pytal provides open-source assessment tools to help quantify the effectiveness of different digital infrastructure strategies, particularly for 4G and 5G deployment.
pytal enables researchers to examine trade-offs in infrastructure decisions relating to technologies, infrastructure sharing, regulation and taxation, with the ultimate aim of helping to connect more people to a faster Internet.
- Oughton, E.J., Comini, N., Foster, V., Hall, J.W., 2022. Policy choices can help keep 4G and 5G universal broadband affordable. Technological Forecasting and Social Change 176, 121409. https://doi.org/10.1016/j.techfore.2021.121409
The aim is to be able to assess representative countries using a spatially-explicit modeling codebase, and then use these estimates to scale the results globally. Firstly, using metrics which affect 4G and 5G infrastructure deployment, a k-means clustering method groups low- and middle-income countries into similar clusters based on GDP per capita, population density and existing 4G deployment.
In Oughton et al. (2021) the analysis includes eight representative countries across the six clusters, including Malawi, Uganda, Kenya, Senegal, Pakistan, Albania, Peru and Mexico. For different types of cellular technologies the cost composition can be obtained.
The aim of achieving universal broadband is highly important for economic development,
therefore pytal helps to quantify the financial cost for this target given different user
capacities (for example, up to <25, <200 or <400 Mbps per user as the peak speed). The financial
cost to society (the private cost + government cost) represents the cost of achieving universal
broadband. This includes any net required government subsidy, after spectrum and taxation
income is accounted for.
Once different strategies have been modeled for representative countries, the results can be scaled within each country cluster to provide global cost estimates. This allows researchers to understand the required investment for each country, or by each country income group. The results are reported for the required percentage of annual GDP which would need to be invested over the next decade.
The recommended installation method is to use conda, which handles packages and virtual environments, along with the conda-forge channel which has a host of pre-built libraries and packages.
Create a conda environment called pytal:
conda create --name pytal python=3.7 gdal
Activate it (run this each time you switch projects):
conda activate pytal
First, install optional packages:
conda install geopandas fiona shapely rtree pyproj
Then install pytal:
python setup.py install
Alternatively, for development purposes, clone this repo and run:
python setup.py develop
You will need numerous input data sets.
First, download the Global Administrative Database (GADM), following the link below and making sure you download the "six separate layers.":
Place the data into the following path data/raw/gadm36_levels_shp.
Then download the WorldPop global settlement data from:
Place the data in data/raw/settlement_layer.
Next, download the nightlight data here:
https://ngdc.noaa.gov/eog/data/web_data/v4composites/F182013.v4.tar
Place the unzipped data in data/raw/nightlights/2013.
Obtain the Mobile Coverage Explorer data from Collins Bartholomew:
https://www.collinsbartholomew.com/mobile-coverage-maps/mobile-coverage-explorer/
Place the data into data/raw/Mobile Coverage Explorer.
Once complete, run the following to preprocess all data:
python scripts/preprocess.py
First run the following initial preprocessing script to extract the necessary files:
python scripts/preprocess.py
Then once you have all preprocessed data, you can run the model:
python scripts/run.py
pytal was jointly developed at the University of Oxford and George Mason University, with funding support from UKRI (EPSRC) and the World Bank 5G Flagship.