Sensitivity enhancement of distributed polarization coupling detection in Hi-Bi fibers

Abstract

The intensity and position of the coupling points in high birefringence (Hi-Bi) fibers can be detected effectively using distributed polarization coupling (DPC) detection. The detection sensitivity can decrease due to mechanical vibration disturbance and environment noise. Thus, a method based on empirical mode decomposition is proposed to detect weak mode coupling points. The simulation and experimental results illustrate that the proposed method can suppress the noise effectively and improve sensitivity significantly. The method can identify coupling points as weak as -60 dB embedded in noise automatically and effectively. The algorithm is applicable for DPC, and the experimental sensitivity is improved by 10 dB.OCIS codes: 060.2420060. , 070.2025060. , 060.2300 High birefringence (Hi-Bi) fibers are used in many areas related to fiber optics, including coherent optical communications, integrated-optic devices, and optical sensors based on interferometric techniques. Distributed polarization coupling (DPC) detection employs white-light interferometry (WLI) in Hi-Bi fibers, and is widely used in the measurements of strain, twist, temperature, and many other physical parameters [1−3] . Due to high spatial resolution and wide dynamic range brought about by the adoption of WLI [4] , the intensity and position of the coupling points can be detected effectively.Mechanical vibration disturbance and environment noise can lead to the degradation of detection sensitivity of DPC [5,6] . Therefore, the weak coupling points are submerged in noise, leading to a wrong diagnosis. To date, phase modulation, differential signal detection, or rotation angle optimization can be employed for weak coupling measurement [7−9] ; however, the complexity of the required system has also increased. In order to increase the detection sensitivity, some signal analysis methods, such as band-pass filtering [10] and the wavelet transforms [11] , have been reported for noise reduction. The main drawback, however, is that the base functions are fixed, and no such function has been proposed to correspond to the features of the acquired signals. In comparison, the empirical mode decomposition (EMD) is a highly efficient technique for processing nonlinear and non-stationary signals, because the procedure is datadriven, adaptive, and not restricted by linearity or priori conception [12] . Gong et al. applied EMD and the threshold method to reduce the noise in the lidar signals [13] . Deng et al. used EMD and local entropy to detect small targets [14] . Here EMD combined with DPC is proposed to enhance the detection sensitivity and realize the distributed weak mode coupling measurement.The scheme of DPC detection is shown in Fig. 1. A superluminescent diode (SLD) emitting at 1 328 nm was used as the light source. An in-line polarizer was fusion spliced in front of the Hi-Bi fiber. Its polarization orientation was aligned to the slow axis of the polarization maintaining fiber (PMF), so that only the slow axis was excited. The output light from the fi...