusters and denotes the function of cost:
(9)
Where, the distance among channel head and is specified by the help of . if current
state becomes the state of the neighborhood, then .
is assumed as total iterations in the final level, where is equal and greater to then the
algorithm can be dismissed otherwise it will go to the level of neighborhood find. Once a
determination of an optimal CHs is finished, BS is utilized to broadcast the message that contains
CH’s-ID for every single node which is present in the network. Anyways, the particular ID of a
node is similar to CH’s ID then it became the CH. Else, the node selects its TDMA-slot for the
process of data transfer and the given node falls dead until the time that has come to the forward
signal to its CH.
As node and its CH has moved out of the communication range with each other, the
transmitted data packet is assumed to be lost from to . We refer a node that is separated
from its CH as DN (detached-node). In this case, the DN differs the quantity of the transmitted
power and then shows data to all given CHs in a network. On the other side, in order to define an
optimal-CH for DN, every single CH sends small messages to BS. This small message contains

60
distance among CH and DN , the energy level of CH i.e., and the speed of
current CH i.e, . Then, BS calculates integrated weights for the CH as follows:
(10)
Where, , , and C is the CHs set within a present round. Moreover, is defined as a
weighted factor and . Besides, simplifies the range of maximum
transmission and defines the speed of max MNs. Since BS is in the area center, we describe
where is defined as the side length of square-area. Thus, an optimal CH for
DN is given by:
(11)
After, BS discovers an optimum CH for DN, it sends messages that contain an optimal CH-ID
to DN, which transmits a joint-request message to the CH Next, the CH allots free-timeslot to
DN. During its allotted timeslot, DN attached itself and sends sensed data to its CH .
4. RESULT AND ANALYSIS
The SNs are powered by batteries, if some of the nodes are fatigued, the network may not work
properly. Therefore, energy constraint has been the main key to operate an efficient WSN. The
cluster-based routing has been proposed with various contributions. In this research, our aim is to
improvise the algorithm for the selection process of CH, increase the throughput and minimize
energy consumption. This section provides the result and analysis at the proposed OCE-CHS
approach. The complete simulation has done in MATLAB 2018a with system configuration; Intel
i5 processor, 12GB RAM, and Windows operating system.
Here, WSN is analyzed under various sensor nodes such as; 700, 800, 1900 and 2000 which
consistently scattered in the 400 x 400-meter squared network coverage area. In addition, real
time environment of network is replicated; where energy level, nodes current position and
velocity known by node. Before exploring further result analysis, let’s have discuss about
LEACH, it’s a well-known hierarchical routing protocol with respect to saving energy with
routing protocols. In order to reduce the WSNs-energy consumption, the protocol of LEACH
routing defines only one CH-node in every single cluster and chooses on a rotation basis, which
is not efficient to manage energy consumption. The proposed approach utilizes a distributed-
clustering algorithm to choose CHs based on the speed of the current node and remaining node
energy. For analysis and evolution purpose; packet success rate, error rate, throughput and energy
consumption has considered. Finally, state-of-art methods such as; ACO, LEACH and energy
balance clustering routing algorithm (EB-CRA) [40] has considered for comparison.
Table 1. Simulation parameters details
Initial Energy 0.1 dB
Optimal Election Probability for CH 0.2
Sensor Nodes 700, 800, 1900 and 2000
Network Size 400 x 400m2
Hyperparameter value 1
Maximum number of polling 500
Existing Approach ACO, LEACH and EB-CRA [40]

61
Figure 2. Cluster formation and CHs Localization
Figure 3. Impact of Polling Number on Packet Success Rate at 700 Nodes

62
Figure 4. Impact of Polling Number on Packet Success Rate at 2000 Nodes
The representation of cluster formation and CHs localization is shown in figure 2, where blue is
dividing the clusters in a 400 x 400-m2
network and yellow covered dot shows the CH for each
cluster.
Figure 5. Impact of Polling Number on Throughput at 700 & 800 Nodes
The impact of Polling Number on Packet Success Rate (PSR) at 700 Nodes is shown in figure 3,
it’s a box plot that represents the standard deviation of 100 poll for PSR values. The x-axis shows
in the interval at 100 polling number. Red line inside box represents the mean value of PSR at

