MCSDecoderNR

class sionna.phy.nr.utils.MCSDecoderNR(*args: Any, precision: Literal['single', 'double'] | None = None, device: str | None = None, **kwargs: Any)

Bases: sionna.phy.utils.misc.MCSDecoder

Maps a Modulation and Coding Scheme (MCS) index to the corresponding modulation order, i.e., number of bits per symbol, and coderate for 5G-NR networks. Wraps decode_mcs_index() and inherits from MCSDecoder.

Inputs:

• mcs_index – […], torch.int32. MCS index.

• mcs_table_index – […], torch.int32. MCS table index. Different tables contain different mappings.

• mcs_category – […], torch.int32. 0 for PUSCH, 1 for PDSCH channel.

• check_index_validity – bool. If True, a ValueError is raised if the input MCS indices are not valid for the given configuration. Defaults to True.

• transform_precoding – […], torch.bool | bool. Specifies whether the MCS tables described in Sec. 6.1.4.1 of [3GPPTS38214] are applied. Only relevant for “PUSCH”. Defaults to False.

• pi2bpsk – […], torch.bool | bool. Specifies whether the higher-layer parameter tp-pi2BPSK described in Sec. 6.1.4.1 of [3GPPTS38214] is applied. Only relevant for “PUSCH”. Defaults to False.

• verbose – bool. If True, additional information is printed. Defaults to False.

Outputs:

• modulation_order – […], torch.int32. Modulation order corresponding to the input MCS index.

• target_coderate – […], torch.float32. Target coderate corresponding to the input MCS index.

Parameters:

• args (Any)

• precision (Literal['single', 'double'] | None)

• device (str | None)

• kwargs (Any)

Examples

from sionna.phy.nr.utils import MCSDecoderNR
import torch

decoder = MCSDecoderNR()
# Scalar input
mod_order, rate = decoder(mcs_index=14, mcs_table_index=1, mcs_category=0)
print(f"Modulation order: {mod_order.item()}, Target rate: {rate.item():.3f}")

# Tensor input
mcs_indices = torch.tensor([10, 14, 20])
mod_orders, rates = decoder(mcs_index=mcs_indices,
                            mcs_table_index=1, mcs_category=0)