WDM-RoF over C+L Bands: Procedural Simulation of an 8-Channel 5G Fronthaul

A compact, open-source, procedural simulator for an 8-channel Wavelength Division Multiplexed Radio-over-Fiber (WDM-RoF) link spanning the C and L optical bands, built for 5G fronthaul/backhaul link studies. Optical channel impairments (dispersion, noise, quality factor) are generated procedurally through analytical functions rather than pre-calibrated datasets, making the model fast, reproducible, and dependency-light compared to commercial tools like OptiSystem or VPItransmissionMaker.

This repository accompanies the paper:

WDM–RoF over C+L Bands: Procedural Simulation of an 8-Channel 5G Fronthaul at 2.5/5/10 Gbps Vignesh Naidu, Aditya Verma, Tharun Ravuru Department of Communication Engineering, Vellore Institute of Technology, Vellore, India

Overview

The simulator models an 8-channel WDM-RoF link:

• 4 C-band wavelengths: 1550, 1551, 1552, 1553 nm
• 4 L-band wavelengths: 1570, 1570.5, 1571, 1571.5 nm
• Line rates: 5, 10, 20 Gbps (see Note on parameters)
• Fiber spans: 30–70 km in 10 km steps
• Modulation: NRZ-coded PRBS on an RF subcarrier upconverted onto an optical carrier
• Channel model: length-dependent Gaussian dispersion + AWGN
• Receiver: envelope detection, low-pass filtering, sampling, BER/Q extraction

For each (bit rate, fiber length, wavelength) combination, the simulator computes BER, Q-factor, and SNR; the extended version additionally computes Shannon channel capacity and FEC/framing-adjusted throughput.

Results consistently show that C-band channels maintain lower BER and higher SNR than L-band channels at the same distance, with near-error-free C-band operation up to 60 km at 10 Gbps, consistent with trends reported in the literature (see the paper's references).

Repository structure

wdm-rof-5g-fronthaul/
├── src/
│   ├── wdm_rof_basic_sim.py       # BER / Q-factor / SNR simulation + plots
│   └── wdm_rof_extended_sim.py    # Adds channel capacity & throughput analysis
├── results/                       # CSV results and generated plots (git-ignored)
├── requirements.txt
├── LICENSE
└── README.md

Getting started

Requirements

• Python 3.9+

pip install -r requirements.txt

Run the simulation

Basic version (BER, Q-factor, SNR + BER-vs-distance plot and eye diagrams):

python src/wdm_rof_basic_sim.py

Extended version (adds channel capacity and throughput analysis):

python src/wdm_rof_extended_sim.py

Each script:

1. Sweeps all combinations of bit rate × fiber length × wavelength.
2. Writes a results table to results/wdm_rof_5g_results_*.csv.
3. Saves plots (BER, SNR, capacity, throughput, eye diagrams) to results/.

Mathematical model (as reported in the paper)

• Chromatic dispersion, approximated as a Gaussian frequency-domain filter with dispersion parameter D = 16 ps/nm·km.
• Target quality factor: Q_tgt(ℓ, λ) = Q0 · exp(-αℓ) · (1 - βλ)
• BER estimate: BER = 0.5 · erfc(Q_tgt / √2)
• OSNR: approximated as SNR - 4 dB
• Channel capacity: C = B · log2(1 + SNR)
• Throughput: T = min(C, B) · (1 - BER) · η, with adaptive FEC/framing efficiency η ∈ [0.6, 0.92] depending on the BER regime.

The Python implementation in this repo realizes this model procedurally: it generates a PRBS sequence, applies NRZ encoding and RF upconversion, passes the signal through a Gaussian dispersion filter sized by fiber length, injects AWGN, then performs envelope detection and low-pass filtering at the receiver to extract BER, Q-factor, and SNR directly from the simulated waveform.

Results at a glance

Bitrate (Gbps)	Fiber (km)	Channel	λ (nm)	Q-factor	BER	SNR (dB)
10	30	Ch-1	1550.0	10.67	7.19e-27	23.10
10	60	Ch-1	1550.0	—	—	—
10	30	Ch-5	1570.0	6.47	4.77e-11	18.88
10	70	Ch-5	1570.0	—	4.19e-02	7.51

Full results are in the paper's Table II; regenerate the complete sweep by running the scripts above.

Note on parameters

The paper's abstract and Table I specify bit rates of 2.5 / 5 / 10 Gbps; the simulation scripts in this repo currently sweep 5 / 10 / 20 Gbps. Update the bitrates list at the top of each script if you want to reproduce the exact rates quoted in the paper.

Future work

• Nonlinear fiber effects (SPM, XPM, FWM)
• Hybrid EDFA–Raman amplification
• Machine-learning-based adaptive FEC and impairment prediction

Citation

If you use this code, please cite:

@inproceedings{naidu2026wdmrof,
  title     = {WDM--RoF over C+L Bands: Procedural Simulation of an 8-Channel 5G Fronthaul at 2.5/5/10 Gbps},
  author    = {Naidu, Vignesh and Verma, Aditya and Ravuru, Tharun},
  booktitle = {Optical Fiber Communications Submission},
  organization = {Department of Communication Engineering, Vellore Institute of Technology},
  year      = {2026}
}

Acknowledgment

The authors thank the faculty and peers of VIT Vellore for their guidance and feedback during this research.

License

MIT — see LICENSE.