close

Вход

Забыли?

вход по аккаунту

?

JP2011232347

код для вставкиСкачать
Patent Translate
Powered by EPO and Google
Notice
This translation is machine-generated. It cannot be guaranteed that it is intelligible, accurate,
complete, reliable or fit for specific purposes. Critical decisions, such as commercially relevant or
financial decisions, should not be based on machine-translation output.
DESCRIPTION JP2011232347
A folding Sagnac sensor array is provided. A folded Sagnac fiber optic acoustic sensor array
(1200) operates in the same manner as a Sagnac interferometer, but uses a common delay path
to reduce dispersive pickup in the downlead fiber. A fiber optic acoustic sensor (716) is used to
detect acoustic waves in the water. By placing the foundation of the array on the same principles
of operation as the Sagnac interferometer, rather than placing it on the Mach-Zehnder
interferometer, the sensor array has a stable bias point and reduced phase noise It allows
broadband signal sources to be used rather than requiring more expensive narrow linewidth
lasers. A number of acoustic sensors (718 (N)) can be multiplexed into the configuration of the
folded Sagnac fiber optic acoustic array. [Selected figure] Figure 1
Acoustic sensor, method and sensor for detecting acoustic signal
[0001]
FIELD OF THE INVENTION This invention is in the field of fiber optic acoustic sensor arrays in
which light propagates in the array and analyzes the effect of the acoustic signal on the light
returning from the array to determine the characteristics of the acoustic signal.
[0002]
Description of the Related Art Fiber optic based acoustic sensors are promising as alternatives to
conventional electronic sensors.
05-05-2019
1
Among those advantages are high sensitivity, large dynamic range, lightness and compact size.
The ability to easily multiplex a large number of fiber optic sensors onto a common bus also
makes them attractive for large arrays. Recently, large scale fiber optic sensor arrays by
successfully incorporating multiple low gain erbium doped fiber amplifiers (EDFAs) into a fiber
optic sensor array to increase the number of sensors that can be supported by a single fiber pair
Has become even more competitive.
[0003]
For sound detection, the selected fiber optic sensor was a Mach-Zehnder interferometric sensor.
In any of the interference sensors, phase modulation is associated with intensity modulation by a
raised cosine function. Due to this non-linear transfer function, sinusoidal phase modulation
produces higher order harmonics. An interferometer biased in quadrature (interfering beams
with π / 2 out of phase) has a maximum response at the first harmonic and a minimum response
at the second harmonic. For this reason, quadrature is the preferred bias point. As the bias point
drifts out of quadrature (eg, due to external temperature changes), the response at the first
harmonic decreases and the response at the second harmonic increases. When the
interferometer is biased at 0 or out of phase, the first harmonic disappears completely. The
reduction of this response at the first harmonic (as a result of the bias point leaving quadrature)
is called signal fading.
[0004]
Because Mach-Zehnder interferometric sensors have unstable bias points, they are particularly
susceptible to the aforementioned signal fading problems. Demodulation of the return signal is
required to overcome signal fading. A typical demodulation technique is a Phase-Generated
Carrier (PGC) scheme that requires a path mismatched Mach-Zehnder interference sensor. (For
example, Multiplexing of Interferometric Sensors Using Phase Carrier Techniques, Journal of
Lightwave Technology, vol. LT-5, No. 7, 1987, 7 by Anthony Dandridge et al. This path imbalance
also results in the conversion of laser phase noise into intensity noise, which limits the
performance of Mach-Zehnder interferometric sensor arrays at low frequencies, and Impose
strict requirements on the line width of the source. This narrow linewidth requirement delays the
development of a 1.55 μm amplified Mach-Zehnder interferometric sensor.
[0005]
05-05-2019
2
It has been found that Sagnac interferometers are widely used in fiber optic gyroscopes. (たとえ
ば、B. See B. Culshaw et al., Fiber optic gyroscopes, Journal of Physics E (Scientific
Instruments), Vol. 16, No. 1, 1983, pages 5-15. It has been proposed that sound waves can be
detected using Sagnac interferometers. (たとえば、E. Fiber-optic acoustic sensor based on
the Sagnac interferometer based on E. Udd, Proceedings of the SPIE-The International Society for
Optical Engineering, Vol. 425, 1983, 90. ~ 91 pages; Sagnac interferometer for underwater sound
detection by Kiell Krakenes et al .; Sagnac interferometer for underwater sound detection: noise
properties, OPTICS LETTERS, Vol. 14, No. 20, 1989. October 15, pp. 1152-1145; and Sverre
Knudsen et al., An Ultrasonic Fiber-Optic Hydrophone Incorporating a Push-Pull, Incorporating a
Push-Pull Transducer into a Sagnac Interferometer Transducer in a Sagnac Interferometer),
JOURNAL OF LIGHTWAVE TECHNOLOGY, Vol. 12, No. 9, September 994 years, see pp. 16961700. 2.) Due to its common path design, Sagnac interferometers are reciprocal and therefore
have stable bias points, which eliminates signal fading and prevents conversion of source phase
noise to intensity noise. Thus, the Sagnac interferometer is not susceptible to phase noise that
limits the Mach-Zehnder interferometric sensor at low frequencies.
[0006]
Anthony Dandridge et al., "Multiplexing of Interferometric Sensors Using Phase Carrier
Techniques with Phase Carrier Technology", Journal of Lightwave Technology, Vol. 5, No. 7, July
1987, 947 to 952
[0007]
One aspect of the present invention is an acoustic sensor that includes an optical pulse source, a
first coupler, a polarization dependent second coupler, an optical delay path, and at least one
detector.
The first coupler couples the light pulse to the first optical path having the first optical path
length and the array of sensors. The array of sensors includes at least a first sensor. The first
sensor is in a second light path having a second light path length different from the first light
path length. A polarization dependent second coupler couples the light pulse of the first
polarization received from the first optical path to the optical delay path and couples the light
pulse of the second polarization received from the array to the optical delay path. The light pulse
of the first polarization coupled to the optical delay path returns from the optical delay path to
the second coupler with a second polarization. The light pulse of the second polarization coupled
to the optical delay path returns from the optical delay path to the second coupler at the first
05-05-2019
3
polarization. The second coupler couples an optical pulse returning from the optical delay path to
the second coupler with a first polarization into the first optical path for propagation to the first
coupler therefrom. The second coupler couples the light pulses returning from the optical delay
path to the second coupler with a second polarization into the array and propagates therefrom to
the first coupler. The first coupler combines the light pulses from the first light path with the
light pulses from the array to interfere with the light pulses traveling an equal distance through
the first light path and the array to produce a detectable output signal Do. The detectable output
signal fluctuates in response to the acoustic energy striking the first sensor. A detector detects
the detectable output signal and generates a detector output signal responsive to variations in
the detectable output signal from the first coupler. Preferably, the array comprises a second
sensor. The second sensor is in a third light path having a third light path length different from
the first light path length and the second light path length. Also preferably, the polarization
dependent second coupler comprises a polarization beam splitter. In the preferred embodiment,
the optical delay path includes a length of optical waveguide and a polarization rotation reflector.
The reflector reflects the light of the first polarized light incident on the reflector as the light of
the second polarized light, and reflects the light of the second polarized light incident on the
reflector as the light of the first polarized light. The reflector preferably comprises a Faraday
rotation mirror. In a particularly preferred embodiment, the first optical path comprises a nonreciprocal phase shifter, and the non-reciprocal phase shifter comprises a first optical path in a
first direction such that the light combined by the first coupler has a phase bias. The light
propagating through and the light propagating through the first optical path in the second
direction experience a relative phase shift. Preferably, in such an embodiment, the third light
path is positioned in parallel with the first light path.
One of the first optical path and the third optical path timed the light pulse by having a
propagation time different from the propagation time of the light propagating through the first
optical path from the light propagating through the second optical path For multiplexing, an
optical delay is included that causes the first optical path to have an optical path length different
from that of the third optical path. Preferably, the non-reciprocal phase shifter includes a first
Faraday rotator, a quarter wave plate, and a second Faraday rotator. The first Faraday rotator, the
quarter-wave plate and the second Faraday rotator are such that light propagating in the first
direction is the first Faraday rotator, then the quarter-wave plate, The light propagating through
the second Faraday rotator and propagating in the second direction is the second Faraday
rotator, then the quarter wave plate, and then the first Faraday rotator. It is positioned to pass
through. Alternatively, the nonreciprocal phase shifter includes a first quarter wave plate, a
Faraday rotator, and a second quarter wave plate. The first quarter-wave plate, the Faraday
rotator and the second quarter-wave plate are such that light propagating in the first direction is
the first quarter-wave plate, and then the Faraday rotator Light passing through the second
quarter-wave plate, and light propagating in the second direction, the second quarter-wave plate,
then the Faraday rotator, and It is positioned to pass through a quarter wave plate.
05-05-2019
4
[0008]
Another aspect of the invention is an acoustic sensor that includes an input light pulse source, an
array of light sensors, a light delay path, a light detector system, and an input / output system.
An input / output system receives input light pulses and passes a first portion of each light pulse
having a first polarization through the array of photosensors in a first direction, then through the
optical delay path, and then Direct to a light detector system. The input / output system passes a
second portion of each light pulse of a second polarization, orthogonal to the first polarization,
through the optical delay path, then through the light sensor array in a second direction, and
then light. Direct to the detector system. A photodetector system receives light pulses of the first
and second polarizations and detects changes in the light pulses caused by perturbations in the
light sensor.
[0009]
Another aspect of the invention is a method of detecting an acoustic signal. The method
comprises the steps of generating an input optical signal and coupling the input optical signal to
at least first and second propagation paths to propagate it in respective first directions. The first
and second propagation paths have first and second optical path lengths, respectively. The first
and second propagation paths output respective first and second output light portions. The first
and second output light portions are output from the first and second propagation paths at
different times according to the differences in the first and second optical path lengths. The
second output light portion is modulated by the acoustic signal striking the second propagation
path. The first light portion is coupled to the delay path at a first polarization and the second
light portion is coupled to the delay path at a second polarization. The delay path outputs a first
delayed light portion corresponding to the first output light portion. The first delayed light
portion has a second polarization. The delay path outputs a second delayed light portion
corresponding to the second output light portion. The second delayed light portion has a first
polarization. The first and second delay light portions are coupled to the first and second
propagation paths and propagate there through in respective second directions opposite to
respective first directions. The first propagation path outputs a first set of return light portions.
The first set of return light portions includes respective return light portions for each of the first
and second delayed light portions. The second propagation path outputs a second set of return
light portions. The second set of return light portions includes respective return light portions for
each of the first and second delayed light portions. The first and second sets of return light
portions are coupled to at least one detector. The return light portions in the first and second sets
05-05-2019
5
of return light portions are obtained from the output light portion and the delayed light portion
traveling the same optical path length and interfere to produce a detectable output signal. The
method selectively detects the detectable output signal to detect only the output signal obtained
from the interference of the light portion propagating in the first propagation path in either the
first direction or the second direction. . The detectable output signal fluctuates in response to the
acoustic signal striking the second propagation path.
[0010]
Another aspect of the invention is a sensor that includes a light source and a first coupler that
couples light to a common path and a sensing array and propagates it in respective first
directions. The sensing array includes a plurality of sensing paths. A polarization dependent
second coupler couples light from the common path and sensing array into the delay path. The
second coupler couples only light of the first polarization from the common path to the delay
path. The second coupler couples only light of the second polarization from the sensing array to
the delay path. The delay path rotates light of a first polarization to a second polarization and
rotates light of a second polarization to a first polarization. The second coupler further couples
the light of the first polarization from the delay path to the common path and couples the light of
the second polarization from the delay path to the sensing array, in each of the second directions.
Propagate to the first coupler. The first couplers provide output light responsive to light
propagating in their respective second directions. A detector receives the output light from the
first coupler and produces an output signal responsive to the interference of the light at the first
coupler. Preferably, the delay path comprises a length of optical fiber and a polarization rotation
reflector. The length of the optical fiber is selected to provide the optical delay time. Light
propagates from the second coupler to the reflector through the optical fiber. The reflector
reflects light into the optical fiber and propagates through the optical fiber to the second coupler.
The reflector further rotates light incident with a first polarization to second polarization and
rotates light incident with a second polarization to first polarization. Preferably, the reflector
comprises a Faraday rotation mirror. Also preferably, the polarization dependent second coupler
includes a polarization beam splitter positioned such that the delay path receives light from the
port of the polarization beam splitter and returns the light to the port of the polarization beam
splitter.
[0011]
The invention is described below in connection with the attached drawings.
[0012]
05-05-2019
6
FIG. 7 is a diagram of an example Sagnac interferometer with a single sensing loop.
FIG. 5 is a diagram of a Sagnac sensor array according to the invention in which each rung of the
sensor array forms a further Sagnac interferometer. FIG. 5 is a diagram of a Sagnac sensor array
that includes an erbium-doped fiber amplifier to regenerate the signal power lost to coupling and
dissipation losses. FIG. 5 is a graphical representation of the frequency response of a Sagnac
interferometer according to the invention compared to three major ocean floor noises. A graph of
the maximum and minimum acoustic signals detectable by a Mach-Zehnder interferometer and
the maximum and minimum acoustic signals detectable by a Sagnac interferometer according to
the invention, showing the relatively constant dynamic range of the Sagnac interferometer over a
wide range of frequencies FIG. FIG. 7 is a graphical representation of minimum detectable
acoustic signal versus frequency for three Sagnac interferometer configurations with different
fiber lengths in hydrophone and delay loop. FIG. 5 is a diagram of a Sagnac interferometer
according to the invention including an additional delay loop to increase the dynamic range of
the interferometer. FIG. 8 is a graphical representation of the dynamic range provided by the
interferometer of FIG. 7; FIG. 7 is a diagram of positioning the interferometer delay loop at the
dry end of the sensor array system. FIG. 5 is a diagram of positioning the interferometer delay
loop at the wet end of the sensor array system. FIG. 9B is a diagram of the annotated Sagnac
interferometer of FIG. 9B showing the lengths used to calculate the effects of phase modulation.
FIG. 7 is an illustration of a technique for winding a delay loop to reduce the effects of sound
waves on the delay loop. FIG. 5 is a diagram of a Sagnac interferometer according to the
invention including an empty rung detecting distributed pick-up noise that may be subtracted
from the signal generated by the sensor. FIG. 5 is a diagram of a Sagnac interferometer according
to the invention that includes a depolarizer to reduce the effects of polarization induced fading.
FIG. 2 is a diagram of a Sagnac interferometer utilizing a frequency division multiplex
transmission scheme. FIG. 15 is a graph showing the generation of a beat signal between the
delayed modulated signal and the return sensor signal in the interferometer of FIG. 14; FIG. 5 is a
diagram of a Sagnac interferometer utilizing a code division multiplex transmission scheme. FIG.
