Intelligent Automation & Soft Computing DOI:10.32604/iasc.2023.027076
Article

Fuzzy with Metaheuristics Based Routing for Clustered Wireless Sensor Networks

Ashit Kumar Dutta1,*, Yasser Albagory2, Majed Alsanea3, Abdul Rahaman Wahab Sait4 and Hazim Saleh AlRawashdeh5

1Department of Computer Science and Information Systems, College of Applied Sciences, AlMaarefa University, Ad Diriyah, Riyadh, 13713, Kingdom of Saudi Arabia
2Department of Computer Engineering, College of Computers and Information Technology, Taif University, Taif, 21944, Kingdom of Saudi Arabia
3Department of Computing, University Name: Arabeast Colleges, Riyadh, 11583, Kingdom of Saudi Arabia
4Department of Archives and Communication, King Faisal University, Al Ahsa, Hofuf, 31982, Kingdom of Saudi Arabia
5Department of Cyber Security, Onaizah Private Colleges, Unaizah, 51911, Kingdom of Saudi Arabia
*Corresponding Author: Ashit Kumar Dutta. Email: drashitkumar@yahoo.com
Received: 10 January 2022; Accepted: 16 February 2022

Abstract: Wireless sensor network (WSN) plays a vital part in real time tracking and data collection applications. WSN incorporates a set of numerous sensor nodes (SNs) commonly utilized to observe the target region. The SNs operate using an inbuilt battery and it is not easier to replace or charge it. Therefore, proper utilization of available energy in the SNs is essential to prolong the lifetime of the WSN. In this study, an effective Type-II Fuzzy Logic with Butterfly Optimization Based Route Selection (TFL-BOARS) has been developed for clustered WSN. The TFL-BOARS technique intends to optimally select the cluster heads (CHs) and routes in the clustered WSN. Besides, the TFL-BOARS technique incorporates Type-II Fuzzy Logic (T2FL) technique with distinct input parameters namely residual energy (RE), link quality (LKQ), trust level (TRL), inter-cluster distance (ICD) and node degree (NDE) to select CHs and construct clusters. Also, the butterfly optimization algorithm based route selection (BOARS) technique is derived to select optimal set of routes in the WSN. In addition, the BOARS technique has computed a fitness function using three parameters such as communication cost, distance and delay. In order to demonstrate the improved energy effectiveness and prolonged lifetime of the WSN, a wide-ranging simulation analysis was implemented and the experimental results reported the supremacy of the TFL-BOARS technique.

Keywords: Type II-fuzzy logic; uncertainty; WSN; energy; lifetime; route selection

1  Introduction

Wireless sensor networks (WSNs) comprised number of sensors using single-hop or multi-hop transmission [1]. It is a self-organized and dynamic network to transmit and gather the data in specific region, lastly deliver the data to base station (BS) [2]. In practical application, the sensors are often deployed in complex environments and energy-constrained, making it complex to preserve the constrained energy [3]. Thus, the major problem for WSN is to reduce the overall power utilizing and balance the network loads, as well as extend lifetime of the network. In order to resolve the issue, the researcher has presented several routing protocols for extending the lifetime of WSN. Now, clustering routing method is widely employed where all the clusters are created by many adjacent nodes, which comprises of various cluster members and cluster head (CH), further it is separated into clusters [4]. Clustering routing protocol has a different types of benefits namely less load, more scalability, more robustness and less energy consumption. The CH of lower-level system is the clustering members of higher-level system and the CH of topmost level network communicates with the BS network.

Various methods use fuzzy logic system (FLS) [5] for handling uncertainty in WSN. But the abovementioned protocol uses type-1 FLS that could not simply model and handle the uncertainty existing in WSNs since they use precise and crisp type-1 FLS (that is., membership function (MF)) that doesn’t enable uncertainty regarding membership value. Since an expansion of concepts of ordinary fuzzy sets (that is, type-1 FS), the model of type-2 FS was presented by Zadeh. A type-2 FS (T2FS) is considered as a fuzzy MF, that is., the membership values for all the elements of FS. The MF of T2FS are 3D and include footprint of uncertainty (FOU) that offer further degrees of freedom which allows us to handle and model directly the uncertainty in WSN [6]. Fig. 1 illustrates the structure of WSN.

