Received: January 16, 2023. Revised: April 13, 2023.

509

Multi Objective Energy Based Improved Jellyfish Swarm Optimization for

Effective Cluster Head Discovery in UWSN

Seema Swamy Gowda1* Ambika Ramalingappa2

1
Department of Electronics and Communication Engineering, Sir M Visvesvaraya Institute of Technology,
Bengaluru, India
2
Department of Electronics and Communication Engineering, BMS Institute of Technology and Management,
Bengaluru, India
* Corresponding author’s Email: [email protected]

Abstract: Underwater wireless sensor networks (UWSNs) have a huge amount of sensors located underwater to collect
data from the underwater scenario. UWSN is considered a promising method for monitoring and exploring an
underwater scenario. Energy-efficient and reliable data broadcasting are considered challenging tasks, because of the
limited energy source of sensors. To address this issue, an energy-efficient cluster head (CH) selection and multi-hop
routing are developed in UWSN. The multi objective energy based improved jellyfish swarm optimization (MOEIJSO)
is proposed to select optimal CHs from normal sensors. The input parameters of MOEIJSO are residual energy,
neighbor node distance, sink distance, and CH balancing factor. Next, multi-hop routing is developed by using ant
colony optimization (ACO) for delivering the data packets. Therefore, the proposed MOEIJSO-ACO method is used
to improve energy efficiency while increasing data delivery. The proposed MOEIJSO-ACO method is evaluated by
using the alive and dead nodes, total energy consumption, data received in base station (BS), and life expectancy. The
existing researches such as metaheuristics-based clustering with a routing (MCR) for UWSN, modified low energy
adaptive clustering hierarchy (MLEACH) and cooperative energy-efficient routing (CEER) approach are used to
compare the MOEIJSO-ACO method. The alive nodes of the MOEIJSO-ACO are 399 for 800 rounds, which is high
when compared to the MCR-UWSN.
Keywords: Ant colony optimization, Energy efficiency, Life expectancy, Multi-hop routing, Multi-objective energy
based improved jellyfish swarm optimization, Underwater wireless sensor networks.

sensors accomplish direct or indirect transmission of

1. Introduction information to the surface sensor (i.e., sink). Next, the
data from the sink is transmitted to an offshore
Earth is occupied by a 3/4 ratio of water using
monitoring center to analyze and study the gathered
seas, lakes, canals, rivers and streams. Since a huge
data [4]. The sensors of UWSN are located
amount of unexplored and hidden resources exist
underwater for evaluating the monitoring features
underwater which require to be discovered whereas
such as density, temperature, pressure and so on [5].
underwater environments are highly sophisticated for
UWSN is extensively used in diverse fields such as
humans to discover. Accordingly, an exploration of
military target tracking, oil/gas spill monitoring,
the underwater environment is possible with the use
submarine detection, offshore exploration, disaster
of wireless technology [1, 2]. UWSN is developing
prevention and so on [6]. The UWSN faces various
technology that is utilized for observing and
issues such as strong network dynamics, expensive
discovering the changes in an aqueous environment
deployment, less available bandwidth and restricted
[3]. UWSN is a wireless network that comprises a
battery energy [7].
group of sensors and autonomous underwater
The battery energy of underwater sensors is
vehicles for collecting and sensing data. These
restricted whereas these batteries cannot be simply
International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 510

