hc/JSCC
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

add Rayleigh channel

Browse Source
This commit is contained in:
chun 2024-06-09 01:46:49 +08:00
parent ff3d377583
commit b7bd3bdd42
1 changed files with 29 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

41
channel.py
View File

@ -11,20 +11,37 @@ class Channel(nn.Module):
self.snr = snr self.snr = snr
def forward(self, z_hat): def forward(self, z_hat):
if z_hat.dim() == 4: if z_hat.dim() not in {3, 4}:
# k = np.prod(z_hat.size()[1:]) raise ValueError('Input tensor must be 3D or 4D')
k = torch.prod(torch.tensor(z_hat.size()[1:]))
sig_pwr = torch.sum(torch.abs(z_hat).square(), dim=(1, 2, 3), keepdim=True) / k # if z_hat.dim() == 4:
elif z_hat.dim() == 3: # # k = np.prod(z_hat.size()[1:])
# k = np.prod(z_hat.size()) # k = torch.prod(torch.tensor(z_hat.size()[1:]))
k = torch.prod(torch.tensor(z_hat.size())) # sig_pwr = torch.sum(torch.abs(z_hat).square(), dim=(1, 2, 3), keepdim=True) / k
sig_pwr = torch.sum(torch.abs(z_hat).square()) / k # elif z_hat.dim() == 3:
# # k = np.prod(z_hat.size())
# k = torch.prod(torch.tensor(z_hat.size()))
# sig_pwr = torch.sum(torch.abs(z_hat).square()) / k
if z_hat.dim() == 3:
z_hat = z_hat.unsqueeze(0)
k = z_hat[0].numel()
sig_pwr = torch.sum(torch.abs(z_hat).square(), dim=(1, 2, 3), keepdim=True) / k
noi_pwr = sig_pwr / (10 ** (self.snr / 10)) noi_pwr = sig_pwr / (10 ** (self.snr / 10))
noise = torch.randn_like(z_hat) * torch.sqrt(noi_pwr) noise = torch.randn_like(z_hat) * torch.sqrt(noi_pwr/2)
if self.channel_type == 'Rayleigh': if self.channel_type == 'Rayleigh':
# hc = torch.randn_like(z_hat) wrong implement before # hc = torch.randn_like(z_hat) wrong implement before
hc = torch.randn(1, device = z_hat.device) # hc = torch.randn(1, device = z_hat.device)
z_hat = hc * z_hat hc = torch.randn(2, device = z_hat.device)
# clone for in-place operation
z_hat = z_hat.clone()
z_hat[:,:z_hat.size(1)//2] = hc[0] * z_hat[:,:z_hat.size(1)//2]
z_hat[:,z_hat.size(1)//2:] = hc[1] * z_hat[:,z_hat.size(1)//2:]
# z_hat = hc * z_hat
return z_hat + noise return z_hat + noise
@ -40,6 +57,6 @@ if __name__ == '__main__':
print(z_hat) print(z_hat)
channel = Channel(channel_type='Rayleigh', snr=10) channel = Channel(channel_type='Rayleigh', snr=10)
z_hat = torch.randn(64, 10, 5, 5) z_hat = torch.randn(10, 5, 5)
z_hat = channel(z_hat) z_hat = channel(z_hat)
print(z_hat) print(z_hat)