Figure 1: WSN structure

In WSN, routing is difficult to process because of some natural features of WSN that distinguish it from transmission systems, for example, Adhoc network [7]. In the beginning, it isn’t possible to create a global addressing system for the placement of several nodes. Thus, traditional IP-related protocol could not be utilized in WSN. Next, different from the transmission systems, most of the WSN application requires flow of the sensed information from various nodes to a certain sink [8]. Then, node senses similar information within the neighborhood of phenomenon, the data traffics has considerable redundancy in it. This redundancy must be applied by the routing approach for improving energy efficacy bandwidth and. AT last, sensors in WSN are extremely controlled interms of processing capacity, transmission power, bandwidth and energy supply. Thus, they needed resource management and careful routing to minimize the overall power utilization [9,10].

This article develops an effective Type-II Fuzzy Logic with Butterfly Optimization Based Route Selection (TFL-BOARS) for clustered WSN. The TFL-BOARS technique applies T2FL technique with distinct input parameters namely residual energy (RE), link quality (LKQ), trust level (TRL), inter-cluster distance (ICD) and node degree (NDE) to select CHs and construct clusters. Moreover, the butterfly optimization algorithm based route selection (BOARS) technique is derived to select optimal set of routes in the WSN. Furthermore, the BOARS technique has computed a fitness function using three parameters such as communication cost, distance and delay. For examining the enhanced energy effectiveness and prolonged lifetime of the WSN, a wide-ranging simulation analysis is carried out.

2  Prior Works on Clustered WSN

Wang et al. [11] introduced an energy-effective clustering routing method. Considering the non-uniform traffic distribution, uneven cluster formation system is presented for energy efficiency and load balancing. Furthermore, a distributed CH rotation method is proposed for balancing power utilization within all the clusters. Since for longer distance communication to BS, we developed a dynamic multi-hop routing method amongst CH nodes-based distance-and-energy-aware cost function for avoiding the energy hole problems. Han et al. [12] an energy-effective clustering routing method has been proposed. To attain the aim of energy preservation, this method considered several clustering factors based on power utilization to choose CHs like number of neighbors through weighting, RE, distance from node to BS and neighbors. Lastly, transform the questions of effective Clustering into the optimization of two variables: weight coefficient W and neighbor transmission range R of cluster factor.

Zhao et al. [13] proposed network architecture and integrates the original power utilization method for constructing a novel methodology for determining the optimum amount of clusters for the overall power utilization reduction. According to the balanced power utilization, we optimized AGglomerative NESting (AGNES) approach includes: (1) the dual-CH (D-CH) divisions of energy balance approach, (2) the node dormancy method and (3) introduction of distance variance. Additionally, the CH priority function is created according to position of node and RE.

Wang et al. [14] introduce a hybrid multiple hop partition-based clustering (HMPBC) routing method that meets interests and needs, fits certain environment, extend lifetime and balance load of the networks. In HMPBC, the single-chain structure within CHs selection and clusters based on RE are functioned using self-organization, as well the region minimal spanning tree model amongst clusters is determined using BS model. Xu et al. [15] presented an energy-effective clustering routing protocol with higher node placement based intercluster routing method (EECRP-HQSND-ICRM) in WSN. Firstly, introduces a formula definition for data redundancy, integrity and validity from the coverage rates as well as presented a node placement method based on two-fold coverage. Next, to fulfill the uniform distribution of CHs, the monitoring fields are separated as the CH is carefully chosen in the separate cells and smaller areas centered on the BS.

3  Design of TFL-BOARS Technique

In this study, a novel TFL-BOARS technique has been developed for clustered WSN to optimally select the CHs and routes in the clustered WSN. Besides, the TFL-BOARS technique incorporates T2FL technique with distinct input parameters to select CHs and construct clusters. Also, the BOARS technique is derived to select optimal set of routes in the WSN.