replaced or charged in the ocean. The energy energy, node coverage, processing power and
consumption of the UWSN is high because of the communication cost were utilized to choose the
higher communication power, restricted bandwidth candidate CH. Further, butterfly optimization was
capacity, higher and variable time delay and higher used for transmitting the data to the receiver. The
bit error rate [8-10]. The limited battery source of the developed MCBOR was used to improve the PDR
node creates the failure nodes over the UWSN [11]. and minimize the loss over the UWSN. Moreover, the
These aforementioned features of UWSN create an developed MCBOR was required to be analyzed in a
effect on sensor data gathering and long-distance large-scale network.
communication. Hence, the routing is required to be Bhattacharjya, [17] developed an energy-efficient
developed as simple and energy efficient for UWSNs. UWSN for reducing energy expenditure and
This indicates that effective and reliable data improving performance in the underwater
broadcasting to the sink node is a highly challenging environment. The developed cluster-based UWSN
task as well as it is one of the essential concerns in (CUWSN) utilized the advantages of CH and multi-
UWSN [12, 13]. A cluster-based data collection is hop data transmission. Here, the CH was chosen
developed for improving the energy efficacy of the based on the residual energy of the sensors over the
WSN [14]. The main objective of the clustering network. The developed CUWSN was used to
approach is used to separate the network into small improve life expectancy based on multi-hop
divisions namely clusters. An entire cluster is transmission. The CH discovery was considered only
represented by using the CH that gathers the observed the residual energy, however, a data transmission
information from sensors and it is broadcasted to the distance was required to be considered in CUWSN.
sink by utilizing the CH as the next hop. The Faheem, [18] developed the dynamic firefly
developed clustering approach is used to minimize mating optimization-based routing protocol namely
the overall routing distance and overhead of sensors FFRP for the UWSN scenario. A stable and reliable
[15]. route was discovered by using the developed FFRP.
The contributions are concise as follows: A balancing of data traffic load using FFRP was used
to reduce energy usage and latency. Accordingly, the
• The MOEIJSO based CH discovery is data transmission over the stable link was used to
developed for enhancing the energy improve the data delivery of the network.
efficiency of UWSN. The IJSO is chosen for Subramani, [19] presented MCR for UWSN to
this research because of its enhanced local enhance energy efficiency. In MCR-UWSN, the
search capacity which is used to obtain in cultural emperor penguin optimizer-based clustering
best CH solutions. (CEPOC) was developed for generating the clusters.
• Further, the ACO-based route discovery is The developed CEPOC considered the node’s motion
used for identifying the route from and average distance for discovering the CH. Further,
transmitter CH to BS. Therefore, the the routing was developed by using grasshopper
MOEIJSO-ACO is used to improve the life optimization according to the list of nodes, energy,
expectancy of UWSN while enhancing the and distance. For an effective CH selection, cluster
data delivery. balancing among the nodes was essential to enhance
the performance.
The paper organization is sorted as follows: Rizvi, [20] presented an energy efficient
Section 2 delivers the related works of the energy- approach i.e., MLEACH for UWSN. The CHs were
efficient data transmission of UWSN. The randomly chosen in this MLEACH for preserving the
MOEIJSO-ACO based data broadcasting is detailed energy in each round. The non ‑ persistent carrier
in section 3 whereas the results are presented in sense multiple access was used to establish the data
section 4. Finally, the conclusion is given in section transmission among CH and its respective node. The
5. direct data transmission from the CH to BS was
resulted in higher energy depletion of nodes.
2. Related work Ahmad [21] developed the cooperative energy-
efficient routing (CEER) approach for generating the
This section provides the existing energy efficient
reliable network and increasing the network lifetime.
data broadcasting-related research in UWSN.
The problem of hotspot was eliminated by
Chenthil and Jesu Jayarin, [16] presented the
minimizing the energy utilization based on sink
multilayer clustering-based butterfly optimization
mobility approach. The network reliability was
routing (MCBOR) for broadcasting the data to the
obtained by using the cooperative technique in
receiver without loss. The list of parameters such as
UWSN. However, the clustering over the network
International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 511

Figure. 1 Block diagram for the MOEIJSO-ACO

was required to be developed for an additional

improvement of lifetime. 3.2 Discovering optimal CH using MOEIJSO
The limitations found from the related works are
inadequate fitness function, higher energy usage and The MOEIJSO is used for discovering the
routing without clustering. In this research, an optimal CHs from the sensors of UWSN. The
appropriate objective measures are used with conventional Jellyfish Swarm Optimization (JSO) is
MOEIJSO to develop an effective clustering and motivated based on jellyfish’s search-feeding
routing for minimizing the energy usage of overall activities and movement patterns through the ocean.
network. The main difference between JSO and IJSO [22] is
that the incorporation of the sinusoidal factor for
improving the local search capacity resulted in the
3. MOEIJSO-ACO method
best optimal solutions. Further, the IJSO is developed
In this research, effective data transmission is as MOEIJSO for selecting optimal CHs from the
obtained by using a MOEIJSO based optimal CH sensors.
selection and ACO based routing in UWSN. The
energy usage of sensors is minimized by developing 3.2.1. Representation and initialization
a clustering along with a multi-hop transmission
An initial solution of MOEIJSO i.e., jellyfish has
using MOEIJSO-ACO. The reduction in energy
utilization helps to improve the life expectancy of the a set of nominee nodes to be chosen as CH. Here,
each jellyfish is initialized with the ID of a random
sensors which results in higher data delivery. The
block diagram of the MOEIJSO-ACO is presented in node between 1 and 𝑁, where the total amount of
Fig. 1. nodes in the underwater scenario is denoted as 𝑁. The
MOEIJSO’s 𝑖 th jellyfish is denoted as 𝑋𝑖 =
3.1 Sensor initialization (𝑋𝑖,1 , 𝑋𝑖,2 , … , 𝑋𝑖,𝑑𝑖𝑚 ) , where 𝑑𝑖𝑚 denotes each
jellyfish’s dimension as amount of CHs.
The nodes are positioned randomly in the WSN
followed by the CHs chosen using MOEIJSO. The 3.2.2. Iterative process
clusters are formed in the UWSN, once the CHs are
chosen from the network. Further, the route from the The IJSO has three different rules that are
transmitter CH to receiver BS is discovered using mentioned as follows: 1) two types of motion exist in
ACO. The optimal CH and route discovery are jellyfish such as following the ocean current motion
explained in the following sections. and moving in the jellyfish population. Here, the
time-controlled mechanism is used to switch among
these motions; 2) Jellyfish moves in the ocean for
searching food, where the jellyfishes are attracted to
International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 512

