The Department of Information Networks Discusses a PhD Dissertation on Enhancing the Accuracy of Node Localization in Marine Wireless Networks

Duhaa Fadill Abbas
The Department of Information Networks at the College of Information Technology held a discussion for the PhD dissertation titled “An Efficient Predicting Method for Marine Ad hoc Networks Nodes Localiztion ”, presented by PhD candidate Ghufran Abdulameer Hussein Ali, under the supervision of Dr. Yusra Hussein Ali. The discussion took place at 9:00 AM on Tuesday, February 25, 2025, in the conference hall of the College of Information Technology.

The dissertation highlights the Ship Ad-Hoc Networks (SANETs) have gained prominence due to the growing need for efficient and reliable wireless communication in maritime
environments. These networks are characterized by dynamic and unpredictable node movements influenced by environmental conditions such as ocean currents and
weather conditions. One of the key critical challenges in these networks is node
localization (i.e., during ship movement), where accurately determining the position
of nodes is essential for optimizing communication, ensuring network stability, and enhancing maritime operations.
To address this problem, the main objective of this study is to enhance the
accuracy of node localization in SANETs by proposing integrated fuzzy clustering
with a predictive model. The solution combines Dynamic Weighted Gradient-Based Possibilistic Fuzzy C-Means (DWGB-PFCM) clustering and a Particle Swarm Optimization (PSO)-enhanced Long Short-Term Memory (LSTM) neural network.
The DWGB-PFCM clustering algorithm effectively groups nodes based on spatial
and dynamic movement patterns, while the PSO-optimized LSTM model accurately predicts future node positions.
For this purpose, an adaptable, robust, and efficient framework that addresses
the challenges posed by node mobility and environmental variability has been
developed. This is achieved through parameter tuning, incorporating environmental data (such as wind speed, humidity, and wave height), and leveraging optimization techniques to improve both clustering and prediction performance.
The framework is applied and tested utilizing diverse datasets, including
Network Simulation 2, Automatic Identification System, and Geomedia data, representing both simulated maritime and real-time scenarios. The obtained results III demonstrate significant improvements in localization accuracy, achieving a high accuracy rate of (94%) and high R² scores with low RMSE, and MAE across all datasets. Additionally, the employed research methodology demonstrates adaptability to dynamic marine conditions, underscoring its potential for real-world maritime environments.
This dissertation offers a trustworthy solution for node localization in
SANETs, enabling the way for enhanced maritime navigation, fleet management,
and search-and-rescue operations. The significant results demonstrate the feasibility of implementing the proposed model in real-time scenarios, enabling reliable communication and streamlined maritime operations.