5 is a schematic diagram of a folded Sagnac acoustic fiber sensor array. FIG. 7 is a graphical
representation of the number of return pulses per time interval showing the separation of time of
signal and noise pulses. FIG. 7 is a diagram of a folded Sagnac acoustic fiber sensor array having
a second delay loop to provide an expanded dynamic range. FIG. 18 is a diagram of a folded
Sagnac acoustic fiber sensor array having a phase modulator and nulling circuit instead of the
reflector of FIG.
FIG. 20 is a view of a further alternative embodiment of FIG. 19 in which two delay loops are
connected to different ports of the coupler. FIG. 7 is a diagram of an alternative embodiment of a
05-05-2019
7
fiber optic acoustic sensor array system using a Faraday rotator mirror. FIG. 7 is a diagram of a
further alternative embodiment of a fiber optic acoustic sensor array that utilizes a nonpolarizing source in combination with a depolarizer, a polarizing beam splitter and a Faraday
rotator mirror. FIG. 7 is a diagram of a further alternative embodiment of a fiber optic acoustic
sensor array that utilizes a non-polarizing source in combination with a depolarizer, a polarizing
beam splitter and a Faraday rotator mirror. FIG. 7 is a diagram of a further alternative
embodiment of a fiber optic acoustic sensor array that utilizes a non-polarizing source in
combination with a depolarizer, a polarizing beam splitter and a Faraday rotator mirror. FIG. 7 is
an illustration of an alternative embodiment of a folded fiber optic acoustic sensor array utilizing
a non-polarizing source in combination with an optical circulator, a 2 × 2 coupler, and a nonreciprocal phase shifter. FIG. 25 is an illustration of an alternative embodiment of a folded fiber
optic acoustic sensor array similar to FIG. 24 with depolarizers placed in the second array input /
output fiber. FIG. 26 is a diagram of a first preferred embodiment of the non-reciprocal π / 2
phase shifter of FIGS. 24 and 25 showing the effect on polarization of light propagating through
the phase shifter in a first direction. FIG. 27 is a diagram of the effect on polarization of light
propagating in a second (opposite) direction through the phase shifter of FIG. FIG. 26 is an
illustration of an alternative preferred embodiment of the non-reciprocal π / 2 phase shifter of
FIGS. 24 and 25 showing the effect on polarization of light propagating through the phase shifter
in a first direction. FIG. 29 is a diagram of the effect on the polarization of light propagating in
the second (opposite) direction through the phase shifter of FIG. 28. Further alternative
embodiments of a folded fiber optic acoustic sensor array utilizing polarization based biasing for
multiple detectors, with each detector having a bias point that is independently settable with the
bias point of the other detector Of the FIG. 31 is an illustration of an alternative embodiment of a
folded fiber optic acoustic sensor array similar to FIG. 30 in which the depolarizers are placed in
the second array input / output fiber. FIG. 31 is a view of an alternative embodiment of the
folded fiber optic acoustic sensor array similar to FIG. 30 with the optical circulator replacing a 2
× 2 coupler. FIG. 35 is an illustration of an alternative embodiment of a folded fiber optic
acoustic sensor array similar to FIG. 32 wherein a depolarizer is placed on the second array input
/ output fiber. FIG. 7 is a diagram of a further alternative embodiment of a folded Sagnac sensor
array including a combined input / output subsystem.
FIG. 35 is an illustration of an alternative embodiment of a folded fiber optic acoustic sensor
array similar to FIG. 34, wherein a depolarizer is placed on the second array input / output fiber.
A diagram of a further alternative embodiment of a folded fiber optic acoustic sensor array
similar to FIGS. 34 and 35, wherein the detector is coupled to the input / output subsystem by an
optical fiber to allow the detector to be remotely located. It is.
[0013]
05-05-2019
8
The invention is described below in connection with an array of acoustic sensors (eg
hydrophones) in a Sagnac loop. Before describing the preferred embodiment, a brief overview of
the operation of a single loop Sagnac acoustic sensor is provided.
[0014]
Single Loop Sagnac Acoustic Sensor A simple Sagnac based acoustic sensor 100 is shown in FIG.
The Sagnac loop is divided into two parts, a delay loop 102 and a hydrophone 104. The delay
loop 102 is simply a long fiber, typically over 1 km. The hydrophone 104 is the portion of the
fiber where sound waves are converted to phase modulation of the optical signal propagating
through the fiber. High responsiveness to acoustic waves is typically achieved by selecting an
optimized coating for the area of the hydrophone 104 fiber and winding the fiber around a
mandrel of suitable composition. (たとえば、J.A. See Optical Fiber sensor coatings, Optical
Fiber Sensors, Proceedings of the NATO Advanced Study Institute, 1986, pp. 321-338, by JA
Bucaro et al. The length of the fiber wound around the hydrophone 104 is typically 10 meters to
100 meters. For example, light from a source 110, such as a superfluorescent fiber source (SFS),
is split by the 3 × 3 coupler 112 into a clockwise (CW) beam and a counterclockwise (CCW)
beam. The operation of 3 × 3 coupler 112 is well known, for example, Sun K. K. et al. Fiber-Optic
gyroscope with [3 × 3] directional coupler by Seam (Sang K. Sheem), Applied Physics Letters,
Vol. 37, No. 10, 1980 November 15, pp. 869-871.
[0015]
Although described herein as using 3 × 3 couplers 112, other couplers (eg, 2 × 2 couplers, 4 ×
4 couplers, etc.) can be used in alternative embodiments of the invention. For example, to use a 2
× 2 coupler, make a Sagnac interferometer using both ports on one side. One port on the other
side is a detection port. The remaining ports are used to launch light into the array and can also
be used as detection ports if couplers or circulators are employed (in a similar manner as done
with fiber optic gyroscopes). In general, use any n × m coupler by making a Sagnac
interferometer using two ports on one side of the coupler and using the port on the other side of
the coupler as a detection port, a launch port, or both can do.
[0016]
05-05-2019
9
After splitting, the CW beam first travels through the delay loop 102 and then through the
hydrophone 104, and the CCW beam travels first through the hydrophone 104 and then through
the delay loop 102. During a time delay Tdelay between the time the CW beam travels the
hydrophone 104 and the time the CCW beam travels the hydrophone 104, the acoustic signal at
the hydrophone 104 and also the acoustically induced phase modulation change. This change in
phase modulation is mapped to the phase difference between the counter-propagating beams,
which are converted to intensity modulation when the beams are recombined at the 3 × 3
coupler. This intensity modulation is then detected by only the first detector 120 and the second
detector 122 or only one of these two detectors.
[0017]
More explicitly, if the acoustic signal induces phase modulation φ h cos (Ω t) in the fiber of
hydrophone 104, then the obtained phase modulation φ int (t) between the interfering beams at
hydrophone 104 is given by .
[0018]
[0019]
However, Tdelay is the advancing time in the delay loop.
Thus, φ int (t) is a function of the hydrophone modulation φ h and the product of the loop delay
T delay and the acoustic modulation frequency Ω.
This is different from the Mach-Zehnder interference sensor in which φint (t) is a function of
only the hydrophone modulation φh. Maximum sensitivity is achieved in the Sagnac loop
acoustic sensor when the product of the acoustic frequency Ω and the time delay Tdelay is an
odd multiple of π (maximum of the first sine term of Equation 1). The acoustic frequency
producing this product π is called the appropriate frequency of the loop, which is the lowest
frequency at which maximum sensitivity is achieved. Most underwater sensing applications relate
to the detection of acoustic frequencies below 10 kHz. In order for the appropriate loop
frequency to be less than 10 kHz, a delay time of at least 50 microseconds and thus a delay loop
length of at least 10 km is required. Thus, the Sagnac acoustic sensor 100 requires a large
amount of fiber for detection of low acoustic frequencies (<10 kHz).
05-05-2019
10
[0020]
The common path design inherent to the Sagnac interferometer has many advantages over the
Mach-Zehnder interferometer in addition to the stable bias point and phase noise removal
already mentioned. The Sagnac interferometer enables the use of a short-bandwidth source with
a short coherence length, such as the superfluorescent fiber source (SFS), which is an example of
an extended spontaneous emission (ASE) source. Such sources are inexpensive and can easily
provide high power. The use of a 3 × 3 coupler has been shown to passively bias the Sagnac
acoustic sensor near quadrature. (サン・K. Fiber-Optic gyroscope with [3 × 3] directional
coupler by Seam (Sang K. Sheem), Applied Physics Letters, Vol. 37, No. 10, 1980 November 15,
pages 868-871; See Low-cost fiber-optic gyroscope by H. Poisel et al., Electronics Letters, Vol. 26,
No. 1, Jan. 4, 1990, pp. 69-70. . The source excess noise, which is the limiting noise source of the
SFS source, can be reduced by subtracting the signals from the two detection ports of the 3 × 3
coupler, which adds to the intensity variation of the phase modulation induction by the
hydrophone. This allows the Sagnac interferometer to approach the limited performance of shot
noise. Sagnac interferometer for underwater sound detection by Kjell Krakenes et al .: Sagnac
interferometer for underwater sound detection: noise properties, OPTICS LETTERS, Vol. 14, No.
20, October 1989. 15 days, pages 1152-1145).
[0021]
Previous work on Sagnac based acoustic sensors has been limited to single sensor configurations.
Because of the inherent advantages of the Sagnac interferometer, the applicant has determined
that it is desirable to replace a large array of Mach-Zehnder interferometric sensors with Sagnacbased sensors. Each of the Sagnac sensors 100 discussed above requires many kilometers of
fiber, making inserting many such sensors into a large array impractical. Research to reduce the
fiber length requirement using a recirculating delay loop has resulted in a sensor that suffers
from high noise but requires much less fiber to incorporate the EDFA into the recirculating loop.
(たとえば、J.T. J. Kringlebotn et al., Sagnac Interferometer Including Recirculating Ring
with Erbium-Doped Fiber Amplifier (Sagnac Interferometer Including A Recirculating Ring With
An Erbium-doped Fiber Amplifier, OFS '92 Conference Proceedings, pp. 6-9 See A novel approach
to reduce the required fiber is described below.
[0022]
05-05-2019
11
Novel Sensor Arrays Based on Sagnac Interferers As described below, Applicants are novel to
reduce the amount of fiber needed for large Sagnac-based arrays by multiplexing multiple
sensors into the same delay loop Systems have been discovered, resulting in a practical Sagnac
sensor array (SSA). As shown in FIG. 2, a Sagnac sensor array 200 in accordance with the present
invention includes an array 210 of hydrophones 212 (i) in a ladder configuration coupled to a
single delay loop 214. For example, FIG. 2 shows a Sagnac sensor array 210 having N
hydrophones 212 (1), 212 (2)... 212 (N) on each rung 216 (1), 216 (2). Show. Each rung 216 (i)
in the Sagnac sensor array 210 comprises a single fiber wound around a respective hydrophone
212 (i). All paths from the 3 × 3 coupler 220 through the delay loop 214 and the array 210
back to the coupler 220 include separate Sagnac interferometers. Thus, for an array of N sensors
212, there are N separate Sagnac interferometers, each of which behaves similarly to the single
loop Sagnac sensor 100 shown in FIG. Each Sagnac interferometer measures the acoustic signal
at discrete points in space, ie, at the location of the hydrophone 212 (i). For example, a Sagnac
interferometer, including delay loop 214 and rung 216 (1), measures the acoustic signal at
hydron 212 (1). Furthermore, each Sagnac interferometer picks up an acoustic signal (eg noise)
elsewhere in the loop, but this noise is advantageously reduced as discussed below.
[0023]
The Sagnac sensor array 200 is most easily understood in a time division multiplexed (TDM)
configuration (non-TDM schemes are discussed later). The source 222 (which may
advantageously include a conventional pulse source or may include a cw source with an external
modulator) generates a light pulse, which is shown in FIG. The Sagnac loop enters through the
third port of and propagates in both the CW and CCW directions. Upon reaching the array 210,
the CCW pulses are split into a train of N separate pulses. At this point, the CW input pulse has
not yet reached the array 210 and is still a single pulse. When the CW pulse reaches the array
210, it is also divided into a train of N pulses. Each pulse in the CW train travels the respective
rung 216 (i) and then returns to the 3 × 3 coupler 220 and interferes with the pulses in the
CCW train that traveled the same rung 216 (i) in the opposite direction. Thus, N pulses are
detected by the first detector 230 and the second detector 232, each pulse being a CW and CCW
pulse of one of the N Sagnac loops (ie the same respective rung). Two pulses that have traveled in
opposite directions through 216 (i). Because the pulses traveling through different combinations
of rungs do not travel the same optical path, such pulses do not simultaneously exist in coupler
220 and therefore do not interfere with each other in coupler 222. The pulse width should be
smaller than the delay difference between adjacent sensors so that pulses from adjacent sensors
do not overlap.
[0024]
05-05-2019
12
As shown in FIG. 3, a low gain Erbium-doped fiber amplifier (EDFA) 240 is advantageously added
to the array portion 210 just as an EDFA is added to the Mach-Zehnder interferometric sensor
array. (たとえば、クレイグ・W. Optimization of large scale fiber sensor arrays incorporating
multiple optical amplifiers by Craig W. Hodgson et al. Part 1: Optimization of Large-Scale Fiber
Sensor Arrays Incorporated Multiple Optical Amplifiers-Part I: Signal- to-Noise Ratio), JOURNAL
OF LIGHTWAVE TECHNOLOGY, Vol. 16, No. 2, Feb. 1998, pp. 218-223; Hodgeson et al.,
Optimization of Large Scale Fiber Sensor Arrays Incorporating Multiple Optical Amplifiers-Part 2:
Optimization of Large-Scale Fiber Sensor Arrays Incorporated Multiple Optical Amplifiers-Part II:
Pump Power, JOURNAL OF LIGHTWAVE TECHNOLOGY 16, No. 2, Feb. 1998, pp. 224-231;
Jefferson L., et al. Novel Fiber Sensor Arrays Using Erbium-Doped Fiber Amplifiers by Jefferson L.
Wagener et al., Novel Fiber Sensor Arrays Using Erbium-Doped Fiber Amplifiers, JOURNAL OF
LIGHTWAVE TECHNOLOGY, Vol. 15, No. 9, 1997 July 1681-1688; and C.I. W. Hodgeson et al.,
Large-Scale Interferometric Fiber Sensor Arrays with Multiple Optical Amplifiers, OPTICS
LETTERS, Vol. 22, No. 21, Nov. 21, 1997, 1651 See pages-1653. 2.) The EDFA 240 increases the
number of sensors that can be supported by a single array 210 by regenerating the signal output
lost to coupling and dissipation losses.
The EDFA is advantageously pumped by one or more pump laser sources 242 via a splitting
coupler 244 and via a first wavelength division multiplexing (WDM) coupler 246 and a second
WDM coupler 248.