3.1 Network and Energy Model

Consider N sensor which is located arbitrarily from network domain for monitoring the location and their physical feature periodically [16]. All sensors have neighboring sensors and it transfers information to most neighboring sensors. It can be considered an immobile sensor with equivalent primary energy. The calculation abilities of all sensors were similar. The symmetric radio connects are assumed that amongst some 2 neighboring sensors. The sink was placed inside the network area. Assume that the maximal broadcast of all sensors is R . The adaptive broadcast was regarded as utilizing distance amongst some 2 adjacent sensors. The 1st-order radio method for analyzing the energy utilization of the presented routing. Assume that 7mm is the size of packet from bits. The energy has required to transmit a m bits of packet across d unit distance amongst sender sensors and among their adjacent sensor is demonstrated as:

ETX(m,d)={m∗Eelect+m∗εfsp∗d2ifd<do,m∗Eelect+m∗εmpf∗d4ifd≥do. (1)

For receiving a m bits of packet, the energy necessity is provided as:

ERX(m)=m∗Eelect, (2)

where Eelect implies the statistics on the energy dissipate to transmits electron per bit. Many factors like acceptable bit-rate, digital coding and modulation affect the Eelect. The εfsp and εmpf represents the require of energy from the free-space as well as multipath environments, correspondingly. If the 2 adjacent sensors to that energy procedure were computed are divided with the distance lesser than or equivalent to (lo=Efsp/Emp) , the radio method executes Eq. (1) else Eq. (2) for calculating the energy required to broadcast the data.

3.2 Steps Involved in TFL Based Clustering Process

At the initial stage, the TFL-BOARS technique incorporates T2FL technique with distinct input parameters namely RE, LKQ, TRL, ICD and NDE to select CHs and construct clusters [17].

In order to choose the CH, a novel method was utilized named fuzzy logic (FL). This FL is 5 input variables such as RE, LKQ, TRL, ICD and NDE were utilized. Then, the resultant feature is contained with probability of developing as CH (PRCH). Primarily, RE represents the RE of node, LKQ demonstrates the quality of connection, TRL illustrates the intensity of stabilities, ICD defines the distance amongst the nodes and NDE indicates the amount of nodes located towards place. T2FL has 4 levels as demonstrated below:

Fuzzifier

It can be utilized for converting the actual input as fuzzified value. In some input features with respect to linguistic attributes were utilized to choose the CH as well as cluster size which is listed in Tab. 1. At this point, there are any linguistic parameters to RE that are low, medium and high. Likewise, the linguistic feature of DBS is near, far, farthest but ND is minimum, moderate, maximum, correspondingly.

Fuzzy rules/Inference engine

The framework of T1FL and T2FL are the same. At this point, the group of 27 rules was implemented. Afterward, a set of fuzzy rules to CH and decided cluster size. So, the rule was offered in Eq. (3).

Rule(i)IFx1isA1(i)ANDx2isA2(i)ANDx3isA3(i)

THENy1isB1(i)ANDy2isB2(i) (3)

where i is i th rule in a fuzzy rule, A1, A2 and A3 are matching fuzzy sets of x1, x2 and x3. The rule base suggestion engine has constrained with 27 rules and created based on Mamdani Inference method. The group of fuzzy rules is provided in Tab. 2. Evaluation of union as well as intersection is important. Fig. 2 demonstrates the process of Type-II Fuzzy Logic technique.

Figure 2: Process of Type-II fuzzy logic

T2FL has been demonstrated as supervised membership function (MF) and inferior MF. This expression is demonstrated in the application of T1FL MF. Then, the space in 2 functions represents the Footprint of Uncertainty (FOU) which explains the T2FL set. Supposing FOU was offered as f. When f є [0, 1] and f →0, so MF is named as T1FL. Once f →0 to 1, afterward T2FL contains distinct range in FOU amongst zero to one. But, the progress of rules from T2FL is same as T1FL which is written as:

Type2FL=PrincipalMF(Type1FL)+FOU (4)

If the PRCH was received, it transmits the message to target node. This message was enclosed with node ID and PRCH measures. The sensor node (SN) that is superior probabilities was chosen as CH and broadcasts CH_WON to equivalent SN less SN receives many CH_WON in neighboring SNs. In this regard, it sends a CH_JOIN message and merges with nearer CH. During the retrieve CH_JOIN message, neighboring CH verifies the provided cluster size from previous to get novel member.

