Script Documentation

Every csv file contains the data to manipulate: as indicated by the filename extension, they contain comma-separated values (strings), each of them representing a complex number (that is formatted as needed within the code in order to have the numbers parsed correctly by Python). Essentially, they contain an m * n table, where m is the number of studied sub-carriers, and n is the number of packets sampled on each sub-carrier. The setup of the m parameter is automatic and depends on the value of the bandwidth of the studied channel set up by the user in main.py (see Section [Main Script](README.md#Main Script)).

The unnecessaryPlots file contains the names of the sub-carriers that do not need to be plotted because they are unused for modulation, therefore are not currently relevant to the purpose of this study. If necessary, it is possible to add other sub-carriers to the list contained in this file. To do so, one can simply add to the file a new line containing the string "SC" followed by the number of the sub-carrier that does not need to be plotted.

It is fundamental that the names of all files and folders (except for those having a .py extension) stay unaltered because they are embedded as strings in the code, therefore their change would cause the script not to compile and run correctly.

Main Script

The main.py file contains the code that, by calling functions contained in other Python files - whose content is described in section Python files content - plots the content of the folders named after the specified csi file.

Every folder contains the following multiple sub-folders:

artificial_increments: contains the outputs of the manipulation of artificially generated data. The code generating the content of this folder is contained in the artificial_trace_processor.py file.
auto-correlation_graphs: auto-correlation graphs plotted using python's statsmodels.graphics.tsaplots.plot_acf function in autocorrelation_plotter.py
auto-correlation_through_formulae: auto-correlation graphs plotted using manually-typed formulae in autocorrelation_plotter.py
best_fits_params and sub-folders: each sub-folder is named after one of the five distributions that best fit the increments of the sub-carriers. Each folder contains the plots of the values of the parameters that characterize the distribution naming the folder. The plot is done through the code in best_fits_param_calculator.py
fit_by_sc: files containing the values of sum of square errors, Akaike and Bayesian Information Criterion for each of the listed distributions, which are the five best-fitting distributions for the selected sub-carrier. The distributions are selected among a list of the ten most common. The code generating this folder is in fitting_by_sc.py
fit_by_sc_2: files containing the values of sum of square errors, Akaike and Bayesian Information Criterion for each of the listed distributions, which are the five best-fitting distributions for the selected sub-carrier. The distributions are selected based on a list specified by the user in main.py. The code generating this folder is in fitting_by_sc.py
fit_increments and sub-folders: files containing the values of sum of square errors, Akaike and Bayesian Information Criterion for each of the listed distributions, which are the five best-fitting distributions for the increments of the selected sub-carrier. The distributions are selected based on a list specified by the user in main.py. The code generating this folder is in fitting_by_sc.py
fit_specific_dists: files containing the output of the fitting of the increments of specific distributions.
histograms: histograms of the sub-carriers plotted by histograms_plotter.py
increments_hist: histograms of the increments of each sub-carrier plotted by increments_plotter.py
merged_plot: plots and information about merged data and merged increments generated by merged_plotter.py
params: standard deviation, kurtosis and skewness for data and increments generated by parameters_calculator.py
time_evolution: plots of the evolution of the amplitude of the packets through time for each sub-carrier generated by time_evolution_plotter.py

Python files content

Below are listed the Python files whose methods are called in main.py:

artificial_trace_processor.py: generates and processes artificial traces
auto-correlation_plotter.py: contains functions that are used to plot graphs representing the auto-correlation process for each sub-carrier
best_fits_param_calculator.py: contains functions that are used to fit the five distributions that best fit the merged increments onto each sub-carrier and to calculate and plot their parameters
fitting_by_sc.py: contains functions that are used to find the best-fitting distribution for each sub-carrier and for the distributions of the increments of each sub-carrier and save each output in a dedicated file
histograms_plotter.py: contains functions that are used to plot a magnitude/relative frequency histogram for each sub-carrier, including functions used to verify whether a process is stationary, divide the column corresponding to a process in batches, and process each batch
increments_plotter.py: contains functions that are used to plot an increment/frequency histogram for each sub-carrier
merged_plotter.py: contains the functions that are used to elaborate the merged increments (i.e. by considering all the increments as if they belonged to a single sub-carrier)
parameters_calculator.py: contains the functions that are used to calculate variance, skewness kurtosis for each sub-carrier and for the distribution of the increments of each sub-carrier
std_deviation_and_kurtosis_plotter.py: contains a function used to plot standard deviation and kurtosis calculated on the increments of each sub-carrier
time_evolution_plotter.py: contains a function that is used to plot a graph representing the evolution of the signal magnitude over time for each sub-carrier

Libraries

The code is written in Python 3.6.5 and uses the following libraries:

The required versions are specified in requirements.txt.

How to run the code

The code can be run by executing main.py using Python 3.6.5.

Notes

The code is not optimized for speed yet. It is intended to be used as the source code to plot the data and graphs described in the previous sections.