[0025]
Because it uses a Sagnac configuration, the Sagnac sensor array 200 has all of the advantages of
the single loop Sagnac based sensor 100 discussed above. The common path design prevents
source phase noise from being converted to intensity noise at the interference coupler 220. The
source 222 may be a fiber ASE (extended spontaneous emission) source (ie, the SFS discussed
above) that provides high power at 1.55 μm inexpensively. Passive bias near quadrature is
achievable for all sensors by using a 3 × 3 coupler 220. The 3 × 3 coupler 220 also provides a
convenient means to detect the two interference outputs at the detectors 230, 232 and use the
outputs of the two detectors to reduce source excess noise. (たとえばK. Sagnac
interferometer for underwater sound detection, showing the use of two detectors combined with
a single Sagnac interferometer by K. Krakenes et al .: Sagnac interferometer for underwater
sound detection: noise properties OPTICS LETTERS, 14: 1989, pages 1152-1154).
05-05-2019
13
[0026]
The characteristics of the novel Sagnac sensor array 200 will be discussed more specifically
below, and then the frequency response and dynamic range obtained from using the Sagnac
interferometer will be discussed in more detail. Subsequently, calculation of the size of the
distributed pickup from the non-hydrophone fiber loop segment is described along with
techniques for reducing the size of this pickup. Polarization is also dealt with below. We will now
discuss a new source of noise introduced by the Sagnac design. Finally, a multiplexing scheme
other than TDM for Sagnac sensor arrays is presented.
[0027]
Although the invention has been described above for a single sensor in each rung 216 (i) of the
array 210, each rung 216 (i) is advantageously, for example, incorporated herein by reference,
March 1997. It is understood that a sub-array having multiple sensors may be included, such as
those described in U.S. patent application Ser. No. 08 / 814,548, filed on Nov. 11 and patented.
(C.W. Hodgeson et al., Large Scale Interference Fiber Sensor Array with Multiple Optical
Amplifiers, Optics Letters, Vol. 22, 1997, 1651-1653; W. Wagner et al., Novel Fiber Sensor
Arrays Using Erbium-Doped Fiber Amplifiers, Journal of Lightwave Technology, Vol. 15, 1997,
pp. 1681-1688; W. Hodgeson et al., Optimization of large scale fiber sensor arrays incorporating
multiple optical amplifiers, Part 1: Signal-to-noise ratio, Journal of Lightwave Technology, Vol.
16, 1998, pp. 218-223; W. Hodgeson et al. Optimization of large scale fiber sensor arrays
incorporating multiple optical amplifiers, Part 2: Pump power, Journal of Lightwave Technology,
Vol. 16, 1998, pp. 224-231).
[0028]
Frequency Response As mentioned above, the Sagnac sensor has a frequency dependent
response given by Equation 1. At frequencies well below the appropriate frequency of the loop,
defined as 1 / (2 · Tdelay), the minimum detectable acoustic signal is proportional to the
reciprocal of the acoustic frequency. This reduction in acoustic sensitivity at low frequencies is of
major interest for Sagnac acoustic sensors. However, it is pointed out that this reduction in
sensitivity at low frequencies is fortunately consistent with the increase in the ocean noise floor
function. (For example, by Sver Kunadsen, Ambient and Optical Noise in Fiber-Optic
Interferometric Acoustic Sensors), Norwegian University of Science and Technology University
Chapter 3, Michelson and Sagnac Interferometers Optical fiber sensors based on: Fiber-Optic
05-05-2019
14
Sensors Based on the Michelson and Sagnac Interferometers: Responsivity and Noise Properties,
1996, pages 37-40. 2.) Ideally, it would be desirable if the smallest detectable acoustic signal of
the array at a given frequency is a constant quantity below the ocean noise floor function at that
frequency. Thus, the minimum detectable acoustic signal will also increase at lower frequencies
to coincide with the increase of the ocean noise floor function. The frequency response of the
Sagnac sensor array 200 of the present invention actually provides a good match between the
ocean noise floor function and the acoustic sensitivity. This is the smallest detectable acoustic
signal for a Sagnac sensor array, given an optical noise floor function of 10 μrad / √Hz, a
hydrophone phase response of 3.2 × 10 <-7> rad / μPa, and a delay loop length of 20 km. Is
plotted as curve 250. (The vertical axis is dB relative to a 1 μrad / √Hz baseline. A further plot
in FIG. 4 is the ocean noise floor function for the three major ocean noise sources at these
frequencies, and the sum obtained from the noise from the three sources. A curve 252 represents
noise from a rough sea, an earthquake, a volcanic eruption or the like. Curve 253 represents light
transmission noise.
Curve 254 represents DSS0 (far-field transmission and storm) noise. Curve 256 represents the
sum of noise floor functions from three major sources (ie, the sum of curves 252, 253 and 254).
(たとえば、ロバート・J. Sea noise background by Robert J. Urick: Ambient noise levels,
Principles of Underwater Sound, 3rd Edition, Chapter 7, McGraw-Hill, 1983, pages 202-236. The
minimum detectable acoustic signal of the Sagnac sensor array 200 is augmented in such a way
as to provide an approximately constant amount of detectable signal below the ocean noise floor
function at all frequencies below 10 kHz. Thus, the frequency dependent response of the Sagnac
sensor array 200 does not interfere with low frequency acoustic detection. The Mach-Zehnder
array exhibits the same tendency as the Sagnac sensor array, i.e. the sensitivity decreases with
decreasing frequency, but in the Mach-Zehnder array the decreasing sensitivity is smaller than in
the Sagnac-based sensor.
[0029]
Both the Mach-Zehnder interferometer and the Sagnac sensor array 200 have similar frequency
dependent responses, but their sources of frequency response are fundamentally different. The
increase in the minimum detectable signal in the Mach-Zehnder Interferometer sensor array is
due to the increase in the optical noise floor function. The cause of this increase in the optical
noise floor function is the phase noise introduced by the path imbalanced Mach-Zehnder
interferometer. Thus, the noise floor function is 10 μrad / √Hz at 10 kHz, but it increases as the
frequency gets lower. In the Sagnac sensor array 200, the increase in the minimum detectable
acoustic signal is due to the sin (.OMEGA.Tdelay / 2) term of Equation 1 and not due to the
increase in the optical noise floor function. The optical noise floor function is 10 μrad / √Hz
05-05-2019
15
constant over the entire frequency range.
[0030]
The significance of this difference can be seen by examining the dynamic range of the MachZehnder interferometric sensor array and the Sagnac sensor array 200 shown in FIG. The
dynamic range of the sensor is bounded by the minimum detectable phase shift and the
maximum detectable phase shift. For interferometric sensors, the maximum detectable phase
shift is bounded by the non-linear response of the interferometer and the minimum detectable
phase shift is bounded by the optical noise floor function. Both Mach-Zehnder interferometric
sensor arrays and Sagnac sensor arrays have a maximum detectable phase shift that is constant
over the acoustic frequency range. However, although the Sagnac sensor array 200 also has a
flat optical noise floor function, it has a flatter detectable minimum phase shift, but the MachZehnder interferometric sensor array is driven by the phase noise introduced by the path
imbalance interferometer The minimum detectable phase shift is increased due to the increase of
the resulting optical noise floor function. Thus, while the Sagnac sensor array 200 has a constant
dynamic range at all acoustic frequencies, the Mach-Zehnder interferometric sensor array has a
decreasing dynamic range at low acoustic frequencies. This is represented in FIG. 5 where the
minimum and maximum detectable acoustic signals (in dB arbitrary units) are plotted for the
Sagnac sensor array 200 and the Mach-Zehnder interferometric sensor array. As shown in FIG. 5,
both arrays have a dynamic range of about 100 dB above 1 kHz, and phase noise does not limit
the Mach-Zehnder interferometric sensor array. At 10 Hz, phase noise affects the Mach-Zehnder
interferometric sensor array, reducing its dynamic range to about 74 dB. On the other hand, the
dynamic range of the Sagnac sensor array 200 remains about 100 dB.
[0031]
It is interesting to look at the frequency response of Sagnac sensor array 200 at frequencies well
below the loop proper frequency as a function of delay loop length and hydrophone
responsiveness. At these frequencies, the sin (.OMEGA.Tdelay / 2) factor of Equation 1 can be
approximated as .OMEGA.Tdelay / 2, indicating that the response of the Sagnac sensor array 200
is proportional to the product of .phi.h and Tdelay. φ h itself is proportional to the amount of
fiber in each hydrophone 212 (i), and T delay is proportional to the amount of fiber in delay loop
214. Therefore, the response at a frequency considerably lower than the loop appropriate
frequency is proportional to the product of the hydrophone fiber length and the delay fiber
length. FIG. 6 is a plot of the minimum detectable acoustic signal for several Sagnac sensor array
configurations, where the product of the fiber length at each hydrophone 212 (i) and the fiber
05-05-2019
16
length at delay loop 214 is constant. However, the relative allocation of fibers between the delay
loop 214 and each hydrophone 212 (i) has changed. For example, curve 260 represents the
frequency response of Sagnac sensor array 200 having 45 km of fibers in its delay loop 214 and
100 meters of fiber in each hydrophone 212 (i), and curve 262 represents 30 km in its delay
loop 214. Curve 264 represents the frequency response of the Sagnac sensor array 200 with the
following fibers and 150 meters of fiber in each hydrophone 212 (i), and curve 264 shows the
fibers in 15 km of its delay loop 214 and 300 in each hydrophone 212 (i). 7 represents the
frequency response of a Sagnac sensor array 200 having meters of fiber. As shown, each Sagnac
sensor array 200 has the same sensitivity at low frequencies, but approaches maximum
sensitivity at different frequencies given by their respective loop appropriate frequencies. Thus,
for a given minimum detectable acoustic signal at low frequency, there is still some freedom in
selecting the fiber length of the delay loop 214 and the hydrophone 212 (i). This freedom can be
used to help Sagnac sensor array 200 meet other criteria, such as minimizing the total amount of
fiber required or minimizing the delay loop length.
[0032]
Increasing the Dynamic Range of the Sagnac Sensor Array As discussed above, the Sagnac sensor
array 200 has a larger dynamic range than the Mach-Zehnder interferometric sensor array at low
acoustic frequencies because it is not affected by phase noise. is there. Ideally, array 200
provides sufficient dynamic range to detect the strongest and weakest acoustic signals that are
likely to occur. This requirement often leads to the required dynamic range of about 150 dB. To
achieve such a large dynamic range in a Mach-Zehnder interferometric sensor array, two
separate sensors, each detecting a portion of the 150 dB total dynamic range and having
different phase response, are required . The obvious disadvantage to this scheme is that it
requires two sensor arrays (ie twice as many hydrophones, rungs, sources and detectors). In fact,
an array capable of supporting N hydrophones can only detect an acoustic signal at N / 2 points.
[0033]
In the Sagnac sensor array 200, a large dynamic range can be achieved without the use of
additional hydrophones 212. Since the phase response in the Sagnac sensor array is a function of
the hydrophone response and the delay loop length as shown in equation 1, the phase response
of the entire array of hydrophones is by modifying the delay loop length It can be changed. As
shown in the modified sensor array 266 in FIG. 7, by simultaneously using two separate delay
loops 214 (1) and 214 (2) of lengths L1 and L2, respectively, the detection range of the array
266 is dramatic. Can be increased. Array 266 now has 2N separate Sagnac loops. Each
05-05-2019
17
hydrophone 212 (i) returns a separate signal for each of the two delay loop paths, and the length
of each delay loop 214 (1), 214 (2) determines the acoustic detection range of that signal. The
acoustic detection range sum of each hydrophone 212 (i) is a combination of the detection
ranges of each of the two Sagnac loop sensors that surround hydrophone 212 (i). The lengths of
L1 and L2 set the sound detection range. The length L1 + L2 is chosen so that the array 266 can
detect the minimum acoustic signal of interest. And the length L1 of the delay loop 214 (1)
positions the detection range of the signal traveling only on this shorter delay loop above the
detection range of the signal traveling on both the delay loops 214 (1) and 214 (2) To be chosen.
In a TDM system, as a result of the insertion of the second loop, the repetition frequency of the
source pulse is halved in anticipation of the time for 2N pulses to return, and of the delay loops
214 (1), 214 (2). The length is chosen such that there is no overlap of pulses. As the repetition
frequency is halved, the dynamic range of each individual signal is reduced by 3 dB. This
reduction is more than offset by the increase in overall dynamic range achieved by piggybacking
the dynamic range of the two separate signals. In FIG. 7, the second delay loop 214 (2) is
positioned such that all light passing through the second delay loop 214 (2) passes through the
first delay loop 212 (1). Alternatively, the two delay loops 214 (1), 214 (2) are optically such that
light passing through the second delay loop 214 (2) does not pass through the first delay loop
214 (1). It is understood that it may be parallel.
In such a case, the fiber length of the second delay loop 214 (2) would have to be the sum of the
first and second lengths (ie, L1 + L2). However, this adjustment is not essential as L1 is much
shorter than L2. The embodiment of FIG. 7 reduces the overall fiber requirements by adding the
length of the first delay loop to the second delay loop.
[0034]
FIG. 8 is an expanded dynamic range enabled by using two delay loops 214 (1), 214 (2) in array
266, wherein the dynamic range of each signal is 100 dB and L1 / L2 Indicates that the ratio is
set to 5000. As shown, the array 266 can now detect the entire dynamic range of interest (about
160 dB range) without increasing the number of hydrophones.
[0035]
In distributed sensing Sagnac sensor array 266, any phase modulation in the interferometer can
be transferred to intensity modulation with interfering 3 × 3 coupler 220. This distributed
sensing across the Sagnac loop is disadvantageous for acoustic sensor arrays. To be practical, the
05-05-2019
18
acoustic sensor array should sample the acoustic signals at several discrete points in space (i.e.
with hydrophones) and return these signals independently. The Mach-Zehnder interferometric
sensor array achieves this because the interferometer is confined to a small space and thus
senses only at that point. In order for the Sagnac sensor array 266 to be practical, the distributed
sensing of the Sagnac loop must be reduced.
[0036]
Most of the fibers of the interferometer make up the delay loop 214, which can be placed in two
positions. The first one comprises a source 222 at the dry end (i.e. outside the water) and
detection electronics (i.e. the detector 230 and the detector 232) as shown in Fig. 9A. Here, the
delay loop 214 can be shut down environmentally to minimize any external modulation.
However, the downlead fibers 270, 272 connecting the wet end to the array portion 210 are part
of the interferometer. The second possibility is to position the delay loop 214 at the wet end (ie,
underwater) as shown in FIG. 9B. Thus, the delay loop 214 is not separable to the same extent as
if it were positioned at the dry end, but the downlead fibers 270, 272, 274 are unsensing as they
are external to the interferometer. The relative size of the downlead and the delay loop
distributed pick up determines which configuration is optimal for a particular application. If the
delay loop 214 is located at the dry end (FIG. 9A), then the downlead fibers 270, 272 will remain
stationary, bending or vibrating these fibers which may lead to very large phase modulations. It
should be noted that physical movement such as must be prevented. These are phase
modulations induced by fiber motion as opposed to acoustically induced phase modulations.