If the SN receives a CM_REJECT message, afterward it re-transmission CM_JOIN message to future CH without assumptions of distant CH that is repeating still a novel CH was explored. Besides, the SN is not be integrated with another CH that endures inside a coverage area ‘R’, afterward, it chooses as CH. Therefore, all SNs go to cluster where the divided SN is from VANET. From the processing of different rounds, for eliminating the premature death of CH, rotation function was implemented by CH. If the RE of CH was superior if related to threshold value, afterward CH rotation exists. But RE of CH exceeds a threshold value, a novel CH is chosen utilizing PRCH. As function remove the initial death of CH and improve the network lifespan.

Type reducer/Defuzzifier

This technique generates a T1FL outcome that is restructured to mathematical outcome afterward the execution of defuzzifier was done.

3.3 Steps Involved in BOARS Technique

Once the clusters are constructed, the BOARS technique is derived to select optimal set of routes in the WSN using three parameters such as communication cost, distance and delay. The BOA imitates the natural performance of butterflies on mating and food source find [18]. This technique implements 2 distinct navigation designs for searching the area. During the exploration stage (r1≤p) , butterfly move near optimum butterflies of colony but during the exploitation stage (r1>p) , butterflies carry out an arbitrary search inside the search spaces by moving near arbitrary butterflies from the colony [19]. These 2 search designs are mathematically expressed as:

if r1≤p , Global search

t+1Xi=tXi+(r22×g∗−tXi)×φi (5)

if r1>p, Local search

t+1Xi=tXi+(r32×tXj−tXk)×φi (6)

where t and t+1 correspondingly represent the present and upgrade state of equivalent variable. Besides, the place of optimum butterflies from the colony was demonstrated by g∗ , and tXi and tXk signifies the places of 2 arbitrarily chosen butterflies; r1,r2 and r3 stands for the 3 arbitrary scalars uniformly selected in zero and one, φi implies the fragrance factor and it can be determined as under:

φi=cIa (7)

where, φi refers the fragrance magnitude to ith butterfly; I and a signifies the intensity of stimulating and fluctuating absorption degree and c has been coefficient. I has connected to main function value and to ith butterflies it can be obtained as f(Xi), where f returns main function value of problem. The coefficients of a and c are chosen in intervals of zero and one, ⋅ p stands for the probability switch that defines the search performance of this technique.

The BOARS method focuses on simultaneously reducing the delay, cost and distance of the BS. Where S indicates a set of ξ(i,j=1,2,...,m) and ξ represent a sojourn point, di,j shows the distance from ξi to ξj , Ci,j represent the cost of moving from ξi to ξj , as well as τi,j signifies the traveling delay from ξi to ξj . The decision parameter 0 is shown below. When the objective function to minimize the delay, cost and distance are shown in the following.

Γi,j={1ifξj isvisitedfromξi0otherwise (8)

C:Min∑im⁡∑jm⁡Ci, jΓi, j (9)

D:Min∑im⁡∑jm⁡di, jΓi, j (10)

T:Min∑im⁡∑jm⁡τi, jΓi, j (11)

when the optimization constraint is ∑im⁡Γi,j=1 for every i and j , the estimated path mustn’t be selected repeatedly (Γi,j+Γj,i≤1) and Γi,j≥1 .

