Uneven clustering and fuzzy logic based energy-efficient wireless sensor networks

Indonesian Journal of Electrical Engineering and Computer Science
Vol. 25, No. 2, February 2022, pp. 1011~1019
ISSN: 2502-4752, DOI: 10.11591/ijeecs.v25.i2.pp1011-1019  1011
Journal homepage: http://ijeecs.iaescore.com
Uneven clustering and fuzzy logic based energy-efficient
wireless sensor networks
Mohammed Adnan Altaha1
, Ahmed Adil Alkadhmawee2
, Wisam Mahmood Lafta3
1
Department of Veterinary Public Health, College of Veterinary, Universitiy of Basrah, Basrah, Iraq
2
Department of English, College of Education for Human Sciences, Universitiy of Basrah, Basrah, Iraq
3
Department of Computer Science, Universitiy of Technology, Baghdad, Iraq
Article Info ABSTRACT
Article history:
Received Jul 17, 2021
Revised Dec 1, 2021
Accepted Dec 9, 2021
Clustering is the fundamental issue in terms of ensuring long-term operation
of wireless sensor networks (WSNs). The problem of hot spots remains the
most prominent research challenge relating to the design of energy-efficient
clustering algorithm. This paper proposed a protocol, namely an uneven
clustering and fuzzy logic-based energy-efficient (UCFLEE), for prolonging
network lifetime. Depending on the communication distance, the UCFLEE
protocol divides the network into uneven clusters for suppressing the hot
spot problem. The fuzzy logic selects the optimal cluster head in accordance
with certain parameters. The advocated method adopts a dynamic energy
threshold to chnage the cluster head. The UCFLEE protocol is dependent on
the iterative deepening A (IDA) star algorithm for identifying the routing
path from the cluster heads to the base station. The IDA-star method is
reliant upon a cost bounded method to select the optimal solution for the
base station. The UCFLEE protocol is tested and subsequently contrasted
with other protocols. The results obtained from the UCFLEE protocol enable
an energy consumption equilibrium, eradicates the hot spot challenge, while
also attaining maximum network lifetime.
Keywords:
Cost bounded
Energy threshold
Fuzzy logic
IDA star algorithm
Uneven clustering
This is an open access article under the CC BY-SA license.
Corresponding Author:
Mohammed Adnan Altaha
Department of computer sciences, College of Veterinary, Universitiy of Basrah
Basrah, Iraq
Email: mohammed.altaha@uobasrah.edu.iq
1. INTRODUCTION
Wireless sensor networks (WSN) is a significant and evolving form of communications network that
may be adopted in order to sense numerous environmental and physical parameters (for example humidity,
smoke, pressure and temperature) [1], [2]. A WSN is formed from integrated and miniaturised sensor nodes,
embedded systems, wireless communications, in addition to other technologies [3]. WSN nodes have limited
energy resource capabilities, whole typically being unreachable and unmanned [4]-[6]. Accordingly,
conserving energy and the means of identifying an energy-efficient strategy for extending network lifetime
have emerged as fundamental challenges in relation to WSN design.
Clustering is the most prominent issue in terms of accommodating the limited resources of sensor
nodes in WSNs, particularly in relation to energy capacity [7], [8]. Clustering aims to diminish the network’s
energy consumption by gathering those nodes possessing equivalent characteristics, or those nodes in close
proximity, to form clusters. The base station (BS) elects the cluster head (CH) per cluster so as to manage the
cluster activities. The CHs are responsible for aggregating the sensed data in order to measure physical
phenomenon of interest from their member nodes. Subsequently, the CHs forwards the aggregated data
directly to the BS or via relay CHs [8], [9].