(Such physical movement is a problem in towed arrays, but not a significant problem in
stationary arrays). Thus, if the delay loop 214 is positioned at the dry end (FIG. 9A), the entire
wet end of the Sagnac sensor array 210 must be stationary. However, if the delay loop 214 is
located at the wet end (FIG. 9B), only the right portion of the 3 × 3 coupler 220 of FIG. 9B
should remain stationary, but this is a downlead fiber 270, This is because 272 and 274 are not
part of the interferometer. If the delay loop is located at the wet end (FIG. 9B), the delay loop
fiber must be desensitized. The delay loop 214 can be made stationary by wrapping the delay
loop fiber around a desensitization cylinder (not shown), thus eliminating fiber movement and
providing acoustic pickup of the main distributed pick-up signal source. Because the sensitivity of
the fiber to acoustically induced phase modulation is easier to suppress than the sensitivity of the
fiber to motion induced phase modulation, the configuration for locating the delay loop 214 at
the wet end (FIG. 9B) is for towed array applications Are preferred and are described in more
detail below.
[0037]
05-05-2019
19
Calculation of acoustic pickup noise induced in the delay loop In this section, an estimate for the
magnitude of the acoustically induced dispersive pickup noise compared to the acoustically
induced hydrophone phase modulation in the Sagnac sensor array 210 of FIG. 9 (b) is derived Ru.
Intensity modulation with dispersive phase modulation resulting from the pickup of acoustic
signals in delay loops and bus fibers (fibers connecting each hydrophone to delay loops and 3 ×
3 couplers) is considered as a noise source. In the following description, one loop of the Sagnac
sensor array comprises only one delay fiber of length Ld, a bus fiber of length Lb, a hydrophone
fiber of length Lh, and a total length L, as shown in FIG. I think that it includes. Also assume that
Ld is longer than Lb and Lh. The phase response of the fiber to the acoustic signal results from
the pressure dependent propagation constant β. In general, the pressure dependent component
of the propagation constant at position l and at time t can be expressed as: β (l, t) = β0R (l) P (l,
t) (2) where β0 is Zero pressure propagation constant, R (l) is normalized fiber phase response,
P (l, t) is pressure as a function of space and time.
[0038]
[0039]
Where P0 is the steady state pressure, Pm is the amplitude of the pressure modulation (assumed
to be independent of l), and θ (l) contains the spatial phase variation of the acoustic wave.
In general, the induced phase difference between interfering beams in the Sagnac loop with
acoustically induced phase modulation from l = l1 to l = l2 is obtained from the following
integration:
[0040]
[0041]
Where v is the speed of light in the fiber and L is the length of the loop.
Substituting equation 3 into equation 4, we get:
05-05-2019
20
[0042]
[0043]
Equation 5 can be used to determine the phase difference between the interfering beams due to
the acoustic modulation of the hydrophone, the bus and the delay fiber.
[0044]
For hydrophone fibers, Equation 5 can be integrated from l1 = ld + lb / 2 to l2 = ld + lb / 2 + lh.
It is assumed that θ (l) is constant over this range (ie, the acoustic wavelength is greater than the
hydrophone dimension).
Also assume that the normalized fiber phase response R (l) is constant and equal to Rh in this
range. Equation 5 determines the phase difference amplitude between interfering beams due to
hydrophone fiber modulation:
[0045]
[0046]
Note that equation 2 fits the equation given in equation 1.
For bus fibers, Equation 5 first integrates from l1 = ld to l2 = ld + lb / 2 and then l1 = L-lb / 2 to
l2 = L to include both upper and lower bus lines. Again, R (l) is assumed to be constant and equal
to Rb for all bus fibers, so θ (l) is constant in the integral of equation 5. The phase difference
amplitude between interfering beams due to fiber modulation is as follows:
[0047]
05-05-2019
21
[0048]
For delay fibers, Equation 5 is integrated from l 1 = 0 to l 2 = l and, as above, θ (l) is constant
over this range (ie, the delay loop coil is Also small), R (l) is constant and equal to Rd across the
integral.
Equation 5 gives the phase difference amplitude between the interfering beams with delay fiber
modulation as follows:
[0049]
[0050]
Equations 6-8 can be used to calculate the relative magnitudes of these phase modulation
amplitudes.
First, for example, JA Bucaro et al. "Optical fiber sensor coatings", optical fiber sensors, and the
Procedures of NATO Advanced Study Institute, 1986, pp. 321-338. As noted, it should be noted
that standard plastic coated fiber has a normalized phase response R of -328 dBre 1 / μPa. For
example, CC Wang et al., "Very high responsivity fiber optic hydrophones for commercial
applications", Proceedings of the SPIE-The International Society for Optical Engineering, Vol. As
described in 2360, 1994, pp. 360-363, a fiber wound around a current hydrophone, fabricated
from an air-backed mandrel, is -298 dB re It has a normalized phase response of 1 / μPa and is
increased by 30 dB over standard fibers. Assuming that the delay loop and bus fiber have the
normalized phase response of a standard plastic coated fiber and the hydrophone fiber is wound
around an air-backed mandrel, the ratio of Rh to Rb or Rd is about 30 dB. Thus, under the
simplified assumption to arrive at equations 6 to 8, the following is obtained:
[0051]
[0052]
05-05-2019
22
Thus, despite the fact that the hydrophone fiber occupies a relatively small amount in the entire
Sagnac loop, the acoustically induced phase modulation in the hydrophone fiber is a delay loop
fiber and bus fiber, even for the farthest hydrophone Greater than acoustically induced phase
modulation in
The following sections describe means for dealing with this level of distributed pick-up noise with
an empty rung.
[0053]
To evaluate the integral in equation 5 for the delay loop fiber, assume that R (l) = Rd, and all l are
smaller than Ld. Since R (l) is constant, there is no contribution to the integral from l = (l−Ld) to
Ld in Equation 5 (since the integrand is an odd function for L / 2) . However, coiling long fibers
results in some dependence on l in R (l) (possibly because the inner layer of the fiber has a
different R than the outer layer). These variations in R (l) increase the delay loop pickup from l =
L-Ld to Ld. In order to reduce this pickup, R (l) needs to be an even function only around L / 2 in
order to make the integrand of equation 5 an odd function with respect to L / 2. Please note. As
shown in FIG. 11, R (l) can be made more symmetrical with respect to L / 2 by winding the delay
loop in such a manner as to arrange the symmetrical points of the fiber loop next to each other.
Such a winding ensures that the points of interest of the delay loops are placed close to one
another, so any variation in R (l) due to the position of the fiber on the coil is possible for L / 2 It
is as symmetrical as possible, so the delayed pick-up approaches as close as possible to the
equation of equation 8. Each Sagnac loop in the Sagnac sensor array has a different L / 2 point,
and only one loop can be wound as shown in FIG. 11, whereby R (all but one of the Sagnac loops)
Note that we introduce small oddness in l).
[0054]
In addition to improving the acoustic sensitivity of the fiber in hydrophones, it is also possible to
suppress the sensitivity of the fiber by applying a metal coating of a specific diameter. (See, for
example, J. A. Buccaro's "Fiber Optic Sensor Coating" cited above). A measured normalized phase
response as low as −366 dB re 1 / μPa has been reported. If such fibers are used in the delay
or bus lines, the ratio of Rh to Rb or Rh to Rd approaches 68 dB (except 30 dB for plastic coated
delay and bus fiber), and the hydrophone induced signal , 38 dB over delay and bus induction
05-05-2019
23
signals.
[0055]
Reducing distributed pick-up noise with empty rungs To further eliminate the distributed pick-up
signal, place hydrong-free empty rungs 300 in the array 210 as shown in FIG. It can be separated
from dispersive pick-up modulation. One empty rung 300 (i) is placed in front of each rung 216
(i) that includes hydrophones 212 (i), referred to as sensing rungs. The fact that the non-sensing
fiber confining the empty rung (i) of each loop is approximately identical to the non-sensing fiber
confining the corresponding sensing rung 212 (i) of the loop is the empty rung 300 (i) and the
corresponding It means that the sensing rungs 212 (i) will have approximately the same
distributed pick-up signal. Treat this empty rung 300 (i) as another sensor in the array 210 and
properly pulse (in TDM) the pulses so that they do not overlap from the empty rung 300 (i) and
the sensing rung 212 (i) Thus, it is possible to measure the distributed pick-up signal appearing
on each sensing rung 212 (i). After detection, this signal can be removed from the sensing rung
signal, leaving only the intensity modulation generated by phase modulation in the hydrophone
fiber. Implementing such a strategy would reduce the duty cycle of the individual signals by half
since 2N rungs would be required for the N sensor array 210.
[0056]
If it is not necessary to de-sensitize the bus portion of array 210, then a single empty rung 300
can be placed in array 210 to measure the distributed pick-up signal associated with delay loop
214, so that In contrast, only N + 1 rungs (N sensing rungs 212 (i) and one empty rung 300) are
required. If one empty rung 300 does not adequately measure the distributed pick-up signal for
each sensing rung 212 (i), then the distributed pick-up signal appearing on each sensing rung
212 (i) is the closest of these empty rungs 300. Additional empty rungs 300 can be added along
the array at regular intervals until sufficient measurements can be made. Using fewer empty
rungs results in higher duty cycles for the individual signals. FIG. 12 shows an extreme example
where empty rungs have been added for every sensing rung.
[0057]
Polarization For the maximum contrast in any interferometric sensor, the polarization state (SOP)
of the interfering beam must be identical when recombined. If they are orthogonal, no
05-05-2019
24
interference occurs and thus no amplitude modulation signal is present. This is called
polarization induced signal fading. Because each sensor in the Sagnac sensor array is a Sagnac
loop, the research conducted on polarization-induced signal fading in the Sagnac fiber gyroscope
applies equally to the Sagnac sensor array. One promising solution is to place a depolarizer in the
Sagnac loop. (For example, K. Boehm et al. "LOW-DRIFT FIBER GYRO USING A
SUPERLUMINESCENT DIODE" using superfluorescent diodes, ELECTRONICS LETTERS, Vol. 17,
No. 10, 1981) May 14, pp. 352-353). The depolarizer ensures that at least half of the optical
power always returns to the 3x3 coupler with the correct SOP. This general strategy yields
constant visibility regardless of loop birefringence. (See, for example, William K. Burns et al.,
"Fiber-Optic Gyroscopes with Depolarized Light", JOURNAL OF LIGHTWAVE TECHNOLOGY, Vol.
10、No. 7, July 1992, pp. 992-999). The simplest configuration uses a non-polarizing source
such as a fiber superfluorescent source and a depolarizer in a loop. As shown in FIG. 13, in the
Sagnac sensor array 200, one depolarizer 310 is placed at a point that is common to all Sagnac
loops. The depolarizer 310 ensures that each sensor 212 (i) has this constant visibility
independent of birefringence, as long as the loop birefringence is constant. This shows that the
treatment of polarization induced signal fading is much simpler than the method used in a MachZehnder interferometer sensor array.
[0058]
Slow changes in birefringence will be sufficiently counteracted by the contradictory nature of
Sagnac interferometers, but birefringence modulation at frequencies in the relevant acoustic
range results in polarization noise. Most birefringence at these frequencies occurs as a result of
physical fiber motion. Thus, the Sagnac loop should remain stationary in order to reduce
polarization noise (and the dispersive pick-up signal).
[0059]
Sources of noise introduced by the use of Sagnac interferometer Thermal phase noise As the
refractive index of the fiber changes with temperature, thermal fluctuations in the fiber cause
phase fluctuations in the light traveling therethrough. Because these rate variations are not
correlated across the length of the fiber, the resulting phase variation is proportional to the
square root of the length. Because the Mach-Zehnder interferometer typically uses less than 100
meters of fiber in each arm, this magnitude of thermal phase noise is negligible. Sagnac
interferometers have so many more fibers in the interferometer that thermal phase noise can be
a limiting noise source. The magnitude of this thermal phase noise in a Sagnac interferometer has
been theoretically described and experimentally confirmed. (Eg, Sverre Knudsen et al.,
05-05-2019
25
"Measurements of Fundamental Thermal Induced Phase Fluctuations in Fibers of a Sagnac
Interferometer", IEEE Photonics, Technology Letters, Vol. 7、No. 1, 1995, pp. 90-93; and Kieler
Krons et al., "Comparison of Fiber-Optic Sagnac and Mach-Zehnder Interferometers for Thermal
Processing in Fibers (Comparison of Fiber-Optic Sagnac and Mach-Zehnder Interferometers with
"Respect to Thermal Processes in Fiber)", JOURNAL OF LIGHTWAVE TECHNOLOGY, Vol. 13、No.
4, April 1995, pp. 682-686). For loops longer than 2 km, thermal phase noise may exceed 1
μrad Hz Hz in the relevant frequency range, which is on the order of the required array
sensitivity.
[0060]
Thermal phase noise can be considered as a distributed pick-up noise source, similar to external
modulation to the delay loop, and thus can be reduced with empty rungs as described above.
Thermal phase noise can also be reduced by shortening the loop length. As described above, by
increasing the hydrophone fiber length by the amount by which the delay loop is shortened, the
loop length can be shortened without changing the low frequency sensitivity. For example, a 40
km delay loop with 50 meters of hydrophone fiber has the same low frequency response as a 20
km delay loop with 100 meters of fiber. However, the latter combination has less thermal phase
noise since the total loop length is almost half.
[0061]
Kerr Effect Induced Phase Noise The car induced phase shift that can occur in Sagnac
interferometers has received great attention for fiber optic gyroscopes. (Eg, RA Bergh et al.,
“Source statistics and the Kerr effect in fiber-optic gyroscopes, OPTICS LETTERS, Vol 7, No. 11,.
Nov. 1982, pp. 563-565; RA Berg et al., "Compensation of the optical Kerr effect in fiber-optic
gyroscopes, OPTICS LETTERS, Vol. 7、No. 6, June 1982, pp. 282-284; and NJ Frigo et al. "Optical
Kerr effect in fiber gyroscopes: Effect of non-monochromatic sources (Optical Kerr effect in fiber
gyroscopes: effecs of nonmonochromatic sources), OPTICS LETTERS, Vol. 8、No. 2, February
1983, pp. 119-121). However, because gyroscopes measure DC levels, the requirements for
gyroscopes and for acoustic sensors are different. The small DC offsets generated by the car
induced phase shift that can limit the fiber gyroscope are not an issue with acoustic sensors. Kerr
induced DC phase shift is not a problem as long as it does not move the bias point too far from
quadrature. Intensity noise at the light source can produce Kerr induced phase noise at the
output. However, this car induced AC phase noise is small as long as the car induced DC phase
shift remains small. The source of car induced phase shift in the Sagnac sensor array is different
from that in the fiber gyroscope. The asymmetry of the Sagnac sensor array causes such a Kerr
05-05-2019
26
phase shift more easily than a nominally symmetric gyroscope. The asymmetry results from the
placement of the asymmetric array portion and the EDFA, where one beam gains gain before
propagating through the delay loop, then suffers losses while the oppositely propagating beam
suffers losses and then Get gain.