4  Experimental Validation

The experimental result analysis of the TFL-BOARS technique takes place with recent methods [12] under varying measures. A comparison study is made with Energy efficient cluster formation (EECF), Energy Aware Clustering Hierarchy Protocol (EACHP) and Well Pattern Optimization with Energy Efficient Chain-based Routing Protocol (WPO-EECRP). Tab. 3 and Fig. 3 demonstrates the number of alive node (NOAN) analysis of the TFL-BOARS system with other models under the node count of 100. The results show that the TFL-BOARS algorithm can attain higher NOAN over the other methods. For instance, with 100 rounds, the TFL-BOARS system has offered an increased NOAN of 100 nodes but the EECF, EACHP and WPO-EECRP algorithms have obtained lower NOAN of 89, 94 and 97 nodes respectively. Likewise, with 300 rounds, the TFL-BOARS algorithm has gained a superior NOAN of 87 nodes while the EECF, EACHP and WPO-EECRP algorithms have resulted in inferior NOAN of 22, 39 and 73 nodes respectively. Moreover, with 600 rounds, the TFL-BOARS algorithm has exhibited improved NOAN of 11 nodes whereas the EECF, EACHP and WPO-EECRP approaches have accomplished reduced NOAN of 0, 0 and 1 node respectively.

Figure 3: Comparison NOAN study of TFL-BOARS technique under 100 nodes

A comparative number of dead node (NODN) analyses of the TFL-BOARS algorithm with existing algorithms under 100 nodes are offered in Tab. 4 and Fig. 4. The results exhibited that the TFL-BOARS system has exhibited maximum lifetime with the minimal NODN. For instance, with 100 rounds, the TFL-BOARS algorithm has provided a least NODN of 0 nodes whereas the EECF, EACHP and WPO-EECRP approaches have offered higher NODN of 11, 6 and 3 nodes. Concurrently, with 300 rounds, the TFL-BOARS technique has provided a least NODN of 13 nodes whereas the EECF, EACHP and WPO-EECRP techniques have offered higher NODN of 78, 61 and 27 nodes. Simultaneously, with 600 rounds, the TFL-BOARS technique has accomplished minimal NODN of 0 nodes whereas the EECF, EACHP and WPO-EECRP techniques have resulted in maximum NODN of 100, 100 and 99 nodes.

Figure 4: Comparison NODN study of TFL-BOARS technique under 100 nodes

Tab. 5 and Fig. 5 illustrate the NOAN analysis of the TFL-BOARS approach with other methods under the node count of 200. The outcomes demonstrated that the TFL-BOARS approach has the offered of achieving superior NOAN over the other methods. For instance, with 100 rounds, the TFL-BOARS algorithm has offered an increased NOAN of 198 nodes whereas the EECF, EACHP and WPO-EECRP systems have obtained lower NOAN of 169, 175 and 194 nodes correspondingly. At the same time, with 300 rounds, the TFL-BOARS algorithm has reached a superior NOAN of 174 nodes whereas the EECF, EACHP and WPO-EECRP techniques have resulted in inferior NOAN of 88, 114 and 157 nodes correspondingly. Besides, with 600 rounds, the TFL-BOARS approach has demonstrated higher NOAN of 87 nodes whereas the EECF, EACHP and WPO-EECRP methods have accomplished reduced NOAN of 0, 4 and 47 nodes correspondingly.

Figure 5: Comparison NOAN study of TFL-BOARS technique under 200 nodes

A comparative NODN analysis of the TFL-BOARS method with existing systems under 200 nodes is offered in Tab. 6 and Fig. 6. The results exhibited that the TFL-BOARS technique has exhibited increased lifetime with the minimal NODN. For instance, with 100 rounds, the TFL-BOARS approach has provided a worse NODN of 2 nodes whereas the EECF, EACHPs and WPO-EECRP techniques have obtainable higher NODN of 31, 25 and 6 nodes. Concurrently, with 300 rounds, the TFL-BOARS algorithm has provided a least NODN of 26 nodes whereas the EECF, EACHP and WPO-EECRP techniques have offered higher NODN of 112, 86 and 43 nodes. Concurrently, with 600 rounds, the TFL-BOARS system has accomplished lesser NODN of 113 nodes whereas the EECF, EACHP and WPO-EECRP methodologies have resulted in increased NODN of 200, 196 and 153 nodes.