63
total 100 polling number. In figure, it clearly seen that mean value of initial 100 poll is 93.2%,
whereas at last 400 to 500 poll we achieve 99.95% of PSR.
Figure 6. Impact of Polling Number on Throughput at 1900 & 2000 Nodes
Figure 7. Error Rate w.r.t Polling Number at 700 & 800 Nodes

64
Figure 8. Error Rate w.r.t Polling Number at 1900 & 2000 Nodes
Similarly, when 2000 nodes were considered for analysis of polling number w.r.t PSR as shown
in figure 4, at 0 to 100 poll we got PSR mean value as 84.5% and at 400 to 500 poll we achieved
98.24 PSR. Both figure 3 and 4 shows that as per increment in polling number the PSR is also
increasing. The Impact of polling number on throughput at 700 & 800 nodes are shown in figure
5, cyan line shows for 700 nodes, while the yellow line represents for 800 nodes. In figure 5, a
growth in throughput can be seen from 100 to 300 polls, afterwards it increasing a bit because of
its optimal reach of space.
Figure 9: Impact of Polling Number on Energy Consumption at 700 Nodes

65
Similarly, the Impact of polling number on throughput at 1900 & 2000 nodes can be seen in
figure 6, at shown 1900 nodes, we have got 58 Kbits/Hz at 500 polling number while at 2000
nodes got 65 Kbits/Hz. Figure 7 shows an error rate w.r.t polling number at 700 & 800 nodes,
where blue bar shows for 700 nodes and yellow bar shows for 800 nodes.
Figure 10. Impact of Polling Number on Energy Consumption at 2000 Nodes
Figure 11. Presentation of energy comparison at different polling number
In figure 7 polling number 1, the error rate is 0.0583 for 700 nodes and 0.0573 error rate for 800
nodes. But increasing in polling number, the error rate is more in high number of nodes. When
we further increase the number of nodes to 1900 and 2000, the error rate w.r.t polling number can
be seen in figure 8. Where at 1 polling number, 0.141 error rate at 1900 nodes and 0.148 error
rate at 2000 nodes.

66
Table 2. Comparison analysis at energy consumption with state-of-art model
Poll No. ACO LEACH EB-CRA OCE-CHS
100 0.228 0.3501 0.1394 0.1215
200 0.2709 0.3 0.1141 0.1022
300 0.2 0.2649 0.0791 0.0763
400 0.1723 0.3248 0.1376 0.0723
500 0.175 0.2511 0.1243 0.059
Figure 12. Latency impact as per polling number at 700 nodes
Figure 13. Latency impact as per polling number at 2000 nodes

67
Figure 9 represents the impact of polling number on energy consumption at 700 nodes. In 0 to
100 polling number, average power consumption is 0.067 J as seen in fig 9, whereas from 400 to
500 polling number average power consumption is 0.0004 J that is very less compared to
previous polling range. Similarly, the impact of polling number on energy consumption at 2000
nodes is shown in figure 10, where average power consumption is 0.15 J at 0 to 100 polling
number and 0.0175 J at 400 to 500 polling numbers. Figure 11 presents energy comparison at
different polling number and the detailed numerical analysis is shown in table 1. For this
particular energy comparison, a WSN model with the SNs is observed and our network model
consists of 200 number of SNs that uniformly distributed in monitoring area of 100 x 100 meter.
At 100 polling number, our proposed OCE-CHS approach has got 0.1215 J, which is 46%, 65%
and 12% less energy consumption compared to ACO, LEACH and EB-CRA techniques. At 400
poll number, our proposed OCE-CHS approach has got 58%, 77% and 47% less energy
consumption compared to ACO, LEACH and EB-CRA techniques. Transmission of data from
source node to destination node is considered in order to achieve the delay requirements.
Therefore, the less latency is required to improve used experience and higher reliability. Network
transmission latency impact as per polling number at 700 nodes is shown in figure 12 and for
2000 nodes is shown in figure 13. While analyzing both the figs 12 and 13, its clearly seen that
normalize the value of time has achieved as per increasing in polling number.
5. CONCLUSIONS
In WSN, the deployment of large and autonomous SNs is a very difficult task to perform. There
are various other challenges as well; EE protocol to minimalize power consumption, high
throughput, and packet transmission. The major advantage of cluster-based WSN are: per cluster
one representative node requires to be involved in the routing method and DA, which
significantly reduces the energy consumption. In addition, it conserves bandwidth of
communication as the SNs requires to interact with CHs and avoids redundant messages in
transmission. In this study, the OCE-CHS approach for SNs is represented to improvise the
throughput, PDR and reduce the average energy-dissipation and error rate. The entire analysis has
done under a large network scenario, where simulation has set up in MATLAB tool with user’s
alteration. The routing protocol such as LEACH protocol, ACO, and EB-CRA has considered for
energy comparison is considered real-time network environment. The analysis of our result has
shown in the above section clearly state that our proposed approach is much more efficient and
stable for data transmission in WSN. Here, our optimized cluster-based approach managed in
terms of energy and it improvises the network scalability. However, this approach is not tested
under a complete 5G environment, so this can be considered as future scope of work.
CONFLICTS OF INTEREST
The authors declare no conflict of interest.
ACKNOWLEDGMENTS
Both authors contributed to the design and implementation of the research, to the analysis of the
results, and to the writing of the manuscript.
REFERENCES
[1] Li Li, Xiaoguang Hu, Ke Chen, Ketai He (2011) “The applications of wiﬁ-based wireless sensor
network in internet of things and smart grid.” In 2011 6th IEEE Conference on Industrial Electronics
and Applications, pages 789–793.