 ISSN: 2502-4752
Indonesian J Elec Eng & Comp Sci, Vol. 25, No. 2, February 2022: 1011-1019
1012
Numerous clustering protocols have been presented with the aim of lengthening the network
lifespan via optimising energy management. Heinzelman et al. [10], the authors designed a first clustering
protocol, namely low energy adaptive clustering hierarchy (LEACH), which adapted the network to evenly
share an energy load among the nodes. The LEACH protocol formed nodal clusters and adopted a local node
as a head of members per cluster. The authors proposed a LEACH-C as the protocol, which enhances the
performance of the LEACH protocol [11]. This protocol involves the BS selecting various CHs and placing
each CH at the centre of a cluster. Liang et al. [12], the authors proposed the PSO-C protocol to provide the
WSN with a higher lifetime. The proposed scheme applies the PSO algorithm as a means of calculating the
optimal CH as well as the fitness function, thus optimising the WSN’s energy efficiency and reducing
consumption. The aforementioned protocols used the single communication approach between the CHs and
the BS. The CHs suffer from preliminary death when they are located at great a distance from the BS [13], [14].
Cengiz and Dag [15] presented a novel protocol called multi-hop low energy fixed clustering
algorithm (MLEFCA), as a means of limiting the energy dissipation. The MLEFCA protocol offers a multi-
hop routing to the BS via electing the closer neighbour CH as a relay node. Selvi et al. [16], the researchers
proposed the honey bee optimization (HBO) technique in order to balance energy consumption, through
selecting the optimum routing path. The HBO technique utilised the enhanced k-means algorithm to form the
clusters, in addition to the HBO algorithm to determine the path to the BS. The balanced residual energy-
LEACH (BRE-LEACH) is an original protocol introduced to expand network lifetime [17]. The BRE-
LEACH protocol depends on the remaining energy to select the best CH. This proposed approach selects the
optimal CH as the root CH. The farthest CHs used the multi-hop path to aggregate data at the root CH. In
multi-hop wireless communication, the CHs nearest to the BS aggregate the data packet from the farther
CHs. The CHs nearest to the BS are exerting additional energy compared with other CHs, as a result of data
dissemination and heavy traffic. This creates a hot spot problem in WSNs and swifter expenditure of energy
by the CHs [18]-[21]. Consequently, selecting the CH and resolving the hot spot problem are the foremost
challenges to account for while designing energy efficient clustering.
For load balance achievement and mitigation of the hot spot problem, this paper advocates a
protocol named uneven clustering and fuzzy logic-based energy-efficient (UCFLEE). Based on the
communication distance, the UCFLEE protocol divides the network area into two sectors of different sizes.
The smaller sector is situated in closer proximity to the BS, whereas the larger sector is located farther away
from the BS. The larger sector is further divided into equal size sectors in accordance with the communication
distance. The proposed protocol utilises fuzzy logic as a means of identifying optimal CHs. The CH change is
dependent on the energy threshold to equally distribute the CHs’ roles between the nodes. The proposed scheme
developed the iterative deepening A (IDA) algorithm to establish the multi-hop path to the BS.
The remainder of this paper is as shown in: section 2 describes the system model. The UCFLEE
protocol is discussed with all its details in section 3, while section 4 details the UCFLEE protocol’s overall
performance following the completion of the simulation trials. In section 6, the conclusions obtained from
this paper are presented.
2. SYSTEM MODEL
2.1. Network model
The network comprises of numerous sensors that are disseminated randomly throughout the
network. The following properties describe the network sensors: i) The BS is immobile and aware of the
nodes locations; ii) The BS used a sufficient amount of resources to manage the network; iii) Nodes are
static, with each sensor having a unique identification while also being unaware of the location; and iv) Initially,
nodes have the same amount of appropriated energy, computation capabilities and communication power.
2.2. Energy model
The node battery is consumed significantly via the data communication process (data transmission
and data reception). The first radio model is used to compute the energy consumed by the nodes [11].
The energy consumed to transmit (𝐸𝑐−𝑡𝑥) and receive (𝐸𝑐−𝑅𝑥) n-bit data over communication distance
d metres may be calculated by (1)-(3):
𝐸𝑐−𝑡𝑥(𝑛, 𝑑) = {
𝑛 × 𝐸𝑒𝑙𝑒𝑐 + 𝑛 × ∈𝑓𝑠 × 𝑑2
𝑛 × 𝐸𝑒𝑙𝑒𝑐 + 𝑛 × ∈𝑚𝑝 × 𝑑2
}
𝑑 ≤ 𝑑𝑇ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑
𝑑 > 𝑑𝑇ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑
(1)
𝐸𝑐−𝑅𝑥(𝑛, 𝑑) = 𝑛 × 𝐸𝑒𝑙𝑒𝑒𝑐 (2)

