Code for Generating Statistical ON-OFF Link Models from Sea Trial Data

This folder contains the code for generating and testing statistical ON-OFF link models from sea trial data. It was used to generate the results for the following paper:

N. Morozs, F. Campagnaro, L. Shen, B. T. Henson, F. Mahieu, Y. Zakharov and P. D. Mitchell, "Statistical ON-OFF Link Modeling Based on Sea Trial Data," in proceedings of WUWNet'24, Sibenik, Croatia, Oct 2024.

This code is published under the MIT License (see full license in the 'LICENSE' file).

Required software:

1) MATLAB (tested on R2018a and R2022a)

***********
Data files:
***********

- 'BerlinMultimodal-06-16.mat' - The Werbellin lake experiment dataset downloaded from the ASUNA database (https://sites.google.com/marsci.haifa.ac.il/asuna/).

- 'on_off_dur_model_werbellin16.mat' - The ON-OFF statistical model generated by the "gen_link_model.m" script, as explained in the paper.

- 'on_off_dur_cdfs_werbellin16.csv' - the same model but in the CSV format (to enable its use with any other programming languages); it contains six CDFs: for three link types (good, average, poor) and two link states (ON and OFF).

***************
MATLAB scripts:
***************

- 'analyse_trial_data.m' - a script to analyse the data from the Werbellin lake experiment and generate the plots from the paper.

- 'gen_link_model.m' - a script to process the ASUNA Werbellin lake dataset and generate statistical link models as described in the paper.

- 'gen_link_realisations.m' - a script to synthesise random link realisations using the statistical model generated by the 'gen_link_model.m' script.

- 'test_on_off_dur_gen_model.m' - to test if the synthesised random link realisations follow the same CDF distributions as the statistical model generated by the 'gen_link_model.m' script.