We experimentally assess the accuracy of a constant output power generalized droop (COP-GD) model combined with transponder implementation penalty to predict transmission performance for single-core long-haul optical transmission systems operating at low channel launch powers and high baud rates. The droop model includes Guided Acoustic Waveguide Brillouin Scattering (GAWBS) and nonlinear interference (NLI) effects along with amplifier noise for signal-to-noise ratio (SNR) calculation in the context of long-haul systems with constant output power amplifiers. We show how the model accuracy is decreased if one of the modeling components is removed, namely, either the droop effect, GAWBS or the transponder implementation penalty. We show that inclusion of implementation penalty and GD is important for accurate performance evaluation, moreover, the transponder implementation penalty has impact both at low and high SNR ends. We compare the model including both effects to experimental transmission data over a wide link SNR range for the first time and demonstrate excellent agreement between modeling predictions and experimental results.

中文翻译：

---

使用 SNR 测量（包括应答器实施惩罚）对广义信号下降模型进行实验评估


我们通过实验评估恒定输出功率广义下垂（COP-GD）模型与转发器实现惩罚相结合的准确性，以预测在低通道发射功率和高波特率下运行的单核长途光传输系统的传输性能。下垂模型包括导声波导布里渊散射 (GAWBS) 和非线性干扰 (NLI) 效应以及放大器噪声，用于在具有恒定输出功率放大器的长距离系统中计算信噪比 (SNR)。我们展示了如果移除其中一个建模组件（即下垂效应、GAWBS 或应答器实现惩罚），模型精度将如何降低。我们表明，包含实施惩罚和 GD 对于准确的性能评估非常重要，此外，转发器实施惩罚对低 SNR 端和高 SNR 端都有影响。我们首次将包括这两种效应在内的模型与宽链路 SNR 范围内的实验传输数据进行比较，并证明建模预测与实验结果之间具有良好的一致性。