Indonesian J Elec Eng & Comp Sci ISSN: 2502-4752 
Uneven clustering and fuzzy logic based energy-efficient wireless sensor … (Mohammed Adnan Altaha)
1013
𝐸𝑐−𝑡𝑜𝑡𝑎𝑙 = 𝐸𝑐−𝑡𝑥(𝑛, 𝑑) + 𝐸𝑐−𝑅𝑥(𝑛, 𝑑) (3)
𝐸𝑒𝑙𝑒𝑐 indicates the electronic circuit’s energy consumption, while either ∈𝑓𝑠 (free space channel)or
∈𝑚𝑝 (𝑚𝑢𝑙𝑡𝑖𝑝𝑎𝑡ℎ 𝑓𝑎𝑑𝑖𝑛𝑔) denote the transmitter amplifier’s energy consumption. Either ∈𝑓𝑠 or ∈𝑚𝑝 are
used depending on the communication distance (𝑑) between the nodes. 𝑑𝑇ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑 refers to the threshold
distance, which may be determined in (4):
𝑑𝑇ℎ𝑟𝑒𝑠ℎ𝑜𝑙𝑑 =
∈𝑓𝑠
∈𝑚𝑝
(4)
3. PROPOSED UCFLEE PROTOCOL
3.1. Sector formation phase
The BS partitions the network area into two sectors of varying size. The smaller sector is situated
nearer to the BS, whereas the larger sector is located at some distance from the BS. According to (5), the BS
determined the smaller sector’s size, where R denotes the maximum range of communication distance. The
clusters’ small size averts the premature death of those nodes in closer proximity to the BS, therefore
resolving the hot spots issue. The larger sector is that beyond the smaller sector. The BS divided the larger
sector into sub-sectors, with each sub-sector’s size being equal to R. In the network area, each sector is
partitioned into clusters of equal width. Each cluster’s width is always equal to the value of sector𝑠𝑚𝑎𝑙𝑙𝑒𝑟 in (5).
Figure 1. presents the sector formation phase. Algorithm 1 clarifies all of the steps involved in the sectors’
formation.
sector𝑠𝑚𝑎𝑙𝑙𝑒𝑟 =
𝑅
2
(5)
Figure 1. Sector formation phase
Algorithm 1. Sector formation phase
Input: Network region dimension (X*Y;
Output: Forming sectors.
Initial 𝒙𝒂𝒙𝒊𝒔 = 𝟎, 𝒚𝒂𝒙𝒊𝒔 = 𝟏𝟎𝟎, 𝒊 = 𝟏, 𝑹
While ( yaxis > 0)
If ( yaxis = 100)
𝒚𝒂𝒙𝒊𝒔−𝒏𝒆𝒘 = 𝒚𝒂𝒙𝒊𝒔 −
𝑹
𝟐
Else
𝒚𝒂𝒙𝒊𝒔−𝒏𝒆𝒘 = 𝒚𝒂𝒙𝒊𝒔 − 𝑹
EndIF
yaxis = yaxis-new
While (xaxis < X)
𝒙𝒂𝒙𝒊𝒔−𝒏𝒆𝒘 = 𝒙𝒂𝒙𝒊𝒔 +
𝑹
𝟐
Clusterid = i
i = i +1
End while
𝒙𝒂𝒙𝒊𝒔 = 𝟎
End while

