• Authors
• Introduction
• Overview
• Usage
• Validation
• Contribute
• Acknowledgements

• Sergio Barrachina-Muñoz
• Francesc Wilhelmi

Both authors have the same contribution to this project.

Komondor is a wireless network simulator that includes novel mechanisms for next-generation WLANs, such as dynamic channel bonding or enhanced spatial reuse. One of the main purposes of Komondor is to simulate the behavior of IEEE 802.11ax-2019 networks, an amendment designed to boost spectral efficiency in dense deployments.

Komondor has been conceived as an open source tool that contributes to the ongoing research in wireless networks, especially regarding the implementation of novel functionalities that are not available in other well-known wireless simulators. In addition, it has been prepared for a simple integration with Machine Learning (ML) modules.

The project is structured as follows:

• Apps: contains auxiliary applications that support the Komondor's core operation. For example, in /Apps we find the "Input Generation" project, which contains a Java application that generates inputs for Komondor in an easy and flexible way.
• Code: contains the core files to compile and run Komondor, including input and output folders.
• Documentation: contains Documentation related to Komondor, such as a Manual, a user's guide, presentations, etc.

Doxygen documentation available here.

The code to run simulations is organized as follows:

• COST: constitutes the Komondor's primitive operation. Here we find the CompC++ library that allows generating discrete event simulations. For further information about COST, please refer to its main website.
• main: contains the core files (komondor.cc, node.h, traffic_generator.h, agent.h, and central_controller.h) that are in charge of orchestrating all the simulation. "komondor.cc" is the main component, which initializes all the other components of "Type II". All these modules are aware of the existence of the simulation time. In addition to the core components, here we find "build_local", a bash script that compiles the libraries for executing the code. Note that the file "compcxx_komondor_main.h" is also required to carry out such a compilation.
• methods: by following clean architecture guidelines, independent methods used by core files are contained in the methods folder. Several libraries are provided according to the nature of their functions. For instance, "backoff_methods.h" contains methods to handle the backoff operation in the Distributed Coordination Function (DCF).
• structures: the Komondor simulator considers several header files to carry out its operation. Among them, we find "wlan.h", which defines the main characteristics of a WLAN (WLAN id, list of associated STAs, etc.). In addition, the "notification.h" structure allows us to define the information to be exchanged between devices.
• learning_modules: here we find the implementation of ML methods that receive feedback about the network's performance in simulation time.
• list_of_macros.h: all the static parameters (e.g., constants) are contained in this file.
• input: contains the input files that allow building the simulation environment.
• output: contains the data generated by Komondor as a result of a given simulation.
• scripts_multiple_executions: contains bash scripts to perform multiple simulations.

An overview of the current modules available in Komondor is shown next:

Detailed installation and execution instructions can be found in the Komondor User's Guide.

In short, to run Komondor, just build komondor_main and then execute it by following the next steps:

$ cd ./Code/main
$ chmod -R 777 <dirname>

WARNING: Following instructions use a pre-built CompC++. The binary Code/COST/cxx is platform-dependent and might not work on your machine.

$ ./build_local

Following instructions require make.

$ make

Run Komondor simulator for the given input information (basic simulation)

$ ./komondor_main INPUT_FILE_NODES OUTPUT_FILE_LOGS FLAG_SAVE_NODE_LOGS FLAG_PRINT_SYSTEM_LOGS FLAG_PRINT_NODE_LOGS SIM_TIME SEED

The inputs are further described next:

• INPUT_FILE_NODES: file containing nodes information (e.g., position, channels allowed, etc.).The file must be a .csv with semicolons as separators.
• OUTPUT_FILE_LOGS: path to the output file to which write results at the end of the execution (if the file does not exist, the system will create it).
• SIMULATION_CODE: name given to the simulation, used to keep track of specific simulations during campaigns.
• FLAG_SAVE_NODE_LOGS: flag to indicate whether to save the node logs into separate files (1) or not (0). If this flag is activated, one file per node will be created.
• FLAG_PRINT_SYSTEM_LOGS: flag to indicate whether to print the system logs (1) or not (0).
• FLAG_PRINT_NODE_LOGS: flag to indicate whether to print the node logs (1) or not (0).
• SIM_TIME: simulation time
• SEED: random seed the user wishes to use

IMPORTANT NOTE (!): Setting FLAG_SAVE_NODE_LOGS to TRUE (1) entails a larger execution time.

