DOI:10.32604/csse.2022.022662
| Computer Systems Science & Engineering DOI:10.32604/csse.2022.022662 | |
| Article |
Provisioning Intelligent Water Wave Optimization Approach for Underwater Acoustic Wireless Sensor Networks
1Department of Computer Science and Engineering, Vidhyaa Vikas College of Engineering and Technology, Tiruchengode, Namakkal, Tamilnadu, India
2Department of Computer Science and Engineering, KSR College of Engineering (Autonomous), Tiruchengode, Namakkal, Tamilnadu, India
*Corresponding Author: M. Manikandan. Email: murugan. manikandan@gmail.com
Received: 14 August 2021; Accepted: 18 October 2021
Abstract: In the Acoustics channel, it is incredibly challenging to offer data transfer for time-sourced applications in an energy-efficient manner due to higher error rate and propagation delay. Subsequently, conventional re-transmission over any failure generally initiates significantly larger end-to-end delay, and therefore it is not probable for time-based services. Moreover, standard techniques without any re-transmission consume enormous energy. This investigation proposes a novel multi-hop energy-aware transmission-based intelligent water wave optimization strategy. It ensures reduced end-to-end while attaining potential amongst overall energy efficiency end-to-end packet delay. It merges a naturally inspired meta-heuristic approach with multi-hop routing for data packets to reach the destination. The appropriate design of this Meta heuristic-based energy-aware scheme consumes lesser energy than the conventional one-hop transmission strategy without re-transmission. However, there is no hop-by-hop re-transmission facilitated. The proposed model shows only lesser delay than conventional methods with re-transmission. This work facilitates extensive work to carry out the proposed model performance with the MATLAB simulation environment. The results illustrate that the model is exceptionally energy-efficient with lesser packet delays. With 500 nodes, the packet delivery ratio of proposed model is 100%, average delay is reduced by 2%, total energy consumption is 8 J, average packet redundancy is 1.856, and idle energy is 6.9Mwh. The proposed model outperforms existing approaches like OSF, AOR, and DMR respectively.
Keywords: Acoustic applications; energy efficiency; network communications; underwater sensor networks; meta-heuristic approach; intelligent water wave optimization
The Earth is occupied by water (ocean) which plays an essential role in sustaining human life. It is an essential role in global development components [1]. Underwater detection and marine surveillance are not simple approaches. Underwater acoustic sensor networks (UWASN) give a higher ability to resolve these issues. The underwater communication and networking technology is flexibly applied over various environments like diverse water depths, coverage distance, and network structure and so on [2]. It is extensively utilized in practical observations to comprehend information interaction among various device-based observations in diverse spatial locations [3]. Underwater acoustic networking and communication technology have recently been an active and hotspot research topic in the marine field [4].
The UWASN link-based reliability is influenced rigorously owing to the underwater acoustic channel’s complexities [5]. The link performance among neighbourhood nodes may reduce over a specific period that summons multi-hop re-transmissions those outcomes in energy dissipation [6]. Moreover, energy provisioning are generally complicated, and constraints to recharge. Therefore, during multi-hop UWASN, energy-efficiency is hugely significant to select an optimal routing to transmit data packets from source to destination [7]. It needs the routing protocol for both energy consumption and the network’s energy consumption. Therefore, balanced energy utilization helps to enhance network lifetime [8].
Usually, energy consumption is the primary metric during the modelling of UWSAN. The principal target of this research is to model multi-hop energy-aware transmission-based water wave optimization to deal with multi-objective constraint in UWSAN [9]. The analysis over routing protocol for multi-hop UWSAN possesses various research priorities. The author in [10] proposes an enhanced vector-based forward routing protocol that adopts the number of re-transmission, location information, and residual energy for periodic information during data forwarding compared to conventional vector-based routing protocol in multi-hop UWSAN. Vector-based forward routing enables the utilization of the energy node and examines the appropriate data transmission. The author in [11] discusses energy-efficient and reliable protocol to enhance the network life by distributing the residual nodes’ energy eventually and finds the most appropriate data transmission among the routing path. The author in [12] discusses the energy-efficient awareness-based routing protocol is initiated. It executes various greedy hopping and location-free packet forwarding approaches that significantly increase network capacity and diminish communication costs.
