A hybrid energy-efficient routing protocol for WSN by integration of bio-inspired, fuzzy logic, and deep learning approaches
Khalid, Zain (2026)
2026
Sähkötekniikan DI-ohjelma - Master's Programme in Electrical Engineering
Informaatioteknologian ja viestinnän tiedekunta - Faculty of Information Technology and Communication Sciences
This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Hyväksymispäivämäärä
2026-04-10
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:tuni-202602242755
https://urn.fi/URN:NBN:fi:tuni-202602242755
Tiivistelmä
Wireless sensor networks (WSNs) consist of small electromagnetic devices deployed to take readings for decision-making and to provide promising solutions across diverse areas of practical life. The fundamental problem of WSNs is energy efficiency due to constrained energy resources of SNs, which impact the reliability and quality of service (QoS). This research aims to improve the efficacy by developing energy-efficient routing protocols, a widely adopted procedure through clustering, cluster head selection, route optimization, and data compression by employing artificial intellegent (AI) techniques.
The proposed model integrates advanced AI approaches that improve energy efficiency in WSNs by reducing redundant data communication and balancing energy consumption across all sensor nodes (SNs). The method applies a fuzzy-based system to generalize data and reduce transmissions from SNs to CHs, thereby saving significant energy. In this model, optimize CHs using the Peacock Tail Optimization (PTO) algorithm, which selects a CH based on the fitness of nodes within each cluster. At the CH, it performs compression using an Autoencoder algorithm, a deep learning technique that reduces the data sent to the BS, helping save energy.
The performance of the model is assessed in MATLAB simulations by using the fuzzy logic and deep learning toolboxes. Benchmark comparisons are drawn with LEACH-CR and GA-PCT routing protocols under scenarios with and without data compression. Energy consumption, number of dead nodes, throughput, and end-to-end delay performance metrics are examined for improvement in network lifetime. Uniform parameters with values are taken for simulation as used by the benchmark models to ensure a fair comparison and validate the results. Obtained results in both compressed and uncompressed cases depict the overall improvement of the proposed model compared to the baseline techniques. The validation of the proposed model regarding the longer lifetime and reduced energy consumption of WSNs has been done based on these evaluation metrics, and it proves the potential of the proposed model for real-world WSN applications.
The obtained results from the simulation prove the validity of the study, showing overall lifetime improvements with the proposed model providing a 75 percent improvement over LEACH-CR and 61 percent over GA-PCT without data compression. When the compression approach is applied, the performance also improves, showing 59.5 percent over LEACH-CR and 47 percent over GAPCT. It is more suitable for energy consumption than the baseline models, and the nodes’ lifetime is improved. The dead time of the first and last nodes is also shorter than that of the baseline models, and the fallout of the designed model is better than that of the baseline models. The delay also remains lower than that of the baseline models. Hence, based on the considered performance evaluation parameters, the proposed model justifies its validity and efficiency, thereby extending the network lifetime and ensuring the reliability and stability of WSNs.
The proposed model integrates advanced AI approaches that improve energy efficiency in WSNs by reducing redundant data communication and balancing energy consumption across all sensor nodes (SNs). The method applies a fuzzy-based system to generalize data and reduce transmissions from SNs to CHs, thereby saving significant energy. In this model, optimize CHs using the Peacock Tail Optimization (PTO) algorithm, which selects a CH based on the fitness of nodes within each cluster. At the CH, it performs compression using an Autoencoder algorithm, a deep learning technique that reduces the data sent to the BS, helping save energy.
The performance of the model is assessed in MATLAB simulations by using the fuzzy logic and deep learning toolboxes. Benchmark comparisons are drawn with LEACH-CR and GA-PCT routing protocols under scenarios with and without data compression. Energy consumption, number of dead nodes, throughput, and end-to-end delay performance metrics are examined for improvement in network lifetime. Uniform parameters with values are taken for simulation as used by the benchmark models to ensure a fair comparison and validate the results. Obtained results in both compressed and uncompressed cases depict the overall improvement of the proposed model compared to the baseline techniques. The validation of the proposed model regarding the longer lifetime and reduced energy consumption of WSNs has been done based on these evaluation metrics, and it proves the potential of the proposed model for real-world WSN applications.
The obtained results from the simulation prove the validity of the study, showing overall lifetime improvements with the proposed model providing a 75 percent improvement over LEACH-CR and 61 percent over GA-PCT without data compression. When the compression approach is applied, the performance also improves, showing 59.5 percent over LEACH-CR and 47 percent over GAPCT. It is more suitable for energy consumption than the baseline models, and the nodes’ lifetime is improved. The dead time of the first and last nodes is also shorter than that of the baseline models, and the fallout of the designed model is better than that of the baseline models. The delay also remains lower than that of the baseline models. Hence, based on the considered performance evaluation parameters, the proposed model justifies its validity and efficiency, thereby extending the network lifetime and ensuring the reliability and stability of WSNs.