 ISSN: 2502-4752
1014
3.2. CH selection phase
The BS utilises the fuzzy logic (FL) model in order to select the optimal CH per cluster. Two input
variables, namely the distance to BS as well as residual energy, are given to the FL model, while the output
variable is produced, namely CH chance. Table 1 presents the tabulation of the membership functions for the
input and output variables. The value range of the distance to BS as well as residual energy parameters are
[0-120] and [0-0.5] respectively, as presented in Figures 2 and 3. The output variable’s value range is [0-1],
as evidenced in Figure 4.
In the FL model, every input variable’s value is transformed into the linguistic variable via the
fuzzification process. Subsequently, if-then rules may be applied in relation to the linguistic variables as a
means of connecting the input parameters and relevant output variables. A total of 25 (52
) if-then rules are
performed depending on two input variables, as Table 2 clarifies. Lastly, by applying the centre of area
method, the defuzzification process enables the output linguistic variables to be transformed into the output
value [22].
The BS uses the dynamic energy threshold (DT) to chnage the CH in each cluster. According to (6),
the BS calculates the 𝐷𝑇 value at the conclusion of each round. The CH changes whether its residual energy
was below the DT value. Algorithm 2 describes the CH selection method.
DT =
1
𝑁
∗ 𝐸𝑡𝑜𝑡𝑎𝑙 ∗ (1 −
𝑟current
𝑅estimated
) (6)
𝐸𝑡𝑜𝑡𝑎𝑙 indicates the total energy during the network’s initial operation. The 𝑟current pertains to the
numeral of the current round, while 𝑅estimated refers to the number of estimated rounds until expiry of all of
the network’s nodes. N denotes the total sensor number. As shown in (7) represents the 𝑅estimated value,
when 𝐸𝐶𝑢𝑟𝑟𝑒𝑛𝑡 is the current round’s energy consumption:
𝑅estimated =
𝐸𝑡𝑜𝑡𝑎𝑙
𝐸𝑐𝑢𝑟𝑟𝑒𝑛𝑡
(7)
Table 1. Membership function for the proposed protocol
Variable Membership function
Distance to BS Very Close (DVC), Close (DC), Medium (DM), Far (DF)Very Far (DVF)
Residual Energy Very High (REVH), High (REH), Medium (REM), Low (REL), Very Low (REVL)
CH Chance Very Strong (CVS ), Strong (CS),Medium (CM), Weak (CW), Very Weak (CVW)
Table 2. Fuzzy rules for the proposed protocol
No. Residual Energy Distance to BS Chance
1 REVL DVF CVW
2 REVL DF CVW
3 REVL DM CVW
4 REVL DC CVW
5 REVL DVC CW
6 REL DVF CVW
7 REL DF CW
8 REL DM CW
9 REL DC CM
10 REL DVC CM
11 REM DVF CW
12 REM DF CW
13 REM DM CM
14 REM DC CM
15 REM DVC CS
16 REH DVF CM
17 REH DF CM
18 REH DM CS
19 REH DC CS
20 REH DVC CVS
21 REVH DVF CS
22 REVH DF CS
23 REVH DM CVS
24 REVH DC CVS
25 REVH DVC CVS

1015
Figure 2. The member function for distance to BS Figure 3. The member function for residual energy
Figure 4. The member function for the output variable
Algorithm 2. CH selection method
Input: Ecurrent , Etotal , rcurrent , Clusterid , N
Output: CH selection in each round
𝑹
𝒆𝒔𝒕𝒊𝒎𝒂𝒕𝒆𝒅=
𝑬𝒕𝒐𝒕𝒂𝒍
𝑬𝒄𝒖𝒓𝒓𝒆𝒏𝒕
𝑫𝑻 =
𝟏
𝑵
∗ 𝑬𝒕𝒐𝒕𝒂𝒍 ∗ ( 𝟏 −
𝒓𝒄𝒖𝒓𝒓𝒆𝒏𝒕
𝑹𝒆𝒔𝒕𝒆𝒎𝒂𝒕𝒆𝒅
)
IF ( rcurretn = 1 )
For each Clusterid do
For each node in Cluster do
- Calculate Fuzzy Value for each node
End For
-Select node to be CH that have best Fuzzy Value
End For
Else
For each Clusterid do
IF ( CHenergy < DT )
For each node in Cluster do
- Calculate Fuzzy Value for each node
End For
-Select node to be CH that have best Fuzzy Value
End IF
End For
End IF