Subsequently, to eliminate more extended path loss and unpredictability over acoustic channels, cooperative transmission is applied to get an optimal solution. Zhou et al. [13] depicts the cooperative superiority of communication systems over P2P solutions and fulfils the need for multi-hop UWSAN eminently. The multi-hop networking environment can reliably enhance bandwidth utilization and enhance underwater acoustic communication systems and bit error rates. However, it expands the network coverage via the relay node cooperation. On the contrary, direct lost-distance/multi-hop transmission diminishes the entire system’s energy consumption [14]. The basic idea towards the optimization approaches like ACO is determined as an intellectual approach for the heuristic process with appropriate feasibility and distributed computing abilities. It is easier to merge with other models. Moreover, the drawback associated with these converges towards a locally optimal solution rather than the global optimum solution. The intelligent swarm algorithm can converge quickly to a globally optimal solution; however, it provides lower precision in determining a globally optimal solution. Jin et al. [15] ACO-FSA integration-based routing algorithm is anticipated for multi-hop UWSAN which merges both ACO and FSA’s benefits. The fusion process possesses virtues; practically, it can diminish prevailing routing protocols, energy consumption, transmission delay and enhance the routing protocols. Moreover, it is still unfeasible to attain reliable data transmission owing to the harsh marine environments (multi-hop UWASN), and more action is considered to enhance the feasibility of the algorithm. The significant contributions of this research are given below:
1) Based on the extensive analysis of the general approaches, this work proposes a novel approach known as the multi-hop energy-aware transmission-based intelligent water wave optimization (MEA-WWO) strategy. It ensures reduced E2E error rate while attaining potential amongst overall energy efficiency and E2E packet delay. MEA-WWO merges a naturally inspired meta-heuristic approach with multi-hop routing for data packets to reach the destination. The randomness of the model is adopted to predict the optimal global solution with a faster convergence speed. The energy consumption of the model is lower than the conventional approaches during multi-hop UWSAN. Specifically, the anticipated MEA-WWO model applies to both cooperative nodes. Owing to the beneficial MEA-WWO model, the anticipated system is reliable towards multi-cases in bandwidth efficiency and transmission delay.
2) The performance of the anticipated MEA-WWO routing algorithms with ACO, AFSA and enhanced ACO based on a tuneable model in diverse network scaling, specifically with network model (small and medium scale) with practical outcomes in multi-hop UWSAN. It is more resourceful for promoting the AI applications during practical execution as the routing model is considered with diverse networking scale w.r.t. energy conservation. The work is organized as follows: Section 2 is an extensive analysis of background studies. Section 3 is the MEA-WWO methodology with optimization for multi-hop UWSAN. Section 4 is numerical results and discussion with the conclusion in Section 5.
This section discusses various existing approaches with network models and various performance metrics that help improve network lifetime, energy efficiency, etc. With the extensive analysis of UWASN, Su et al. [16] propose a novel chain-based routing model. This model reduces the data packet transmission over the network and improves network efficiency. The consequences of packet size and number of data packets on energy consumption are evaluated. The UWASN-based grid technology is examined by Jin et al. [17]. The energy consumption of various energy nodes is evaluated, and network performance is examined. Zhuo et al. [18] have investigated a MAC protocol by analyzing the scalable, distribution, and energy-efficient protocol for more extended UWASN medium and propagation delay. The simulation carried out by the proposed model shows no longer propagation delay and general E2E delay. The transmission-based energy waste and the duty cycle, and the number of nodes over collision are examined. Yan et al. [19] discuss the lifetime maximization by packet size optimization in UWASN is examined. Similarly, packet size and transmission size is jointly considered to enhance the network-lifetime. Chen et al. [20] discuss the link-layer realistic model to reduce the network model’s energy consumption. The energy analysis for designing the routing protocols is cluster-based capacity based UWASN research. Here, the stochastic geometry-based network model is examined for examining the cluster-based capacity. The simulation outcomes determine whether the model is theoretically analyzed and verified to examine the error rate causes. The energy-based routing protocol and avoidable void model over UWASN are examined by Ahmad et al. [21]. Flooding, void, and cycle transmission has to be avoided in UWASN. The energy consumption, network lifetime, packet delivery ratio, and transmission delay are performance measures utilized to examine the functionality of UWASN. E2E delay, PDR, network density, and energy consumption are discussed efficiently. The energy hole avoidance in UWASN using load balance is examined in Alkindi et al. [22]. Load balancing characteristics, live nodes, residual energy, and dead nodes are determined for computational purposes. The energy harvesting and lifetime analysis over UWASN are carried out by Azam et al. [23].
