光通信
信号处理
计算机科学
物理
分解
光学
电子工程
信噪比(成像)
光学滤波器
偏振模色散
干扰(通信)
集成光学
算法
数学
光放大器
光纤
相位噪声
波分复用
噪音(视频)
估计理论
光调制幅度
物理光学
传输(电信)
相位调制
超快光学
随机过程
统计物理学
光学性能监测
非线性光学
干涉测量
激光器
调制(音乐)
作者
Huazhi Lun,Lei Shi,Xingyu Wang,Huicun Yu,Jie Tang
标识
DOI:10.1109/jlt.2026.3721504
摘要
Machine learning (ML) has shown strong performance for Quality of Transmission (QoT) estimation in optical networks, but reliability under out-of-distribution (OOD) conditions remains a key operational concern. We propose an uncertainty-aware QoT framework that decomposes predictive uncertainty into aleatoric and epistemic components using a heteroscedastic neural predictor with a post-hoc Laplace approximation. The Laplace epistemic-uncertainty step can be fitted offline around a trained heteroscedastic predictor, while online inference remains deterministic. Across multiple OOD scenarios, the decomposition reveals complementary roles: aleatoric uncertainty mainly tracks modeled noise intensity, while epistemic uncertainty reflects model unfamiliarity under distribution shift. The empirical results indicate that uncertainty behavior differs clearly between familiar and unfamiliar regimes, which has direct implications for safe network operation. These support the conceptual two-regime operational policy: use aleatoric uncertainty for adaptive margining in trusted ML regions, and trigger fallback to a physics-based baseline when epistemic uncertainty is high. Overall, the results demonstrate the practical value of uncertainty decomposition for safer QoT estimation.
科研通智能强力驱动
Strongly Powered by AbleSci AI