Enhancement of Quality of Service in Underwater Wireless Sensor Networks

International Journal of Computer Networks & Communications (IJCNC) Vol.17, No.2, March 2025
DOI: 10.5121/ijcnc.2025.17204 53
ENHANCEMENT OF QUALITY OF SERVICE IN
UNDERWATER WIRELESS SENSOR NETWORKS
Vinayprasad M S and Jayaram M N
Department of Electronics and Communication Engineering, JSS Science and
Technology University, Mysore, Karnataka.
ABSTRACT
Underwater Wireless sensor network (UWSN) has become a main topic in the research of underwater
communication with more research challenges. One of the main issues in the UWSN communication
process is Quality of Service (QoS). Therefore, for enhancing the QoS in the UWSN a novel Clustering
Hello routing based Honey Badger GoogleNet (CHbHBG) model is proposed. Primarily, the required
sensor hubs are placed in the underwater communication environment. Further, the energy usage of each
node is monitored and energy-efficient cluster head is selected by the proposed mechanism. Moreover, the
data rate resources were predicted and allocated at the channel using the fitness process of the model. The
optimal allocation process improves the QoS in the network. To prove the efficacy of the system, the
metrics including throughput, network lifetime, latency, energy consumption, PDR, transmission loss, and
path creation time are validated and compared with the recent models. The developed model attained the
higher network performance as 99.72% PDR, 949.2kbps throughput, 4004.31s network lifetime, and
230.84J energy consumption.
KEYWORDS
Underwater Wireless Sensor Networks, Routing protocol, Quality of Service(QoS), Cluster head, Honey
badger optimization, Hello routing
1. INTRODUCTION
UWSNs are a new and innovative way to monitor the aquatic environment in real-time. They
have the potential to revolutionize a wide range of time-sensitive applications, such as tracking
aquatic animals, collecting oceanographic data, preventing natural disasters, and monitoring
environmental conditions [1]. Even though land is the principal habitat and location for human
activity, it only makes up around 20% surface of the Earth. On the other hand, water bodies cover
about 80% surface of the Earth. Similar to space exploration, ocean exploration offers humanity
both essential opportunities and difficult problems" The underwater surroundings of the deep sea
is almost impossible to explore. However, the advent of UWSNs, which are made up of
Numerous independent and self-organizing sensor nodes (SN) has enabled group monitoring
tasks and conduct research on the seabed's natural resources [2,3,4].
UWSN has emerged as a ground-breaking technology. These networks, comprising small,
strategically deployed SN in aquatic environments, have enabled the scientific community and
various industries to delve into the depths of our oceans, unveiling their mysteries and harnessing
valuable information. However, the successful operation of UWSNs hinges on the critical
imperative of enhancing Quality of Service (QoS). This imperative arises from the unique
challenges of underwater communication, where signals contend with signal attenuation,
reflection, and dispersion. These challenges lead to increased latency, limited bandwidth, and the
risk of data loss [5,34].

54
Many UWSNs and SNs are powered by finite energy sources, necessitating energy-efficient
protocols to prolong their operational lifespans [6]. Ensuring the reliability and integrity of
transmitted data is also paramount, especially in applications like marine research and offshore
monitoring, where data accuracy is non-negotiable. Furthermore, low latency and excellent QoS
are required for certain real-time applications. To compound the complexity, UWSNs must adapt
to the ever-changing underwater environment, characterized by shifting water conditions and
dynamic marine life [7,35]. Energy efficiency is also important in UWSNs because SNs are
typically battery-operated and have restricted energy supplies [8]. Maintaining the applications'
QoS requirements while minimizing energy usage requires the implementation of QoS-aware
protocols and algorithms [9].
Several techniques can be used to improve QoS in energy-efficient UWSNs [10]. These
techniques may be divided into two categories: data-oriented techniques and network-oriented
techniques. Data-oriented methods concentrate on minimizing the quantity of data that must be
sent. Data aggregation, filtering, and compression may be used to achieve this [11]. Network-
oriented techniques focus on improving the efficiency of the network itself. This can be done
through routing protocols, medium access control (MAC) protocols, and power management
schemes [12].
Energy-efficient underwater communication systems, adaptive routing protocols, and machine
learning algorithms (ML) possess the capacity to improve QoS in UWSNs while still conserving
energy [13]. ML can be used to design intelligent QoS-aware protocols and algorithms that can
learn from the data and adapt to the shifting circumstances of the underwater world [14]. This
could help to improve QoS performance by enabling the network to make better decisions about
how to allocate resources [15]. Cross-layer design can be used to design QoS-aware protocols
and algorithms that take into account the relationships among the network stack's various layers
[16]. This could help to improve QoS performance by enabling the network to make more
informed decisions about how to optimize resource allocation and routing [17]. Improving QoS
in energy-efficient UWSNs is a challenging task, but it is essential for many applications. The
developed approach is exposed in Fig. 1. The primary significance of this work is summarized as
follows,
 Initially, the needed amount of UWSN was created with the limited data rate.
 Consequently, novel CHbHBG was designed with sufficient features for allocating the
resources.
 The cluster head (CH) was selected according to the energy consumption of every SN
capability.
 Moreover, the required data resources of each UWSN node were predicted, and resources
were allocated.
 Then, the QoS has been measured to estimate the designed model's robustness.
 Finally, all network metrics were measured and compared to standard models.
The novelty of this research lies in integrating and modifying the GoogleNet and Hello routing
protocol with the Honey Badger optimization algorithm. The optimization capabilities of the
Honey Badger with the GoogleNet network enhanced the QoS and helped in identifying the
optimal CH forwarder for a Node to enhance the QoS with an optimal data rate resource. The
proposed model can adapt to changing underwater conditions, such as signal strength and noise
level variations, improving the accuracy of routing decisions over time.