Sherubha et al. [24] discuss the minimization of hole problem and energy consumption during data collection. The protocol is evaluated in a cooperative, optimal, and energy-efficient manner. This energy-efficient relay model is chosen to make the energy-efficiency model more efficient for UWASN. As well, E2E delay, energy consumption, PDR, dead, and throughput are examined. Uniform nodes distribution over the 2D scenario is deployed where interference and noise are considered as UWASN. The localization interference and energy minimization hole for UWASN routing is examined. The total amount of packets attained from the sink, number of dead nodes, packet drops and the total energy consumption is further discussed by Sasirekha et al. [25]. Sherubha et al. [26] anticipate a novel coverage control model adopted with control scheduling and probabilistic detection approach to enhance the network lifetime. Also, diverse approaches are known as Network Connectivity Restoration through inserting and adjusting the model and Network Connectivity Restoration through adjusting that deals with the issues over coverage problem in an underwater sensor network [27]. The anticipated model provides better performance during coverage ratio and fulfils the connectivity with lesser time complexity.
Yildiz et al. [28] anticipate damage prediction approach for wind turbines actuator and WSNs. With a fuzzy logic model, the functionality over damage prediction state is examined appropriately over the wind turbines. For better performance analysis, the accidents are eliminated, and delays and maintenance cost with power generation is further reduced. Wang et al. [29] provide an efficient way for underwater acoustic-based clustering approaches, and comparative examination is done with various performance metrics. Some non-conventional communication approaches demonstrate the features of specific vital characteristics. Sandeep et al. [30] anticipate a novel optimization approach via the Integer Linear Programming model. The initiation with these Link-Level transmission-based power control methods with the most acceptable power level was selecting all ACK frames and data over each link. The transmission power and packet size are considered for enhancing the network lifetime. Based on diverse underwater models; the outcomes are optimized and examined for analyzing the time complexity. Tab. 1 depicts the comparison of diverse optimization approaches.
Here, a multi-hop energy-aware mechanism is performed with nodes like sensor nodes (SN), relay nodes (RN), and destination nodes (DN). The nodes are specified by ‘S’, Nh−1, and D, respectively. The relay among the nodes are provided by RNi = (i = 1, 2, 3, …, Nh−1), and the broadcasting nature is provided based on Ci where i = 1, 2, …, Nh. The hopping-based transmission is expressed in Nh UWSAN. It is explained below:
1) The nodes are half-duplex and cannot transmit/receive data simultaneously;
2) The transmitted data from one node is received by successive neighbourhood relay nodes, i.e., Ri and Ri−2; and the transmission from the source is heard only by neighbourhood relay nodes RNi−1 and Ri. With these assumptions, data transmission over every hop should not influence the successive hops.
3) When the distance (relative) among the neighbourhood relay nodes RN and the RN and cooperative nodes’ distance over certain coverage regions are provided with better data decoding. It is assumed that r1 = 3 m and r2 = 4 m is the boundary regions for decoding the data. The transmission distance from i → i + 1 is represented as d1. These nodes help in constraint transmission power and distance transmission. Thus, the relay node selection within these ranges r2 and cooperative node selection need to be located to complete the transmission task efficiently. Fig. 1 depicts the transmission mechanism for multi-hop UWASN.
