Parrot dandelion optimizer-based routing and spiking neural network enabled intrusion detection in underwater IoT
4 affiliations
- 1Associate Professor, Department of Artificial Intelligence and Machine Learning , Rajalakshmi Engineering College
- 2Associate Professor, Department of Computer Science and Engineering, Saveetha Engineering College
- 3Associate Professor, Department of Computer Science and Engineering, Easwari Engineering College
- 4Assistant Professor, Department of CSE, SRM Institute of Science and Technology
Abstract
Underwater Internet of Things (IoT) is essential to monitor and gather real-time data in the marine environment. Due to the complicated and dynamic nature of this environment, underwater IoT networks face various security threats, such as eavesdropping, resource exhaustion, flooding, and blocking attacks. To overcome these issues, several intrusion detection systems (IDS) have been developed. However, the conventional IDS for underwater IoT are not effective at identifying threats owing to many false positives that take more time for computation. In this research, Parrot Dandelion Optimizer based Deep Spiking Neural Network (PDO_Deep SNN) is presented to detect intrusion in underwater IoT. Initially, an underwater IoT system model is simulated, and next, routing is done. Routing is done employing Parrot Dandelion Optimizer (PDO) by considering fitness parameters, like energy, trustworthiness, delay, selfishness, distance, and honesty. However, PDO is devised by incorporating Parrot Optimizer (PO) with Dandelion Optimizer (DO). Next, intrusion detection is conducted at the underwater thing layer based on the following steps. Firstly, the input log file is acquired from the BoT-IoT dataset. Afterwards, data is normalized using tanh-estimator normalization and then, feature selection is performed by Mutual Information. Lastly, intrusion is detected utilizing Deep Spiking Neural Network (Deep SNN), and it is tuned employing a PDO that is designed by combining PO with DO. Additionally, PDO attained better experimental results with an end-to-end delay, distance, residual energy, bandwidth, failure rate, and energy overhead of 0.672 ms, 5.930 m, 31.575 J, 7.251 Hz, 0.489, and 0.495. Furthermore, PDO_Deep SNN acquired a True Positive Rate (TPR) of 91.904% and a True Negative Rate (TNR) of 91.980%.
Related
- Asymmetric class-conditional routing in a federated quantum mixture-of-experts improves imbalanced intrusion detection in edge industrial IoT networks
- An EfficientNet hybrid deep learning detection framework of network intrusion with an optimized mixture-of-experts architecture
- Attention-based intrusion detection using transformer networks on UNSW-NB15
- An energy-aware and lightweight spiking neural network framework for real-time intrusion detection in IoT networks
- Causality-driven hybrid intrusion detection with entropy-aware feature learning and federated intelligence
BibTeX
@article{ayyadurai2026parrot,
title = {{Parrot dandelion optimizer-based routing and spiking neural network enabled intrusion detection in underwater IoT}},
author = {Ayyadurai M and Nithya Savarimuthu and Vishnupriya Gurunathan and Sujatha Kesavan},
journal = {Scientific Reports},
year = {2026},
month = aug,
doi = {10.1038/s41598-026-64120-8},
url = {https://doi.org/10.1038/s41598-026-64120-8},
note = {CC BY},
}
RIS
TY - JOUR
TI - Parrot dandelion optimizer-based routing and spiking neural network enabled intrusion detection in underwater IoT
AU - Ayyadurai M
AU - Nithya Savarimuthu
AU - Vishnupriya Gurunathan
AU - Sujatha Kesavan
PY - 2026
DA - 2026/08/01
JO - Scientific Reports
DO - 10.1038/s41598-026-64120-8
UR - https://doi.org/10.1038/s41598-026-64120-8
AB - Underwater Internet of Things (IoT) is essential to monitor and gather real-time data in the marine environment. Due to the complicated and dynamic nature of this environment, underwater IoT networks face various security threats, such as eavesdropping, resource exhaustion, flooding, and blocking attacks. To overcome these issues, several intrusion detection systems (IDS) have been developed. However, the conventional IDS for underwater IoT are not effective at identifying threats owing to many false positives that take more time for computation. In this research, Parrot Dandelion Optimizer based Deep Spiking Neural Network (PDO_Deep SNN) is presented to detect intrusion in underwater IoT. Initially, an underwater IoT system model is simulated, and next, routing is done. Routing is done employing Parrot Dandelion Optimizer (PDO) by considering fitness parameters, like energy, trustworthiness, delay, selfishness, distance, and honesty. However, PDO is devised by incorporating Parrot Optimizer (PO) with Dandelion Optimizer (DO). Next, intrusion detection is conducted at the underwater thing layer based on the following steps. Firstly, the input log file is acquired from the BoT-IoT dataset. Afterwards, data is normalized using tanh-estimator normalization and then, feature selection is performed by Mutual Information. Lastly, intrusion is detected utilizing Deep Spiking Neural Network (Deep SNN), and it is tuned employing a PDO that is designed by combining PO with DO. Additionally, PDO attained better experimental results with an end-to-end delay, distance, residual energy, bandwidth, failure rate, and energy overhead of 0.672 ms, 5.930 m, 31.575 J, 7.251 Hz, 0.489, and 0.495. Furthermore, PDO_Deep SNN acquired a True Positive Rate (TPR) of 91.904% and a True Negative Rate (TNR) of 91.980%.
SN - 2045-2322
ER -