Performance and Improvement Analysis of the Underwater WSN Using a Diverse Routing Protocol Approach

,


Introduction
Although water covers a large portion of our planet, much of it is still unknown [1][2][3][4]. Exploration in this region has signifcantly increased recently. In addition to being endowed with abundant natural resources, it has contributed signifcantly to the development of ships, oil pipelines, and the military. In environments like oceans, seas, and the like, which contain enormous amount of naturally occurring data, this is the researcher's top priority. By developing a variety of UWSN protocols for the underwater environment, the developers have been able to gather and analyse a signifcant amount of data to some extent [5][6][7][8]. Research in this area is built on the fndings of earlier studies, as described in the background information below. Te general structure of underwater wireless communication is depicted in Figure 1.

Routing Protocol Abbreviations.
Ad-hoc on-demand distance vector (AODV) Zone routing protocol (ZRP) Dynamic source routing protocol (DSR) fsheye state routing (FSR) Dynamic MANET on-demand routing protocol (DYMO) location-aided routing (LAR) Optimized link state routing protocol (OLSR) Source tree adaptive routing (STAR) Source tree adaptive routing-optimum routing approach (STAR-ORA) Source tree adaptive routing-least overhead routing approach (STAR-LORA) Ad-hoc wireless networks known as wireless sensor networks (WSNs) are used to provide a wireless communication setup, such as for underwater wireless communication [9]. Routing protocols for ad-hoc networks include AODV, DSR, DYMO, LAR1, Bellman-Ford, OLSR, fsheye, STAR-ORA, ZRP, and STAR-LORA, among others [10].
Te mobility model shows the movement of nodes as well as how their positions, speeds, and accelerations alter over time. When researching a new network protocol, it is crucial to simulate and assess the protocol's performance. In protocol simulation, the mobility model and the communicating trafc pattern are two crucial variables. Mobility models are used to describe user movement patterns. Te state of mobile services is described by the trafc model [11][12][13][14][15]. Figure 2 illustrates a realistic 3D underwater wireless communication scenario with a variety of nodes. Te potential challenges posed by the surrounding subsurface environment must be given the attention and consideration they require when thinking about the use of underwater sensor networks [16][17][18]. Te host conditions present numerous difculties, including three-dimensional topology and continuous node movement. In addition, a lot of underwater applications, like those used for detection or rescue missions, tend to be ad-hoc in nature [19][20][21][22][23][24].
As depicted in Figure 3, a wireless sensor network (WSN) is made up of spatially dispersed autonomous (selforganized) sensors that monitor environmental or physical conditions such as temperature, sound, vibration, pressure, motion, or pollutants and cooperatively transmit their data through the network to a central location. Modern networks are bidirectional, allowing for the control of sensor activity as well. Military use of wireless sensor networks, such as battlefeld surveillance, served as the impetus for their development. Today, these networks are employed in a wide range of commercial and consumer applications, including machine health monitoring, process control, and industrial process monitoring. Te WSN is composed of "nodes," which can range in number from a few to thousands and are each connected to one or more sensors [1]. Figure 3 depicts the main parts of a sensor node, which include the following subsystems: a communication (transceiver) subsystem, a computing (processing) subsystem, a sensor subsystem, and a unifed power supply system.

