ertk.pytorch.utils.add_signal_awgn

ertk.pytorch.utils.add_signal_awgn(x: Tensor, snr: float = 20) → Tensor

Add additive white Gaussian noise (AWGN) to a signal or batch of signals.

Parameters:

x: torch.Tensor

Input tensor of shape (signal_len,) or (batch_size, signal_len).

snr: float

Target signal-to-noise ratio.

Returns:

augmented: torch.Tensor

Augmented tensor of the same shape as the x.