Figure 1: Transmission mechanism for multi-hop UWASN
The next ith hops are performed among the nodes with nodes-level synchronization. The nodes receive a transmission from RNi, and the relationship between the input and output nodes are expressed as in Eq. (1):
Here, RNi+1 is a vector form of relay nodes with sub-carriers; ‘l’ is transmitted symbol vector,
Here,
The underwater communication process is based on acoustic waves (transmission), where energy consumption process is hugely different. It is initiated as follows: 1) P0 is the lowest level of power where the packets are decoded successfully towards the receiver; 2) U(d) is the power attenuation over the distance ‘d’. It is mathematically expressed as in Eq. (3):
Specifically, U(d) is a physical quantity that is associated with propagation and transmission frequency. It is given as in Eqs. (4) & (5):
Here, γ(f) is the absorption coefficient (dB/km). Based on the propagation condition, the ‘m’ value varies. If m → 1, then it is related to the shallow water channel with appropriate wave propagation. When the value is 2, then it is related to the deepwater channel. Based on this consideration, we take m = 1.5, and ‘f’ is frequency. The empirical formula is given as in Eq. (6):
Therefore, the optimal frequency is related to the path with the determination of ‘d’, where the distance measures from i → j. Then, in case of no further re-transmission in every hop, the energy consumed during transmission is expressed as in Eq. (7):
Here, ‘T’ is transmitting time without any generality loss. It is considered that P0 = 1 watt and the data length is 1024 bits for every time, and the data rate is 106 bps; transmitting time is T = 6.5 seconds at the transmitter.
3.2 Multi-Hop Energy-Aware Transmission Based on Intelligent Water Wave Optimization (MEA-WWO)
The proposed MAE is integrated with WWO to provide a global solution with a feedback mechanism, randomness, and supportive nodes. Here, multi-hop energy-aware transmission is adopted to find the sequence of routing and some nodes are considered supportive nodes where routing table (final) is of two forms. The energy consumption during the routing strategy needs to be reduced (see Fig. 2), where the single-hop functionality of the proposed model is given as follows. It is shown in Algorithm 1 to demonstrate the functionality of the supportive nodes and the neighbourhood nodes. Similarly, the multi-hop functionality is measured using intelligent water wave optimization. It is shown in Algorithm 2.
Figure 2: Flow diagram of MAE-WWO
3.3 Intelligent Water-Wave Optimization (IWWO)
Intelligent water-wave optimization is inspired by a shallow wave model for handling constant optimization problems [31–33]. The individual wave possesses a height of ‘h’ and wavelength ‘λ’. The searching space is associated with a certain region. The fitness solution is evaluated for handling the maximal optimization issue and is inversely proportional to seabed depth. The waves are expressed as
Here, ( − 1, 1) is random numbers of range [−1, 1] with uniform distribution, L(a) is the length of search space. In this scenario, when the x′ is superior to ‘x’, then ‘x’ is substituted to x′. Similarly, the height is set as hmax. Else, the values remain the same. It is mathematically expressed as in Eq. (9):
Here, ‘f’ is the fitness function, fmax and fmin is maximal and minimal values of wavelength reduction coefficient, and ε avoids division by zero. The current waveform is expressed mathematically as in Eq. (10):
Here, N is a Gaussian random function with appropriate mean and SD. After refraction, x → x′ and height are reset to hmax, where the λ is expressed as
where β is the breaking coefficient, when there is no change with the wave, it remains the same. Else,
4 Numerical Results and Discussion
Here, the simulation is carried out in a MATLAB environment. The proposed MEA-WWO routing model’s performance is evaluated and compared with existing approaches to provide a better trade-off. Due to the lower bandwidth ability, the collision probability is increased at the receiver side. Therefore, collision detection and avoidance are not so efficient. Thus, some investigators like adopt the aloha method for long propagation delay computation. The simulation environment is deployed with 10–500 nodes over 10 * 10 * 10 kilometres with ten sink nodes uniformly distributed over the ocean surface at equal distance. Based on the consideration, the sink remains stationary at SNs and water’s surface, which moves towards the X-Y plane direction (2D random walk). The nodes movement is 1 to 3 m/s towards the horizontal direction, where the vertical direction is negligible. The performance of the proposed MAE-WWO is evaluated against the prevailing routing approaches like optimal selective forwarder (OSF), adaptive opportunistic routing (AOR) and distance measure routing (DMR) protocol. The simulation is conducted to analyze the metrics like E2E delay, PDR, energy-efficiency, and redundant transmission. The simulation is performed multiple times, and 20 runs are plotted. Node density varies from 100–500 nodes over the deployed region. The nodes are deployed randomly over the network. Disk model is extensively used over the communication range, and the link over the transmitting packets gives PDR 100%, else PDR is 0%. However, the model gets closer proximity to energy consumption. Fig. 3 depicts the topological view of node placement. Fig. 4 shows the sink and source/destination node generation. Fig. 5 depicts the energy consumption during a simple routing mechanism. Fig. 6 depicts the energy consumption during the multi-hop routing mechanism.
