Aerial CUDA-Accelerated RAN 24-1 Download PDF

This module contains classes related to the Physical Uplink Shared Channel, PUSCH.

class aerial.phy5g.pusch.pusch_rx.PuschRx

PUSCH receiver pipeline.

This class implements the whole PUSCH reception pipeline from the received OFDM post-FFT symbols to the received transport block (along with CRC check).

__init__(cell_id, num_rx_ant, num_tx_ant, num_ul_bwp=273, num_dl_bwp=273, mu=1, enable_cfo_correction=0, enable_to_estimation=0, enable_pusch_tdi=0, eq_coeff_algo=1, ldpc_kernel_launch=aerial.pycuphy.PuschLdpcKernelLaunch.PUSCH_RX_ENABLE_DRIVER_LDPC_LAUNCH)

Initialize PuschRx.

  • cell_id (int) – Physical cell ID.

  • num_rx_ant (int) – Number of receive antennas.

  • num_tx_ant (int) – Number of transmit antennas.

  • num_ul_bwp (int) – Number of PRBs in a uplink bandwidth part. Default: 273.

  • num_dl_bwp (int) – Number of PRBs in a downlink bandwidth part. Default: 273.

  • mu (int) – Numerology. Values in [0, 3]. Default: 1.

  • enable_cfo_correction (int) –

    Enable/disable CFO correction:

    • 0: Disable (default).

    • 1: Enable.

  • enable_to_estimation (int) –

    Enable/disable time offset estimation:

    • 0: Disable (default).

    • 1: Enable.

  • enable_pusch_tdi (int) –

    Time domain interpolation on PUSCH.

    • 0: Disable (default).

    • 1: Enable.

  • eq_coeff_algo (int) –

    Algorithm for equalizer coefficient computation.

    • 0 - ZF.

    • 1 - MMSE (default).

    • 2 - MMSE-IRC.

  • ldpc_kernel_launch (PuschLdpcKernelLaunch) – LDPC kernel launch method.

Return type


run(rx_slot, num_ues, slot=0, num_dmrs_cdm_grps_no_data=2, dmrs_scrm_id=41, start_prb=0, num_prbs=273, dmrs_syms=None, dmrs_max_len=2, dmrs_add_ln_pos=1, start_sym=2, num_symbols=12, scids=None, layers=None, dmrs_ports=None, rntis=None, data_scids=None, mcs_tables=None, mcs_indices=None, code_rates=None, mod_orders=None, tb_sizes=None, rvs=None, ndis=None)


This runs the cuPHY PUSCH receiver pipeline for a single UE group sharing the same time-frequency resources, i.e. having the same PRB allocation, and the same start symbol and number of allocated symbols. Default values get filled for the parameters that are not given.

  • rx_slot (numpy.ndarray) – A tensor representing the receive slot buffer of the cell.

  • num_ues (int) – Number of UEs in the UE group.

  • slot (int) – Slot number.

  • num_dmrs_cdm_grps_no_data (int) – Number of DMRS CDM groups without data [3GPP TS 38.212, sec]. Value: 1->3.

  • dmrs_scrm_id (int) – DMRS scrambling ID.

  • start_prb (int) – Start PRB index of the UE group allocation.

  • num_prbs (int) – Number of allocated PRBs for the UE group.

  • dmrs_syms (List[int]) – For the UE group, a list of binary numbers each indicating whether the corresponding symbol is a DMRS symbol.

  • dmrs_max_len (int) – The maxLength parameter, value 1 or 2, meaning that DMRS are single-symbol DMRS or single- or double-symbol DMRS.

  • dmrs_add_ln_pos (int) – Number of additional DMRS positions.

  • start_sym (int) – Start OFDM symbol index for the UE group allocation.

  • num_symbols (int) – Number of symbols in the UE group allocation.

  • scids (List[int]) – DMRS sequence initialization for each UE [TS38.211, sec].

  • layers (List[int]) – Number of layers for each UE.

  • dmrs_ports (List[int]) – DMRS ports for each UE. The format of each entry is in the SCF FAPI format as follows: A bitmap (mask) starting from the LSB where each bit indicates whether the corresponding DMRS port index is used.

  • rntis (List[int]) –

  • data_scids (List[int]) – Data scrambling IDs for each UE, more precisely dataScramblingIdentityPdsch [TS38.211, sec].

  • mcs_tables (List[int]) – MCS table to use for each UE (see TS 38.214).

  • mcs_indices (List[int]) – MCS indices for each UE.

  • code_rates (List[float]) – Code rate for each UE. This is the number of information bits per 1024 coded bits.

  • mod_orders (List[int]) – Modulation order for each UE.

  • tb_sizes (List[int]) – TB size in bytes for each UE.

  • rvs (List[int]) – Redundancy versions for each UE.

  • ndis (List[int]) – New data indicator per UE.


A tuple containing:

  • np.ndarray: Transport block CRCs.

  • List[np.ndarray]: Transport blocks, one per UE.

Return type

np.ndarray, List[np.ndarray]

Previous PDSCH
Next LDPC 5G
© Copyright 2024, NVIDIA. Last updated on Jun 6, 2024.