kaira.channels.FlatFadingChannel

Inheritance diagram for FlatFadingChannel

class kaira.channels.FlatFadingChannel(fading_type: str, coherence_time: int, k_factor: float | None = None, avg_noise_power: float | None = None, snr_db: float | None = None, shadow_sigma_db: float | None = None, *args: Any, **kwargs: Any)

Bases: BaseChannel

Flat fading channel with configurable distribution and coherence time.

Models a wireless channel where the fading coefficient remains constant over a specified coherence time and then changes to a new independent realization. This represents blockwise fading commonly used in communications analysis [Tse and Viswanath, 2005] [Rappaport, 2024].

Mathematical Model:

y[i] = h[⌊i/L⌋] * x[i] + n[i] where L is the coherence length, h follows a specified distribution, and n ~ CN(0,σ²)

Parameters:

• fading_type (str) – Distribution type for fading coefficients (‘rayleigh’, ‘rician’, or ‘lognormal’)

• coherence_time (int) – Number of samples over which the fading coefficient remains constant

• k_factor (float, optional) – Rician K-factor (ratio of direct to scattered power), used only when fading_type=’rician’

• avg_noise_power (float, optional) – The average noise power σ²

• snr_db (float, optional) – SNR in dB (alternative to avg_noise_power)

• shadow_sigma_db (float, optional) – Standard deviation in dB for log-normal shadowing, used only when fading_type=’lognormal’

Example

>>> # Create a flat Rayleigh fading channel with coherence time of 10 samples
>>> channel = FlatFadingChannel('rayleigh', coherence_time=10, snr_db=15)
>>> x = torch.complex(torch.ones(100), torch.zeros(100))
>>> y = channel(x)  # Output with block fading effects

Methods

__init__	Initialize the Flat Fading channel.
forward	Apply flat fading and noise to the input signal.
get_config	Get a dictionary of the channel's configuration.

Attributes

k_factor
avg_noise_power
snr_db
shadow_sigma_db

__init__(fading_type: str, coherence_time: int, k_factor: float | None = None, avg_noise_power: float | None = None, snr_db: float | None = None, shadow_sigma_db: float | None = None, *args: Any, **kwargs: Any)

Initialize the Flat Fading channel.

Parameters:

• fading_type (str) – Distribution type (‘rayleigh’, ‘rician’, ‘lognormal’).

• coherence_time (int) – Samples over which fading is constant.

• k_factor (float, optional) – Rician K-factor (for ‘rician’).

• avg_noise_power (float, optional) – Average noise power σ².

• snr_db (float, optional) – SNR in dB (alternative to avg_noise_power).

• shadow_sigma_db (float, optional) – Shadowing std dev in dB (for ‘lognormal’).

• *args – Variable length argument list passed to the base class.

• **kwargs – Arbitrary keyword arguments passed to the base class.

k_factor: float | None

shadow_sigma_db: float | None

get_config() → Dict[str, Any]

Get a dictionary of the channel’s configuration.

This method returns a dictionary containing the channel’s parameters, which can be used to recreate the channel instance.

Returns:

Dictionary of parameter names and values

Return type:

Dict[str, Any]

avg_noise_power: float | None

snr_db: float | None

forward(x: Tensor, *args: Any, csi=None, noise=None, **kwargs: Any) → Tensor

Apply flat fading and noise to the input signal.

Parameters:

• x (torch.Tensor) – The input tensor.

• *args – Additional positional arguments (unused).

• csi (Optional[torch.Tensor]) – Pre-computed channel state information (fading coefficients). If provided, these coefficients are used instead of generating new ones.

• noise (Optional[torch.Tensor]) – Pre-generated noise tensor. If provided, this noise is added.

• **kwargs – Additional keyword arguments (unused).

Returns:

The output tensor after applying fading and noise.

Return type:

torch.Tensor