the position with more food, and 3) Food position and The location update of class A motion with
objective value defines the amount of discovered sinusoidal adaptive factor is expressed in Eq. (6).
food.
𝑋𝑖 (𝑡 + 1) = 𝑋𝑖 (𝑡) + 𝑆 × 𝛾 × 𝑟𝑎𝑛𝑑(0,1) ×
3.2.2.1. Following the movement of ocean current
(𝑢𝑏 − 𝑙𝑏 ) (6)
The movement way of current is denoted as
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗ Where lower and upper bound are denoted as 𝑙𝑏
𝑡𝑟𝑒𝑛𝑑 that defines the average of the overall vector
and 𝑢𝑏 respectively; jellyfish’s movement
sum from every individual current location of
coefficient is denoted as 𝛾 where it is taken as 0.1.
jellyfish to the current finest location. Eq. (1) shows
Accordingly, the range of motion is mainly based on
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗ for the direction of motion.
the current 𝑡𝑟𝑒𝑛𝑑 𝑙𝑏 , 𝑢𝑏 and 𝛾.
The jellyfish accomplishes the motion of class B
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑡𝑟𝑒𝑛𝑑 = 𝑋 ∗ − 𝐷𝐹 (1) and approaches the discrete jellyfish with huge
amount of food at its position and collects with food
Where the optimal location is denoted as 𝑋 ∗ and as its target. The location update of class B motion is
the difference between the current jellyfish placed in expressed in Eq. (7).
an optimal location and the average location of all
jellyfish is denoted as 𝐷𝐹 is expressed in Eq. (2). 𝑋𝑖 (𝑡 + 1) = 𝑋𝑖 (𝑡) + ⃗⃗⃗⃗⃗⃗⃗⃗
𝑠𝑡𝑒𝑝 (7)

𝐷𝐹 = 𝑒𝑐 𝜇 (2) Where, 𝑖 th jellyfish’s step length with the

⃗⃗⃗⃗⃗⃗⃗⃗ which is
direction of motion is denoted as 𝑠𝑡𝑒𝑝
Where, the parameter which defines the food expressed in Eq. (8).
attractiveness to jellyfish is denoted as 𝑒𝑐 which is
expressed in Eq. (3) and the average location of all ⃗
⃗⃗⃗⃗⃗⃗⃗⃗ = 𝑟𝑎𝑛𝑑(0,1) × 𝐷
𝑠𝑡𝑒𝑝 (8)
jellyfish is denoted as 𝜇.
Where, 𝑖th jellyfish’s motion direction is denoted
𝑒𝑐 = 𝛽 × 𝑟𝑎𝑛𝑑(0,1) (3) ⃗ which is shown in Eq. (9).
as 𝐷
Where, the distribution coefficient is denoted as
𝛽. Eq. (4) shows the jellyfish’s location update for 𝑋𝑗 (𝑡) − 𝑋𝑖 (𝑡) , 𝑖𝑓 𝑓(𝑋𝑖 (𝑡)) ≥ 𝑓 (𝑋𝑗 (𝑡))
following ocean current motion. ⃗ ={
𝐷
𝑋𝑖 (𝑡) − 𝑋𝑗 (𝑡) , 𝑖𝑓 𝑓 (𝑋𝑗 (𝑡)) ≥ 𝑓(𝑋𝑖 (𝑡))
⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗⃗
𝑋𝑖 (𝑡 + 1) = 𝑋𝑖 (𝑡) + 𝑟𝑎𝑛𝑑(0,1) × 𝑡𝑟𝑒𝑛𝑑 (9)
(𝑡) ∗
= 𝑋𝑖 + 𝑟𝑎𝑛𝑑(0,1) × (𝑋 − 𝛽 × 𝑟𝑎𝑛𝑑(0,1) × 𝜇)
(4) Where, 𝑖 th and 𝑗 th jellyfish’s location are
represented as 𝑋𝑖 (𝑡) and 𝑋𝑗 (𝑡) respectively; the
Where, 𝑋𝑖 (𝑡) is jellyfish’s current location and objective function according to the 𝑋 is denoted as 𝑓.
𝑋𝑖 (𝑡 + 1) is location obtained after updating in
3.2.2.3. Time control mechanism
search space.
The time control mechanism is used for
3.2.2.2. Group movements
performing the switching among three modes of
The jellyfish’s group motion is categorized into movement of jellyfish where the time control
passive (i.e., Class A) and active motion (i.e., Class function 𝑐(𝑡) is expressed in Eq. (10).
B). The movement of Class A occurs, when the
𝑡
swarm of jellyfish is generated whereas the 𝑐(𝑡) = |(1 − 𝑇) × (2 × 𝑟𝑎𝑛𝑑(0,1) − 1)| (10)
movement of Class B is carried out after the Class A
movement. The class A motion is accomplished as Where, the current iteration is denoted as 𝑡 ;
the jellyfish traveling around their current location. In maximum iterations are denoted as 𝑇 and the random
IJSO, the sinusoidal adaptive factor (𝑆) shown in Eq. number is generated in the range of [0, 1]. If 𝑐(𝑡) ≥
(5) is used for improving the local search capability.
0.5, then the jellyfish movement is controlled in
𝜋(2𝑇+𝑡) ocean currents; otherwise, the jellyfish movement is
𝑆 = 1 + sin 2𝑇
(5) controlled by the intra-group motion of jellyfish.