 ISSN: 2502-4752
1016
3.3. Data routing phase
The BS adopts the iterative deepening A* (IDA-star) method to discover the optimal multi-hop path
from the CHs. The IDA-star method enables the establishment of the shortest path with the least memory
usage based on iterative deepening [23]-[25]. Furthermore, the IDA-star algorithm determines the evolution
function of cluster heads 𝑓(𝐶𝐻𝑠) in accordance with (8). The IDA-star algorithm is reliant upon two
parameters to calculate the f value, namely energy level (𝐸𝐶𝐻) and the distance to the BS (𝑑𝐶𝐻−𝐵𝑆). The IDA-
star algorithm uses the cost bounded (Costbounded) value to determine the optimal solution to the BS, which
is expressed by (9).
𝑓(𝐶𝐻𝑠) = 𝑑𝐶𝐻−𝐵𝑆 + 𝐸𝐶𝐻 (8)
Costbounded = smallest ( 𝑓(𝐶𝐻) ) (9)
The value of Costbounded is the f value of the CH for the initial state. Per new level, the Costbounded
is the smallest f value among all the CHs that exceeded the previous Costbounded of the preceding level. The
CH collects data from the sensor nodes. Subsequently, the CH with an f value that exceeded the cost bounded
is added to the list called the ‘previous list’. In this previous list, the CH with a larger f-value is added to the
optimal path list, enabling its selection as the next hop. The IDA star continues until the optimal path has
been guaranteed based on attaining the BS.
Having completed the routing path, the CH that has the information sends the route request (RREQ)
message to the next CH in the optimal path. The CH waits for the route reply (RREP) message. Having
delivered the RREP message, the information is sent to the next CH. This process repeats from the next CH
in the routing path, until the information has been delivered to the BS. Following each round, the BS checks
the possibility of the current path sending further information or not, by comparing the energy per CH that
exists in this path with the DT value. If the DT value exceeds the energy of CH, then the BS adopts the FL
model to identify the new CH in the cluster.
4. PERFORMANCE EVALUATION
The UCFLEE protocol’s performance is evaluated by conducting simulation experiments. The
simulation was undertaken utilising the MATLAB environment. 100 nodes were spread to the sensing region
100 𝑚 ∗ 100 𝑚. The precise BS position was 100 𝑚 ∗ 50 𝑚 of the network area. Table 3 presents further
details of all the adopted simulation parameters. The proposed UCFLEE is compared with two widely
recognised clustering protocols, namely BRE-LEACH [17] and PSO-C .
[
22
] All protocols’ performance
analyses are informed by the evaluation variables, for example network lifetime and total residual energy per
round. The performance of the UCFLEE protocol, BRE-LEACH and PSO-C may be described as follows,
depending on the above factors.
Table 3. The simulation parameters of WSNs
Parameter Value
Area 100 𝑚 ∗ 100 𝑚
N 100 nodes
𝑹 40 𝑚
Position of BS 100 𝑚 ∗ 50 𝑚
Initial amount of energy 0.5 J
Data Packet 4000 bit
∈𝒎𝒑 0.13 bit/m4
𝑬𝒆𝒍𝒆𝒄 50 nJ/bit
∈𝒇𝒔 10 pJ/bit/m2
4.1. Network lifetime
The time interval between beginning the network operation to the death of the last node is
represented as the network lifetime [26]. Figure 5 presents the network lifetime performance for the UCFLEE
protocol as well as other protocols. This figure evidences that for BRE-LEACH and PSO-C, every node had
died by 5000 and 6739 rounds respectively. Contrastingly, for the UCFLEE protocol, only 53 nodes died at
9000 rounds. Therefore, the UCFLEE approach contributes to lengthening the network lifetime to a greater
extent than the BRE-LEACH and PSO-C protocols, by 64% and 56% respectively. The DT concept to alter
the CHs and the FL model in order to select the optimal CHs is the principal reason for the UCFLEE protocol
being more effective than other related protocols with regard to the network lifetime.

