malrabeiah / VABT

Vision-Aided Beam Tracking

Beam Tracking Using Previously-Observed Beam Sequences

The baseline solution for the ViWi-Vision Aided Beam Tracking (ViWi-BT) task. Please see ViWi Vision-Aided mmWave Beam Tracking: Dataset, Task, and Baseline Solutions for more details.

Dependecies

1. Python 3.7 and later.

2. PyTorch 1.3 and later (with torchvision).

3. Numpy 1.6 or later.

4. Pandas.

5. NVIDIA GPU with CUDA (10.XX).

Running instructions

These scripts assume the ViWi-BT dataset in .csv formate. The visual data, traain_set.csv, and val_set.csv files all need to be included in the same root directory as the scripts. To run the script, do the following:

1. In the script main_beam.py, adjust the training and validation hyperparameters as needed.

2. Run main_beam.py

The results will be saved in a ".mat" file.

Citation

If you plan to use these scripts or part of them, please cite the following paper:

@misc{alrabeiah2020viwi,
    title={ViWi Vision-Aided mmWave Beam Tracking: Dataset, Task, and Baseline Solutions},
    author={Muhammad Alrabeiah and Jayden Booth and Andrew Hredzak and Ahmed Alkhateeb},
    year={2020},
    eprint={2002.02445},
    archivePrefix={arXiv},
    primaryClass={eess.SP}
}