68
[2] Bressan Nicola, Bazzaco Leonardo, Bui Nicola, (2010) “The deployment of a smart monitoring
system using wireless sensor and actuator net-works.” In 2010 First IEEE International Conference
on Smart Grid Communications, pages 49–54.
[3] Rezaei Zahra, Mobininejad Shima (2012) “Energy saving in wire-less sensor networks.” Int Journal
Comp Sci Eng Surv 3(1): Comp Net 38(4):393–422.
[4] Rault Tifenn, Bouabdallah Abdelmadjid, Challal Yacine (2014) “Energy eﬃciency in wireless sensor
networks: A top-down survey.” Comp Net 67:104–122
[5] S. Srivastava, M. Singh and S. Gupta, (2018) "Wireless Sensor Network: A Survey," International
Conference on Automation and Computational Engineering (ICACE), Greater Noida, India, pp. 159-
163.
[6] Zungeru Adamu Murtala, Ang Li-Minn, Seng Kah Phooi (2012) “Classical and swarm intelligence-
based routing protocols for wire-less sensor networks: A survey and comparison.” Journal Net Comp
Appl 35(5):1508–1536.
[7] Vijayalakshmi K, Anandan P (2019) “A multi objective tabu particle swarm optimization for
eﬀective cluster head selection in wsn.” Cluster computing 22(5):12275–12282.
[8] Solaiman Basma (2016) “Energy optimization in wireless sensor networks using a hybrid k-means
pso clustering algorithm.” Turkish Journal of Electrical Eng Comp Science 24(4):2679–2695.
[9] T. Kaur and D. Kumar, (2018) "Particle Swarm Optimization-Based Unequal and Fault Tolerant
Clustering Protocol for Wireless Sensor Networks," in IEEE Sensors Journal, vol. 18, no. 11, pp.
4614-4622, 1.
[10] S. Jain and N. Agrawal, (2020) "Development of Energy Efficient Modified LEACH Protocol for IoT
Applications," 12th International Conference on Computational Intelligence and Communication
Networks (CICN), Bhimtal, India, pp. 160-164.
[11] Obaidat, M.S.; Misra, S. (2014) “Principles of Wireless Sensor Networks”; Cambridge Univ. Press:
Cambridge, UK.
[12] Alnawafa, E.; Marghescu, I. (2018) “New energy efficient multi-hop routing techniques for wireless
sensor networks: Static and dynamic techniques”. Sensors, 18, 1863.
[13] Jadoon, R.; Zhou, W.; Jadoon, W.; Ahmed Khan, I. (2018) “RARZ: Ring-zone based routing protocol
for wireless sensor networks.” Appl. Sci., 8, 1023.
[14] Mahfuz, S., & Yesmin, F. (2014). “An approach to wireless nano sensing network routing protocol
based on energy and lifetime.” Int J Sci Technol Res, 3(2), 127-130.
[15] Heinzelman, W. R., Chandrakasan, A., & Balakrishnan, H. (2000). “Energy-efficient communication
protocol for wireless microsensor networks.” Proceedings of the 33rd annual Hawaii international
conference on System sciences IEEE, pp.10.
[16] Kim, D. S., & Chung, Y. J. (2006, June). Self-organization routing protocol supporting mobile nodes
for wireless sensor network. In Computer and Computational Sciences, IMSCCS'06. First
International Multi-Symposiums, Vol. 2, pp. 622-626.
[17] Chiang, S.; Huang, C.; Chang, K. (2007) “A Minimum Hop Routing Protocol for Home Security
Systems Using Wireless Sensor Networks” IEEE Press: New York, NY, USA, Volume 53, pp. 1483–
1489, ISBN 1558–4127.
[18] Ho, J.; Shih, H.; Liao, B. (2012) Chu, S. “A ladder diffusion algorithm using ant colony optimization
for wireless sensor networks.” Inf. Science, 192, 204–212.
[19] Suh, Y.; Kim, K.; Shin, D.; Youn, H. (August 2015) “Traffic-Aware Energy Efficient Routing
(TEER) Using Multi-Criteria Decision Making for Wireless Sensor Network.” In Proceedings of the
2015 5th International Conference on IT Convergence and Security (ICITCS), Kuala Lumpur,
Malaysia, 24–27.
[20] Migabo, M. E., Djouani, K., Kurien, A. M., and Olwal, T. O. (2015). “A stochastic energy
consumption model for wireless sensor networks using GBR techniques.” AFRICON, 1–5. doi:
10.1109/AFRCON.2015.7331987.
[21] Heinzelman, W. R., Chandrakasan, A., and Balakrishnan, H. (2000). “Energy-efficient
communication protocol for wireless microsensor networks,” in Proceedings of the 33rd Annual
Hawaii International Conference on System Sciences IEEE, 10. doi: 10.1109/HICSS.2000.926982.
[22] Kim, D., Ha, S., and Choi, Y. (1998). “K-hop cluster-based dynamic source routing in wireless ad-
hoc packet radio network,” in 48th
IEEE Vehicular Technology Conference, VTC 98, Vol. 1, Ottawa,
ON, 224–228. doi: 10.1109/VETEC.1998.686541.

