The effects of optical sources with different laser linewidths on Brillouin optical time domain reflectometry (BOTDR) are investigated numerically and experimentally. Simulation results show that the spectral linewidth of spontaneous Brillouin scattering remains almost constant when the laser linewidth is less than 1 MHz at the same pulse width; otherwise, it increases sharply. A comparison between a fiber laser (FL) with 4-kHz linewidth at 3 dB and a distributed feedback (DFB) laser with 3-MHz linewidth is made experimentally. When a constant laser power is launched into the sensing fiber, the fitting linewidths of the beat signals (backscattered Brillouin light and local oscillator (LO)) is about 5 MHz wider for the DFB laser than for the FL and the intensity of the beat signal is about a half. Furthermore, the frequency fluctuation in the long sensing fiber is lower for the FL source, yielding about 2 MHz less than that of the DFB laser, indicating higher temperature/strain resolution. The experimental results are in good agreement with the numerical simulations.

Polarization influences on the performance of multi-wavelength Brillouin Er³⁺-doped fiber laser are investigated by adjusting the polarization controller (PC) in the fiber loop mirror (FLM), where the linear laser cavity is composed of a fiber-tailed mirror and an FLM, and the stimulated Brillouin scattering (SBS) and the Er³⁺-doped fiber amplification (EDFA) simultaneously serve as the cavity gain. We realized 1–7 Brillouin laser lines by increasing the 980 nm pump power. For the first-order Brillouin laser, the signal–noise ratio (SNR) and optical intensity present a sinusoidal envelope; the conversion efficiency changes significantly from 0.56465 dBm/mw to 0.44975 dBm/mw by adjusting the ring’s angle in the PC; the first-order SBS thresholds are 20.4 mw, 36.1 mw and 28.5 mw at different angles θ₂ = 36°, 276° and 300°, respectively; flatness between the two Brillouin lasers change obviously from 2.863 dB to 41.801 dB with different ring angles; the second-order Brillouin laser is suppressed and disappears finally at Δθ₂ = −64° to −84° and 106°~136° angle variation. For the fifth-order Brillouin laser, the highest-order Brillouin laser line is seriously suppressed until it disappears at some angle variations similarly. The powers and wavelength stabilities for one-, three- and seven-wavelength Brillouin fiber lasers were measured in 1 h, and the obtained Er³⁺-doped multi-wavelength Brillouin fiber laser (MWBFL) worked stably during that time, but the stabilities become worse with higher SBS orders.

Accurate measurement of optical fiber dispersion is important for optical processing, such as optical communication, optical fiber-based quantum state generation, etc. In this paper, we propose a new method for measuring optical fiber dispersion that can accurately measure the dispersion of long optical fibers in a wide spectral range. By using a femtosecond laser and unbalanced Mach–Zehnder interferometer with a phase-locking system, the coherence length of the optical source and stability of dispersion measurement have been improved. Thus, we overcome the shortcomings of traditional dispersion measurement methods that are difficult to make compatible with long fibers and wide spectral ranges. Using this method, we measure the dispersion parameters of two pieces of dispersion shifted fiber (DSF) with lengths of 200 m and 300 m, respectively, in the wavelength range 1500–1600 nm. Measurement results show that the zero-dispersion wavelength (ZDW) and the zero-dispersion slope (ZDS) of the 200 m DSF are 1535.53 ± 0.05 nm and 0.034 ± 10⁻⁴ ps nm^–2 km⁻¹, respectively, while the ZDW and ZDS of the 300 m DSF are 1551.26 ± 0.06 nm and 0.040 ± 10⁻⁵ ps nm^–2 km⁻¹, respectively. We test the stability of the system by using the root-mean-square (RMS) fluctuations of six measurements of the dispersion parameter D(λ) for the two pieces of fiber; the RMS values are both less than 0.02 ps nm^–1 km⁻¹. Our research is valuable for measurement of dispersion for long fibers within a wide spectral range, considering the effects of fiber uniformity.

Dual tapers based on a graded-index (GI) two-mode (TM) fiber are proposed for refractive index (RI) sensing in this work. Sensitivity of 1433.94 nm/RIU with R ² =0.98 is achieved using the direct peak-finding method, where the sensitivity enhancement originated from two guided modes and corresponding two sets of excited cladding modes. To improve wavelength demodulation accuracy, spatial frequency concentration algorithm is presented to filter out the influence of optical intensity fluctuations and mode noise via a bandpass filter, and yielding the sensing sensitivity 1669.61 nm/RIU and R ² =0.9699 after data processing. Furthermore, empirical mode decomposition (EMD) is utilized to decompose transmission spectra according to their noise features. By calculating correlation coefficients between each intrinsic mode function and the raw spectrum, noise components are eliminated to smooth the spectra. The EMD-processed spectra produce the sensing sensitivity of 1458.52 nm/RIU with R ² =0.9930 through peak-finding again. Experimental results demonstrate that the GI two-mode fiber enhances the sensing sensitivity of tapered fiber sensors, while the EMD algorithm improves the sensing accuracy. The results in this work could provide technical support for the practical sensing applications of tapered fiber sensors.