International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 513

3.2.3. Derivation of multiple objective functions for balancing over the UWSN.
MOEIJSO
𝐴𝑁
Multiple objectives used in this MOEIJSO for 𝑓4 = ∑𝑑𝑖𝑚
𝑖=1 𝑑𝑖𝑚 − 𝐶𝑀𝑗 (15)
selecting the optimal CHs are residual energy (𝑓1 ),
neighbor node distance (𝑓2 ), sink distance (𝑓3 ) and Where the amount of alive nodes is denoted as
CH balancing factor (𝑓4 ). Eq. (11) shows the multiple 𝐴𝑁.
objective functions used in MOEIJSO for improving The above-mentioned objective values are
the searching process during the CH discovery. utilized for selecting adequate CHs from normal
sensors. The remaining energy of the sensors is used
𝑓 = 𝜓1 × 𝑓1 + 𝜓2 × 𝑓2 + 𝜓3 × 𝑓3 + 𝜓4 × 𝑓4 (11) to compute the possibility of a failure node during the
data broadcasting because the failure node needs to
Where, 𝜓1 − 𝜓4 is the weight value assigned to be avoided for increasing the data delivery. On the
each objective value. The multiple objectives are other hand, the distance factors and CH balancing
defined as follows: factor is used for improving the energy efficiency of
the UWSN.
• The energy usage of CH is essential in
3.3 Cluster formation
UWSN, due to the different processes such as
information collection, aggregation, and After selecting the CHs using MOEIJSO, the
distribution over the network. For an cluster members are assigned to their respective CHs.
effective transmission, the node with huge The distance and residual energy are considered in a
remaining energy is desired as a CH and Eq. potential function shown in Eq. (16) for creating the
(12) shows the computation of remaining clusters.
energy.
𝐸𝐶𝐻
1 𝑃𝑜𝑡𝑒𝑛𝑡𝑖𝑎𝑙 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 (𝑁𝑖 ) = (16)
𝑓1 = ∑𝑑𝑖𝑚
𝑖=1 𝐸 (12) 𝑑𝑖𝑠(𝑁𝑖 ,𝐶𝐻)
𝐶𝐻𝑖

