feat: Add BER/FER benchmark for nrLDPCDecode

[AlonNaor22]

Summary

• Full encode-modulate-noise-demodulate-decode pipeline benchmark for nrLDPCDecode
• Generates random data, LDPC encodes, QPSK modulates, adds AWGN noise, QPSK demodulates, LDPC decodes, then measures BER and FER
• Pytest verifies zero errors at high SNR (5 dB Eb/N0)
• Standalone __main__ sweeps Eb/N0 from -1 to 5 dB and prints a BER/FER table showing the expected LDPC waterfall curve

Example output

LDPC BER/FER Benchmark
  Base graph: 2, K=2560, F=36, N=12800, rate=0.200
  Modulation: QPSK, 10 frames/point

  Eb/N0 (dB)          BER          FER
----------------------------------------
         0.0     0.370721       1.0000
         1.0     0.340927       1.0000
         2.0     0.002179       0.2000
         3.0     0.000000       0.0000
         5.0     0.000000       0.0000

Files changed

• tests/test_nrLDPCDecode_BER.py (new)

Test plan

• pytest tests/test_nrLDPCDecode_BER.py passes (1 passed in 4.6s)
• Standalone sweep produces expected LDPC waterfall behavior

Closes #50

🤖 Generated with Claude Code

[catkira]

nice work, can u also generate a BER plot put it in a docs folder and reference it in README.md?

[AlonNaor22]

Sure i will get on it later today בתאריך יום ו׳, 6 בפבר׳ 2026, 5:48 PM, מאת Benjamin Menküc ‏< ***@***.***>:

…

*catkira* left a comment (catkira/py3gpp#77) <#77 (comment)> nice work, can u also generate a BER plot put it in a docs folder and reference it in README.md? — Reply to this email directly, view it on GitHub <#77 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/A6CXKSGCZJKLS5N4BVKHUJL4KSZVVAVCNFSM6AAAAACUHXFRGKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZTQNRRGE2DSMRZGU> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[catkira]

or ur agent :)

[AlonNaor22]

i mean in this day and age is feels wrong to ignore this

[catkira]

I agree, I use it a lot myself