55
The article is arranged as follows, the merits and the limitations of the prevailing models are
debated in section 2, the main problem that motivated the present research is discussed in section
3, the process of novel method developed in this research is described in section 4, the
comparison of the outcomes are displayed in section 5, and the research's conclusion is presented
in Section 6.
2. RELATED WORKS
Ashwini et al. [18] proposed an Energy Optimization utilizing Routing Optimization (EORO) for
UWSNs is a cutting-edge routing technique. Swarm intelligence is used by this optimization to
choose the optimum transmit node for the transfer of data while maximizing QoS and energy
efficiency. Each forwarder node's residual energy, packet transfer capacity, node connectedness,
and distance are the four components that the optimization takes into account while calculating
fitness. To prevent packet collisions, and reduce consumption of energy as well as delay while
increasing throughput, these configurations are carefully selected. According to experimental
findings, it surpasses conventional routing approaches based on throughput, energy use, delay, as
well as Packet Delivery Ratio. It may be less scalable than other routing protocols, as the number
of parameters to tune increases with the size of the network.
Xingsi et al. [19] developed the enhanced artificial bee colony (EABC)- based clustering that
uses less energy with a cross-layer based on the Optimisation algorithm Harris-Hawks routing
strategy for WSN and k-medoids to attain enhanced quality-of-service (QoS) efficiency. QoS-
performance measurements are computed across several nodes. According to the experiment's
findings, it is more QoS-efficient than the existing approach. By lowering the packet-loss ratio,
increasing throughput, and lowering end-to-end delay, it boosts the effectiveness as well as
efficiency of WSNs. It has a higher overhead than some existing algorithms because it needs to
exchange more control messages with neighbor nodes to coordinate routing decisions.
Pramod et al. [20] developed an innovative, effective statistical channel framework for
computation. They sought to create an ideal channel model, transmission framework, as well as
protocol for routing for UWSN communication that was efficient in terms of energy delay-
tolerant, and QoS-enhanced. The suggested channel model aimed to construct an effective and
accurate channel model for audio communication by incorporating both random topological
alterations and fundamental physics notions of audio propagation. In contrast to conventional
channel models, the suggested statistical model can offer more precise and immediate replies.
The recommended routing protocol with repeated buffer flush improves this capability when used
in vast UWSNs where QoS-focused highly essential Information exchange and energy-efficient
(EE) transmission are important. It can be computationally expensive to implement, especially in
low-power sensor nodes.
Pramod et al. [21] proposed a systematic random linear network coding (SRLNC) transmission-
based routing technique for UWSNs. The proposed approach is designed to improve throughput,
computational complexity, delay, and energy efficiency in UWSNs. It achieves this by using
several techniques, including, a low Galois field and low coefficient vector, minimal
requirements for duplicate messages, An effective pre-coding approach in terms of computation,
Improved FEC-based decoding, and iterative buffer flushing. In addition to being more EE and
delayed resilient than current methods, the suggested SRLNC-routing strategy is also more robust
and dependable. It can be computationally expensive to implement, especially in low-power
sensor nodes.
Mamta and Nitin [22] proposed an energy-efficient localization technique for UWSNs using
mobility and propagation delay prediction. The system model is made up of regular nodes (RN)