Figure 3: Topological view
Figure 4: Node creation
Figure 5: Node connectivity (lesser energy consumption)
Figure 6: Node connectivity (higher energy consumption)
PDR is depicted as the reception of packets successfully over the final destination. It is evaluated as the total amount of packets received to the total transmitted packets. It is mathematically expressed as in Eq. (12):
The average number of PDR at diverse node density is demonstrated. The outcomes depict the constant increase towards the PDR as the increase in nodes’ density and the sustainability in MEA-WWO is due to the network partitioning. The division of forwarding area redundant packets are successfully suppressed when the PDR is successful with the reception of data packets distinctively. With proactive approach to attain the neighbourhood information, the optimization model’s local maxima should be avoided. As illustrated in Fig. 4, the PDR is doubled when the number of nodes = 150. It is observed that the ratio of PD is gradually increased with number of nodes. In the contrary, MEA-WWO selects the neighbourhood node based on multi-hop energy-aware information from source node. However, it adopts the control parameter to select suitable forwarder node. Based on this parameter, it is observed that neighbourhood nodes’ choice over the cyclic manner leads to quicker energy dissipation, which makes the network nodes prone (void region). Also, inefficient data forwarding maximizes duplicate packets with the least packets over the destination region (Tab. 2). Hence, it is proven that PDR is higher when compared to other models like OSF, AOR and DMR. The transmissions of packets are reliable. Due to the existing model’s lesser distinctive nature, the PDR of those models is lesser compared to other approaches.
E2E is depicted as the packets’ total time to reach destination, where the average time after data generation till it reaches the destination. It is evaluated in seconds. The E2E delay of the proposed MEA-WWO is reduced due to avoiding the void region (See Tab. 3). The proposed MEA-WWO exploits AOR to choose backup nodes responsible during the fall of the back procedure. Hence, the computational time is less when compared to another model. The average E2E delay of existing OSF, AOR and DMR is higher due to the added computational time required during the forwarding region. It partitions the communication range into various regions, and these regions are managed to reduce the duplication of nodes transmission. When the nodes are deployed over the forwarded packets after node priority failure, it incorporates unnecessary delay. It is the cause behind the higher E2E delay.
On the contrary, MEA-WWO model shows lesser E2E delay when compared to the other models. The forwarding regions are chosen selectively, and the unnecessary nodes are terminated from the deployment region and terminate the same transmission process over the receiver side. With the avoidance of these nodes, surplus energy is being saved when transmitting the packets via the intermediate nodes. It is proven that E2E delay is more sustainable when the node density increases. The concept behind these E2E delay computation increases the transmission range with avoidance of collision with sparse environment. Number of nodes over the communication region requires enormous time to transmit the corresponding destination. The forwarding nodes are chosen based on the primary and the secondary forwarding regions with lesser holding time. Therefore, it is concluded that the results attained shows higher advantages corresponding to the communication range. Thereby, it diminishes the holding time of data packets. Thus, delay among the intermediate nodes is reduced with the proposed MEA-WWO model.
4.3 Energy Consumption Comparison
Energy consumption is depicted as total energy dissipated during the data packets transmitted from source to destination. It is composed of transmitting, receiving, and idle nodes that participate in the energy communication process. It is mathematically expressed as in Eq. (13):
Based on Eq. (13), E (t) is transmitter energy, E(r) is receiver energy, E(idle) is the energy consumed when the node is in the idle region. Fig. 5 depicts the energy consumption of the proposed MEA-WWO with existing OSF, AOR and DMR. The graph shows that the battery utilization is slightly increased with the nodes deployment (See Tab. 4). The reductions of redundant packets are suppressed with the partitioning of the transmission region. Here, nodes may directly discard the packets to fulfil the packet transmission successfully. The avoidance of packet energy consumption reduces the computational time is compared to existing approaches. The prediction of nodes holding time over the intermediate regions is lesser than the prevailing approaches. The claim of the proposed MEA-WWO model is validated with the simulation environment. It efficiently increases the network lifetime.