69
[23] Intanagonwiwat, C., Govindan, R., and Estrin, D. (2000). “Directed diffusion: a scalable, and robust
communication paradigm for sensor networks,” in Proceedings of the 6th Annual International
Conference on Mobile Computing, and Networking, Rome:IEEE, pp. 56–67.
[24] Heinzelman, W. R., Kulik, J., and Balakrishnan, H. (1999). “Adaptive protocols for information
dissemination in wireless sensor networks,” in Proceedings of the 5th Annual ACM/IEEE
International Conference on Mobile Computing, and Networking IEEE, Seattle, WA.
[25] Lindsey, S., and Raghavendra, C. S. (2002). “PEGASIS: Power Efficient Gathering in Sensor
Information Systems,” in Proceedings of IEEE Aerospace Conference: IEEE.
[26] Zonouz, A. E., Xing, L., Vokkarane, V. M., and Sun, Y. (2016). “Hybrid wireless sensor networks: a
reliability, cost and energy-aware approach.” IET Wirel. Sensor Syst. 6, 42–48. doi: 10.1049/iet-
wss.2014.0131.
[27] Stojmenovic and X. Lin, (1999) “GEDIR: loop-free location based routing in wireless networks,” in
Proceedings of the International Conference on Parallel and Distributed Computing and Systems,
Boston, Mass, USA, pp. 109–115.
[28] Y. Xu, J. Heidemann, and D. Estrin, (2001) “Geography-informed energy conservation for ad hoc
routing,” in Proceedings of the 7th Annual International Conference on Mobile Computing and
Networking, Rome, Italy, pp. 215–222.
[29] B. Chen, K. Jamieson, and H. Balakrishnan (2001) “Span: an energy-efficient coordination algorithm
for topology maintenance in Ad Hoc wireless networks,” in Proceedings of the 7th ACM
International Conference on Mobile Computing and Networking, pp. 315–322, July 2001.
[30] Y. Yu, D. Estrin, and R. Govindan, (2001) “Geographical and energy aware routing: a recursive data
dissemination protocol for wireless sensor networks,” UCLA Computer Science Department, Los
Angeles, Calif, USA, Tech. Rep. UCLA/CSD-TR-01-0023
[31] F. Kuhn, R. Wattenhofer, and A. Zollinger, (2003) “Worst-case optimal and average-case efficient
geometric ad-hoc routing,” in Proceedings of the 4th ACM International Symposium on Mobile Ad
Hoc Networking and Computing, Annapolis, Md, USA, pp. 267–278.
[32] Ennaciri A., Erritali M., & Bengourram J (2019). “Load balancing protocol (EESAA) to improve
quality of service in wireless sensor network.” Procedia Computer Science, 151, 1140–1145.
[33] Sharawi, M. & Emary, E. (2017). “Impact of grey wolf optimization on WSN cluster formation and
lifetime expansion.” In 2017 Ninth international conference on advanced computational intelligence