56
surface buoys (SB) and anchor nodes (AN). SB floats on the water's surface, AN Ordinary nodes
are widely dispersed at different depths, and AN drift occurs at different depths. Every AN speed
is assessed and recorded throughout each localization session using the AN mobility estimation
approach. The precise localization is thus made possible by anticipating and adjusting for the
propagation delay. Overall efficiency in energy use has increased, according to simulation results.
The accuracy of the proposed scheme can be affected by the accuracy of the anchor node
mobility prediction.
3. SYSTEM MODEL WITH PROBLEM
Applications including tactical surveillance, ocean pollution monitoring, and water resource
exploitation were improved by the UWSN environment's increasing popularity. The network
consists of SN and the surface sink hub (SH) for data transmission to the base station (BS).
Acoustic communication is a technology used for underwater communication. However, energy
maintenance and improving the quality of service are the key processes for data reliability at
UWSN. QoS is the main problem that impacts the availability of the UWSN service. The
architecture of the UWSN with their problems is shown in Fig.1.
Sensor node
GPS signal
Acoustic signal
Satellite
Base
station
Surface sink node
 Poor Quality of service
 Higher energy consumption
 Less network life time
Fig.1. UWSN architecture and problems
Several research works were implemented for the UWSN; however, the QoS became less in
many cases due to the limited data rate of each node. Considering this, the QoS improvement is
considered for this study. In addition, for efficient prediction and QoS improvement, deep
learning with an optimization procedure is used. In addition, here the QoS is improved by
maximizing the throughput and network lifetime. For better network performance, data rate
resources were estimated and assigned to every node. Finally, the comparison is made with the
traditional approaches.
4. PROPOSED METHODOLOGY
The main objective of this work is to maximize throughput and network longevity to maximize
QoS. Hence, a novel Clustering hello-routing-based Honey Badger Googlenet (CHbHBG)
Mechanism has been introduced. Here, offering the required data rate resources for all the UWSN
nodes are considered the major process. Finally, the throughput ratio and network longevity were
measured to determine the UWSN network's efficiency. Besides, all network parameters were
measured and compared with all traditional studies. Fig. 2 illustrates the steps of the proposed
technique.

57
UWSN
CHbHBG
Offering data
resources
Data sharing
Improving
QoS
Performance metrics
Fig.2 Proposed methodology
4.1. Node Initialization
The proposed method aims to allocate the data rate resources in the radio acoustic link of the
UWSN to send the data packets that the node detected without interruption, which leads to
increased QoS. Here the resources are allocated by the combined process of Honey Badger
optimization [23] and GoogleNet [24]. The process is begun by the deployment of required nodes
in the UWSN environment. The required node initialization process is shown in Eqn. (1).
}
,.......
3
,
2
,
1
{ z
n
X
X n 
 (1)
Here the Underwater network (UN) is indicated as X , n
X specifies the network nodes and z
indicates the overall quantity of nodes. The nodes are initialized based on the concept of Google
Net. The nodes gather the information underwater and send it to the SN are positioned at the
water's surface using an acoustic and radio connection. According to the link quality and
availability, the transmission is done for the data from the SN.
After, the node initialization process, the proposed model utilized the Google Net structure for
the selection of the best CH and offering data resources. GoogleNet is a type of deep
convolutional neural structure. The features such as connectivity, residual energy levels, the
distance between the nodes, and sink node (SN) distances were extracted from the sensor nodes'
data by the Google Net function to generate the search area for the CH selection. Here the Google
Net includes the layers such as layers for input, convolution, pooling, inception, completely
connected, and output. Utilizing these features extracted by Google Net, the honey badger
function identifies the best optimal CH. The CH selection process is described as follows.
4.2. Cluster Head Selection
An energy-efficient cluster head is chosen to transfer the data with minimal energy usage after
the positioning of the nodes. Here the CH is selected by evaluating the energy usage of each
node. The energy-efficient CH was selected based on the intensity-defining function of the
Honey badger. The selection CH process is explained in Eqn. (2).

58
2
4
)
(
i
n
c
d
X
E
r
S


 (2)
Here, c
S denotes the CH selection variable, r is the random number ranging from 0 to 1,
)
( n
X
E expresses the node's energy consumption (EC) in UWSN, and
2
i
d denotes the distance
of the nodes. Here 


n
X
N
N
n
n R
X
X
E
1
)
( , N
R denotes the remaining energy in each node. The
Google Net function keeps track of each node's energy usage. Therefore, the CH is chosen based
on its distance from the sink node and the quantity of energy it utilizes.
4.3. Offering Data Resources
The selected CHs obtain the information from the sub-node and transmit it to another CH or BS.
During the data transmission, to enhance the QoS, optimal data rate resources must be allocated
at the radio link. However, to allocate the resources, the data rate level should be determined
initially. The data rate level determination has been carried out based on the density factor
updating process of the Honey badger and is explained in Eqn. (3).