The energy consumption of the general approaches is notably higher when compared to the proposed model. The cause behind the higher energy consumption over the existing model determined as the fall back in every probable direction to predict the necessary route to reach destination. Therefore, inefficiency over the existing models is because of the higher redundant data. Also, in OSF, the holding time is shorter due to the successive hops towards the forwarding region. The node with higher priority fails during the packet propagation towards the destination. The fewer reception packets are compared with the transmission packets as the nodes initiate the transmission over the channel. It causes a rise in data loss and collision. The inefficiency encountered over the prevailing models leads to higher energy consumption when compared to MEA-WWO, respectively.
The average packet redundancy is depicted as the total packets received at the destination with the avoidance of redundant packets to total packets generated from the network. It is computed by plotting the graph and evaluating the mathematical expression as in Eq. (14):
Here, TPR is total packets received, RP is a redundant packet, and TP is total packets. From Fig. 6, the sustainable nature of the model over the energy efficiency is comparatively higher, where the proposed MEA-WWO model shows efficiency towards the packet redundancy (See Tab. 5). The packet forwarding region is partitioned to suppress packet transmission over hidden nodes of the source region. Tab. 6 depicts the idle energy consumed during packet size transmission (512 bytes, 1024 bytes, 2048 bytes, and 4096 bytes).
With the exclusion of forwarding nodes, the packets are received. The packets are broadcasted due to their mobility over the acoustic environment. Thus, the model predicts the holding time, neighbourhood nodes, and generated packets which commences from communication range after packet broadcast. Packet redundancy increases with the rise of total nodes. On the contrary, the existing OSF, AOR, and DMR models do not reduce the redundant data packets, specifically over the hidden terminals. It is not evaluated due to the lesser amount of distinctive packets. The holding time of OSF, AOR, and DMR is significantly smaller that quickly expires, which is similar to that of the re-transmission process. Fig. 7 depicts the convergence rate of the anticipated model. The cause behind this issue is the packet redundancy drawback encountered over the prevailing models.
Figure 7: Convergence rate of MAE-WWO
In this research, we proposed a novel MEA-WWO model for routing with a multi-hop energy-aware model. The proposed model is analyzed with various metrics like PDR, average E2E delay, average packet redundancy, and energy consumption efficiently. The proposed model’s performance is compared with three different approaches like OSF, AOR, and DMR. The multi-hop functionality is performed over 10–150 nodes, where the simulation outcomes show a slight difference in energy consumption compared with other approaches. The energy consumption of the proposed MAE-WWO is lesser than other models significantly. The performance of the proposed MAE-WWO is notably higher even when the nodes range increases.
Additionally, the proposed model reduces the computational complexity as the environment is set for simulation purpose. This complexity is still acceptable in the marine environment. The proposed model gives lesser E2Edelay, higher PDR, lesser energy consumption, and reduced packet redundancy, where it is practically applicable to the medium-scale real-time environment. The model proves the efficiency of the multi-hop energy-aware transmission efficiently with increased network lifetime. The major research constraints are the analysis with the optimization approach and simulation with sensor placement. In future, the model is further extended with practical implementation, which motivates to suit the acoustic network application with various scaling factors like (small scale, medium scale, and large-scale, respectively).
Funding Statement: The authors received no specific funding for this study.
Conflicts of Interest: The authors declare that they have no conflicts of interest to report regarding the present study.