Figure 6: Comparison NODN study of TFL-BOARS technique under 200 nodes

A detailed RE analysis of the TFL-BOARS technique with other methods under 100 nodes is provided in Fig. 7. The experimental values reported the supremacy of the TFL-BOARS technique with the higher RE. For instance, with 100 rounds, the TFL-BOARS approach has obtainable an increased RE of 43J whereas the EECF, EACHP and WPO-EECRP methodologies have reached lesser RE of 38J, 38J and 41J correspondingly. Also, with 300 rounds, the TFL-BOARS method has achieved a superior RE of 29J whereas the EECF, EACHP and WPO-EECRP methodologies have resulted in inferior RE of 13J, 15J and 24J correspondingly. In addition, with 500 rounds, the TFL-BOARS methodology has exhibited improved RE of 14J whereas the EECF, EACHP and WPO-EECRP techniques have accomplished reduced RE of 0, 0 and 10J correspondingly.

Figure 7: Comparison RE study of TFL-BOARS technique under 100 nodes

Finally, a detailed RE analysis of the TFL-BOARS technique with other algorithms under 200 nodes is offered in Fig. 8. The experimental values reported the supremacy of the TFL-BOARS technique with the higher RE. For instance, with 100 rounds, the TFL-BOARS technique has offered a higher RE of 184J whereas the EECF, EACHP and WPO-EECRP approaches have obtained reduced RE of 159J, 167J and 185J respectively. Followed by, with 300 rounds, the TFL-BOARS system has reached a superior RE of 132J whereas the EECF, EACHP and WPO-EECRP techniques have resulted in inferior RE of 63J, 99J and 114J correspondingly. Eventually, with 500 rounds, the TFL-BOARS system has exhibited improved RE of 72J whereas the EECF, EACHP and WPO-EECRP methods have accomplished reduced RE of 22J, 27J and 58J correspondingly. From the above mentioned result analysis, it is ensured that the presented model has the ability to prolong the lifetime of the WSN.

Figure 8: Comparison RE study of TFL-BOARS technique under 200 nodes

5  Conclusion

In this study, a novel TFL-BOARS technique has been developed for clustered WSN to optimally select the CHs and routes in the clustered WSN. Besides, the TFL-BOARS technique incorporates T2FL technique with distinct input parameters namely RE, LKQ, TRL, ICD and NDE to select CHs and construct clusters. Followed by, the BOARS technique is derived to select optimal set of routes in the WSN. In addition, the BOARS technique has computed a fitness function using three parameters such as communication cost, distance and delay. For examining the enhanced energy effectiveness and prolonged lifetime of the WSN, a wide-ranging simulation analysis is carried out. The comparative experimental results reported the supremacy of the TFL-BOARS technique over the recent approaches. In the future, the lifetime of WSN can be further prolonged by the design of data compression approaches.

Acknowledgement: The authors deeply acknowledge the Researchers supporting program (TUMA-Project-2021-27) Almaarefa University, Riyadh, Saudi Arabia for supporting steps of this work.

Funding Statement: This research was supported by the Researchers Supporting Program (TUMA-Project-2021-27) Almaarefa University, Riyadh, Saudi Arabia. Taif University Researchers Supporting Project number (TURSP-2020/161), Taif University, Taif, Saudi Arabia.

Conflicts of Interest: The authors declare that they have no conflicts of interest to report regarding the present study.

References

 1.  S. Arjunan and P. Sujatha, “Lifetime maximization of wireless sensor network using fuzzy based unequal clustering and ACO based routing hybrid protocol,” Applied Intelligence, vol. 48, no. 8, pp. 2229–2246, 2018. [Google Scholar]

 2.  J. Uthayakumar, T. Vengattaraman and S. Arjunan, “An efficient near lossless image compression algorithm using dissemination of spatial correlation for remote sensing color images,” Wireless Personal Communications, vol. 122, pp. 2963–2994, 2021. [Google Scholar]

 3.  G. K. Nigam and C. Dabas, “ESO-LEACH: PSO based energy efficient clustering in LEACH,” Journal of King Saud University-Computer and Information Sciences, vol. 33, no. 8, pp. 947–954, 2021. [Google Scholar]

 4.  S. Arjunan and S. Pothula, “A survey on unequal clustering protocols in Wireless Sensor Networks,” Journal of King Saud University-Computer and Information Sciences, vol. 31, no. 3, pp. 304–317, 2019. [Google Scholar]

