PolarEncoder

class sionna.phy.fec.polar.PolarEncoder(frozen_pos: numpy.ndarray, n: int, *, precision: str | None = None, device: str | None = None, **kwargs)

Bases: sionna.phy.block.Block

Polar encoder for given code parameters.

This block performs polar encoding for the given k information bits and the frozen set (i.e., indices of frozen positions) specified by frozen_pos.

Parameters:

• frozen_pos (numpy.ndarray) – Array of int defining the n-k frozen indices, i.e., information bits are mapped onto the k complementary positions.

• n (int) – Defining the codeword length.

• precision (str | None) – Precision used for internal calculations and outputs. If None, precision is used.

• device (str | None) – Device for computation (e.g., ‘cpu’, ‘cuda:0’). If None, device is used.

Inputs:

bits – […, k], torch.float. Binary tensor containing the information bits to be encoded.

Outputs:

cw – […, n], torch.float. Binary tensor containing the codeword bits.

Notes

As commonly done, we assume frozen bits are set to 0. Please note that - although its practical relevance is only little - setting frozen bits to 1 may result in affine codes instead of linear code as the all-zero codeword is not necessarily part of the code any more.

Examples

import torch
from sionna.phy.fec.polar import PolarEncoder
from sionna.phy.fec.polar.utils import generate_5g_ranking

k, n = 100, 256
frozen_pos, _ = generate_5g_ranking(k, n)
encoder = PolarEncoder(frozen_pos, n)

bits = torch.randint(0, 2, (10, k), dtype=torch.float32)
codewords = encoder(bits)
print(codewords.shape)
# torch.Size([10, 256])

Attributes

property k: int

Number of information bits.

property n: int

Codeword length.

property frozen_pos: numpy.ndarray

Frozen positions for Polar decoding.

property info_pos: numpy.ndarray

Information bit positions for Polar encoding.

Methods

build(input_shape: Tuple[int, ...]) → None

Build and check if k and input_shape match.

Parameters:

input_shape (Tuple[int, ...])