By choosing the appropriate position for the EDFA in the delay loop, it is possible to balance
these asymmetries and eliminate the Kerr induced phase shift. The particular one depends on the
exact array configuration and which multiplexing scheme is used.
[0062]
Nonlinear phase modulation arising from EDFA The population inversion occurring in the EDFA
induces a phase shift in the signal light passing therethrough (e.g. M.J.F. Digonnet et al., "For lowpower all-optical switching Resonantly Enhanced Nonlinearity in Doped Fibers: Overview (A
Review), OPTICAL FIBER TECHNOLOGY, Vol. 3、See No. 1, pp. 44-64). This phenomenon has
been used to make all-optical interferometer switches. In a Sagnac sensor array, the EDFAs in the
interferometer generate non-linear phase shifts through the same mechanism. Variations in
population inversion due to pump or signal power variations result in phase modulation being
converted to intensity noise.
[0063]
To estimate the magnitude of this noise source, one must first determine how the population
inversion will respond to pump and signal power variations. This is relatively straightforward
than done by the velocity equation for the erbium system.
[0064]
[0065]
Where N1 and N2 are the lower and excited state collective densities, respectively, N0 is the
collective density sum, I is the intensity, σ is the cross section, and Aeff is the effective mode
region in the fiber , Τ2 is the lifetime of level 2.
05-05-2019
27
Subscripts p and s indicate pump and signal, respectively, and superscripts a and e indicate
absorption and emission, respectively.
[0066]
Dividing N1, N2, Ip, Is into their steady-state and time-varying components, and then substituting
this into Equation 12 to combine Equation 12 and Equation 11, the results are as follows:
[0067]
[0068]
Here, the subscript ss indicates a steady state value, and the time varying component is written
as an explicit function of time (N2 = N2 <ss> + N2 (t)).
Assuming that N2 (t) is smaller than N2 <ss>, the last two terms of Equation 13 can be ignored.
Ip (t) = Ip <m> sin (fpt) and Is (t) = Is <m> sin (fst) (where Ip <m> and Is <m> are Ip (t) and Is (t)
Denoting the modulation amplitudes respectively, fp and fs respectively represent the pump and
signal modulation frequencies) and solving the resulting differential equation yields:
[0069]
[0070]
If λp = 1480 nm, λs = 1550 nm, and Ip <ss> = 1 W, and assuming a typical erbium-silica cross
section, equations 14 and 15 are simplified as follows:
[0071]
[0072]
05-05-2019
28
Pump induced population inversion (Equation 17) is analyzed first.
It is assumed that Is <ss> = 1 mW, Ip <ss> = 1 W and Ip <m> / Ip <ss> = 10 <−6> / √Hz (120 dB
/ √Hz electronic SNR) Then, at a frequency sufficiently lower than 4.3 kHz, | N2 (fp) | N2 <ss> =
9 × 10 <−10> √Hz <−1>.
To convert this number into phase modulation, one can use the fact that a pump power of 10
mW absorbed in an erbium-doped fiber induces a phase shift of about 7 radians at 1550 nm.
(See, for example, M. J. F. Digone et al., "Resonantly Enhanced Nonlinearity in Doped Fiber for
Low-Power All-Optical Switching: An Overview", OPTICAL FIBER TECHNOLOGY, Vol. 3、No. 1,
January 1997, pp. 44-64). Using simulation, a pump power of 10 mW absorbed into a typical
erbium-doped fiber results in a small signal gain of about 6 dB at 1550 nm, which is the gain
required by each amplifier in an array with distributed EDFAs It is close. (For example, Craig W.
Hodgson et al. "Optimization of large scale fiber sensor arrays incorporating multiple optical
amplifiers-Part 1: Optimization of Large-Scale Fiber Sensor Arrays Incorporating Multiple Optical
Amplifiers-Part I: Singal-to-Noise Ratio "; Craig W. Hodgeson et al." Optimization of large scale
fiber sensor arrays incorporating multiple optical amplifiers-Part 2: Pump power (Part II) "Pump
Power"; Jefferson L. Wagener et al. "Novel Riber Sensor Arrays Using Erbium-Doped Fiber
Amplifiers" using erbium-doped fiber amplifiers; See W. Hodgeson et al. "Large scale
interferometric fiber sensor array with multiple optical amplifiers"). Thus, each amplifier provides
a DC phase shift of about 7 radians. Since the nonlinear phase shift is proportional to the higher
state population N2, we can write ΔN2 / N2 <ss> = Δφ / φ <ss>. Using this relationship and
Equation 17 again for Is <ss> = 1 mW, Ip <ss> = 1 W, Ip <m> / Ip <ss> = 10 <-6> / √Hz and fs
<4.3 kHz, The low frequency phase noise induced by each EDFA is (7 radians) × (9 × 10
<−10>) √Hz <−1> = 6.3 × 10 <−9> rad / √Hz.
Assuming that there are a total of 500 such amplifiers, and the phase modulations from all of the
500 amplifiers are coherently summed, the sum of the pump noise induced phase shifts is
estimated to be 3.2 μrad / √Hz. The target phase noise floor function is typically set to 1 μrad
/ √Hz, and non-linear phase noise induced by the EDFA due to pump power fluctuations
approximates the required phase noise floor function, but more It shows that it is not remarkably
large. In practice, the phase modulation of the amplifier does not add coherently, so the value of
3.2 μrad / √Hz is low.
05-05-2019
29
[0073]
The calculation of the phase shift induced by the signal power variation is more complicated,
because the signal power not only has intensity noise but is also modulated by the multiplexing
scheme. Considering again the case of TDM, in general, while a given pulse travels through a
particular EDFA, there may or may not be pulses propagating in the opposite direction
simultaneously through that EDFA. . Considering the worst case that there are always pulses
propagating in the opposite direction, Is <m> is twice the intensity noise of each pulse. For
amplifiers, Is <m> is typically 1.5 to 2 times the intensity noise of each pulse. Assuming that the
signal light has an electronic SNR of 120 dB / √Hz at the acoustic frequency (ie, Is <m> / Is <ss>
= 10 <-6> / √Hz), this value is Ip <ss> = 1 W and Inserting into Equation 18 with Is <m> = 2 mW,
| N2 (fs) | / N2 <ss> is approximately 2.4 × 10 <-9 >> Hz <-1> at frequencies lower than 4.3 kHz
The phase noise induced by signal strength noise in each EDFA is calculated to be 1.68 × 10
<−8> rad / √Hz. Again, assuming 500 amplifiers and coherent summation of all EDFA induced
phase modulations, the total EDFA induced phase noise in each pulse is 8.4 μrad / √Hz, which
again limits the performance of the Sagnac sensor array Level that could be However, more
accurate calculations require more detailed studies that take into account the exact timing of the
multiplexing scheme and the array.
[0074]
Multiplexing Schemes in Sagnac Arrays Time Division Multiplexing So far, it has been assumed
that Sagnac sensor arrays are operated in a TDM configuration. Note that in the Sagnac sensor
array, the source requirements for such a TDM system are not as stringent as that of the MachZehnder interferometer sensor array in a TDM configuration. The reason for this is the
broadband source used in Sagnac sensor arrays. In a Mach-Zehnder Interferometer sensor array,
light from adjacent rungs is coherent due to the narrow linewidth source, and therefore
extremely high extinction ratios are required at the input pulse to prevent multipath coherent
interference. This high extinction ratio requirement is achieved by placing a large number of
modulators in series, which results in a complex, lossy and expensive source. In Sagnac sensor
arrays, the extinction ratio does not have to be very high because the broadband source
eliminates the possibility of multipath coherent interference. Furthermore, the narrow linewidths
required by the Mach-Zehnder Interferometer sensor array do not allow the use of a pulsed laser
source instead of a continuous wave (cw) laser source that is externally modulated by a lithium
niobate intensity modulator. In the Sagnac sensor array, the source may be constructed using
either an externally modulated continuous wave ASE source, a pulsed ASE source, or some
combination thereof. Again, this is because the Sagnac sensor array does not require a narrow
05-05-2019
30
linewidth source. Although the invention does not require a narrow linewidth source, it should be
understood that the Sagnac sensor array of the invention can also be used with narrow linewidth
sources, such as lasers.
[0075]
Frequency Division Multiplexing By using a broadband source, Sagnac sensor arrays can also
operate in non-TDM configurations without the need for design changes or additional sources.
Frequency Division Multiplexing (FDM) is commonly used in Mach-Zehnder Interferometer
sensor arrays using a Phase-Generated Carrier (PGC) scheme, but is also compatible with Sagnac
sensor arrays. FIG. 14 shows a basic Sagnac sensor array 400 using the FDM scheme. A fiber
superfluorescent source (SFS) 402 (or other broadband source such as, for example, an LED)
generates input light. Chirped intensity modulation is applied to the input light via an intensity
modulator 404 controlled by a chirped frequency generator 406. Modulated light enters sensor
array 410 via 3 × 3 coupler 412. The light passes through the delay loop 414 and a plurality of
sensing rungs 416 (i) with respective sensors 418 (i). An empty rung (not shown) may also be
included if desired. After passing through the delay loop 414 and the rung 416 (i), the light exits
the sensor array 410 via the coupler 412 and is detected by the detector 420, which responds to
the detected light with the light output signal. Generate The electrical output signal from detector
420 is mixed in mixer 422 with the same chirp frequency delayed by delay 424 which delays the
chirp frequency by time Δt. In the setup shown in FIG. 14, the output of mixer 422 is provided
to spectrum analyzer 426. In an embodiment of operation, the output of mixer 422 is provided to
a signal processing subsystem (not shown), which analyzes the output of mixer 422 to
regenerate the acoustic signal that enters array 410.
[0076]
The signals returning from the sensors 418 (i) in the various rungs 416 (i) are further delayed
with respect to the delayed chirp frequency. This is illustrated in the graph of FIG. 15 with the
original chirp frequency 450, the delayed chirp frequency 452 from delay 424, the chirp return
signal 460 from the first rung, the chirp return signal 462 from the second rung, and the third
As the chirp return signal 464 from the rung. In mixer 422, separate beat frequencies fb 1470,
fb 2472 and fb 3474, respectively (shown in FIG. 14) are formed between mixing chirp
frequency 452 and each of the signals returning from the various rungs in Sagnac sensor array
410. . (For example, SF Collins, SF et al., “A Multiplexing Scheme for Optical Fiber
Interferometric Sensors Using FMCW Generated Carrier”, OFC '92 Conference Proceedings , pp.
209-211). Although only three chirp return signals 460, 462, 464 are shown in FIG. 15, it is
05-05-2019
31
contemplated that up to N return signals may be provided, where N is the number of rungs in the
array 410. The chirp return signal from the Nth rung provides a beat frequency fb N at mixer
422.
[0077]
As shown in the spectral output diagram of FIG. 14, the acoustic modulation of the signal appears
as upper side bands 480, 481, 482 and lower side bands 484, 485, 486 relative to the beat
frequency. The advantage of this FDM scheme is that the requirements for array timing are much
less than for TDM systems. TDM systems require specific delays between adjacent rungs to
prevent pulse overlap, which can present severe technical problems. In FDM, variations in fiber
length shift the beat frequency but do not cause overlap between the signals as long as the beat
frequencies are separated by twice the acoustic detection range. This is achieved by selecting the
appropriate char plate. Unlike TDM systems, all paths always return light, which can lead to
phase noise between different non-coherent signals. Broadband ASE sources minimize the
magnitude of this phase noise. (For example, Moslehi, "Analysis of Optical Phase Noise in an
Optical Fiber System Using a Laser Source with Arbitrary Coherence Time, Fiber-Optic Systems
Employing a Laser Source with Arbitrary Coherence Time"). Journal of Lightwave Technology,
Vol. LT-4, No. 9, September 1986, pp. 1334-1351).
[0078]
Code Division Multiplexing Code Division Multiplexing (CDM) has recently attracted attention for
use in sensor arrays. (E.g., AD Kersey et al., "Code-division Multiplexed Interferometric Array
With Phase Noise Reduction And Low Crosstalk" with reduced phase noise and low crosstalk)
OFS '92 Conference Proceedings, pp. 266H.-269; and HS Al-Raweshidy et al. "Spread spectrum
technique for passive multiplexing of interferometric fiber sensors" SPIE , Vol. 1341 See Fiber
Optics '90, pp. 342-347). As shown in the Sagnac sensor array 600 of FIG. 16, in CDM, the input
light from the fiber superfluorescent source 602 (or other broadband source such as, for
example, an LED) is pseudo-randomly generated by the code generator 606 Modulated in the
intensity modulator 604 according to the code. Modulated light is provided to interferometer
loop 608 via 3 × 3 coupler 610 and propagates through delay loop 614 and a plurality of rungs
616 (i) in array 612. In the illustrated embodiment, each rung 616 (i) includes a respective
sensor 618 (i). It may also include an empty rung (not shown) if desired. The light returns from
the 3 × 3 coupler 610 and is detected by detector 620. The electrical output of detector 620 is
provided to correlator 622 along with the output of code generator 606, which is delayed by a
delay 624 for a period of τ cor. The bit duration for the pseudorandom code is less than the
05-05-2019
32
propagation delay between adjacent rungs in array 612. If .tau.cor is equal to one of the loop
travel times .tau.j through each rung 616 (i), then the signal returning from this sensor in rung
616 (i) is correlated to the delayed pseudorandom code.
[0079]
[0080]
The correlation process multiplies the detected signal by 1 or -1 (or depending on whether the
correlation code is on or off, for example (or the signal at electronic gate 630, non-inverted of
differential amplifier 632, and Gated to the inverting input).
The output of the differential amplifier at line 634 is the correlated output. The signal is then
time averaged over a period tavg equal to the duration of the code. The uncorrelated signal is
time averaged to zero, thereby separating the signal from the sensor 618 (i). τ cor is scanned to
retrieve the signals from all sensors sequentially.
[0081]
The advantage of CDM over TDM is that there is no need to accurately control the delay between
sensors. Any loop delay τj can be accepted, where | τj−τj ± 1 |> τbit (where τbit is the
duration of the pulse in the code). The correlation requires that τ j be known, but this is easily
measured. In FDM, the use of a broadband source has the advantage of reducing the phase noise
caused by adding all the signals.
[0082]
Above, a novel design for acoustic sensor arrays based on Sagnac interferometers has been
described. The main advantage of this design is the use of a common path interferometer. This
prevents the conversion of source phase noise to intensity noise, which is common in MachZehnder interferometer sensors, and allows the use of cheap, high power ASE sources or other
broadband sources. The response of the Sagnac sensor array as a function of acoustic frequency
is shown to match the ocean noise floor. This design also makes it possible to dramatically
05-05-2019
33
increase the dynamic range without adding hydrophones by using one additional very short
delay loop. Techniques for eliminating polarization induced signal fading are described above.