1017
Figure 5. The network life time performance
4.2. Residual energy
Figure 6 presents the results of the total residual energy for the three approaches. After 7000 rounds,
the energy in BRE-LEACH and PSO-C is completely consumed, whereas the proposed protocol preserves
over 23% of its energy at 7000 rounds. With the UCFLEE strategy, the nodes’ energy is depleted more
slowly compared with other protocols, meaning that it enables effective conservation of the nodes’ remaining
energy. The DT contributes to continuing the CH for multiple rounds without change and distributing the
traffic load between the nodes, thus saving greater energy. Moreover, the Costbounded in IDA-star facilitated
the diminishing of the nodes’ traffic load by expanding only those CHs with high f- value. Evidently, the
proposed UCFLEE protocol can achieve effective equilibrium of energy consumption, keeping the majority
of nodes alive to a greater extent than the related protocol.
Figure 6. The total residual energy for for all of the protocols the three protocols
5. CONCLUSION
A new protocol for WSNs called UCFLEE has been presented in this paper. The UCFLEE protocol
considers the issues of minimising energy dissipation and load balancing. The UCFLEE protocol contributed
to minimising the hot spot problems and facilitated the identification of efficient routing to the base station.
The concepts of altering and selecting cluster heads are employed to decrease energy dissipation and to
balance the nodal loads. The threshold concept is engaged to enable all nodes to consume an equivalent
amount of energy. The extensive experimentations confirm that the UCFLEE scheme significantly decreases
the node’s energy consumption, while enhancing network lifetime to a greater extent than the previous
protocols.

