An Energy Efficient Cluster Head Selection in UWSN through Optimization of Convolutional Neural Network with One-to-One Optimizer
Computer and Decision Making: An International Journal · 09.11.2026 · 3 (7S) · 128-135
DOI: 10.1922/comdem.v3i7s.19202 ↗Kurz gefasst
Fachartikel in der Zeitschrift Computer and Decision Making: An International Journal, online erschienen am 09.11.2026. Verfasst von Palkin Sharma und 1 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
Selecting cluster heads in underwater wireless sensor networks (UWSNs) is a sequential decision problem because acoustic transmission costs, residual energy and cluster membership change as data are collected. This study revisits a hybrid convolutional neural network (CNN) and One-to-One Based Optimizer (OOBO) framework using explicit decision labels, a reproducible learning-rate search and a frequency-dependent acoustic-link approximation. The original galley contained incompatible classification metrics and inconsistent simulation horizons; therefore, its numerical claims are not treated as validated observations. An independent illustrative experiment instead uses 1,000 rule-labeled synthetic candidates, a 600/200/200 training/validation/test split and a 150-node network evaluated over 3,000 rounds in five deployments. The tuned CNN attains 93.5% test accuracy and 89.92% F1-score, compared with 91.0% and 86.76% for the fixed-learning-rate CNN. In the network experiment, first-node death occurs at mean rounds 1,236.4 for OOBO-CNN, 1,233.0 for CNN, 1,255.8 for a HEED-style comparator and 991.8 for a LEACH-style comparator. Thus, classifier improvement does not translate into a clear network-lifetime advantage. The work clarifies the decision mechanism and its limits, while identifying the field calibration and full-protocol benchmarking needed before deployment claims are justified.
Zitieren
APA
Sharma, P., & Sharma, V. (2026). An Energy Efficient Cluster Head Selection in UWSN through Optimization of Convolutional Neural Network with One-to-One Optimizer. Computer and Decision Making: An International Journal, 3(7S), 128-135. https://doi.org/10.1922/comdem.v3i7s.19202
Vancouver
Sharma P, Sharma V. An Energy Efficient Cluster Head Selection in UWSN through Optimization of Convolutional Neural Network with One-to-One Optimizer. Computer and Decision Making: An International Journal. 2026;3(7S):128-135. doi:10.1922/comdem.v3i7s.19202
BibTeX
@article{Sharma2026,
author = {Sharma, Palkin and Sharma, Vikrant},
title = {An Energy Efficient Cluster Head Selection in UWSN through Optimization of Convolutional Neural Network with One-to-One Optimizer},
journal = {Computer and Decision Making: An International Journal},
year = {2026},
volume = {3},
number = {7S},
pages = {128--135},
doi = {10.1922/comdem.v3i7s.19202}
}
EUVIMED