1. R. Khan, S. H. Ahmed and D. Kim, “AUV-Assisted energy-efficient clustering in underwater wireless sensor networks,” in Proc. IEEE Global Communication Conf., Abu Dhabi, United Arab Emirates, pp. 1–7, 2018. [Google Scholar]
2. H. Zhang, S. L. Wang and H. X. Sun, “A low complexity clustering optimization algorithm for underwater sensor networks,” in Proc. IEEE Int. Conf. on Signal Processing, Communications and Computing, Hong Kong, pp. 1–6, 2016. [Google Scholar]
3. S. Dhongdi, A. Bhandari, J. Singh, S. Kachhadia and V. Joshi, “Joint clustering and routing protocol for 3-d underwater acoustic sensor net- work,” in Proc. Tenth Int. Conf. on Ubiquitous and Future Networks, Prague, Czech Republic, pp. 415–420, 2018. [Google Scholar]
4. Y. D. Chen, D. R. Wu, W. Chen and K. P. Shih, “A self-adaptive cooperative routing protocol for underwater acoustic sensor networks,” in OCEANS-MTS/IEEE, Washington, DC, USA, pp. 1–5, 2015. [Google Scholar]
5. J. Wang, S. Zhang, G. Gao, L. Cai, X. Zuo et al., “A clustering algorithm of underwater acoustic sensor networks based on hierarchical 3D mesh,” in Proc. Int. Conf. on Networking and Network Applications (NaNA), Xi’an, China, pp. 80–85, 2018. [Google Scholar]
6. R. Webster, K. Munasinghe and A. Jamalipour, “Murmuration inspired clustering protocol for underwater wireless sensor networks,” in Proc. IEEE Int. Conf. on Communications (ICC), Kansas City, MO, USA, pp. 1–6, 2018. [Google Scholar]
7. G. Yang, L. Dai and Y. Lei, “A secure and energy balanced clustering protocol for underwater wireless sensor networks,” in Proc. Int. Conf. on Security, Pattern Analysis, and Cybernetics (SPAC), Jinan, China, pp. 193–198, 2018. [Google Scholar]
8. G. Ahmed, X. Zhao, M. M. S. Fareed and M. Z. Fareed, “An energy-efficient redundant transmission control clustering approach for underwater acoustic networks,” Sensors, vol. 19, no. 19, pp. 1–19, 2019. [Google Scholar]
9. J. Wang, G. Gao, P. Qu, W. Chen, S. Zhang et al., “A software-designed clustering mechanism for underwater acoustic sensor networks,” IEEE Access, vol. 7, pp. 121742–121754, 2019. [Google Scholar]
10. S. M. Ghoreyshi, A. Shahrabi, T. Boutaleb and M. Khalily, “Mobile data gathering with hop-constrained clustering in underwater sensor networks,” IEEE Access, vol. 7, pp. 21118–21132, 2019. [Google Scholar]
11. G. Han, H. Wang, S. Li, J. Jiang, and W. Zhang, “Probabilistic neighborhood location-point covering set-based data collection algorithm with obstacle avoidance for three-dimensional underwater acoustic sensor networks,” IEEE Access, vol. 5, pp. 24785–24796, 2017. [Google Scholar]
12. R. Zhao, H. Long, O. A. Dobre, X. Shen, T. M. N. Ngatched et al., “Time reversal based MAC for multi-hop underwater acoustic networks,” IEEE Systems Journal, vol. 13, no. 3, pp. 2531–2542, 2019. [Google Scholar]
13. Z. Zhou, B. Yao, R. Xing, L. Shu and S. Bu, “E-CARP: An energy efficient routing protocol for UWSNs in the internet of underwater things,” IEEE Sensors Journal, vol. 16, no. 11, pp. 4072–4082, 2016. [Google Scholar]
14. M. A. Rahman, Y. Lee and I. Koo, “EECOR: An energy-efficient cooperative opportunistic routing protocol for underwater acoustic sensor networks,” IEEE Access, vol. 5, pp. 14119–14132, 2017. [Google Scholar]
15. Z. Jin, Y. Ma, Y. Su, S. Li and X. Fu, “A Q-learning-based delay-aware routing algorithm to extend the lifetime of underwater sensor networks,” Sensors, vol. 17, no. 7, pp. 1–15, 2017. [Google Scholar]
16. Y. Su, R. Fan, X. Fu and Z. Jin, “DQELR: An adaptive deep Q-network-based energy- and latency-aware routing protocol design for underwater acoustic sensor networks,” IEEE Access, vol. 7, pp. 9091–9104, 2019. [Google Scholar]
17. Z. Jin, Q. Zhao and Y. Su, “RCAR: A reinforcement-learning-based routing protocol for congestion-avoided underwater acoustic sensor networks,” IEEE Sensors Journal, vol. 19, no. 22, pp. 10881–10891, 2019. [Google Scholar]