STEP 2-1: Run Komondor simulator with intelligent agents

Alternatively, and in order to indicate the usage of agents, the console input must add the following extra information:

$ ./komondor_main INPUT_FILE_NODES INPUT_FILE_AGENTS OUTPUT_FILE_LOGS FLAG_SAVE_NODE_LOGS FLAG_SAVE_AGENT_LOGS FLAG_PRINT_SYSTEM_LOGS FLAG_PRINT_NODE_LOGS FLAG_PRINT_AGENT_LOGS SIM_TIME SEED

The agent's operation has been summarized at README_agents.

There is an input file that is required for basic Komondor's execution. Input files are located at the "input" folder:

• input_nodes_conf.csv: define parameters such as the node ID, the node location, etc.
• agents.csv: define parameters used by the agents' operation. Refer to README_agents.

Apart from the input nodes file, different models are loaded through the "config_models" file (located here).

Regarding the output ("output" folder), some logs and statistics are created at the end of the execution.

Debugging:

$ apt-get install gdb valgrind 

Komondor v.2.0 has been validated by means of ns-3 and SF-CTMN, and Bianchi analytical models. The presentation of the validation can be found at "S. Barrachina-Muñoz, F. Wilhelmi, I. Selinis & B. Bellalta. Komondor: a Wireless Network Simulator for Next-Generation High-Density WLANs. 2018". Additional resources are available in this repository, in the folder /Documentation/Validation. Files used for each simulation tool can be found:

1. ns-3: execution script and instructions, together with the simulation environment used for validation.
2. Komondor: input "nodes" and "system" files. Release pointing to v.2.0 must be used.
3. SF-CTMN: simulation environment, to be executed through the SF-CTMN framework, available at https://github.com/sergiobarra/SFCTMN
4. Bianchi: Matlab files emulating the Bianchi model ("Bianchi, G., Fratta, L., & Oliveri, M. (1996, October). Performance evaluation and enhancement of the CSMA/CA MAC protocol for 802.11 wireless LANs. In Personal, Indoor and Mobile Radio Communications, 1996. PIMRC'96., Seventh IEEE International Symposium on (Vol. 2, pp. 392-396). IEEE.") can be found, which simulate the throughput achieved by each WLAN in each of the proposed scenarios.

An automated regression test is available to ensure that the development of new features does not affect the previous implementation, which was validated in Barrachina-Muñoz, S., Wilhelmi, F., Selinis, I., & Bellalta, B. (2019, April). Komondor: a Wireless Network Simulator for Next-Generation High-Density WLANs. In 2019 Wireless Days (WD) (pp. 1-8). IEEE.

In order to execute the regression test, go to ./Komondor/Code/input and run the script named "script_regression_validation_scenarios.sh". This script will take the inputs from the "validation" folder, execute the corresponding simulations, and compare the output with the expected results (i.e., the results obtained for the aforementioned paper).

The output of the regression test will be displayed in the console. In case of success, the following output should be observed:

Before executing the regression test, it is important to ensure that "simulation_ix_output_script" in config_models is set to 10 and "path_loss_model" to 4.