(ICACI), Doha, pp. 157–162.
[34] Jadhav, A. R. & Shankar, T. (2017). “Whale optimization-based energy-efficient cluster head
selection algorithm for wireless sensor networks.” Neural and Evolutionary Computing.
arXiv:1711.0938 9.
[35] Yahiaoui, S., Omar, M., Bouabdallah, A., Natalizio, E., & Challal, Y. (2018). “An energy efficient
and QoS aware routing protocol for wireless sensor and actuator networks.” AEU - International
Journal of Electronics and Communications, 83, 193–203.
[36] Tianshu, W., Gongxuan, Z., Xichen, Y., & Ahmadreza, V. (2018). “Genetic algorithm for energy-
efficient clustering and routing in wireless sensor networks.” Journal of Systems and Software, 146,
196–214.
[37] Y. Sun, W. Dong and Y. Chen (2017) "An improved routing algorithm based on ant colony
optimization in wireless sensor networks" IEEE Communication. vol. 21, no. 6, pp. 1317-1320.
[38] E. Amiri, H. Keshavarz, M. Alizadeh, M. Zamani and T. Khodadadi, (2014) "Energy efficient routing
in wireless sensor networks based on fuzzy ant colony optimization", Int. J. Distrib. Sensor Network,
vol. 10, no. 7.
[39] S. Okdem and D. Karaboga, (2009) "Routing in wireless sensor networks using an ant colony
optimization (ACO) router chip", Sensors, vol. 9, no. 2, pp. 909-921.
[40] Y. Xu, Z. Yue and L. Lv, (2019) "Clustering Routing Algorithm and Simulation of Internet of Things
Perception Layer Based on Energy Balance," in IEEE Access, vol. 7, pp. 145667-145676.

70
AUTHORS
SHIVSHANKER. P. BIRADAR Completed B.E (Electronics & Communication
Engineering), M. Tech (Communication System) and perusing Ph.D. from VTU,
Belagavi, India. Research work at congestion control in sensor network using routing
techniques. The area of interest in wireless communication, digital processing, etc.
Dr.T.S.Vishwanath Ph.D. in Control System, Institution: Dr. MGR Educational and
Research Institute University Chennai India, M. E. in Control System, Institution:
WCE Sangli, BE in Electrical Engineering, Institution: PDA College of Engineering
Gulbarga. Research work at “Robust H∞ control and analysis” application to
electrical drives. Member of MIE, ISTE, IETE and IJEET.

OPTIMIZED CLUSTER ESTABLISHMENT AND CLUSTER-HEAD SELECTION APPROACH IN WSN

More Related Content

What's hot

Similar to OPTIMIZED CLUSTER ESTABLISHMENT AND CLUSTER-HEAD SELECTION APPROACH IN WSN

More from IJCNCJournal

Recently uploaded

OPTIMIZED CLUSTER ESTABLISHMENT AND CLUSTER-HEAD SELECTION APPROACH IN WSN