18. X. Zhuo, M. Liu, Y. Wei, G. Yu, F. Qu et al., “AUV-Aided energy-efficient data collection in underwater acoustic sensor networks,” IEEE Internet Things Journal, vol. 7, no. 10, pp. 10010–10022, 2020. [Google Scholar]
19. J. Yan, Y. Gong, C. Chen, X. Luo and X. Guan, “AUV-Aided localization for internet of underwater things: A reinforcement learning-based method,” IEEE Internet Things Journal, vol. 7, no. 10, pp. 9728–9746, 2020. [Google Scholar]
20. Y. Chen, X. Jin, L. Wan, X. Zhang and X. Xu, “Selective dynamic coded cooperative communications for multi-hop underwater acoustic sensor networks,” IEEE Access, vol. 7, pp. 70552–70563, 2019. [Google Scholar]
21. I. Ahmad, U. Ashraf and A. Ghafoor, “A comparative QoS survey of mobile ad hoc network routing protocols,” Journal of the Chinese Institute of Engineers, vol. 39, no. 5, pp. 585–592, 2016. [Google Scholar]
22. Z. Alkindi, N. Alzeidi and B. A. A. Touzene, “Performance evolution of grid based routing protocol for underwater wireless sensor networks under different mobile models,” International Journal of Wireless & Mobile Networks (IJWMN), vol. 10, no. 1, pp. 13–25, 2018. [Google Scholar]
23. I. Azam, N. Javaid, A. Ahmad, W. Abdul, A. Almogren et al., “Balanced load distribution with energy hole avoidance in underwater WSNs,” IEEE Access, vol. 5, pp. 15206–15221, 2017. [Google Scholar]
24. P. Sherubha, S. P. Sasirekha, V. Manikandan, K. Gowsic and N. Mohanasundaram, “Graph based event measurement for analyzing distributed anomalies in sensor networks,” Sådhanå (2020), vol. 45, no. 1, pp. 1–5, 2020. [Google Scholar]
25. S. P., Sasirekha, A. Priya, T. Anitha and P. Sherubha, “Data processing and management in IoT and wireless sensor networks,” IOP Journal of Physics: Conference Series, vol. 1712, no. 1, pp. 1–8, 2020. [Google Scholar]
26. P. Sherubha and N. Mohanasundaram, “An efficient network threat detection and classification method using ANP-mVPS algorithm in wireless sensor networks,” International Journal of Innovative Technology and Exploring Engineering (IJITEE), vol. 8, no.11, pp. 1597–1606, 2019. [Google Scholar]
27. P. Sherubha and N. Mohanasundaram “An efficient intrusion detection and authentication mechanism for detecting clone attack in wireless sensor networks,” Journal of Advanced Research in Dynamical & Control Systems, vol. 11, no. 5, pp. 55–68, 2019. [Google Scholar]
28. H. U. Yildiz, V. C. Gungor and B. Tavli, “Packet size optimization for lifetime maximization in underwater acoustic sensor networks,” IEEE Transactions on Industrial Informatics, vol. 15, no. 2, pp. 719–729, 2019. [Google Scholar]
29. Z. Wang and B. Wang, “A novel node sinking algorithm for 3D coverage and connectivity in underwater sensor networks,” Ad Hoc Networks, vol. 56, pp. 43–55, 2017. [Google Scholar]
30. D. N. Sandeep and V. Kumar, “Review on clustering, coverage and connectivity in underwater wireless sensor networks: A communication techniques perspective,” IEEE Access, vol. 5, pp. 11176–11199, 2017. [Google Scholar]
31. Y. J. Zheng, “Water wave optimization: A new nature-inspired metaheuristic,” Computers and Operations Research, vol. 55, pp. 1–11, 2015. [Google Scholar]
32. C. Narmatha, “A new neural network-based intrusion detection system for detecting malicious nodes in WSNs,” Journal of Computational Science and Intelligent Technologies, vol. 1, no. 3, pp. 1–8, 2020. [Google Scholar]
33. S. L. Sarah and L. A. Mahmoud, “A novel intrusion detection system in WSN using hybrid neuro-fuzzy filter with ant colony algorithm,” Journal of Computational Science and Intelligent Technologies, vol. 1, no. 1, pp. 1–8, 2020. [Google Scholar]
 | This work is licensed under a Creative Commons Attribution 4.0 International License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. |