 



max
exp
)
(
t
t
X
D
c
P n
dr (3)
Here the data rate prediction variable is expressed as dr
P , c denotes the constant value,
)
( n
X
D indicates the data present in each node, t defines the current iteration, and max
t
indicates the maximum number of data rate prediction iterations. At this phase, the data rate level
required for sending the data is predicted. Further, to allocate the data rate, the flow rate of the
link should be non-negative and the allocation is defined in Eqn. (4). The data rate allocation
process is earned by the digging phase of the Honey Badger.
d
P
h
F
D
h
s
K
l dr
n
c
n
n
i
dr 







  (4)
Here dr
l denotes the data rate allocation function, i
K denotes the SN, c
D is the amount of data
present in each node of the UWSN, n
h is the selected cluster heads, F denotes the honey badger
that alters the allocation, n
s indicates the other sub-nodes,  is the data rate allocation variable,
and d represents the nodes' distance information.
4.4. Data Sharing
Following the method for allocating resources, the data from the nodes are broadcasted to the
destination using the Hello routing function. The sub-nodes have a small amount of data to
transmit while the CHs have a large amount of data. Once, the data rate has been allocated the
specific sensor nodes start to broadcast the data. In the hello routing protocol, the Hello message
that includes information is sent to the nearer neighbor node and maintains an internal
neighborhood table. When the cluster head receives the data, the neighborhood table is updated.
In this way, the information Nodes are sent to the sink node from the sensor via the selected CHs

59
and then to the base station. The functions and process utilized in the proposed CHbHBG are
explained in pseudocode format in Algorithm.1. The Algorithm inhibits all the expressions that
are considered to complete the process of the QoS enhancement.
Algorithm.1. CHbHBG
Start
{
Node initialization ()
{
n
X
X ,
int
z
n
X n ,.......
3
,
2
,
1


//nodes are placed in the created UWSN environment
}
The proposed CHbHBG model is designed.
// The necessary elements are included in the protocol design.
{
Cluster Head selection ()
{
i
c d
E
r
S ,
,
,
int
//CH selection variables are initialized
The energy consumption of each node is monitored.
2
4
/
. i
n
c d
E
X
r
S 


//CH is selected based on energy consumption rate
}
Data rate Resource prediction ()
{
max
,
,
,
,
int t
t
D
c
P
dr
//data rate prediction variables are initialized
)
/
exp(
)
( max
t
t
X
D
c
P n
dr 



//data rate required to send each node’s data was predicted
}
Data rate resource allocation ()
{
d
F
D
h
s
K
l c
n
n
i
dr ,
,
,
,
,
,
,
int 
//resource allocation variables are initialized
d
P
h
F
D
h
s
K
l dr
n
c
n
n
i
dr 







 )
( 
//required data rates were allocated in the transmission path
}
Data from each node are broadcasted using the Hello routing function.
}
Stop
5. RESULTS AND DISCUSSION
The proposed CHbHBG using the MATLAB platform Primarily the required nodes are initialized
in the network environment. Further, a novel CHbHBG model is proposed for the enhancement
of QoS in UN communication, and the performance metrics are evaluated and compared with
other prevailing methods. The simulation parameters required for designing the underwater
sensor network are listed in Table 1

60
Table 1. Simulation parameters
Parameter Description
Platform MATLAB
Network area 2
2000
2000 m

Idle power 0.008 W
Nodes 50,100,150,200,250,300,350,400,450,500
Mobility of nodes Random
Sink hub 10
Network topology 2D
Packet size 1000 bytes
Antenna Omni antenna
Range of transmission 70m
Wireless channel Acoustic and radio
Frequency 15KHz
Communication medium Wireless
Receiving power 0.1 W
Initial energy 1000 J
Transmitting power 0.5W
5.1. Case Study
This study was conducted to understand the simulation process of the proposed CHbHBG model
with step-by-step simulation results. The necessary nodes are first set up in the simulated
environment. The node initialization process for 50 and 100 nodes is displayed in Fig.3. The
initialized sensor nodes collect the information of the underwater.
Fig.3. Node Initialization (a) 50 nodes, (b) 100 nodes

61
Further, to transfer the information from nodes to SH, an optimal cluster node is selected. Here
the cluster nodes are selected on their EC rate. Using the Honey Badger function nodes
consuming lower energy are selected as CH. The cluster selection process of the presented
CHbHBG model for 50 and 100 nodes is shown in Fig.4. Here the nodes indicate the selected
cluster heads. For 50 nodes network, 4 nodes are selected as CH, and for the network with 100
nodes, 10 nodes are selected as cluster heads.
Fig.4. CH selection (a) 50 nodes (b) 100 nodes
After, the CH selection process, the data rate of each node needs to be identified for the resource
allocation process. The data rate prediction process for 50 and 100 nodes in the simulation
environment has been shown in Fig.5. Here the data rate required for Every node has a prediction
and is shown.

62
Fig.5. Data rate resource prediction and allocation (a) 50 nodes (b) 100 nodes
After the optimal resource allocation, the shortest path between the source and the sink hub is
identified and the data is transmitted using the hello routing process. The data broadcasting
process from the source to SH through the selected cluster heads for 50 and 100 nodes is
displayed in Fig.6.

