Optical fiber is often viewed merely as the modern plumbing powering our connected world, replacing older twisted pair copper connections in data transmission networks. However, due to some unique physical characteristics, fiber optics can and are targeted at a range of distributed sensing applications. Applications that benefit from fiber’s economy combined with an inherent ability to make localized measurements over considerable distances. Metrology modalities divide into three primary sensing approaches:

• Distributed Temperature Sensing (DTS)
• Distributed Acoustic Sensing (DAS)
• Distributed Strain Sensing (DSS)

Fiber optic cables act as waveguides that allow concentrated light (provided by either a laser, a VCSEL or LED source) to travel with modest attenuation, long distances. For carrying signals, fibers benefit from being physically small and lightweight. Moreover, they are immune to electromagnetic interference ​

Science shows that impurities arising in fibers during their manufacture causes light traversing their length to experience low levels of scattering. Three scattering types (figure 1) are described by their physical manifestation and the methods needed to detect them:

• Rayleigh, Brillouin and Raman scattering​
  

Photon scattering is a random probabilistic process. Scattering may be elastic (as in Rayleigh scatter) implying scattered photons maintain the same wavelength. Inelastic Brillouin and Raman scatter sees scattered photons experiencing sizable wavelength shifts coupled to an energy level transition at the atomic level. The wavelength change is denoted either as Stokes or anti-Stokes in the case of decreased wavelength. ​

The basic operation of distributed fiber sensing (fig. 2) requires illuminating a length of optical fiber by repetitive pulses of coherent light. The equipment used is referred to as an interrogator (fig. 3). Each pulse is subject to a transmission delay as it works its way along the fiber. At individual positions throughout the fiber a highly attenuated backscatter signal arises and will return to the source in a known time (propagation delay) determined by its time-of-flight. By sampling the return path signal with a suitable light coupled digitizing system, local physical characteristics including temperature, mechanical strain, and even acoustic energy can be detected - delivering a host of economic metrology applications. ​

Beyond economic considerations, what are the critical, top-level performance factors on which DFOS systems are judged? Two factors stand out:

• Range
• Resolution

Range describes the maximum length of fiber over which the measurement system remains capable of delivering useful metrology – essentially the sensitivity. The boundary conditions for detection range are related to the dynamic range of the specific measurement approach used, the fiber attenuation constant as well as the probe pulse width and illumination signal power. Such diversity of variables leads to significant differences in implementations and capabilities.

Resolution determines how closely spaced each measurement point along the fiber can be. For a given fiber length this is wholly determined by the probe pulse width. Shorter pulses lead to finer spatial resolution at the cost of reduced photon energy and hence dynamic range; attaining an optimal pulse width for a certain application becomes a question of engineering tradeoffs.

Some DFOS systems (e.g. Brillouin optical time domain reflectometry - BOTDR) are specifically designed to maximize detection range up to and beyond 100 km. Equally, some industrial applications may only need a few tens of meters range, but centimeter precise spatial resolution. ​

Closely related to traditional OTDR using Rayleigh scatter is phase-sensitive OTDR (or Φ-OTDR) which uses a fine-line, stable frequency laser source (usually of a 1550 nm wavelength) which enables a class of distributed detection known as DAS. This system works by measuring interference intensity spikes emerging from the pulse illuminated FUT (fiber under test).

In DAS (fig. 4), the laser source light (ω0) is split into two optical branches. One path generates a pulse modulated probe signal using an acousto-optic modulator (AOM) driven by an RF generator. Each pulse of photon energy will range typically between 10 to 100 ns in duration. Pulse duration determines the minimum spatial resolution of the probe signal. Moreover, the AOM adds a small frequency shift (Δω) to light exiting the AOM (i.e. ω0 + Δω). The resultant signal is amplified before being launched into the sensing fiber under test (FUT) via an optical circulator.

Time dependent Rayleigh backscattered light emerging from the FUT is mixed with a reference signal provided by the second branch before being directed to a balanced heterodyning receiver where a pass band filter extracts backscatter phase information (fig. 2). ​

The most important aspect of DFOS applications is to understand that:

• Back scatter, irrespective of the specific type to be detected, comprises low-level signals demanding high system dynamic range to detect (ADC resolution > 10-bit).
• The short light pulses (typically between 5 ns & 20 µs) used for fiber metrology mean high-speed digitizers are essential for their effective capture. Thus, sample rates in the order of a few giga-samples may be necessary.
• The selection of a suitable measurement pulse width is determined by considering the fiber length and assessing the pulse flight time.

In addition, DFOS system designers need to pay attention to the design of suitable optical front end pulse generation and detection circuits - complementing the high-end performance offered by a selected digitizer.

Beneficially for Teledyne SP Devices' digitizers, spare open FPGA resources enable a master control system to be built able to manage the pulse data acquisition & characterization process as well as supplying pulsed laser control. ​