Advanced KNN-based cost-efficient algorithm for precision localization and energy optimization in dynamic underwater...

Area of Science:

• Robotics and Autonomous Systems
• Machine Learning
• Oceanography

Background:

• Underwater sensor networks face significant challenges in localization accuracy, energy efficiency, and operational costs due to dynamic environments.
• Existing methods often struggle with the complexities of underwater acoustics and node mobility, leading to performance degradation.

Purpose of the Study:

• To develop and evaluate a cost-efficient, K-Nearest Neighbors (KNN)-based machine learning algorithm for optimizing underwater context acquisition using sensor nodes.
• To address and minimize localization errors, energy consumption, and time costs in underwater sensor networks.
• To enhance the accuracy and efficiency of sensor node localization in dynamic underwater conditions.

Main Methods:

• A K-Nearest Neighbors (KNN)-based machine learning algorithm was proposed and implemented for underwater sensor node localization.
• The algorithm was designed to optimize context acquisition by predicting node orientation and mapping shortest distances.
• Effectiveness was validated through real-time experiments in a water tank and simulations using Ns-3.37 with the Aqua-sim model.

Main Results:

• Achieved a localization accuracy of 99.98%, significantly improving upon previous methods.
• Reduced the localization error rate from 4.59m to a minimal value (specific value noted as '[Formula: see text]m' in the abstract).
• Demonstrated a reduction in localization energy consumption to 0.0045J and introduced the localization time cost rate at 0.06762s.

Conclusions:

• The proposed KNN-based cost-efficient method offers an innovative and practical solution for enhancing underwater sensor node localization.
• The algorithm effectively minimizes localization errors, energy consumption, and time costs, making underwater environmental exploration more feasible.
• The study emphasizes the real-time implementation and effectiveness of the KNN approach in dynamic underwater scenarios.