63
Fig.6. Data transmission (a) 50 nodes (b) 100 nodes
The efficiency of the developed model is evaluated regarding the network throughput, latency,
PDR, lifetime, EC, and transmission loss.
5.2. Comparative Analysis
The working robustness in the created UWSN environment has been measured in terms of
throughput, network lifetime, latency, EC, PDR, transmission loss, path creation, and simulation
time. The metrics values attained for the present method are compared with the results of the
prevailing models. The prevailing models that are considered to continue the comparative section
are Whale optimized UN (WOUN) [25], Energy Efficient Multilayer Clustering Protocol
(EEMCP) [26], Multipath Energy Aware Protocol (MEAP) [27], Ant Colony based Routing
Protocol (ACbRP) [28], Time-Based Reliable Link (TBRL) [29], Depth Clustering Routing
Model (DCRM), Transmission Adaptive Depth Routing (TADR) and Collision Aware Routing
Protocol (CARP) [30], Elman Bat-based Hello Routing (EBbHR) [31].

64
5.2.1. Network Lifetime
The duration of the active phase that sends and receives the data packet until the node dies is
validated as network lifetime. In other words, the period of the placed nodes when they run out of
energy to send packets is defined as network lifetime. Here the network lifespan is measured in
Seconds.
Fig.7. Network lifetime comparison
The comparison of network lifetime for increasing node numbers is illustrated in Fig.7. Here the
entire network lifespan of the developed approach is made a comparison with conventional
literature such as WOUN, TADR, and CARP. The network lifetime attained for the WOUN
method is the 1980s, 1798s, 1699s, 1557s, and 1380s for 100 to 500 nodes. Similarly, for the
methods TADR and CARP, the attained network lifetime for different nodes are 720s, 730s,
809s, 863s, and 884s; 714s, 750s, 804s, 859s, and 898s. The network lifetime attained for the
proposed model with varying nodes such as 100, 200, 300, 400, and 500 are 2494.2882s,
3196.6437s, 3744.5088s, 3782.608s and 4004.3192s. Here the lifespan of the created network is
higher than the traditional network models.
5.2.2. Latency
The time taken for departed packets to reach the SH is measured as latency. It is the typical delay
of all information packets flowing at the target from various sources. The time efficiency of the
network can be calculated by latency or delay. The latency calculation is shown in Eqn. (5).
packets
total
time
deliver
packet
L  (5)
The comparison of latency values for varying nodes is shown in Fig.8. Here the latency values of
the recommended CHbHBG model are compared with the conventional models such as TBRL
and EEMCP. For the varying nodes the model TBRL attained the delay rate of 369.4ms,
319.6ms, 240.1ms, 198.6ms, and 159.4ms; also, the model EEMCP scored 610ms, 512ms,
436ms, 338ms, and 253ms. The model proposed model’s delay rate for nodes such as 100, 200,
300, 400 and 500 are 161.8484ms, 121.2393ms, 92.4923ms, 79.5943ms and 67.9194ms. The
delay rate of the CHbHBG model is less than the existing models.

65
Fig.8. Latency Comparison
5.2.3. Energy Consumption
The total energy consumed by the network to gather and transmit data packets to the BS is known
as EC. The size of the packet affects the EC of the network and the distance it needs to be sent. In
the present research, energy usage is measured in Joules (J).
Fig.9. Energy consumption comparison
The energy usage comparison for the various methods is displayed in Fig.9. The EC values for
the prevalent methods, namely WOUN, TBRL, and ACbRP, opposed to this, the suggested
method use 48.4993 J, 113.1326 J, 165.496 J, 199.7867 J, and 230.8441 J for 100 to 500 nodes.
The resulting values of CHbHBG uses less energy than other existing methods for all different
numbers of nodes.
5.2.4. Throughput
One of the key metrics for the network efficiency measurement is throughput. The ratio of the
total percentage of data packets sent to the SH to the time it takes for the packets to arrive at the
SH is known as throughput. It validates the capacity that how much data can be processed in the
created UWSN at a specific time. The formulation for throughput ratio calculation is shown in
Eqn. (6).