The Sagnac sensor array also makes it possible to use several multiplexing schemes in a simpler
form than a standard Mach-Zehnder array. These features make the Sagnac sensor array design a
very promising alternative to Mach-Zehnder interferometer based sensor arrays.
[0083]
Folded Sagnac Sensor Arrays FIGS. 17-20 show an alternative embodiment of a distributed
acoustic sensor array based on the Sagnac effect, which has an architecture that has been
modified to reduce the distributed pickup from the downlead fiber. In particular, FIG. 17 shows a
basic folded Sagnac acoustic fiber sensor array 700 that includes a source 702, a first detector
704, and a second detector 706. Preferably, source 702, first detector 704 and second detector
706 are positioned at the dry end of sensor array 700 (e.g., shore or ship).
[0084]
Source 702 generates a light pulse, which is coupled with 3 × 3 coupler 710 via downlead fiber
708. As shown, the 3 × 3 coupler is positioned at the wet end (eg, near the seabed). The 3 × 3
coupler 710 has a first output port coupled to one end of a common fiber rung (Rung 0) 712 and
a second output port coupled to a first array input / output fiber 714 of the array 716, the first
The three output ports are terminated non-reflectively. Approximately 33% of the light from
source 702 is coupled to each of the first and second ports of the 3 × 3 coupler so that
approximately 33% of the light propagates to common fiber rung 712 and approximately 33% of
the light is arrayed Propagating to 716 As noted above, although illustrated here as a 3 × 3
coupler 710, other n × m couplers (eg, 2 × 2 couplers, 4 × 4 couplers, etc.) are described in
the example of FIG. 17 and below. It can be used in alternative embodiments of the invention.
[0085]
Array 716 includes a plurality of rungs 718 (i) (ie, 718 (1), 718 (2)... 718 (N)), which include a
first array input / output fiber 714 and a second array input / Coupled with the output fiber 720.
Each rung 718 (i) includes a respective acoustic sensor (ie, a hydrophone) 722 (i). Array 716
preferably includes a distributed erbium-doped fiber amplifier (EDFA) 724 as described in
connection with FIG. (The pump source for EDFA 724 is not shown in FIG. 17). Although
05-05-2019
34
described herein with respect to array 716, other array configurations may be advantageously
employed in the present invention.
[0086]
The second array input / output fiber 720 couples the array 716 to the first port of the 2 × 2
coupler 730. The second end of common rung (rung 0) 712 is coupled to the second port of 2 ×
2 coupler 730. Although array 716 is described herein as including a plurality of sensors 722 (i),
the invention is also applicable to sensor systems having only a single sensor 722.
[0087]
The third port of 2 × 2 coupler 730 is terminated non-reflectively at termination 732. The
fourth port of 2 × 2 coupler 730 is coupled to delay loop downlead fiber 740. The delay loop
downlead fiber 740 couples the fourth port of the 2 × 2 coupler to the first end of the delay
loop 750. The delay loop 750 is positioned either at the dry end or at the wet end as shown. The
second end of the delay loop 750 is coupled to the reflector 752, whereby light exiting the
second end of the delay loop 750 is reflected to the delay loop 750 and propagates through the
delay loop 750 and the delay loop downlead fiber 740 propagates back to the fourth port of the
2 × 2 coupler 730. The light returned from the loop down lead fiber 740 is split by the 2 × 2
coupler 730 and substantially equal portions propagate through the common rung 712 and the
array 716, both portions towards the 3 × 3 coupler 710. To propagate. The two parts are
combined in the 3 × 3 coupler 710 where light pulses traveling the same distance on the array
716 and the common rung 712 interfere but light pulses traveling different distances do not. The
signal generated by the interference is output from the 3 × 3 coupler 710 as the first and
second output signals, and is transmitted to the first detector 704 via the first detector downlead
fiber 770, respectively. It propagates towards the second detector 706 via the downlead fiber
772. Detectors 704, 706 generate electrical output signals, which are analyzed in a conventional
manner by an electronic device (not shown) to regenerate the acoustic signal that enters sensor
722 (i). As described below, signals interfering within the 3 × 3 coupler 710 return to each
sensor 722 (i) at different times, so that time division multiplexing, frequency multiplexing, code
division multiplexing, etc., as described above Can be divided by Non-interfering signals are
ignored because they do not produce detectable output signals.
[0088]
05-05-2019
35
The embodiment of FIG. 17 inserts a depolarizer (not shown) into one of the first segments 712,
714 or 720 with a non-polarizing source, as described above for the Sagnac interferometer. It
can be further deformed. Such an embodiment is described below in connection with FIGS. 23A,
23B and 23C.
[0089]
Here, the light in a single pulse from source 702 is tracked through sensor array 700. Source
pulses from the source 702 are output and travel toward the source down lead 708 and travel
from the 3 × 3 coupler 710 to the common rung 712 and the array 716. Together, the common
rungs 712 and the N rungs 718 (i) in the array 716 provide N + 1 separate paths for the source
pulse to travel to the 2 × 2 coupler 730. Because there are N + 1 separate paths for the source
pulse to travel, the source pulse is split into N + 1 separate pulses, which travel through the 2 ×
2 coupler 730 through the delay loop down lead 740 to the delay loop 750 Do. After passing
through the delay loop 750, the N + 1 pulse is reflected by the reflector 752 and then passes
back through the delay loop 750 back to the delay loop down lead 740 to the 2 × 2 coupler 730
at the wet end as still separate N + 1 pulses. To reach. Again, each of the N + 1 pulses is divided
into N + 1 pulses in common rung 712 and N rung 718 (i). After passing back through common
rung 712 and rung 718 (i), the (N + 1) <2> pulse is synthesized in 3 × 3 coupler 710 and then
the detector down leads 770, 772 are flipped back to dry. , Where the pulses are detected and
analyzed by the first and second detectors 704,706.
[0090]
Since there are (N + 1) <2> possible combinations in the path from source 702 to reflector 752
back to detectors 704, 706, there are (N + 1) <2> return pulses. The only pulses that interfere in
a usable manner are the pairs of pulses traveling in exactly the same path length and in the
opposite order. For the following description, the pulses are identified by two numbers, the first
number identifies the path from the source 702 to the reflector 752, and the second number the
pulses from the reflector 752 to the detectors 704, 706. Identify the route back to For example,
pulse 0, 1 travels through common rung (rung 0) 712, then through delay loop 750 to reflector
752, back through delay loop 750, and then through rung 718 (1). Pulses 1, 0 first pass through
rung 718 (1) and then through delay loop 750 to reflector 752, back through delay loop 750
and then pass through common rung (rung 0) 712. Since the distance traveled by pulses 0 and 1
is the same as the distance traveled by pulses 1 and 0, pulses 0 and 1 interfere with pulses 1 and
0 when they are combined by 3 × 3 coupler 710, thus A common path interferometer is defined
05-05-2019
36
in the same manner as the Sagnac interferometer of S. (ie, the folded Sagnac interferometer).
Acoustic sensing is caused by the hydrophone 722 (1) placed on the rung 1 responsive to
acoustic modulation. The interference pulses 0, 1 and 1, 1 pass through the hydrophone 722 (1)
at different times, thus picking up the phase difference due to the time-varying acoustic
modulation of the hydrophone 722 (1). In 3 × 3 coupler 710, this phase difference is converted
to intensity modulation and sent to detectors 704, 706 through detector down leads 770, 772.
The same effect occurs for pulses 0, 2 and 2, 0, pulses 0, 3 and 3, 0 etc.
[0091]
Because the folded Sagnac interferometer is a common path, the source 702 may have a short
coherent length, which means that interference occurs only between pulses that have passed
approximately the same path. Thus, the pulses i, j only interfere with the pulses j, i. As mentioned
above, there are N interferometers of interest (pulses 0, i interfering with pulses i, 0, where i = 1
to N). There are also many other interferometers that do not include common rung (Rung 0) 712
(eg, pulses 1, 2 interfering with pulses 2, 1; pulses 1, 3 interfering with pulses 3, 1 etc.). Such
interference pulses give noise to useful pulses, referred to herein as noise pulses. These noise
pulses carry two types of noise. As with all pulses, they can carry additional shot noise, ASE
signal beat noise (in an amplified array), phase noise, etc., but they increase the noise detected.
Noise pulses (such as pulses 1 and 2 that interfere with pulses 2 and 1) that form an undesired
interferometer may also carry intensity modulation due to interferometric sensing of the acoustic
wave. This intensity modulation is an undesired signal and can be considered as a noise source.
These undesired interferometers have couplers 780 (1) -780 (N) as interference points, with
rungs 718 (1) -718 (N) coupled to the first input / output fiber 714 of the array 716. It is
important to note that while having, the signal pulses interfere at the 3 × 3 coupler 710.
Because the noise pulse interferes before reaching the 3 × 3 coupler 710, intensity modulation
of the noise pulse is provided symmetrically to both detectors 704 and 706. However, signal
pulses that interfere at the 3 × 3 coupler 710 result in asymmetric intensity modulation. Thus,
by differentially amplifying the current from detectors 704, 706, the intensity modulation of the
signal pulse is increased and the intensity modulation of the noise pulse is decreased, thereby
reducing the noise contribution of the undesired interferometer.
[0092]
In order to completely remove all the noise added by these noise pulses, it is possible to separate
the relevant pulses from the noise pulses by using a time division multiplexing scheme and
selecting the delay length appropriately . In particular, the optical path length from the 3 × 3
05-05-2019
37
coupler 710 through the common rung 712 to the 2 × 2 coupler 730 is chosen to correspond to
the propagation time τ. The optical path length of the fiber portion from the 3 × 3 coupler to
the 2 × 2 coupler 730 through the coupler 710 (1), the first rung 718 (1), and the
corresponding coupler 790 (1) is (N + 1) τ Is chosen to be A portion of the optical path length is
a common path from 3 × 3 coupler 710 to coupler 780 (1) and from coupler 790 (1) to 2 × 2
coupler 730, and a portion of the optical path length is rung 718 (1) Pass through. The optical
path lengths through each of the rungs 718 (i) are preferably selected to be approximately equal.
The total optical path length from the 3 × 3 coupler 710 to the 2 × 2 coupler 730 through the
second rung 718 (2) is 2 × from the 3 × 3 coupler 710 through the first rung 718 (1) by τ
The optical path length from coupler 780 (1) to coupler 780 (2) and the total optical path length
from coupler 790 (2) to coupler 790 (1) are τ so as to be longer than the total length of the
optical path to coupler 2 Is selected. (Ie, the total length of the optical path between the two
couplers 710, 730 through the second rung 718 (2) is (N + 2) τ). The length of each subsequent
additional optical path is chosen to be τ. Thus, the travel time of light from 3 × 3 coupler 710
through rung 718 (i) to 2 × 2 coupler 730 is defined as the delay time Ti of rung 718 (i).
[0093]
Following the above description, Ti is determined by the optical path length through the rung as
follows: Ti = τ i = 0 (for common rung 712) Ti = (N + i) τ 1 ≦ i ≦ N (each sensing rung 718
From (1), 718 (2), etc.) From the above, it can be seen that the optical path length passing the
furthest rung N is (N + N) τ or 2Nτ.
[0094]
The duration of each pulse is chosen to be less than or equal to τ.
Thus, as shown in FIG. 18, the first pulse 800 returning to the 3 × 3 coupler 710 travels from
the source 702 through the common rung 712 (ie, rung 0) to the reflector 752 and to the
detectors 704, 760. Is a pulse back to This pulse has a total propagation time of 2τ (for
propagation time comparison, through the delay loop 750 to the reflector 752 and the
propagation time of each returning pulse is ignored, which means that the propagation time is all
pulses And acts only as an offset (not shown) to the timing diagram of FIG. The next set of pulses
810 back to the detectors 702, 706 are pulses through the common rung 712 in one direction
and through the sensing rung 718 (i) in the opposite direction (ie, pulses 0, 1 and 1, 0; 0, 2 and
2, 0; 0, 3 and 3, 0 to 0, N and N, 0). They have respective propagation times 2τ + Nτ, 3τ +
Nτ, 4τ + N + τ, to (N + 1) τ + Nτ. Thus, all useful pulses are received between time (N + 2) τ
and time (2N + 2) τ (including the duration τ of the last pulse received). In contrast,
05-05-2019
38
interference pulses (i.e., pulses 1, 1, 1, 2, and 2, 1, 1, 3, and 3, 1, ... 2, 2, 3, 3 and 3) which move
sensing rung 718 (i) in both directions. , 2 ... etc) are received as a set of pulses 820 between
time 2 (N + 2) τ and time (4N + 1) τ. Thus, the signal pulse is separated from the noise pulse.
[0095]
For example, in FIG. 18, the number of return pulses as a function of time is plotted for N = 50.
As shown, a single pulse is received at time 2τ. Thereafter, there are no pulses received during
the interval 3τ to 52τ. Then, between 52τ and 102τ, two pulses are received during each
time interval. The noise pulse then returns from time 102τ to time 201τ. In this manner, the
signal pulse is separated from the noise pulse at the correct time, thus preventing the noise pulse
from adding noise to the signal pulse. The electronics (not shown) are easily synchronized to
monitor only the pulses received between time 52τ and time 102τ.
[0096]
It should be noted that source 702 may be activated to send the next pulse at a time interval of
150τ relative to the previous pulse, but in response to the next pulse, the interval of 0τ to 50τ
may be associated with the preceding source pulse. Of the noise pulse that is returned in
response to the .tau. Thus, the next set 830 of useful pulses can begin to arrive at time 201.
Thus, the embodiments of FIGS. 17 and 18 have an overall duty cycle of about one third of the
usable signal information.
[0097]
An advantage of the folded Sagnac acoustic fiber sensor 700 over the Sagnac loop shown in the
previous drawings is that the delay fiber 750 does not sense modulation. Distributed downlead
pickup is a potentially serious limitation to Sagnac acoustic fiber sensors because the downlead is
often quite long and subject to large motion and vibration. In the folded Sagnac acoustic fiber
sensor 700, the source 708 and detector down leads 770, 772 do not sense as they originate
outside the interferometer. The delay loop down lead 740 does not sense because all interference
pulses are split with a small time delay (about 1 microsecond) to travel the same fiber, and thus
undergo the same perturbations. Any low frequency (below about 1 MHz) modulation to the
delay loop downlead and the delay loop itself is considered substantially equal by both
interference pulses and thus does not contribute to the phase difference. Array portion 716 and
05-05-2019
39
common rung 712 constitute the only sensing fiber in interferometer 700.
[0098]
As shown in FIG. 17, a remotely pumped distributed erbium doped fiber amplifier (ETFA) 724
may be positioned across the array 716 to regenerate power as described above.
[0099]
A 3 × 3 coupler 710 is used to passively bias each sensor 722 (i) near quadrature and allow
source noise removal.
