Ocean Wave Simulation and Prediction

Files
TR Number
Date
2018-09-10
Journal Title
Journal ISSN
Volume Title
Publisher
Virginia Tech
Abstract

WiFi can provide network coverage for users on land at anytime and anywhere, but on the sea, the wireless communication scenes change dramatically due to the signals are non-existence. Although some techniques (e.g. satellite, undersea fiber, microwave communication) have been used in marine communication, they are either too expensive with very small bandwidth, or too limited in its coverage range. We propose to develop a marine wireless mesh network which is formed by low cost buoyed wireless base stations to provide broadband connectivity for users on the sea.

Ocean wave simulation and prediction are key technologies in developing marine mesh network, because marine environments are dramatically different from terrestrial environment. The ocean waves have characteristics of rhythmic oscillations and the line of sight between two communication nodes is often blocked by them. Therefore, we have to develop a new wave-state-aware networking protocol which is suitable for marine environments. Ocean wave simulation technology can simulate this kind of dynamic environments and provide a test platform for the development of marine mesh network. Ocean wave prediction technology can improve the throughput of marine wireless network. Thus, they are indispensable technologies in developing marine mesh network.

In this thesis, we designed an ocean wave measurement method, two ocean wave prediction methods, and an ocean wave simulation method.

Firstly, we designed an accelerometer-based ocean wave measurement method. It can measure the real time wave height accurately.

Secondly, we designed an Elman-neural-network-based ocean wave prediction method for nonlinear waves. It has a higher prediction accuracy than other neural network methods in nonlinear wave prediction.

Thirdly, we designed a multiple-linear-regression-based ocean wave prediction method for linear waves. It has a higher prediction accuracy and less time consumption than other methods in linear wave prediction.

Finally, we implemented and improved a spectrum-based ocean wave simulation method which is originally proposed by Tessendorf. It can present the movement of ocean waves realistically and in real time.

To sum up, above four methods provide an effective test platform and technical support for the development of our marine mesh network.

Description
Keywords
Ocean Wave Simulation, Ocean Wave Prediction, Ocean Wave Spectrum, Fast Fourier Transform, Neural Network, Multiple Linear Regression, Oscillation Measurement
Citation
Collections