66
time
simulation
packets
arrived
T  (6)
Fig.10. Throughput comparison
The throughput comparison with the existing approaches is shown in Fig.10. The prevailing
model WOUN scored the throughput as 188Kbps, MEAP scored 299Kbps, and TBRL scored
26.9Kbps. However, the throughput rate of the proposed network model is greater than other
schemas which are valued at 949.2Kbps for 500 nodes.
5.2.5. Packet Delivery Ratio (PDR)
The PDR is determined by the ratio of adding together all of the data packets that arrive at the SH
to the packets that leave the source nodes. The ratio of all transmitted packets from the source to
all packets reached at the SH is used to calculate it. The formula for calculating average PDR is
found in Eqn. (7).
sent
packets
total
packets
arrived
PDR  (7)
Fig.11. PDR comparison

67
The PDR for the suggested routing protocol was 91.4128%, 95.2683%, 96.5743%, 97.342%, and
99.721% for 100 to 500 nodes respectively. In this case, the PDR of CHbHBG is more than that
of the other methods in use. The comparison graph is displayed in Fig.11.
5.2.6. Transmission Loss
Propagation delay is the term used to describe the loss of transmission in the underwater acoustic
network. The lower the transmission loss increases the network communication efficiency. The
transmission loss calculation is given in Eqn. (8).
3
10
*
*
log
20 

 t
t
LT  (8)
Here the distance between the receiver and sender is indicated as t measured in meters, is
indicates the coefficient of absorption computed in dB/km.
Fig.12. Transmission loss comparison
Fig.12. compares the transmission loss with the current protocol, TBRL. The suggested
CHbHBG approach obtained a relatively low frequency of transmission loss for each variable
when compared to the TBRL method. 71.445 dB, 127.009 dB, 223.259 dB, 387.31 dB, 601.59
dB, 798.782 dB, 1021.908 dB, 1302.295 dB, 1616.716 dB, and 1918.469 dB are the transmission
losses that the proposed CHbHBG achieves for various frequencies, including 500 Hz, 1500 Hz,
2500 Hz, 3500 Hz, 4500 Hz, 5500 Hz, 6500 Hz, 7500 Hz, 8500 Hz, and 9500 Hz. Compared to
other protocols, the presented routing model provided a very low transmission loss.
The entire performance of the designed model in the UWSN environment is compared with the
current techniques to highlight performance improvement. The recent techniques considered for
the comparison are Elman Bat-based Hello Routing (EBbHR) [31], Q-learning based Routing
Protocol (QbRP) [32], and Route Focusing based Protocol (RFbP) [33]. These methods were
tested in the same environment in the MATLAB platform, and efficiency metrics were evaluated.
The performance comparison with these methods is shown in Table 2.

68
Table 2. CHbHBG performance
Methods Throughput
(Kbps)
Network Lifespan
(sec)
PDR(%) Energy usage
(J)
Delay
(ms)
QbRP 258.5 1212.7 92.45 903.5 397.6
RFbP 203 1084.45 86.62 1286.6 443.4
EBbHR 696 2984.53 98.9 350.85 72.51
Proposed 949.2 4004.31 99.721 230.844 67.919
The innovative CHbHBG achieved maximum Throughput and network lifetime of 949.2 Kbps
and 4004.31sec, respectively. Furthermore, it requires less energy and has a short latency.
Furthermore, the ideal routing path is immediately developed with a short distance, making
transmitting data easy. The CHbHBG process focuses on the enhancement of QoS between
sensor hubs employed within the created UWSN, improving network efficiency and reliability.
The integration of Clustering Hello routing with Honey Badger GoogleNet identified the optimal
QoS routing path for data transmission. It leads to better routing decisions and enhances the
overall network performance. By enhancing the efficient CH for the routing and optimizing
mechanism, the CHbHBG model reduces the network's overall energy consumption.
6. CONCLUSION
A novel CHbHBG model is developed to enhance the QoS of the UWSN through cluster head
selection and resource allocation. Initially, the sensor nodes were deployed in the UWSN
communication environment. The nodes are communicated via radio and acoustic links. The
presented model monitors the energy use for each node to determine, which is the most energy-
efficient CH for the routing. Moreover, the data rate resources required to send the data of each
node were predicted and allocated according to the honey badger's fitness procedure.
Subsequently, the data are broadcasted by the routing process to the sink node. Thus, the QoS of
the network can be enhanced by the presented model. The recommended model attained the
higher network performance as 99.721% PDR, 949.2kbps throughput, 4004.31s network lifetime,
and 230.84 J EC. The performance of the model is improved from the existing models. Therefore,
the presented system is better suitable for the QoS enhancement in UWSN. Additionally, the
protocol maximizes throughput, network longevity, less energy consumed and exhibits low
Latency, further contributing to maximizing efficiency and QoS. One of the key strengths of the
CHbHBG protocol is its ability to establish an optimal and reliable routing path with a guaranteed
data rate, making data transmission more efficient and reliable. This is a crucial advantage in
UWSNs, where the underwater environment can pose significant challenges to communication.
In the future, the research will be extended to propose a highly scalable, energy-efficient, and
most importantly secure routing method for underwater sensor nodes with better optimization
strategies to function well in real-time challenges.
CONFLICTS OF INTEREST
The authors declare no conflict of interest.
REFERENCES
[1] Kaveripakam S, Chinthaginjala R. Energy balanced reliable and effective clustering for underwater
wireless sensor networks. Alexandria Engineering Journal. 2023 Aug 15;77:41-62.
[2] Akyildiz IF, Pompili D, Melodia T. Underwater acoustic sensor networks: research challenges. Ad
hoc networks. 2005 May 1;3(3):257-79.