Noise rejection occurs when each detector 704, 706 is biased with the opposite slope (since the
aspects of the signal output from the 3 × 3 coupler 710 are phased with respect to each other),
and thus each detection While the phase modulation affects the intensity asymmetrically in the
detector, source excess noise affects the intensity symmetrically in each detector. Thus,
differentially amplifying the detector output adds to the intensity variation induced by phase
modulation, and source intensity noise is removed in the same manner as the signal is removed
from the unwanted interferometer Ru.
[0100]
With respect to FIGS. 17 and 18, it is understood that similar time division multiplexing effects
can be achieved by providing a longer optical path length through common rung 712 and a
shorter optical path length through sensing rung 718 (i). I want to be For example, the common
rung 712 can be selected to advantageously have an optical path length 2Nτ (ie, T0 = 2N), and
the optical path length through the rungs can be selected to be preferably τ, 2τ, 3τ, ... Nτ .
This can be summarized as follows: Ti = 2Nτ i = 0 (for common rung 712) Ti = iτ 1 ≦ i ≦ N
(for each sensing rung 718 (1), 718 (2), etc.) Thus , The first signal to be returned has a light
propagation time of 2τ (again, except for the propagation time through delay loop 750, which is
common to all signals), but this is the first rung 718 (1) in both directions. Is the time required to
go through. The longest delay of any signal passing through one of the sensing rungs 718 (i) in
both directions is 2N for the signal pulse traveling in both directions through the farthest sensing
rung 718 (N). The first available signal to return is from the common rung 712 through the
reflector 752 and back to the first sensing rung 718 (1) and from the first sensing rung 718 (1)
through the common rung Signal resulting from interference with the signal back to 712. The
05-05-2019
40
interference signal arrives at a later time (2N + 1) τ than the last undesired signal. The last
available signal arrives at time (2N + N) τ (ie 3Nτ). Finally, the signal generated by the pulse
back and forth to reflector 752 in common rung 712 arrives at time 4Nτ, which is well
separated from the available interference signal.
[0101]
It is desirable for the acoustic sensor to have the largest possible dynamic range (range of
detectable acoustic modulation amplitudes). The noise performance of the array sets the
minimum detectable phase modulation and the interferometer's non-linear response function
sets the maximum detectable phase modulation (approximately 1 rad) without using
demodulation techniques such as phase generation carrier techniques. Ru. In a Mach-Zehnder
sensor, the mapping from acoustic modulation to phase modulation is a function of hydrophone
response only. Thus, limitations on detectable phase modulation and the mapping of this acoustic
modulation to phase modulation determine the extent of acoustic modulation that the sensor can
detect.
[0102]
In the folded Sagnac acoustic fiber sensor array, the mapping from acoustic modulation to phase
modulation is a function of both the responsiveness of each hydrophone (sensor) 722 (i) and the
length of the delay loop 750. Thus, by changing the length of the delay loop 750, the dynamic
range of the sensor 722 (i) can be adjusted without changing the hydrophone 722 (i) itself. In
addition, if two reflectors 752 (1) and 752 (2) are used, each sensor 718 (i) may have two
different delay loops 750 (1) and 752 as shown in sensor 850 of FIG. It may have (2). This allows
two signals, each sensor 722 (i) having a different dynamic range, to be returned as described
above with reference to FIGS. 7 and 8, thereby causing each sensor 722 (i) to Greatly increase
the total dynamic range. The penalty is the reduction of the duty cycle for each signal by a factor
of 1 / (number of delay loops).
[0103]
FIG. 20 shows a sensor 900 that implements a phase nulling technique similar to that used in
fiber gyroscopes. The delay loop reflector 752 of FIG. 17 is not used in the sensor 900 of FIG.
Rather, the pulse returns via downlead 910 to a previously unused port of 2 × 2 coupler 730.
05-05-2019
41
An optical isolator 912 is inserted in the return down lead 910 to prevent light from traveling
the delay loop 750 in both directions. The sensor 900 of FIG. 20 behaves similarly to the sensor
700 of FIG. 17 with a reflector 752. However, sensor 900 allows phase modulator 920 to be
added and inserted into return down lead 910. The phase modulator 920 individually applies a
phase shift to each pulse when activated. By applying the detected pulse shift to the phase
modulator 920 via the differential amplifier 922, the phase change is nulled, and the necessary
given phase shift in the phase modulator 920 becomes a signal. In this phase nulling method, the
dynamic range of array 900 is limited to only the largest phase shift that phase modulator 920
can provide.
[0104]
FIG. 21 shows a further alternative to the embodiment of FIG. 19, where two delay loops 750 (1)
and 750 (2) are not connected to the same delay loop downlead. Rather, the first end of the first
delay loop 750 (1) is connected to the first delay loop down lead 740 (1), which is similar to that
of FIG. Connected to port 4 The second end of the first delay loop 750 (1) is coupled to the first
reflector 752 (1) as described above. The first end of the second delay loop 750 (2) is coupled to
the third port of the 2 × 2 coupler 730 via the second delay loop down lead 740 (2) and the
second delay loop The second end of 750 (2) is coupled to a second reflector 752 (2). About half
of the light from the 2 × 2 coupler 730 is coupled to each of the down leads 740 (1), 740 (2).
The light of each downlead 740 (1), 740 (2) is delayed in the respective delay loop 750 (1), 750
(2) and reflected back to the 2 × 2 coupler 730 as described above. The reflected light is
coupled to common rung 712 and array 716. The delay loops 750 (1) and 750 (2) have a delay
of 2 × 2 coupler 730 for all N + 1 pulses propagating from the fourth port of 2 × 2 coupler 730
to first delay loop 750 (1). It is selected not to overlap simultaneously with any of the N + 1
pulses propagating from the third port to the second delay loop 750 (2). Thus, the embodiment
of FIG. 21 provides similar functionality to the embodiment of FIG. 19; however, the embodiment
of FIG. 21 is combined and discarded at the third port of the 2 × 2 coupler 730 of FIG. Use the
sun's light.
[0105]
FIG. 22 shows an alternative embodiment of a fiber optic acoustic sensor system 1000 using a
folded Sagnac sensor array. In system 1000, source 1004 is coupled by X-polarizer 1008 to a
first port of 2 × 2 polarization maintaining coupler 1006. Detector 1002 is connected by Xpolarizer 1010 to the second port of 2 × 2 coupler 1006. By combining the light from the fiber
towards the source 1004, a second detector (not shown) may be advantageously included in the
05-05-2019
42
embodiment of FIG. X-polarizer 1008 only passes light from source 1004 having a first
polarization (eg, X-polarization). Thus, polarization maintaining coupler 1006 receives light with
X polarization from source 1004 and couples the light through the third port to common rung
1020 and through the fourth port to sensor array 1022. The sensor array 1022 has a similar
structure to the sensor array 716 of FIG. 17, and similar elements are numbered accordingly.
[0106]
It should be noted that the two X-polarizers 1008, 1010 may be replaced by one or more Xpolarizers at alternative locations in the system 1000.
[0107]
The common rung 1020 is coupled to the first port of a second polarization maintaining 2 × 2
coupler 1032 via an X-polarizer 1030.
Light propagating through the array 1022 first passes through the depolarizer 1034 and then
through the first input / output fiber 714. Depolarizer 1034 combines substantially equal
amounts of X polarized light into X polarized light and Y polarized light. Thus, about 50% of the
light propagates in the array 1022 as X polarized light, and about 50% propagates in the array
1022 as Y polarized light.
[0108]
After passing through the rungs of the array 1022, light travels through the second input /
output fiber 720 and the Y-polarizer 1040 to the second port of the second coupler 1032. The Y
polarizer 1040 puts only Y polarized light into the second coupler 1032. Coupler 1032 combines
the light from array 1022 with the light from common rung 1020. Approximately half of the
light entering coupler 1032 is coupled to light absorption termination 1042 through the third
port of coupler 1032 and approximately half of the light is coupled to downlead fiber 1050 to
provide a first delay loop 1052. Propagating light to the end.
[0109]
05-05-2019
43
The light passes through the delay loop 1052 to the Faraday rotator mirror (FRM) 1054. The
operation of the Faraday rotator mirror 1054 is well known and will not be described in detail
here. Basically, when light is incident on the Faraday rotation mirror 1054 with one polarization,
it is reflected to orthogonal polarization. Thus, X polarized light passing through the common
rung 1020 is reflected as Y polarized light and Y polarized light passed through the array is
reflected as X polarized light.
[0110]
The reflected light returns to delay 1052 and enters the fourth port of coupler 1032. Light is
coupled to common rung 1020 and array 1022. The X-polarizer 1030 in the common rung
passes only the light with X polarization that originally propagated through the array 1022.
Similarly, the Y-polarizer 1040 passes only Y-polarized light that has originally propagated the
common rung 1020.
[0111]
After propagating through the array 1022, the Y polarized return light is depolarized by the
depolarizer 1034 to produce both X polarized and Y polarized light. Light from common rung
1020 enters the third port of coupler 1006 and light from depolarizer 1034 enters the fourth
port of coupler 1006. The light is combined in the coupler and X-polarized light from the two
ports, traveling the same optical path length, interferes and is coupled to the first and second
ports. The portion coupled to the second port propagates through X polarizer 1010 to detector
1002 where an interference signal is detected.
[0112]
It should be understood that only light that originally traveled to and from the Faraday rotation
mirror 1054 in a different path interferes with the coupler 1006. The only light that is allowed to
propagate through the common rung 1020 in the reflected direction is the X-polarized light
originally propagated through the array 1022 as Y-polarized light. Similarly, the only light
allowed to propagate through any of the rungs of array 1022 in the reflected direction is the Ypolarized light originally propagated through common rung 1020 as X-polarized light.
Potentially interfering light travels in both directions through the rung and can not generate the
noise signal described above in connection with the above-described embodiment. Thus, each of
05-05-2019
44
the pulses originally generated in the array 1022 from the reflected pulse that originally traveled
the common rung 1020 is a single one of the pulses originally generated in the array 1022 and
propagated through the common rung 1020 after it is reflected. Can only interfere. Thus, in the
embodiment of FIG. 22, it is not necessary to include an additional delay to separate the usable
signal pulses from the noise pulses.
[0113]
Figures 23A, 23B and 23C illustrate a further alternative embodiment of the invention. The
sensor array 1100 of the embodiment of FIGS. 23A, 23B and 23C is similar to the sensor array
700 of the embodiment of FIG. 17, and like elements are numbered accordingly. The
embodiment of FIGS. 23A, 23 B and 23 C includes a non-polarizing source 1102. The 2 × 2
coupler 730 of FIG. 17 is replaced by a polarizing beam splitter (PBS) 1104 in FIGS. 23A, 23B
and 23C. Using polarization beam splitter 1104 saves about 6 dB of power as compared to
coupler 730 of FIG. 17 and coupler 1130 of FIG. The reflector 752 of FIG. 17 is replaced with a
Faraday rotation mirror (FRM) 1106 similar to the Faraday rotation mirror 1054 of FIG. The 3 ×
3 couplers 710 of FIGS. 23A, 23B and 23C need not be polarization maintaining couplers.
[0114]
Each of FIGS. 23A, 23 B and 23 C includes a depolarizer 1110. In FIG. 23A, the depolarizer 1110
is located on the first array input / output fiber 714. In FIG. 23B, depolarizer 1110 is located on
common rung 712. In FIG. 23C, the depolarizer 1110 is located on the second array input /
output fiber 720.
[0115]
In the example of FIG. 23A, light from unpolarized source 1102 enters 3 × 3 coupler 710 and
approximately equal portions are coupled to common rung 712 and first array input / output
fiber 714. As described above in connection with FIGS. 3 and 17, the use of a 3 × 3 coupler
provides a passive bias close to quadrature. Light propagating in the first array input / output
fiber 714 passes through the depolarizer 1110, which is substantially half of the light entering
the array in one polarization (eg, X polarization) is orthogonally polarized (eg, Y polarization) )
And likewise has the effect of causing half of the light entering the array with Y polarization to be
combined with X polarization. Thus, after depolarizer 1110, half of the X-polarized light was
05-05-2019
45
generated in X-polarization and the other half of the X-polarized light was generated in Ypolarization. Similarly, after depolarizer 1110, half of the Y-polarized light was generated with Ypolarization and the other half of the Y-polarized light was generated with X-polarization.
Effectively, the depolarizer 1110 scrambles the unpolarized light.
[0116]
Light passes through the array 716 in the manner described above in connection with the other
embodiments. Light exiting array 716 propagates through a second array input / output fiber
720 to a first port 1121 of polarizing beam splitter 1104. Polarization beam splitter 1104 splits
the incident light into two orthogonal polarizations (ie, X polarization and Y polarization). For the
purpose of this description, polarizing beam splitter 1104 acts like a polarization dependent
mirror oriented at 45 °, and light entering one port 1121 with one polarization (eg X
polarization) is a second port 1122 It is assumed that light reflected back into the first port 1121
with the other polarization (eg, Y polarization) is transmitted to the third port 1123. In the
illustrated embodiment, light exiting the second port 1122 is absorbed non-reflectively by the
termination 732. Y-polarized light exiting third terminal 1123 propagates through delay loop
downlead fiber 740 and through delay loop 750 to Faraday rotator mirror 1106. Note that this Y
polarized light from array portion 716 travels through depolarizer 1110, half of which was
originally X polarized and half of which was originally Y polarized. I want to be As mentioned
above, the Faraday rotator mirror 1106 combines incident light into orthogonal polarizations.
Thus, Y polarized light is coupled to X polarized light.
[0117]
The X-polarized light reflected by the Faraday rotation mirror 1106 returns through the delay
loop 750 and the delay loop downlead fiber 740 to the third port 1123 of the polarizing beam
splitter. Because the light is currently in x-polarization, the light is not transmitted to the first
port 1121, but is reflected to the fourth port 1124. Thus, Y polarized light originally incident on
the polarizing beam splitter from the array 716 is coupled to the common rung 712 and is
returned to the 3 × 3 coupler 710 with X polarization.
[0118]
Unpolarized light propagating from the 3 × 3 coupler 710 via the common rung 712 to the
05-05-2019
46
polarizing beam splitter 1104 enters the polarizing beam splitter 1104 via the fourth port 1124.
The component of the Y-polarized light is transmitted to the second port 1122 and is nonreflectively terminated by the termination device 732. The light component of X polarization is
reflected to the third port 1123 and propagates to the Faraday rotation mirror 1106 via the
delay loop downlead fiber 740 and the delay loop 750. (The reason for including the depolarizer
1110 can now be understood. Since only X-polarized light from common rung 712 is coupled to
delay loop downlead fiber 740, the light coupled from array 716 to delay loop downlead fiber
740 by depolarizer 1110 is: It is ensured that some light that is originally X polarized is also
included. ) The Faraday rotator mirror 1106 reflects light as Y polarized light, which propagates
through the delay loop and downlead fiber to the third port 1123 of the polarizing beam splitter
1104.
