Jammer-Loc

This repository includes the code and the Dataset Link for the paper
“Machine and Deep Learning for Indoor UWB Jammer Localization.”

Usage Example

Run the hyperparameter optimization process:

python hpo.py \
  --task classification \
  --mode actual \
  --framework ml \
  --config config.yaml \
  --trials 50 \
  --study-name my_study

1. Models and metrics must be saved by main.py in the expected format.
2. Optuna results are saved to hpo.db by default.
3. All command-line options are defined in hpo.py using argparse. Refer to that file for descriptions and additional flags.

Classification and Regression on Source and Testing it on the Test Set

To run classification or regression on the source dataset and test it on the test set, you can use:

python main.py --config config.yaml

1. Outputs and metrics will be saved to the directory defined by cfg.output.directory.
2. All model checkpoints and metrics will be saved per model (e.g., outputs/simple_nn/).
3. In the event of a runtime failure, detailed logging information — including structured error messages and stack traces — will be recorded to facilitate systematic debugging and root cause analysis.

Domain Alignment

File names for the domain alignment methods correspond directly to their respective names in the publication.

Path Setup

Make sure you adjust the paths according to your project structure in the config.yaml file.

Datasets

Due to size constraints, the raw datasets are not included in this repository.
They can be downloaded from .

However, to facilitate immediate code execution, two Parquet files are provided in this repository for the source and target domains. The compressed NumPy array files (.npz), containing the data for domain alignment models, can be accessed through the link above.