Speaker
Description
Aerial optical fiber cables in telecommunications access networks are highly susceptible to damage from environmental factors such as strong winds and tree contact. Conventional fault localization relies on optical time-domain reflectometry to identify approximate fault sections, followed by visible light injection to pinpoint exact breakpoints through visual observation of light leakage. However, this approach fails when fibers break internally due to bending stress or lateral pressure while the external sheath remains intact, which is a frequently encountered scenario in field maintenance. This limitation has motivated the exploration of alternative diagnostic techniques capable of detecting internal faults non-destructively.
Infrared thermography offers a promising approach by detecting heat generated at fiber breakpoints. When optical signals propagate through a fiber with an internal break, light leaks from the damaged core and converts into thermal energy, producing localized heating at the cable surface. While our previous work utilizing steady-state thermography successfully demonstrated breakpoint localization without damaging the cable sheath, it required optical input power exceeding +3 dBm. This power level poses a critical practical limitation: since typical communication signals operate at 0 dBm or below, accidental injection into an operational fiber at such elevated power levels risks damaging active communication equipment. Reducing the required optical input power is therefore essential for safe field deployment.
As a solution, we propose applying lock-in thermography to this fiber fault diagnosis application to overcome the power limitation inherent in steady-state measurements. By injecting intensity-modulated laser light into the fiber, periodic heating occurs at the breakpoint, enabling high-sensitivity detection through synchronous thermal imaging. This approach extends the application of lock-in thermography to optical fiber infrastructure maintenance.
Our experimental setup utilized a 1550-nm Fabry-Perot laser diode modulated at frequencies of 0.1, 0.5, and 0.9 Hz, with maximum optical input power varied from –8 to 0 dBm. Drop cable samples (cross-section: 2.25 × 1.59 mm) with intentionally broken internal fibers and intact sheaths were imaged at 30 Hz for up to 120 seconds followed by temporal Fast Fourier Transform (FFT) processing for each pixel to extract the modulation frequency component.
While steady-state thermal imaging showed no discernible temperature increase even at 0 dBm input power, lock-in detection successfully revealed clear amplitude peaks at breakpoint locations. The lowest modulation frequency tested (0.1 Hz) proved optimal for this application, providing larger amplitude signals and wider detection range, as spatial resolution is less critical than detectability for fault localization. Extended measurement time progressively improved the signal-to-noise ratio.
Quantitative evaluation using region-of-interest analysis demonstrated that with a 60-second measurement time at 0.1-Hz modulation, breakpoints could be clearly identified at –2 dBm optical input power. This represents a reduction exceeding 5 dB compared to steady-state thermography requirements, bringing the detection threshold close to normal communication signal levels and substantially mitigating equipment damage risks during field diagnostics. These results demonstrate that lock-in thermography can be effectively applied to non-destructive optical fiber fault diagnosis in telecommunications infrastructure.
