A brief summary of the performance limitations of fiber optic gyroscopes interrogated with a broadband light source is presented, followed by a description of the status of current research efforts to push back these limitations by using an air-core fiber instead of a conventional fiber in the sensing coil, and/or coherent instead of incoherent light.
Current status of the fiber optic gyroscopeFirst demonstrated shortly after the advent of single-mode fibers in the mid-1970s, the fiber-optic gyroscope (FOG) is one of the oldest and most successful fiber sensors. Commercialized since around 1985, it is widely used for applications as varied as ship navigation, stabilization, and north finding. This success is rooted in part in the fact that the FOG's sensing element is a Sagnac loop, a two-wave common-path interferometer that is inherently reciprocal and thus highly stable against most external perturbations. Equally instrumental is the use of broadband (incoherent) light to interrogate it. Incoherent light was shown very early on that it essentially eliminates the three detrimental non-reciprocal effects that take place in the fiber loop and contribute to most of the noise and/or drift in the output of the FOG, namely the intensity-dependent nonlinear Kerr effect, coherent backscattering, and polarization coupling. A fourth limitation is the Shupe effect. When the temperature of the sensing coil is subjected to a transient temperature gradient, the two signals counter-propagating in the loop experience slightly different time-dependent phase shifts, and the output drifts. It was also critical to the success of the FOG that the Shupe effect was effectively mitigated by special coil windings (e.g., quadrupolar), as well as other solutions such as modeling.Unfortunately, the adoption of a broadband source introduced two limitations. First, broadband sources such as Er-doped superfluorescent fiber sources (SFS) suffer from excess noise that typically far exceeds shot noise, which limits the FOG's minimum detectable rotation rate. Second, the FOG scale factor, which relates the measured output signal to the rotation rate, must be extremely stable (~1 ppm) for aircraft inertial navigation, the holy grail of gyroscope applications. This scale factor being inversely proportional to the mean wavelength of the light, the latter must have a comparable stability, which is difficult to achieve with broadband light. In addition, even after applying the aforementioned corrective measures the drift due to the Shupe effect is still difficult to control in practice. These hurdles are compounded by the need to keep the cost, size, and energy consumption of the FOG low enough to be competitive with the ring laser gyroscope, which is accurate and stable enough to navigate aircrafts.
Air-core fiber optic gyroscopeIn the last decade two new directions have been investigated to push back these fundamental limitations. The first one is replacing the conventional (solid-core) single-mode fiber used in the sensing loop by a photonic-bandga...