 ISSN: 2502-4752
1018
REFERENCES
[1] Z. Fei, B. Li, S. Yang, C. Xing, H. Chen, and L. Hanzo, “A survey of multi-objective optimization in wireless sensor networks:
metrics, algorithms, and open problems,” IEEE Communications Surveys & Tutorials, vol. 19, no. 1, pp. 550–586, 2017, doi:
10.1109/COMST.2016.2610578.
[2] J. Yan, M. Zhou, and Z. Ding, “Recent advances in energy-efficient routing protocols for wireless sensor networks: A review,”
IEEE Access, vol. 4, pp. 5673–5686, 2016, doi: 10.1109/ACCESS.2016.2598719.
[3] P. K. Mishra and S. K. Verma, “FFMCP: feed-forward multi-clustering protocol using fuzzy logic for wireless sensor networks
(WSNs),” Energies, vol. 14, no. 10, p. 2866, May 2021, doi: 10.3390/en14102866.
[4] G. Anastasi, M. Conti, M. Di Francesco, and A. Passarella, “Energy conservation in wireless sensor networks: A survey,” Ad Hoc
Networks, vol. 7, no. 3, pp. 537–568, May 2009, doi: 10.1016/j.adhoc.2008.06.003.
[5] A. A. Alkathmawee, L. Feng, and I. S. Alshawi, “Prolonging the lifetime of wireless sensor networks using LPA-star search
algorithm,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 1, no. 2, p. 390, Feb. 2016, doi:
10.11591/ijeecs.v1.i2.pp390-398.
[6] S. Randhawa and S. Jain, “MLBC: multi-objective load balancing clustering technique in wireless sensor networks,” Applied Soft
Computing, vol. 74, pp. 66–89, Jan. 2019, doi: 10.1016/j.asoc.2018.10.002.
[7] S. E. Khediri, N. Nasri, R. U. Khan, and A. Kachouri, “An improved energy efficient clustering protocol for increasing the life
time of wireless sensor networks,” Wireless Personal Communications, vol. 116, no. 1, pp. 539–558, Jan. 2021, doi:
10.1007/s11277-020-07727-y.
[8] A. S. Rostami, M. Badkoobe, F. Mohanna, H. Keshavarz, A. A. R. Hosseinabadi, and A. K. Sangaiah, “Survey on clustering in
heterogeneous and homogeneous wireless sensor networks,” The Journal of Supercomputing, vol. 74, no. 1, pp. 277–323,
Jan. 2018, doi: 10.1007/s11227-017-2128-1.
[9] C. Donati-Martin, “Stochastic integration with respect to q Brownian motion,” Probability Theory and Related Fields, vol. 125,
no. 1, pp. 77–95, Jan. 2003, doi: 10.1007/s00440-002-0224-4.
[10] W. B. Heinzelman, A. P. Chandrakasan, and H. Balakrishnan, “An application-specific protocol architecture for wireless
microsensor networks,” IEEE Transactions on Wireless Communications, vol. 1, no. 4, pp. 660–670, Oct. 2002, doi:
10.1109/TWC.2002.804190.
[11] N. M. A. Latiff, C. C. Tsimenidis, and B. S. Sharif, “Energy-aware clustering for wireless sensor networks using particle swarm
optimization,” in 2007 IEEE 18th International Symposium on Personal, Indoor and Mobile Radio Communications, 2007,
pp. 1–5, doi: 10.1109/PIMRC.2007.4394521.
[12] H. Liang, S. Yang, L. Li, and J. Gao, “Research on routing optimization of WSNs based on improved LEACH protocol,”
EURASIP Journal on Wireless Communications and Networking, vol. 2019, no. 1, p. 194, Dec. 2019, doi: 10.1186/s13638-019-
1509-y.
[13] A. Rozas and A. Araujo, “An application-aware clustering protocol for wireless sensor networks to provide QoS management,”
Journal of Sensors, vol. 2019, pp. 1–11, Sep. 2019, doi: 10.1155/2019/8569326.
[14] E. Alnawafa and I. Marghescu, “New energy efficient multi-hop routing techniques for wireless sensor networks: static and
dynamic techniques,” Sensors, vol. 18, no. 6, p. 1863, Jun. 2018, doi: 10.3390/s18061863.
[15] K. Cengiz and T. Dag, “Multi-hop low energy fixed clustering algorithm (M-LEFCA) for WSNs,” in 2016 IEEE 3rd
International Symposium on Telecommunication Technologies (ISTT), pp. 31–34, Nov. 2016, doi: 10.1109/ISTT.2016.7918080.
[16] M. Selvi, C. Nandhini, K. Thangaramya, K. Kulothungan, and A. Kannan, “HBO based clustering and energy optimized routing
algorithm for WSN,” in 2016 Eighth International Conference on Advanced Computing (ICoAC), Jan. 2017, pp. 89–92, doi:
10.1109/ICoAC.2017.7951751.
[17] I. Daanoune, A. Baghdad, and A. Balllouk, “BRE-LEACH: A new approach to extend the lifetime of wireless sensor network,” in
2019 Third International Conference on Intelligent Computing in Data Sciences (ICDS), Oct. 2019, pp. 1–6, doi:
10.1109/ICDS47004.2019.8942253.
[18] S. K. Singh, P. Kumar, and J. P. Singh, “Correction to: An energy efficient protocol to mitigate hot spot problem using unequal
clustering in WSN,” Wireless Personal Communications, vol. 102, no. 2, pp. 2091–2091, Sep. 2018, doi: 10.1007/s11277-018-
5921-0.
[19] A. A. Jasim, M. Y. I. Idris, S. R. Bin Azzuhri, N. R. Issa, M. T. Rahman, and M. F. B. Khyasudeen, “Energy-efficient wireless
sensor network with an unequal clustering protocol based on a balanced energy method (EEUCB),” Sensors, vol. 21, no. 3,
p. 784, Jan. 2021, doi: 10.3390/s21030784.
[20] N. Islam, S. Dey, and S. Sampalli, “Energy-balancing unequal clustering approach to reduce the blind spot problem in wireless
sensor networks (WSNs),” Sensors, vol. 18, no. 12, p. 4258, Dec. 2018, doi: 10.3390/s18124258.
[21] A. A. Alkadhmawee, M. A. Altaha, and W. Mahmood Lafta, “Unequal clustering algorithm with IDA* multi-hop routing to
prevent hot spot problem in WSNs,” Indonesian Journal of Electrical Engineering and Computer Science, vol. 20, no. 1, p. 445,
Oct. 2020, doi: 10.11591/ijeecs.v20.i1.pp445-453.
[22] Z. Saletic, D. M. Velasevic, and N. E. Mastorakis, “Analysis of basic defuzzification techniques,” in Proceedings of the 6th WSES
international multiconference on circuits, systems, communications and computers, 2002, pp. 247–252.
[23] R. E. Korf, “Depth-first iterative-deepening,” Artificial Intelligence, vol. 27, no. 1, pp. 97–109, Sep. 1985, doi: 10.1016/0004-
3702(85)90084-0.
[24] K. Khantanapoka and K. Chinnasarn, “Pathfinding of 2D & 3D game real-time strategy with depth direction A∗
algorithm for multi-layer,” in 2009 Eighth International Symposium on Natural Language Processing, Oct. 2009, pp. 184–188,
doi: 10.1109/SNLP.2009.5340922.
[25] T. Cazenave, “Partial move A*,” in 2010 22nd IEEE International Conference on Tools with Artificial Intelligence, Oct. 2010,
pp. 25–31, doi: 10.1109/ICTAI.2010.79.
[26] G. Smaragdakis, I. Matta, and A. Bestavros, “SEP: A stable election protocol for clustered heterogeneous wireless sensor
networks Heterogeneous WSN o Nodes have different energy levels,” in Second International Workshop on Sensor and Actor
Network Protocols and Applications (SANPA 2004), 2004, pp. 251–261.