69
[3] Wang Q, Li J, Qi Q, Zhou P, Wu DO. A game-theoretic routing protocol for 3-D underwater
acoustic sensor networks. IEEE Internet of Things Journal. 2020 Apr 17;7(10):9846-57.
[4] Wang B, Zhang H, Zhu Y, Cai B, Guo X. Adaptive Power-Controlled Depth-Based Routing
Protocol for Underwater Wireless Sensor Networks. Journal of Marine Science and Engineering.
2023 Aug 9;11(8):1567.
[5] Shah SM, Sun Z, Zaman K, Hussain A, Ullah I, Ghadi YY, Khan MA, Nasimov R. Advancements
in Neighboring-Based Energy-Efficient Routing Protocol (NBEER) for Underwater Wireless Sensor
Networks. Sensors. 2023 Jun 29;23(13):6025.
[6] Ibraheem MK, Mohamed MB, Fakhfakh A. Energy Optimization Efficiency in Wireless Sensor
Networks for Forest Fire Detection: An Innovative Sleep Technique.
[7] Kaveripakam S, Chinthaginjala R, Anbazhagan R, Alibakhshikenari M, Virdee BS, Khan S, Pau G,
See CH, Dayoub I, Livreri P, Abd-Alhameed R. Enhancement of precise underwater object
localization. Radio Science. 2023;58(9):1-29.
[8] Cui Y, Zhu P, Lei G, Chen P, Yang G. Energy-Efficient Multiple Autonomous Underwater Vehicle
Path Planning Scheme in Underwater Sensor Networks. Electronics. 2023 Aug 3;12(15):3321.
[9] Kaur N, Verma S, Jhanjhi NZ, Singh S, Ghoniem RM, Ray SK. Enhanced QoS-aware routing
protocol for delay sensitive data in Wireless Body Area Networks. IEEE Access. 2023 Sep 4.
[10] Alablani IA, Arafah MA. EE-UWSNs: A joint energy-efficient MAC and routing protocol for
underwater sensor networks. Journal of Marine Science and Engineering. 2022 Apr 1;10(4):488.
[11] Messinis GM, Hatziargyriou ND. Review of non-technical loss detection methods. Electric Power
Systems Research. 2018 May 1;158:250-66.
[12] Kanellopoulos D, Sharma VK, Panagiotakopoulos T, Kameas A. Networking Architectures and
Protocols for IoT Applications in Smart Cities: Recent Developments and Perspectives. Electronics.
2023 May 31;12(11):2490.
[13] Shah SM, Sun Z, Zaman K, Hussain A, Ullah I, Ghadi YY, Khan MA, Nasimov R. Advancements
in Neighboring-Based Energy-Efficient Routing Protocol (NBEER) for Underwater Wireless Sensor
Networks. Sensors. 2023 Jun 29;23(13):6025.
[14] Wijesekara PA, Gunawardena S. A Review of Blockchain Technology in Knowledge-Defined
Networking, Its Application, Benefits, and Challenges. Network. 2023 Aug 30;3(3):343-421.
[15] Khasawneh M, Azab A, Alrabaee S, Sakkal H, Bakhit H. Convergence of IoT and Cognitive Radio
Networks: A Survey of Applications, Techniques, and Challenges. IEEE Access. 2023 Jul 10.
[16] Arifuzzaman BM. Cross-Layer Design in the Internet of Things (IoT): Issues and Possible
Solutions.
[17] Zia K, Chiumento A, Havinga PJ. AI-enabled reliable QoS in multi-RAT wireless IoT networks:
prospects, challenges, and future directions. IEEE Open Journal of the Communications Society.
2022 Oct 19;3:1906-29.
[18] Gavali AB, Kadam MV, Patil S. Energy optimization using swarm intelligence for IoT-Authorized
underwater wireless sensor networks. Microprocessors and Microsystems. 2022 Sep 1;93:104597.
[19] Xue X, Shanmugam R, Palanisamy S, Khalaf OI, Selvaraj D, Abdulsahib GM. A hybrid cross layer
with harris-hawk-optimization-based efficient routing for wireless sensor networks. Symmetry.
2023 Feb 7;15(2):438.
[20] Basavaraju PH, Lokesh GH, Mohan G, Jhanjhi NZ, Flammini F. Statistical channel model and
systematic random linear network coding based qos oriented and energy efficient uwsn routing
protocol. Electronics. 2022 Aug 18;11(16):2590.
[21] Basavaraju PH, Lokesh GH, Mohan G, Jhanjhi NZ, Flammini F. Statistical channel model and
systematic random linear network coding based qos oriented and energy efficient uwsn routing
protocol. Electronics. 2022 Aug 18;11(16):2590.
[22] Nain M, Goyal N. Energy efficient localization through node mobility and propagation delay
prediction in underwater wireless sensor network. Wireless Personal Communications. 2022
Feb;122(3):2667-85.
[23] Hashim FA, Houssein EH, Hussain K, Mabrouk MS, Al-Atabany W. Honey Badger Algorithm:
New metaheuristic algorithm for solving optimization problems. Mathematics and Computers in
Simulation. 2022 Feb 1;192:84-110.
[24] Anand R, Shanthi T, Nithish MS, Lakshman S. Face recognition and classification using
GoogleNET architecture. InSoft Computing for Problem Solving: SocProS 2018, Volume 1 2020
(pp. 261-269). Springer Singapore.

