[Solved] Codebook learning (self-supervised) data generation method  

Hi Jakub,

Thanks for your interest in this paper. Based on your description, I think most likely you didn’t save the generated dataset from DeepMIMO by activating the ‘-v7.3’ flag of the Matlab save funtion. So, if the following error message is similar to what you got:

Traceback (most recent call last):
File "D:\ProgramFiles\Anaconda3\lib\site-packages\h5py\_hl\files.py", line 406, in __init__
fid = make_fid(name, mode, userblock_size,
File "D:\ProgramFiles\Anaconda3\lib\site-packages\h5py\_hl\files.py", line 173, in make_fid
fid = h5f.open(name, flags, fapl=fapl)
File "h5py\_objects.pyx", line 54, in h5py._objects.with_phil.wrapper
File "h5py\_objects.pyx", line 55, in h5py._objects.with_phil.wrapper
File "h5py\h5f.pyx", line 88, in h5py.h5f.open
OSError: Unable to open file (file signature not found)

Then, you can try to write ‘-v7.3’ in the Matlab save function, i.e.,

save('your_dataset_name.mat', 'dataset', '-v7.3') % dataset is the variable of your generated channel data

I hope this can solve your problem. Let me know if you still couldn't get the right results.

Good luck,

Yu

Hello,

I have faced this issue during my first step of downloading the scenario, plz could you tell me what the problem?!

I1 scenario for Mapping channels work. Also, is there an extended work in this research?

Hi @Muhammad Alrabeiah,

The download of the large file may occasionally fail. If that happened, the downloaded data is erroneous and you may see such errors.
You may try to re-download through the scenario page. Also, you may use this alternative link for I1_2p5, where each file is separately available and you can download in parts if needed.

Regards,
Umut