Where, 𝐸𝐶𝐻𝑖 denotes the remaining energy of the 3.4 Route discovery using ACO
𝑖th CH
The route among the transmitter CH and receiver
BS is discovered by ACO, once the clustering is done
• The distance between the sensors shown in
in the network. The initial weight value of each path
Eq. (13) is the neighbor node distance and the
is calculated according to the distance between the
distance between the CH and BS shown in Eq.
nodes. The node transition rule is the possibility of
(14) is the sink distance. The energy
selecting 𝑚 as relay CH for the 𝑙th CH computed by
depletion of the sensor is proportional to the
ant 𝑛 is expressed in Eq. (17).
distance among the sensors. Therefore, the
CH with a lesser broadcast distance is
[𝜏𝑙𝑚 (𝑡)]𝛼 [𝜂𝑙𝑚 (𝑡)]𝛽
preferred as CH. 𝑛 𝑖𝑓 𝑚𝜖𝒩𝑛
𝑃𝑙𝑚 = {∑𝑜𝜖𝒩𝑛 [𝜏𝑙𝑜 (𝑡)]𝛼 [𝜂𝑙𝑜 (𝑡)]𝛽 (17)
𝐶𝑀 0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
𝑓2 = ∑𝑑𝑖𝑚 𝑗
𝑗=1 (∑𝑖=1 𝑑𝑖𝑠(𝑁𝑖 , 𝐶𝐻𝑗 )⁄𝐶𝑀𝑗 ) (13)
𝑛
Where, 𝑃𝑙𝑚 denotes the probability of node
𝑓3 = ∑𝑑𝑖𝑚 selection; 𝜏𝑙𝑚 and 𝜂𝑙𝑚 denotes the pheromone
𝑖=1 𝑑𝑖𝑠(𝐶𝐻𝑖 , 𝐵𝑆) (14)
intensity and heuristic value respectively; 𝛼 and 𝛽
Where, the number of cluster members for 𝑗th are used to control the relative importance of 𝜏𝑙𝑚 and
cluster is denoted as 𝐶𝑀𝑗 ; distance from node 𝑖 to CH 𝜂𝑙𝑚 , and the set of CHs 𝑛 that doesn’t visit the ant is
𝑗 is denoted as 𝑑𝑖𝑠(𝑁𝑖 , 𝐶𝐻𝑗 ) and distance from the denoted as 𝒩𝑛 .
The real ant’s foraging process is replicated by
CH 𝑖 to receiver BS is denoted as 𝑑𝑖𝑠(𝐶𝐻𝑖 , 𝐵𝑆).
artificial ants. The rule of node transition is used to
select the next relay CH when the transmitter CH is
• In UWSN, there is a possibility that some big required for transmitting the data. The path is retraced
clusters are formed along with the small to the transmitter CH, when the relay CH i.e., ant
clusters. Therefore, the CH balancing factor reached the BS. Accordingly, the pheromone value of
shown in Eq. (15) is used for balancing the
the path is updated according to the pheromone
clusters which result in improved energy
International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 514

update rule expressed in Eq. (18) which includes Table 1. Simulation parameters
pheromone evaporation and reinforcement. The Parameter Value
pheromone evaporation and reinforcement minimize Number of sensors 50
or maximize the path’s pheromone respectively. Location of BS 50, 50, 50
Network size 100 × 100 × 50𝑚3
𝑛𝑒𝑤
𝜏𝑙𝑚 𝑜𝑙𝑑
= (1 − 𝜌)𝜏𝑙𝑚 𝑛
+ ∑𝐴𝑛=1 ∆𝜏𝑙𝑚 (18) The initial energy of sensors 5J
Size of packet 4000 s
Where, the amount of ants in ACO is denoted as
𝐴 and the pheromone decay coefficient is 𝜌. Eq. (19)
shows the computation of pheromone quantity.
𝑛
∆𝜏𝑙𝑚
𝑄
𝑖𝑓 𝑡ℎ𝑒 𝑎𝑛𝑡 𝑛 𝑡𝑟𝑎𝑣𝑒𝑙𝑙𝑒𝑑 𝑟𝑜𝑢𝑡𝑒 (𝑙, 𝑚)
= {𝑎𝑛
0 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
(19)

Where, the constant value is denoted as 𝑄 and the

objective function of the route is 𝑎𝑛 that is shown in
Eq. (20). The objective function of ACO uses the
residual energy and sink distance
1 Figure. 2 Alive nodes vs rounds
𝑎𝑛 = 𝜀1 × ∑𝑑𝑖𝑚
𝑖=1 + 𝜀2 × ∑𝑑𝑖𝑚
𝑖=1 𝑑𝑖𝑠(𝐶𝐻𝑖 , 𝐵𝑆) (20)
𝐸𝐶𝐻𝑖

Where, 𝜀1 and 𝜀2 are the weight parameters

assigned to objective values considered in route
discovery. The failure node is avoided by considering
the remaining energy whereas the distance is
considered for minimizing the energy by generating
the route with less transmission distance.