70
[25] Rathore RS, Sangwan S, Mazumdar S, Kaiwartya O, Adhikari K, Kharel R, Song H. W-GUN:
Whale optimization for energy and delay-centric green underwater networks. Sensors. 2020 Mar
3;20(5):1377.
[26] Khan W, Wang H, Anwar MS, Ayaz M, Ahmad S, Ullah I. A multi-layer cluster based energy
efficient routing scheme for UWSNs. IEEE Access. 2019 Jun 10;7:77398-410.
[27] Ahmed M, Salleh M, Channa MI, Rohani MF. RMEER: reliable multipath energy efficient routing
protocol for underwater wireless sensor network. International Journal of Electrical and Computer
Engineering (IJECE). 2018 Dec;8(6):4366-73.
[28] Zahra N, Khawatmi S, Fawaz Y. Energy Efficient Routing Protocol in UWSNs using ACO
Algorithm. International Research Journal of Engineering and Technology (IRJET). 2021 May;8(5).
[29] Ali T, Irfan M, Shaf A, Saeed Alwadie A, Sajid A, Awais M, Aamir M. A secure communication in
IoT enabled underwater and wireless sensor network for smart cities. Sensors. 2020 Aug
2;20(15):4309
[30] Sher A, Khan A, Javaid N, Ahmed SH, Aalsalem MY, Khan WZ. Void hole avoidance for reliable
data delivery in IoT enabled underwater wireless sensor networks. Sensors. 2018 Sep
28;18(10):3271.
[31] Vinayprasad M S & Jayaram. M N (2024). Intelligent Efficient Routing and Localization in the
Underwater Wireless Sensor Network to Improve Network Lifetime (EBbHR). International journal
of Computer Networks & Communications. 16. 01-21. 10.5121/ijcnc.2024.16501.
[32] Nandyala CS, Kim HW, Cho HS, (2023) “QTAR: A Q-learning-based topology-aware routing
protocol for underwater wireless sensor networks”, Computer Networks, Vol. 222, pp 109562.
[33] Rajshekhar SA, Biradar A, (2024) “An Efficient Framework for Localization Based Optimized
Opportunistic Routing Protocol in Underwater Acoustic Sensor Networks- (RFbP)”, SN Computer
Science, Vol. 5, No. 5, pp 520.
[34] Gulista Khan and R.K. Dwivedi, (2018) "Energy Efficient Routing Algorithm for Void Avoidance
in UWSN Using Residual Energy and Depth Variance", International journal of Computer
Networks & Communications.
[35] Sihem Souiki et. al "FUZZY BASED CLUSTERING AND ENERGY EFFICIENT ROUTING
FOR UNDERWATER WIRELESS SENSOR NETWORKS" International Journal of Computer
Networks & Communications (IJCNC) Vol.7, No.2, 2015.
AUTHORS
Vinayprasad M S is currently working as an Assistant Professor in the Department of
Electronics and Communication Engineering, JSS Science and Technology, Mysuru. He
completed his M. Tech in Networking and Internet Engineering from SJCE in the year
2014. He is currently pursuing his Ph.D. at JSS Science and Technology University. His
research area of interest includes Networking, Storage Networking, Communication and
Internet of Things.
Dr. Jayaram M N is currently an Associate Professor at the Sri Jayachamarajendra
College Engineering, he has teaching experience of more than 31 years in the areas of
mobile and optical Communication. He obtained his PhD at the University of Mysore and
ME in IISC, Bangalore and participated in several high-profile conferences. His research
area of interest includes underground communication, microwave and antennas, RF and
Optical Communication.

Enhancement of Quality of Service in Underwater Wireless Sensor Networks

More Related Content

Similar to Enhancement of Quality of Service in Underwater Wireless Sensor Networks (20)

More from IJCNCJournal (20)

Recently uploaded (20)

Enhancement of Quality of Service in Underwater Wireless Sensor Networks