get_exit_analytic#

sionna.phy.fec.utils.get_exit_analytic(pcm: numpy.ndarray, ebno_db: float) → Tuple[numpy.ndarray, numpy.ndarray, numpy.ndarray][source]#

Calculates analytic EXIT curves for a given parity-check matrix.

This function extracts the degree profile from the provided parity-check matrix pcm and calculates the EXIT (Extrinsic Information Transfer) curves for variable nodes (VN) and check nodes (CN) decoders. Note that this approach relies on asymptotic analysis, which requires a sufficiently large codeword length for accurate predictions.

It assumes transmission over an AWGN channel with BPSK modulation at an SNR specified by ebno_db. For more details on the equations, see [tenBrink] and [tenBrinkEXIT].

Parameters:
Outputs:

  • mi_a – The a priori mutual information.

  • mi_ev – The extrinsic mutual information of the variable node decoder.

  • mi_ec – The extrinsic mutual information of the check node decoder.

Notes

This function assumes random, unstructured parity-check matrices. Thus, applying it to parity-check matrices with specific structures or constraints may result in inaccurate EXIT predictions. Additionally, this function is based on asymptotic properties and performs best with large parity-check matrices. For more information, refer to [tenBrink].

On this page

This Page