w5688414 / bfnn

this is the pytorch implementation of the paper: Beamforming Design for Large-Scale Antenna Arrays Using Deep Learning

BFNN

this is the pytorch implements of the paper: Beamforming Design for Large-Scale Antenna Arrays Using Deep Learning

说明

源代码的地址为：https://github.com/TianLin0509/BF-design-with-DL

requirements

• torch 1.4
• ubuntu 16.04

注意

• pytorch 对复数操作不友好，目前把复数的两部分直接拼接，然后参与后面的计算，不能保证接近性能方面tf的代码

bfnn's Issues

ffbn_test.py gives a RuntimeError

Dear Professor,

I am trying to run ffbn_test.py but receiving this error. I will be more than grateful if you can kindly provide guidance so that I can reproduce the results.

[tensor(0.0182, grad_fn=), tensor(0.0528, grad_fn=), tensor(0.1363, grad_fn=), tensor(0.3000, grad_fn=), tensor(0.5750, grad_fn=), tensor(1.0037, grad_fn=), tensor(1.6343, grad_fn=), tensor(2.4982, grad_fn=), tensor(3.5960, grad_fn=)]
Traceback (most recent call last):
return gca().plot( File "C:\ProgramData\Anaconda3\envs\pytorch\lib\site-packages\matplotlib\axes_axes.py", line 1668, in plot
lines = [*self._get_lines(*args, data=data, **kwargs)]
File "C:\ProgramData\Anaconda3\envs\pytorch\lib\site-packages\matplotlib\axes_base.py", line 311, in call
yield from self._plot_args(
File "C:\ProgramData\Anaconda3\envs\pytorch\lib\site-packages\matplotlib\axes_base.py", line 494, in _plot_args y = check_1d(xy[1])
File "C:\ProgramData\Anaconda3\envs\pytorch\lib\site-packages\matplotlib\cbook_init.py", line 1382, in _check_1d
return np.atleast_1d(x) File "<array_function internals>", line 180, in atleast_1d
File "C:\ProgramData\Anaconda3\envs\pytorch\lib\site-packages\numpy\core\shape_base.py", line 65, in atleast_1d ary = asanyarray(ary)
File "C:\ProgramData\Anaconda3\envs\pytorch\lib\site-packages\torch_tensor.py", line 956, in array
return self.numpy()
RuntimeError: Can't call numpy() on Tensor that requires grad. Use tensor.detach().numpy() instead.