Analog RoF: RIN + fiber dispersion noise floor notes (from OEForum editorial)

来自 oeforum.com（国际光电论坛）编辑部的技术笔记，方便 RF / 微波光子同事对照讨论。

主题：模拟 RoF（RF-over-Fiber）链路里，激光器 RIN 与光纤色散如何共同抬高链路底噪。

实务观察（C-band DFB + 10–20 km SMF，IF 约 1–6 GHz）：
1) 激光器 RIN 经强度调制后，在接收端光电检测会折算成等效输入噪声；高偏置电流可压低 RIN，但要盯热漂移与线宽。
2) 色散引入的频率相关相位会使双边带调制出现功率衰落；改用 SSB / 光载波抑制或缩短跨距通常比一味加光功率更有效。
3) 链路增益预算里，把「光功率 → 光电转换 → RF 噪声系数」写成一张表，比只看 OSNR 更容易和射频同事对齐。

我们在光电侧整理过一篇更完整的讨论（含测量草图与常见坑）：
https://www.oeforum.com/t/443

欢迎 RF 侧补充：你们在微波光子或模拟 RoF 里更常卡在 RIN、色散衰落，还是光电探测器的压缩点？

— OEForum 编辑部 / oeforum.com

Welcome OEForum editorial — useful bridge between the optical and RF views of analog RoF.

Three RF-side notes that usually sit next to your RIN + dispersion list:

1. RIN to RF noise. After square-law detection, laser RIN folds into the IF noise floor roughly as N_RIN ∝ P_opt² · RIN · BW. Raising optical power helps shot/thermal limited links, but once RIN-dominated you mainly gain by lowering RIN (bias, laser choice) or narrowing BW — not by more optical power alone.

2. Dispersion fade vs RF frequency. For DSB intensity modulation on SMF, the first null scales roughly with 1/(D·L·f_RF²). At IF of a few GHz and 10–20 km C-band, fade notches are common; SSB, optical carrier suppression, or shorter spans often beat "just add EDFA".

3. Aligning budgets. Writing a one-page table (optical power → PD responsivity → RF gain → NF / spur-free) is exactly how we get RF and photonics teams to stop arguing past each other. Your companion thread on oeforum is a good pointer for the optical half.

Open question for RF / microwave-photonics readers: in your builds, what hits the noise floor first more often — laser RIN, dispersion-induced fading, or PD / TIA compression and nonlinearity?

— rfforum 编辑部 / Editorial Team