4. Results and discussion

The outcomes of the MOEIJSO-ACO are detailed
in this section. The implementation and simulation of
the MOEIJSO-ACO are developed in MATLAB
R2020b where the system uses the i5 processor and
6GB of RAM. The underwater scenario is designed
with 50 sensors deployed in the area of 100 × 100 × Figure. 3 Dead nodes vs rounds
50𝑚3. The base station for this UWSN scenario is 4.1.1. Alive & dead nodes
located in the coordinates of 50, 50, 𝑎𝑛𝑑 50 whereas
the sensors are equipped with 5J of energy. Table 1 The amount of nodes with sufficient energy to
shows the simulation parameters considered for this perform the communication is alive nodes whereas
MOEIJSO-ACO method. dead nodes don’t have sufficient energy to
accomplish the data transfer over the network. The
4.1 Performance analysis alive nodes and dead node comparison for
MOEIJSO-ACO with LEACH and DEEC are shown
The MOEIJSO-ACO method is analyzed by
in Fig. 2 and Fig. 3 respectively. The alive nodes of
using alive and dead nodes, total energy consumption,
MOEIJSO-ACO are increased than the LEACH and
data received in BS and life expectancy. Classical
approaches such as Low Energy Adaptive Clustering DEEC by minimizing the energy usage of the nodes.
The energy usage of nodes for MOEIJSO-ACO is
Hierarchy (LEACH) and Distributed Energy
reduced based on effective balancing between
Efficient Clustering (DEEC) are used to evaluate the
clusters and shortest path discovery.
MOEIJSO-ACO method.

International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 515

Figure. 4 Total energy consumption vs rounds

Figure. 7 Packet loss ratio

Figure. 8 Life expectancy

Figure. 5 Data packets received in BS vs rounds

Figure. 9 Number of rounds for energy exhausted vs

nodes

analysis shows that the MOEIJSO-ACO achieves

Figure. 6 Maximum data packets received in BS
lesser energy consumption than the LEACH and
4.1.2. Total energy consumption DEEC. An optimal CH selection using MOEIJSO
along with the multi hop routing developed by ACO
The energy depleted while transmitting and is used to reduce energy usage. The ACO-based
receiving the data is total energy consumption. Fig. 4 shortest route discovery is also used to minimize the
displays the energy expenditure analysis for node’s energy usage.
MOEIJSO-ACO with LEACH and DEEC. This

International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 516

4.1.3. Data packets received in BS and PLR 4.2 Comparative analysis

The amount of data packets collected by the BS The existing research of MCR-UWSN [19],
and its maximum value are illustrated in Figs. 5 and MLEACH [20] and CEER [21] are used for
6 where the MOEIJSO-ACO was compared with evaluating the MOEIJSO-ACO method. Here, the
LEACH and DEEC. Further, the graph for PLR is comparison is made for 400 nodes, 50 nodes and 225
shown in Fig. 7. This analysis shows that the nodes variations. The MCR-UWSN [19] is taken for
MOEIJSO-ACO has high data delivery than the 400 nodes comparison, MLEACH [20] is taken for
LEACH and DEEC. The mitigation of failure nodes 50 nodes comparison and CEER [21] is taken for 225
in MOEIJSO based CH selection and reducing the nodes comparison. Tables 2, 3 and 4 show the
energy consumption of nodes through ACO based comparison of MOEIJSO-ACO with MCR-UWSN
multi hop data broadcasting helps to increase the [19], MLEACH [20] and CEER [21] respectively.
amount of data packets gathered in BS. The life expectancy comparison of MOEIJSO-ACO
is shown in Table 5. The graph for life expectancy
4.1.4. Life expectancy comparison of MOEIJSO-ACO with CEER [21] is
Life expectancy represents the time at which the shown in Fig. 10. This comparison shows the
nodes active during the transmission. There are three MOEIJSO-ACO has improved performance than the
different parameters such as first node die (FND), MCR-UWSN [19], MLEACH [20] and CEER [21].
half node die (HND) and last node die (LND) are The usage of energy is minimized in MOEIJSO-ACO
used while analyzing the life expectancy. based on balancing between clusters and shortest path
The life expectancy comparison for MOEIJSO- discovery. The multi hop routing developed by ACO
ACO with LEACH and DEEC are displayed in Fig. is used to achieve the reliable data transmission while
8. Further, the number of rounds for energy exhausted reducing the energy. Thus, the life expectancy of the
is shown in Fig. 9 in terms of nodes. This analysis MOEIJSO-ACO is improved than the MCR-UWSN
depicts that the MOEIJSO-ACO has a better lifetime [19], MLEACH [20] and CEER [21].
than the LEACH and DEEC.

