Skip to content

Very-Fancy/ldpc-quant

Repository files navigation

ldpc-quant

Quantization for LDPC-decoder node messages

Estimation of performance of data transmission system with different parameters:

  • unquantized AWGN-channel and unquantized Min-Sum (MS) decoder

  • uniformly quantized AWGN-channel and discrete MS (6 bits)

  • non-uniformly quantized AWGN-channel and discrete MS (4 bits)

Running the simulation:

simulate(2, 0,'codes/LDPC_5G_K120_R075.alist', 20,[1 1.5 2 2.5 3 3.5 4 4.5 5], 6, 4, 10)

Parameters:

  • M = 2 for BPSK

  • thrf = 0 for calculting the new decoding treshold and quantization bounds (or provide the location of file with computed decoding threshold)

  • ldpc_filename - alist file location

  • snr_array - snr values to simulate

  • q_uniform - number of bits for uniform quantization

  • q_nonuniform - number of bits for non-uniform quatization (< q_uniform)

  • fe - number of frames with error (the same for unquantized, uniform and non-uniform variants)

About

Quantization for LDPC-decoder node messages

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published