1019
BIOGRAPHIES OF AUTHORS
Mohammed Adnan Altaha Complete his Bachelor's degree in Computer Science
from the College of Education for Pure Science, University of Basrah, Basrah, Iraq (2009),
completed his master’s degree in Computer Science from the college of Science, University of
Basrah, Basrah, Iraq (2018), currently works as a lecture. Assist. in the College of Veterinary,
University of Basrah, Basrah, Iraq, published several scientific researches in computer science.
He can be contacted at email: mohammed.altaha@uobasrah.edu.iq.
Ahmed Adil Alkadhmawee is a lecturer at Basrah University, Iraq. He holds
an M. Sc degree in Computer Engineering at Huazhong University of Science and Technology
in China. He is research areas are Wireless Sensor Network, Machine Learning and Deep
Learning. He has authored more than 11 publications: 1 proceeding and 10 journals, with 3 H-
index and more than 19 citations. He can be contacted at email: ahmedadel@uobasrah.edu.iq.
Wisam Mahmood Lafta Was born in Baghdad, Iraq. Received a BSc in
computer science from the University of Technology; the MSc at Huazhong University of
Science and Technology in China. He is currently a faculty member in the computer science
department, University of Technology, Baghdad, Iraq. He has some important published
papers in international journals and a reviewer at some international journals. He can be
contacted at email: wisam.m.lafta@uotechnology.edu.iq.

Uneven clustering and fuzzy logic based energy-efficient wireless sensor networks

Recommended

More Related Content

Similar to Uneven clustering and fuzzy logic based energy-efficient wireless sensor networks (20)

More from nooriasukmaningtyas (20)

Recently uploaded (20)

Uneven clustering and fuzzy logic based energy-efficient wireless sensor networks