 5.  S. Arjunan, S. Pothula and D. Ponnurangam, “F5N-based unequal clustering protocol (F5NUCP) for wireless sensor networks,” International Journal of Communication Systems, vol. 31, no. 17, pp. e3811, 2018. [Google Scholar]

 6.  R. V. Kulkarni, A. Forster and G. K. Venayagamoorthy, “Computational intelligence in wireless sensor networks: A survey,” IEEE Communications Surveys and Tutorials, vol. 13, no. 1, pp. 68–96, 2011. [Google Scholar]

 7.  J. Kim, S. Park, Y. Han and T. Chung, “CHEF: Cluster head election mechanism using fuzzy logic in wireless sensor networks,” in 2008 10th Int. Conf. on Advanced Communication Technology, Gangwon, Korea (Southpp. 654–659, 2008. [Google Scholar]

 8.  I. Gupta, D. Riordan and S. Sampalli, “Cluster-head election using fuzzy logic for wireless sensor networks,” in 3rd Annual Communication Networks and Services Research Conf. (CNSR’05), Halifax, NS, Canada, pp. 255–260, 2005. [Google Scholar]

 9.  S. S. Kumar, M. N. Kumar and V. S. Sheeba, “Fuzzy logic based energy efficient hierarchical clustering in wireless sensor networks,” International Journal of Research and Reviews in Wireless Sensor Networks (IJRRWSN), vol. 1, no. 4, pp. 53–57, 2011. [Google Scholar]

10. A. KumarSingh, N. Purohit, S. Goutele and S. V. S. Verma, “An energy efficient approach for clustering in WSN using fuzzy logic,” International Journal of Computer Applications, vol. 44, no. 18, pp. 8–12, 2012. [Google Scholar]

11. Z. Wang, X. Qin and B. Liu, “An energy-efficient clustering routing algorithm for WSN-assisted IoT,” in 2018 IEEE Wireless Communications and Networking Conf. (WCNC), Barcelona, pp. 1–6, 2018. [Google Scholar]

12. G. Han and L. Zhang, “WPO-EECRP: Energy-efficient clustering routing protocol based on weighting and parameter optimization in WSN,” Wireless Personal Communications, vol. 98, no. 1, pp. 1171–1205, 2018. [Google Scholar]

13. Z. Zhao, K. Xu, G. Hui and L. Hu, “An energy-efficient clustering routing protocol for wireless sensor networks based on agnes with balanced energy consumption optimization,” Sensors, vol. 18, no. 11, pp. 3938, 2018. [Google Scholar]

14. C. Wang, Y. Zhang, X. Wang and Z. Zhang, “Hybrid multihop partition-based clustering routing protocol for WSNs,” IEEE Sensors Letters, vol. 2, no. 1, pp. 1–4, 2018. [Google Scholar]

15. K. Xu, Z. Zhao, Y. Luo, G. Hui and L. Hu, “An energy-efficient clustering routing protocol based on a high-QoS node deployment with an inter-cluster routing mechanism in WSNs,” Sensors, vol. 19, no. 12, pp. 2752, 2019. [Google Scholar]

16. R. Varun, R. Gangwar, O. Kaiwartya and G. Aggarwal, “Energy-efficient routing using fuzzy neural network in wireless sensor networks,” Wireless Communications and Mobile Computing, vol. 2021, no. 2, pp. 1–13, 2021. [Google Scholar]

17. W. X. Xie, Q. Y. Zhang, Z. M. Sun and F. Zhang, “A clustering routing protocol for wsn based on type-2 fuzzy logic and ant colony optimization,” Wireless Personal Communications, vol. 84, no. 2, pp. 1165–1196, 2015. [Google Scholar]

18. S. Arora and S. Singh, “Butterfly optimization algorithm: A novel approach for global optimization,” Soft Computing, vol. 23, no. 3, pp. 715–734, 2019. [Google Scholar]

19. A. Mortazavi and M. Moloodpoor, “Enhanced butterfly optimization algorithm with a new fuzzy regulator strategy and virtual butterfly concept,” Knowledge-Based Systems, vol. 228, pp. 107291, 2021. [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.