Simplifying OMNeT++ Model Installations with opp_env

Do you like using the pip command for installing Python packages? Pip makes package installation a breeze. Now imagine if you could install OMNeT++ simulation models and model frameworks just as easily.

The Problem: Manual Installation Hassles

Installing many OMNeT++ models and model frameworks can be tedious. They often require:

• A specific OMNeT++ version (which might not be the latest one)
• Dependencies like INET, Veins, or Simu5G in correct versions
• Adjusting variables in config files
• Sourcing environment scripts
• Importing projects into the IDE and checking cross-references

Wouldn’t it be great if all of this could be automated?

Overcoming Compatibility Issues and “Bitrot”

Compiling older models can be challenging due to bitrot. Changes in compilers, system headers, or library versions can disrupt the build process and require time-consuming fixes.

Enter opp_env

We developed opp_env to solve these problems. It can:

• Download and install simulation models, including the appropriate OMNeT++ version.
• Provide a ready-to-use shell environment for running simulations.
• Set up a development environment from a remote Git repository (useful for model developers).

With an internal database of over 70 models, opp_env offers a streamlined approach to handle these installations automatically.

Example Session with opp_env

Here’s a quick demonstration:

1. List the available models:

    $ opp_env list
    ...
    simproctc                   2.0.2
    simu5g                      1.2.2  1.2.1  1.1.0  git
    simulte                     1.2.0  1.1.0  0.9.1
    soa4core                    20240124
    ...
   

2. Get information on specific simulation model:

    $ opp_env info simu5g-latest
   

   This shows essential details like download URL and dependencies:

    simu5g-1.2.2: 5G NR and LTE/LTE-A user-plane simulation model
    Requires:
    - inet: 4.5.2 / 4.5.1 / 4.5.0
    - omnetpp: 6.1.0 / 6.0.3 / 6.0.2
    ...
   

3. Set up a model and open a shell environment:

    $ mkdir my_workspace
    $ cd my_workspace
    $ opp_env init
    $ opp_env install simu5g-latest
    $ opp_env shell
   

   After the install command, the workspace directory has the following directories:

    omnetpp-6.0.3/      inet-4.5.2/      simu5g-1.2.2/
   

   From the active shell, you can start running simulations immediately:

    $ cd simulations/NR/standalone
    $ ./run
   

Streamlined Workflow with One Command

To condense the above steps, use the following all-in-one command:

$ opp_env shell -w my_workspace --init --install simu5g-latest

This creates a working environment with just one command!

Installing opp_env

To try this yourself:

1. Install opp_env via pip:

    $ pip install opp_env
   

2. Set up Nix: You’ll need a recent version of Nix, which is best installed from their website.

Why Nix Matters

opp_env relies on the Nix package manager to ensure compatibility. Nix provides the appropriate versions of the compiler, libraries, and other packages needed to ensure that models build and run without requiring manual adjustments. This emphasis on repeatable builds is crucial for creating reproducible simulations, which are vital for scientific publications and research.

Learn More

After installing opp_env, use -h (or --help) to access general help, or opp_env <subcommand> -h for detailed guidance on a specific subcommand (e.g., opp_env install -h). The help includes extensive descriptions and practical examples for all available options.

For installation instructions and more detailed information, check out the project’s GitHub page: opp_env on GitHub.