Table 2. Comparison of MOEIJSO-ACO with MCR-UWSN

Performances Methods Number of rounds
200 400 600 800 1200
Alive nodes MCR -UWSN [19] 400 400 400 400 0
MOEIJSO-ACO 400 399 399 399 394
Dead nodes MCR -UWSN [19] 0 0 0 0 400
MOEIJSO-ACO 0 1 1 1 6
Total Energy MCR-UWSN [19] 3 12 22 42 100
Consumption (%) MOEIJSO-ACO 2.12 4.2 7.15 9.53 11.54

Table 3. Comparison of MOEIJSO-ACO with MLEACH

Performances Methods Number of rounds
500 1000 1500 2000 2500
Alive nodes MLEACH [20] 29 9 3 1 0
MOEIJSO-ACO 50 50 50 49 47
Dead nodes MLEACH [20] 21 41 47 49 50
MOEIJSO-ACO 0 0 0 1 3
Total Energy Consumption (%) MLEACH [20] 42 81 96 99 100
MOEIJSO-ACO 11.63 22.38 33.39 44.01 52.49

Table 4. Comparison of MOEIJSO-ACO with CEER

Performances Methods Number of rounds
2000 4000 6000 8000 10000
Alive nodes CEER [21] 223 204 182 145 134
MOEIJSO-ACO 225 225 225 188 153
Dead nodes CEER [21] 2 21 43 80 91
MOEIJSO-ACO 0 0 0 37 72
Total Energy Consumption (%) CEER [21] 23 38 49 63 71
MOEIJSO-ACO 17.33 29.07 36.12 58.19 65.14

International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 517

Table 5. Life expectancy comparison of MOEIJSO-ACO

Life 400 nodes 50 nodes 225 nodes
expectancy MCR-UWSN MOEIJSO- MLEACH MOEIJSO- CEER MOEIJSO-
[19] ACO [20] ACO [21] ACO
FND 852 182 110 1890 1700 6100
HND 1121 4185 600 4027 10200 11054
LND 1187 9089 1500 4851 12310 13008

Figure. 10 Life expectancy comparison of MOEIJSO-ACO with CEER

validation, resources, writing—original draft

5. Conclusion preparation, writing—review and editing.
In this research, energy-efficient CH and route
References
discovery are performed for improving the energy
efficiency of UWSN. The MOEIJSO is used for [1] S. Karim, F. K. Shaikh, B. S. Chowdhry, Z.
selecting the optimal CHs from the normal sensors by Mehmood, U. Tariq, R. A. Naqvi, and A. Ahmed,
optimizing it with residual energy, neighbor node “GCORP: Geographic and cooperative
distance, sink distance and CH balancing factor. Next, opportunistic routing protocol for underwater
ACO-based routing is developed for discovering the sensor networks”, IEEE Access, Vol. 9, pp.
optimal route from the transmitter CH to BS. The 27650-27667, 2021.
sensor’s energy usage is decreased by balancing the [2] M. Ismail, M. Islam, I. Ahmad, F. A. Khan, A.
clusters and identifying the route with a lesser B. Qazi, Z. H. Khan, Z. Wadud, and M. A.
transmission distance. The reduction in energy usage Rakhami, “Reliable path selection and
is used to increase the UWSN life expectancy. On the opportunistic routing protocol for underwater
other hand, data delivery is enhanced by improving wireless sensor networks”, IEEE Access, Vol. 8,
energy efficiency and avoiding node failure. The pp. 100346-100364, 2020.
MOEIJSO-ACO achieved better performances than [3] S. Chinnasamy, J. Naveen, P. J. A. Alphonse, C.
the MCR-UWSN, MLEACH and CEER. The alive Dhasarathan, and G. Sambasivam, "Energy-
nodes of the MOEIJSO-ACO are 399 for 800 rounds, Aware Multilevel Clustering Scheme for
which is high when compared to the MCR-UWSN. Underwater Wireless Sensor Networks", IEEE
In the future, modern optimization can be used to Access, Vol. 10, pp. 55868-55875, 2022.
enhance UWSN performances. [4] L. Alsalman and E. Alotaibi, "A Balanced
Routing Protocol Based on Machine Learning
Conflicts of interest for Underwater Sensor Networks", IEEE Access,
Vol. 9, pp. 152082-152097, 2021.
The authors declare no conflict of interest.
[5] T. R. Chenthil and P. J. Jayarin, “An energy-
efficient distributed node clustering routing
Author contributions protocol with mobility pattern support for
For this research work all authors' have equally underwater wireless sensor networks”, Wireless
contributed in Conceptualization, methodology, Networks, Vol. 28, No. 8, pp. 3367-3390, 2022.
International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40
Received: January 16, 2023. Revised: April 13, 2023. 518

