光学
物理
光纤
图像处理
信号处理
计算机科学
材料科学
光时域反射计
折射率
分束器
光纤激光器
作者
Yi Qin,Ruijiang Wang,Jishen Hou,Guijun Hu
摘要
Spatial resolution is a key performance metric of optical time-domain reflectometry (OTDR), but it is typically constrained by hardware factors such as the launched pulse width and receiver bandwidth. To overcome this limitation, this work proposes a model-discrimination-based super-resolution method for OTDR, which reformulates the conventional hardware-limited spatial-resolution problem as one of model selection and parameter inversion in single-fault and two-fault scenarios. Specifically, theoretical backscattering models are established for single-fault and adjacent two-fault cases, and the Bayesian information criterion (BIC) is employed to select the model that best fits the observed signal. When the two-fault model is preferred, the fault spacing, namely, the spatial resolution, is further retrieved through inversion. On this basis, a quantitative relationship is established between the spatial-resolution limit and system parameters such as pulse width, bandwidth, and noise. Furthermore, a BIC-based dense multi-fault region detection mechanism is developed to assess the adequacy of the two-fault model in explaining the observed signal, thereby preventing multi-fault cases from being misclassified as two-fault cases. Numerical simulations show that, for the typical parameter settings considered in this work, the adopted engineering criterion for stable two-fault discrimination corresponds to a spatial resolution of approximately 1/25 of the conventional pulse-width-limited resolution. Experimental results demonstrate that, under the tested system configuration, a spatial resolution of 0.5 m can be achieved over a 10 km fiber link.
科研通智能强力驱动
Strongly Powered by AbleSci AI