Related Works.
Tis section of the article examines a previous work from the standpoint of network architecture, as well as the multiple performance measures that support the concept of extending network life.
Bhattacharjya et al. investigated a universal wireless sensor network (UWSN) using a grid topology [25,26]. Tis study investigates energy usage across a variety of energy modalities, as well as network performance.
EAVARP, a void-avoidant and energy-conscious routing system, is investigated by Wang et al. for wireless sensor networks [27]. UWSN is impervious to transmission, vacancy, and fooding cycles. UWSN performance parameters include packet delivery ratio, network lifetime, energy utilisation, and transmission latency. In terms of mobility, Alkindi et al. investigated a grid-based routing technique for UWSN [28]. Energy usage, network density, packet delivery ratio, and latency are all discussed.
An optimal, collaborative, and resource-saving strategy is being examined [23]. In order to create a UWSN that consumes less energy, a relay node is chosen. Te indices of dead and end-to-end delays, dead packet delivery ratio, and energy usage are investigated. Te nodes are organized in a two-dimensional environment with a regular distribution. In the case of the UWSN, both interference and noise are considered. Khan et al. examines interference-free localization routing for ultrawideband sensor networks (UWSNs), with the goal of minimizing the energy hole [23]. It specifes the total number of dropped packets, the total number of dead nodes, a packet received at the sink, the total amount of energy used, and the total number of dropped packets.
Te main contribution of the manuscript is highlighted as follows:  (i) To design and implement a source tree adaptive routing-least overhead routing approach (STAR-LORA) protocol for underwater wireless sensors (UWSN) network (ii) To compare the STAR-LORA routing protocol with standard routing protocols in the literature, namely, AODV, DSR, DYMO, LAR1, Bellman-Ford, OLSR, fsheye, STAR-ORA, and ZRP (iii) To analyse the energy efciency of the routing protocol with increasing number of underwater wireless sensor nodes (iv) To analyse the tradeof between average transmission delay, average jitter, utilisation rate, and energy in transmit and receive modes (v) To recommend an appropriate routing protocol based on the targeted performance metric for the underwater wireless sensor network Te remainder of this article is organized as follows. Te network scenario is covered in Section 2. Te performance parameter is presented in Section 3. We present the fndings of our results in Section 4. A comparison of results and discussion are presented in Section 5. In Section 6, we fnally conclude.

Network Scenario
With CBR as a deployment application, there are accessible existing networks available. In the proposed network, FTP and VBR are taken into account alongside CBR, and the parameters for all three FTP, CBR, and VBR applications are then compared [14,26,[29][30][31][32]. Te proposed scenario has a 1500 by 1500 square meter design in the QualNet 7.1 Simulator. Te fle transfer protocol (FTP), constant bit rate (CBR), and variable bit rate (VBR) applications are connected by both 60 and 120 nodes, of which 15 are node devices, 25 are ship devices, and 20 are sensor devices for 60 nodes and similarly 30 are node devices, 50 are ship devices, and 40 are sensor devices. Te simulation lasts for 500 seconds in total. Random way-point mobility with a minimum speed of 1.5 m/sec and a maximum speed of 3 to 10 m/ sec is the node mobility model that has been selected. Following AODV as the initial routing protocol are DSR, DYMO, LAR1, Bellman-Ford, OLSR, fsheye, STAR-ORA, ZRP, and STAR-LORA. After fnishing the test, the graphs in the simulator were considered. Te required performance metrics are thus obtained, including the average transmission delay, average jitter, percentage of utilisation, as well as the energy used in transmit, receive modes. Te proposed underwater wireless communication scenario with multiple nodes in X-Y and 3D visualization is shown in Figures 4 and 5, respectively, for 60 nodes. Runtime proposed scenario for the underwater wireless communication with various nodes in both X-Y and 3D visualization is shown in Figures 6 and 7, for 120 nodes [18,22].

Performance Parameters
OPNET, OMNeT, MATLAB, QualNet, and other simulation tools, among others [25,28,[33][34][35][36], commonly take the design of UWSN into account. Te proposed network is created in the QualNet simulator using a number of userfriendly UWSN design parameters. Te performance indicators for the UWSN network in various applications are listed as shown in Figure 8.
Energy consumption: Te energy used by nodes to send data from their point of origin to their point of destination. Average transmission delay: Average transmission delay is the amount of time it takes for an information to successfully travel from its source to its destination [37]. Average jitter: Tis is the time diference between individual packets as a result of route changes or network Journal of Computer Networks and Communications congestion. A routing protocol should be lower to function more efciently. A network's congestion, routing modifcations, or timing drift can all increase jitter by delaying the transmission of individual packets [38].
Percentage of utilisation: Te proportion of packets that are successfully transferred from the transmitting node to the receiving node is known as the throughput of a communication channel [39].

