What do snowflakes, fingerprints, and optical fiber have in common? Each has an individually unique pattern that sets it apart from all other individual units. As the pattern of each snowflake and fingerprint can be seen, the pattern that underlies optical fiber can be seen using Rayleigh scatter that is back-reflected when a wave of light propagates down the core of an optical fiber. The Rayleigh scatter is generated by minute fluctuations in the density and geometry of the waveguiding core of the fiber. Modern, very high resolution reflectometry can be used to measure the Rayleigh backscatter from the fiber and thus map the imperfections that cause it. Once this map, or fingerprint, is stored, it can be used in some very interesting applications from distributed fiber-optic sensing to network security and intrusion prevention.

 

Overview of Fiber Fingerprinting

 Fiber fingerprinting can be accomplished due to imperfections in optical fiber. Each minute piece of fiber reflects back light, which is known as Rayleigh backscatter. In the past, this backscatter level has been too small to notice with standard optical equipment, but with recent technology advances, it can be seen and utilized with commercially available optical test equipment known as an Optical Backscatter Reflectometer. For more on how optical backscatter reflectometry works, check out the OBR Product page.  

Figure 1: Imperfections in the fiber lead to Rayleigh Backscatter

Using the OBR, Luna’s researchers found that the Rayleigh backscatter provides a static, repeatable pattern that can act as a “fingerprint” to the fiber, as seen in the first half of the graph in Figure 2 below. In this example, a piece of optical fiber was spliced and scanned three separate times. Each time, a small portion of fiber was cut off and then it was respliced. Notice how before the splice, the pattern of the fiber is identical and repeatable. However, after the splice, the pattern shifts slightly as portion of the fiber was removed.  

Figure 2. A length of optical fiber can be “fingerprinted” by measuring the fiber’s Rayleigh backscatter profile. Notice the repeatability of the Rayleigh signature prior to the splice.

This type of static, repeatable measurement is consistent for all optical fiber and provides the fingerprint on which we can base multiple applications.

Applications

Because of the ability to “fingerprint” optical fiber, Luna researchers have found ways to utilize this as a tool for multiple applications, including distributed temperature and strain sensing and for security within optical networks

Security

A communications infrastructure secure from threats of intrusion and espionage is a key element in the overall outlook of network security. Monitoring a fiber optic network presents a particularly difficult monitoring challenge due to the fact that fiber tapping methods can be made to be nearly undetectable. Methods of intrusion detection that involve either monitoring or conditioning the data stream work to protect a link in the presence of an intrusion event but do not provide information about the location or nature of the intrusion. Luna researchers found that using fiber fingerprints, you can monitor the network in situ for the types of changes associated with modern, hard to detect optical taps. This technique is not only capable of real-time monitoring of whether or not a fiber network has been breached by a difficult-to-detect source, is also capable determining the location and nature of the breach point in the network.

In the picture below, we show how this is implemented. Using the “fingerprint” from the original fiber as the key, we can cross-correlate it with a measurement of the same segment that had been tampered with.

A correlation operation between the “key” and the corresponding section of the fiber network measured at a later time will determine whether or not the network has been breached:

{key}⊗{tampered segment}

From that analysis, we see that the “strength” of the correlation peak determines the extent to which the fiber segment has remained intact. Breaches such as breaks are easily located by interrogating the OTDR-like picture of network integrity, and low-loss breaches or taps, such as splices, splitters and couplers are detected based on the correlation amplitude.

Figure 3. Correlation level between key and tampered segment of fiber

In the past, it has been difficult to know precisely if and where an intrusion event has occurred in an optical network. By utilizing the fiber fingerprint, you can monitor the network for the types of changes associated with modern, hard to detect optical taps, knowing precisely where it occurred.

Distributed Temperature and Strain Sensing

Not only can the fiber fingerprinting be utilized for security, but also for distributed sensing. Based on the method discussed above, Luna researchers have found that for a given fiber, the scatter amplitude of Rayleigh backscatter as a function of distance is a random but static property of that fiber and can be modeled as a long, weak Fiber Bragg Grating with a random period. Changes in the local period of the Rayleigh scatter caused by an external stimulus (like temperature or strain) in turn cause changes in the locally reflected spectrum. This spectral shift can then be calibrated to form a distributed temperature and strain sensor.

In Figure 4, we demonstrate this method. The top graph (a) shows the Rayleigh backscatter amplitude along 5 mm fiber segment for a heated (solid) and unheated (dotted) measurement scan. The middle graph, (b), shows the wavelength spectra of these segments, and then (c) shows the cross-correlation of these spectra with reference (unheated) spectrum.

Figure 4 a).The scatter amplitude along a 5 mm fiber segment for heated (solid) and unheated (dotted) measurement scan. b) Wavelength spectra of these segments. c) Cross-correlation of these spectra with reference (unheated) spectrum.

Using this cross-correlation between the reference “fingerprint” and the measurement, the fiber itself becomes a distributed sensor. Using standard telecom-grade fiber, you no longer are required to use specialty fiber or gratings for sensing, as shown in the example below which used SMF-28 fiber with the optical backscatter reflectometer.

Figure 5. Wavelength shift as a function of distance along fiber with four “hot spots” and one cold. The inset shows that high spatial resolution is possible.

Using this commercially available technology, fiber fingerprints have enabled the fiber itself to become the sensor. With up to 2 mm resolution over hundreds of meters, +/- 0.1 °C resolution at 1 cm resolution, wide dynamic range (dictated by material parameters) , and scalability ( the method was recently shown to measure Rayleigh scatter in fibers up to 100km), the future of distributed sensing could change forever.

Summary

Just as an individual’s fingerprint sets him or her apart from others, so to does an optical fiber’s “fingerprint”. Using the pattern that underlies optical fiber due to Rayleigh scatter being back-reflected, researchers can harness the power of the fiber fingerprint in real-world applications such as distributed sensing and network security and intrusion protection.