[6] Z. Liu, X. Jin, Y. Yang, K. Ma, and X. Guan, Sensor Networks", IEEE Access, Vol. 9, pp.
"Energy-Efficient Guiding-Network-Based 36730-36747, 2021.
Routing for Underwater Wireless Sensor [16] T. R. Chenthil and P. J. Jayarin, “An Energy-
Networks", IEEE Internet of Things Journal, Aware Multilayer Clustering-Based Butterfly
Vol. 9, No. 21, pp. 21702-21711, 2022. Optimization Routing for Underwater Wireless
[7] K. Hao, Y. Ding, C. Li, B. Wang, Y. Liu, X. Du, Sensor Networks”, Wireless Personal
and C. Q. Wang, “An energy-efficient routing Communications, Vol. 122, No. 4, pp. 3105-
void repair method based on an autonomous 3125, 2022.
underwater vehicle for UWSNs”, IEEE Sensors [17] K. Bhattacharjya, S. Alam, and D. De,
Journal, Vol. 21, No. 4, pp. 5502-5511, 2020. “CUWSN: energy efficient routing protocol
[8] J. Wu, X. Sun, J. Wu, and G. Han, "Routing selection for cluster based underwater wireless
strategy of reducing energy consumption for sensor network”, Microsystem Technologies,
underwater data collection", Intelligent and Vol. 28, No. 2, pp. 543-559, 2022.
Converged Networks, Vol. 2, No. 3, pp. 163-176, [18] M. Faheem, R. A. Butt, B. Raza, H. Alquhayz,
2021. M. W. Ashraf, S. Raza, and M. A. B. Ngadi,
[9] T. Hussain, Z. U. Rehman, A. Iqbal, K. Saeed, “FFRP: dynamic firefly mating optimization
and I. Ali, “Two hop verification for avoiding inspired energy efficient routing protocol for
void hole in underwater wireless sensor network internet of underwater wireless sensor
using SM‐AHH‐VBF and AVH‐AHH‐VBF networks”, IEEE Access, Vol. 8, pp. 39587-
routing protocols”, Transactions on Emerging 39604, 2020.
Telecommunications Technologies, Vol. 31, No. [19] N. Subramani, P. Mohan, Y. Alotaibi, S.
8, p. e3992, 2020. Alghamdi, and O. I. Khalaf, “An Efficient
[10] H. Alasarpanahi, V. Ayatollahitafti, and A. Metaheuristic-Based Clustering with Routing
Gandomi, “Energy‐efficient void avoidance Protocol for Underwater Wireless Sensor
geographic routing protocol for underwater Networks”, Sensors, Vol. 22, No. 2, p. 415, 2022.
sensor networks”, International Journal of [20] H. H. Rizvi, S. A. Khan, and R. N. Enam,
Communication Systems, Vol. 33, No. 6, p. “Clustering Base Energy Efficient Mechanism
e4218, 2020. for an Underwater Wireless Sensor Network”,
[11] K. Latif, N. Javaid, I. Ullah, Z. Kaleem, Z. A. Wireless Personal Communications, Vol. 124,
Malik, and L. D. Nguyen, “DIEER: Delay- No. 4, pp. 3725-3741, 2022.
Intolerant Energy-Efficient Routing with Sink [21] I. Ahmad, I. T. Rahman, A. Zeb, I. Khan, M. T.
Mobility in Underwater Wireless Sensor B. Othman, and H. Hamam, “Cooperative
Networks”, Sensors, Vol. 20, No. 12, p. 3467, Energy-Efficient Routing Protocol for
2020. Underwater Wireless Sensor Networks”,
[12] M. Zhang and W. Cai, "Energy-Efficient Depth Sensors, Vol. 22, No. 18, p. 6945, 2022.
Based Probabilistic Routing Within 2-Hop [22] D. D. Yuan, M. Li, H. Y. Li, C. J. Lin, and B. X.
Neighborhood for Underwater Sensor Ji, “Wind power prediction method: Support
Networks", IEEE Sensors Letters, Vol. 4, No. 6, vector regression optimized by improved
pp. 1-4, 2020. jellyfish search algorithm”, Energies, Vol. 15,
[13] B. Priyalakshmi and S. Murugaveni, “Emperor No. 17, p. 6404, 2022.
Penguin Optimized Q Learning Method for
Energy Efficient Opportunistic Routing in
Underwater WSN”, Wireless Personal
Communications, 2022.
[14] A. Chaaf, M. S. A. Muthanna, A. Muthanna, S.
Alhelaly, I. A. Elgendy, A. M. Iliyasu, and A. A.
A. E. Latif, “Energy-efficient relay-based void
hole prevention and repair in clustered multi-
AUV underwater wireless sensor network”,
Security and Communication Networks, Vol.
2021, p. 9969605, 2021.
[15] S. Karim, F. K. Shaikh, K. Aurangzeb, B. S.
Chowdhry, and M. Alhussein, "Anchor Nodes
Assisted Cluster-Based Routing Protocol for
Reliable Data Transfer in Underwater Wireless
International Journal of Intelligent Engineering and Systems, Vol.16, No.3, 2023 DOI: 10.22266/ijies2023.0630.40