One of the main purposes of Komondor is to serve as an academic/educational tool. In what follows, we list the projects in which Komondor has been used as a simulation tool:

• Wilhelmi, F., Bellalta, B., Szott, S., Kosek-Szott, K., & Barrachina-Muñoz, S. (2025, May). Coordinated Multi-Armed Bandits for Improved Spatial Reuse in Wi-Fi. In 2025 IEEE International Conference on Machine Learning for Communication and Networking (ICMLCN) (pp. 1-6). IEEE. (Download the paper here)
• Wilhelmi, F., Galati-Giordano, L., & F., G. (2025, March). “It's Your Turn”: A Novel Channel Contention Mechanism for Improving Wi-Fi's Reliability. In 2025 IEEE Wireless Communications and Networking Conference (WCNC) (pp. 1-6). IEEE. (Download the paper here)
• Wilhelmi, F., Szott, S., Kosek-Szott, K., & Bellalta, B. (2024). Machine Learning and Wi-Fi: Unveiling the Path Toward AI/ML-Native IEEE 802.11 Networks. IEEE Communications Magazine. [(Download the paper here)](https://arxiv.org/abs/2412.03076
• Wilhelmi, F. & Bellalta, B. (2020). ITU AI for 5G Challenge. UPF/ITU-T AI Challenge - Improving the capacity of IEEE 802.11 WLANs through Machine Learning.
• Wilhelmi, F., Carrascosa, M., Cano, C., Jonsson, A., Ram, V., & Bellalta, B. (2021). Usage of Network Simulators in Machine-Learning-Assisted 5G/6G Networks. IEEE Wireless Communications Magazine.
• Wilhelmi, F., Barrachina-Muñoz, S., Bellalta, B., Cano, C., Jonsson, A., & Ram, V. (2020). A flexible machine-learning-aware architecture for future WLANs. IEEE Communications Magazine, 58(3), 25-31.
• Wilhelmi, Francesc, Sergio Barrachina-Muñoz, and Boris Bellalta. "Spatial Reuse in IEEE 802.11ax WLANs." Computer Communications (2020).
• Barrachina-Muñoz, Sergio, et al. "Online Primary Channel Selection for Dynamic Channel Bonding in High-Density WLANs." IEEE Wireless Communications Letters (2019).
• Barrachina-Muñoz, Sergio, et al. "Komondor: a wireless network simulator for next-generation high-density WLANs." Wireless Days (WD), 1-8 (2019).
• Barrachina-Muñoz, Sergio, et al. "Dynamic Channel Bonding in Spatially Distributed High-Density WLANs." IEEE Transactions on Mobile Computing (2019).
• SECTOR, STANDARDIZATION. FOCUS GROUP ON MACHINE LEARNING FOR FUTURE NETWORKS INCLUDING 5G ML5G-I-125-R3. StandICT.
• Wilhelmi, Francesc, et al. "Potential and pitfalls of multi-armed bandits for decentralized spatial reuse in WLANs." Journal of Network and Computer Applications 127 (2019): 26-42.
• Barrachina-Muñoz, Sergio, et al. "To overlap or not to overlap: Enabling channel bonding in high-density WLANs." Computer Networks 152 (2019).
• Wilhelmi, F., Barrachina-Muñoz, S., & Bellalta, B. (2019, October). On the Performance of the Spatial Reuse Operation in IEEE 802.11 ax WLANs. In 2019 IEEE Conference on Standards for Communications and Networking (CSCN) (pp. 1-6). IEEE. IEEE Conference on Standards for Communications and Networking (CSCN).

If you want to contribute, please contact francisco.wilhelmi@upf.edu

More details in CONTRIBUTING.md

2025-2027: This work is supported by the CHIST-ERA Wireless AI 2022 call MLDR project (ANR-23-CHR4-0005), partially funded by AEI under project PCI2023-145958-2, by Wi-XR PID2021-123995NB-I00 and TRUE-Wi-Fi PID2024-155470NB-I00 (MCIU/AEI/FEDER,UE), and by MCIN/AEI under the Maria de Maeztu Units of Excellence Programme (CEX2021-001195-M). 2019-2020: This work has been partially supported by a Gift from the Cisco University Research Program (CG#890107, Towards Deterministic Channel Access in High-Density WLANs) Fund, a corporate advised fund of Silicon Valley Community Foundation.