C++ and Python laboratory tools for telecommunications-network design. The repository implements blocking-probability calculations, multi-rate traffic models, and alternative-system dimensioning scenarios.
Source code | Module guide | MIT license | Report an issue
- Erlang B blocking-probability calculations with a stable recursive form.
- Kaufman-Roberts state distribution for multi-rate traffic streams.
- Alternative-resource dimensioning with equivalent-changes and Fredericks-Hayward methods.
- Scenario exploration in C++ and Python, including tables and generated plots.
There is no binary release. The simplest route is to run a Python scenario from source.
-
Clone the repository and create a Python environment.
git clone https://github.com/haribo841/PST.git cd PST python -m venv .venv .\.venv\Scripts\Activate.ps1 python -m pip install numpy matplotlib
-
Adjust the parameters near the top of
PST_C03.pyif needed. -
Run the alternative-system scenario.
python PST_C03.py
-
Review the generated text table and
porownanie_metod.pngin the working directory.
Generated from the unmodified default scenario: primary capacities Ci = [10, 8, 12], direct alternative traffic Aalt = 0, and blocking targets 0.01 and 0.05.
Traffic per capacity unit a |
Equivalent changes, B=0.01 | Fredericks-Hayward, B=0.01 | Equivalent changes, B=0.05 | Fredericks-Hayward, B=0.05 |
|---|---|---|---|---|
| 0.2 | 1 | 2 | 1 | 2 |
| 0.8 | 8 | 13 | 6 | 9 |
| 1.3 | 22 | 27 | 18 | 21 |
The complete output table contains all 12 traffic points. These are reproducible outputs of the educational implementation, not a validation of the models for production network planning.
| Environment | Support |
|---|---|
| Python 3.12+ | The current regression suite was verified on Windows with Python 3.13.1. Linux and macOS were not tested locally. |
| Windows with Visual Studio 2022 and C++ Desktop Development tools | The solution uses C++20 and includes all four modules. See build and scenario commands. |
| Binary release | Not published; use the verified Python quick start above. |
The module guide maps each source file to its model and output, and explains how to select a scenario.
Run the Python checks with python -m unittest discover -s tests -v after installing numpy matplotlib pandas. The C++ test runner compiles and runs numerical regressions with Visual Studio. See the guide for the C04 model's assumptions and safe output-filename restrictions.
The project code and documentation are available under the MIT license. Dependencies retain their own licenses. For questions, corrections, or reproducible issues, use GitHub Issues.
The previous one-line README is preserved in the README archive.