Results and Discussion
In this section, the deployment of sensor nodes in 3D for the proposed UWSN network scenario with sensor nodes and anchor nodes is shown in Figure 9. We have considered the data packet size as 50-100 bytes, transmission energy � 48 dB, propagation speed � 3 × 10^8 m/s, random waypoint mobility with 10 m/s, bit rate � 1 Mbps, transmission range � 50 m, and depth of nodes � 20 m (Max) [40,41].
In addition, the investigational results for various routing protocols including AODV, DSR, DYMO, LAR1, Bellman-Ford, OLSR, fsheye, STAR-ORA, ZRP, and STAR-LORA with FTP, CBR, and VBR applications are observed for transmit and receive mode power consumption as displayed in Figures 10(a) and 11(t) for both 60 and 120 nodes. As seen from Figures 10(a)-10(t) and Figures 11(a)-11(t) for both 60 and 120 nodes, more power consumption in transmit and receive mode is observed in Bellman-Ford, OLSR, and LAR1 routing protocols. Te STAR-ORA, ZRP, and STAR-LORA routing protocols display relatively less power consumption due to the hybrid routing features of these protocols in UWSN scenario.
Te fndings from measuring the proposed UWSN network's performance parameters in FTP, CBR, and VBR applications for the deployment of 60 and 120 nodes, respectively, are given. Te UWSN network's performance metrics for FTP, CBR, and VBR applications are as follows. Utilizing as little transmit energy as possible is the aim of UWSN. It is impossible for other routing protocols to match AODV's speed and dependability.  Figures 14 and 15 for the amount of received energy they use with FTP, CBR, and VBR applications for 60 and 120 nodes, respectively. Te AODV routing protocol uses 76.4 percent less receive energy in CBR application than the other routing protocols shown in Tables 1 and 2, including DSR, DYMO, LAR1, Bellman-Ford, OLSR, fsheye, STAR-ORA, ZRP, and STAR-LORA. When dealing with larger packet sizes than other routing protocols, the AODV routing protocol uses signifcantly less receive energy than those other routing protocols. When receiving data, UWSN should use the least amount of energy possible. Performance-wise, the AODV routing protocol outperforms other routing protocols.  Tables 1 and 2, the AODV routing protocol produces an average delay in the CBR application that is 88.6% less than that of other routing protocols.

Conclusion
Along with the exploration of an underwater environment, other aspects are also evolving, such as the monitoring of underwater resources, the investigation of parameters, and the planning of military action. Te extent of battery power is the primary focus of the network because the UWSN can only carry out certain tasks. Tis study compares the performance of the routing protocols AODV, DSR, DYMO, LAR1, Bellman-Ford, OLSR, fsheye, STAR-ORA, ZRP, and STAR-LORA in UWSN networks with variable deployment applications such as FTP, CBR, and VBR for 60 and 120 nodes, respectively. Te metrics such as average transmission delay, average jitter, utilisation rate, and energy used in transmit and receive modes were all tracked. Te simulation results show that when compared to the DSR, DYMO, LAR1, Bellman-Ford, OLSR, fsheye, STAR-ORA, ZRP, and STAR-LORA routing protocols, the AODV routing protocol generates the least overall energy with a slight variation of additional nodes as well as 88.6 percent less average transmission delay. In addition, compared to the AODV, DSR, DYMO, LAR1, Bellman-Ford, OLSR, fsheye, STAR-ORA, ZRP, and STAR-LORA routing protocols, the fsheye routing protocol achieves a 91.4 percent higher percentage of utilisation. Te average jitter produced by STAR-LORA is 85.3 percent lower than that of the other routing protocols for 60 and 120 nodes.

Data Availability
Te data used to support the fndings of this study are available from the corresponding author upon request (head.research@bluecrest.edu.lr).

Conflicts of Interest
Te authors declare that they have no conficts of interest to report regarding the present study.

Authors' Contributions
K. Sathish conceptualized the study, performed data curation and formal analysis, proposed methodology, provided software, and wrote the original draft. C. V. Ravikumar supervised, visualized, and investigated the study and performed formal analysis. Asadi Srinivasulu applied a plagiarism checker and document remover.
A. Rajesh visualized the study, proposed methodology, and edited and reviewed the manuscript. Olutayo Oyeyemi Oyerinde administered the project and acquired funding.