[0119]
Y-polarized light incident on the third port 1123 of the polarizing beam splitter 1104 is
transmitted to the first port 1121 and thus to the second array input / output fiber 720. The Y
polarized light propagates through the array 716 to the first array input / output fiber 714 and
then through the depolarizer 1110 to the 3 × 3 coupler 710. Depolarizer 1110 operates to
convert approximately 50% of the Y polarized light into X polarized light. The X polarized light
from the depolarizer 1110 interferes with the X polarized light from the common rung 712. The
resulting combined light is detected by detector 704 or detector 706 according to the phase
relationship between the interfering optical signals of 3 × 3 coupler 710.
[0120]
It should be noted that the X-polarized light entering the 3 × 3 coupler 710 from the depolarizer
1110 travels the same path length as the X-polarized light from the common rung 712. For
example, light propagating through common rung 712 first propagates in X polarization through
common rung 712 and then propagates in Y polarization through array 716. On the other hand,
light propagating through the array 716 first propagates in Y polarization through the array 716
and then propagates in X polarization through the common rung. Because the two "backpropagating" optical signals are of the same polarization as they propagate through
corresponding portions of the interference path, the propagation lengths are the same, except for
the effects of incident noise sensed by the array 716. .
[0121]
05-05-2019
47
It should be understood that the termination 732 coupled to the second port 1122 of the
polarizing beam splitter 1104 may be replaced by a second delay loop (not shown) and a second
Faraday rotator mirror (not shown) It is possible to provide a second interference path of light
that interferes with polarization. By adjusting the delay provided by the second delay loop, the
return signal from the second interference path can be made not to overlap with the return
signal from the first interference path.
[0122]
The example of FIG. 23B is similar to the example of FIG. 23A, except that the depolarizer 1110
is positioned in the common rung 712. The effect of the depolarizer 1110 of FIG. 23B is to
combine (1) a portion of the light in the common rung 712 returning from the polarizing beam
splitter 1104 with a single polarization (eg X polarization) into orthogonal polarization (2 2.)
Scramble unpolarized light traveling from the 3 × 3 coupler 710 through the common rung 712
towards the polarizing beam splitter 1104). This ensures that the light will interfere when it is
recombined by the 3 × 3 coupler 710 (the same reason as adding the depolarizer 1110 to the
fiber 714 of FIG. 23A).
[0123]
The example of FIG. 23C is also similar to the example of FIG. 23A, except that the depolarizer
1110 is positioned in the second array input / output fiber 720. The embodiment of FIG. 23C is
functionally equivalent to the embodiment of FIG. 23A. This is because it does not matter
whether light passes through the array portion 716 and then through the depolarizer 1110 or
through the depolarizer 1110 and then through the array portion 716. Thus, the functionality of
the embodiment of FIG. 23C is substantially the same as that of the embodiment of FIG. 23A, as
described above.
[0124]
FIG. 24 illustrates a further alternative embodiment of the present invention, wherein the folded
Sagnac sensor array 1200 is connected as shown in the array 1100 of FIG. 23A, polarization
beam splitter (PBS) 1104, Faraday rotation A mirror (FRM) 1106 and a depolarizer 1110 are
05-05-2019
48
included. The other components of FIG. 23A are also numbered as before. Unlike array 1100 of
FIG. 23A with 3 × 3 coupler 710, folded Sagnac sensor array 1200 has polarization maintaining
(PM) 2 × 2 coupler 1220 connected in the same manner as 2 × 2 coupler 1006 of FIG. . One
port of 2 × 2 coupler 1220 is connected to the first port of optical circulator 1222 via a first
polarizer 1224. The second port of the optical circulator 1222 is connected to a first detector
1226. The third port of optical circulator 1222 is connected to non-polarizing source 1228 (eg,
intensity modulated fiber superfluorescent source). The second port of the 2 × 2 coupler 1220
is connected to a second detector 1230 via a second polarizer 1232. Detectors 1226 and 1230
and non-polarizing source 1228 are connected to circulator 1222 by standard fibers (not
polarization maintaining). Polarizers 1224 and 1232 are coupled to polarization maintaining
coupler 1220 via a polarization maintaining fiber such that polarizers 1224 and 1232 are
aligned with the same axis of polarization maintaining 2 × 2 coupler 1220. Alternatively, if a
polarization source is used instead of non-polarization source 1228, the polarization source (not
shown) is connected to a polarization maintaining circulator (not shown) by a polarization
maintaining fiber and the polarization maintaining circulator is a polarization maintaining fiber
To the polarizer 1224. The polarization maintaining component is connected such that polarized
light from the source passes through the polarizer 1224. The connection from the polarization
maintaining circulator to the detectors 1226 and 1230 is provided by a standard fiber (not
polarization maintaining).
[0125]
Folded Sagnac sensor array 1200 further includes non-reciprocal phase shifter 1250. The phase
shifter 1250 is via a first optical fiber 1252 having a first end 1254 and a second end 1256 and
via a second optical fiber 1258 having a first end 1260 and a second end 1262. It is coupled to
the common rung 712. The first end 1254 of the first optical fiber 1252 is coupled via a first
coupler 1264 to a common rung 712 proximate to the 2 × 2 coupler 1220. The first end 1260
of the second optical fiber 1258 is coupled to the common rung 712 proximate to the polarizing
beam splitter 1104 via a second coupler 1266. The respective second ends 1256, 1262 of the
first and second optical fibers 1252, 1258 are coupled to the phase shifter 1250, as described
below in connection with FIGS.
[0126]
Preferably, the common rung 712, the first fiber 1252 and the second fiber 1258 are
polarization maintaining (PM) fibers and the first coupler 1264, the second coupler 1266 and the
2 × 2 coupler 1220 are polarization maintaining (PM) ) Is a coupler. Also preferably, about 50%
05-05-2019
49
of the light is in the common rung while the first coupler 1264 and the second coupler 1266
couple about 50% of the light entering the common rung 712 in either direction to the phase
shifter 1250. It is a 50/50 coupler that stays in Thus, the non-reciprocal phase shifter 1250 and
associated fibers form a second rung 1268 parallel to the common rung 712.
[0127]
Preferably, one of the rungs 712, 1268 (eg, common rung 712) is a delay element (eg, delay) that
introduces a time delay into one rung sufficient to prevent overlapping of pulses propagating
through the rung. Loop 1269). Thus, the light returning from the sensor array 716 to the 2 × 2
coupler 1220 includes two pulses for each sensor, spaced in time from each other. One pulse
contains the combined light that passes through the common rung 712 in each direction. The
other pulse contains the combined light that passes through the non-reciprocal phase shifter
1250 in each direction. It should be understood that the light pulses passing through phase
shifter 1250 in one direction and the light pulses passing through common rung 712 in the
other direction have substantially different propagation times and do not overlap in coupler
1220. Thus, they do not interfere.
[0128]
Light passing through the common rung 712 in one direction does not undergo any phase shift
in the common rung 712 relative to light passing through the common rung in the other
direction. Thus, the relative phase bias of the combined light passing through the common rung
712 in both directions is zero. However, as described later, the nonreciprocal phase shifter 1250
introduces a shift of light in one direction to the light in the other direction. In particular, in the
preferred embodiment, phase shifter 1250 introduces a relative phase shift of π / 2 between
light in two directions. Thus, light propagating through phase shifter 1250 in both directions and
entering coupler 1220 combines with π / 2 phase bias at coupler 1220.
[0129]
Those skilled in the art will understand that the 50% coupler 1220 in the interference
configuration shown in FIG. 24 is such that the return light at the two input ports interferes with
the coupler and their relative phase difference is 0, 2π, 4π, etc It is recognized that the return
light is coupled to the output port corresponding to the original input port, and the return light is
05-05-2019
50
coupled to the other output port when the relative phase difference of light is π, 3π, 5π, etc.
Will. When the return light has a relative phase difference that is not a multiple of π, part of the
return light is output from both ports. For example, when the relative phase difference is an odd
multiple of π / 2 (eg, π / 2, 3π / 2, etc.), about 50% of the return light is coupled to each
output port. By providing two independent propagation paths, each detector 1226, 1230
receives two signals which are spaced in time, so that they can be detected separately. Because
one signal has zero phase bias and one signal has a π / 2 phase bias, if one signal senses little
perturbation, the other signal is very sensitive to perturbation, and The reverse is also true. It
should be understood that it is possible to include additional rungs parallel to the common rung
712 and having different amounts of relative phase shift to provide different phase biases to the
pulses.
[0130]
FIG. 25 illustrates an alternative configuration of a folding Sagnac sensor array 1200 ′
substantially similar to the folding Sagnac sensor array 1200 of FIG. In the folded Sagnac sensor
array 1200 ′ of FIG. 25, the depolarizer 1110 is located at the second array input / output fiber
720 rather than the first array input / output fiber 714. Because of the conflicting structure of
sensor array 716, repositioning depolarizer 1110 in fiber 720 does not change the overall
operation of folding Sagnac sensor array 1200 'relative to the operation of folding Sagnac sensor
array 1200. Thus, the operation of the folded Sagnac sensor array 1200 'will not be described in
detail here.
[0131]
The embodiments of FIGS. 24 and 25 include the sensor array 716 described above in detail. It
should be understood that other configurations of amplified sensor arrays may be used instead
of the sensor array 716 of the embodiment of FIGS. 24 and 25.
[0132]
FIG. 26 illustrates a first preferred embodiment of the non-reciprocal π / 2 phase shifter 1250
of FIGS. 24 and 25. As illustrated in FIG. 26, the phase shifter 1250 includes a first collimating
lens 1270, a first 45 ° faraday rotator 1272, a quarter wave plate 1274, and a second 45 °
faraday rotator 1276. And a second collimating lens 1278. In the illustrated embodiment, the
05-05-2019
51
first Faraday rotator 1272, the second Faraday rotator 1276 and the quarter wave plate 1274
comprise commercially available large optical devices, but advantageously, optical fibers or other
A wave tube device may be included. Collimating lenses 1270, 1278 are positioned proximate to
the second ends 1256, 1262 of PM fibers 1252, 1258 to focus light from fiber ends 1256, 1262
to the Faraday rotators 1272, 1276 respectively and , Focus the light from the Faraday rotators
1272, 1276 on the fiber ends 1256, 1262. Each of the Faraday rotators 1272 and 1276 rotates
its polarization of light entering the Faraday rotator with its polarization at a particular angle,
thereby rotating the polarization by a predetermined amount relative to the original angle And
operate in a known manner to make a new angle. For example, in the preferred embodiment,
each Faraday rotator 1272, 1276 rotates the polarization of the incident light by 45 ° in the
counterclockwise (ccw) direction. Therefore, as shown in FIG. 26, the light emitted from the end
1256 of the PM fiber 1252 whose polarization is horizontally directed is rotated 45 °
counterclockwise by the first Faraday rotator 1272 The polarization, when emerging from the
first Faraday rotator 1272, is oriented at an angle of 45 ° clockwise to the original orientation.
[0133]
A quarter wave plate 1274 is positioned between the two Faraday rotators 1272, 1276. The
quarter wave plate 1274 has a first birefringence axis 1280 and a second birefringence axis
1282 orthogonal to it. Light propagating in polarized light oriented along one birefringence axis
(eg, first birefringence axis 1280) is oriented along the other birefringence axis (eg, second
birefringence axis 1282) It has a slower propagation speed than light propagating in polarized
light. The quarter-wave plate 1274 is, for example, oriented such that the first birefringence axis
1280 makes 45 ° clockwise with respect to the vertical direction, so that the light emerging
from the first Faraday rotator 1272 is It is oriented along one birefringence axis 1280 and
oriented orthogonal to the second birefringence axis 1282. Due to the difference in propagation
velocity along the two axes, the quarter wave plate 1274 is polarized along the first birefringence
axis 1280 to the light polarized along the second birefringence axis 1282 Introduce a π / 2 or
90 ° phase shift to the emitted light. Thus, according to this example, light originally propagated
with horizontally polarized light rotated to align with the first birefringence axis 1280 is 90 for
any light propagating along the second birefringence axis 1282. Produces a relative phase shift
of °.
[0134]
After passing through the quarter wave plate 1274, the light passes through the second Faraday
rotator 1276 and is again rotated 45 ° in a counterclockwise direction. Light emerging from the
05-05-2019
52
second Faraday rotator 1276 passes through the second collimating lens 1278 and is focused on
the second end 1262 of the second PM optical fiber 1258. It should be understood from the
above description that any light output from the first PM optical fiber 1252 with horizontal
polarization enters the second PM optical fiber 1258 with vertical polarization. As described
above, light entering the second PM optical fiber 1258 in vertically polarized light will have
propagated along the slow birefringence axis 1280 of the quarter wave plate 1274 and along the
fast birefringence axis 1282 It will produce a relative phase difference of π / 2 to the
propagating light.
[0135]
As shown by the description, non-reciprocal phase shifter 1250 operates in a non-reciprocal
manner for the operation of Faraday rotators 1272, 1276. As described above, the light passing
through the Faraday rotators 1272 and 1276 from the first PM fiber 1252 to the second PM
fiber 1258 is counterclockwise by each rotator with respect to the light propagation direction
shown in FIG. Is rotated 45 °. If the Faraday rotator is reciprocal, light propagating in the
opposite direction through the Faraday rotators 1272, 1276 is also rotated in a counterclockwise
direction relative to the direction of light propagation. However, because the Faraday rotator is
nonreciprocal, the light is rotated in the opposite direction (ie, clockwise with respect to the
direction of light propagation). The nonreciprocal effect is illustrated in FIG. 27 for light passing
from the second end 1262 of the second PM fiber 1258 through the nonreciprocal phase shifter
1250 to the second end 1256 of the first PM fiber 1252. Looking at FIG. 27, again, the rotation
appears to be in a counterclockwise direction, but it should be noted that the light is now
propagating towards the viewer. Thus, the vertically polarized light emitted from the second end
1262 of the second PM optical fiber 1258 passes through the second collimating lens 1278 and
the second Faraday rotator 1276 to form a quarter wave plate. It is rotated in alignment with the
1274 second (fast) birefringence axis 1282. Thus, light of originally vertical polarization does not
undergo relative retardation as it propagates through the quarter wave plate 1274. After passing
through the quarter wave plate 1274, the light passes through the first Faraday rotator 1272,
whereby the light is further rotated by 45 ° with respect to the horizontal polarization. The light
is then focused onto the second end 1256 of the first PM optical fiber 1252 through the first
collimating lens 1270.
[0136]
From the above, horizontally polarized light passing from the first PM fiber 1252 to the second
PM fiber 1258 through the nonreciprocal phase shifter 1250 in the first direction is the slow
05-05-2019
53
birefringence axis 1280 of the quarter wave plate 1274. It can be seen that it propagates
through and experiences a relative phase delay of 90.degree. Or .pi. / 2. Horizontally polarized
light propagating in a first direction is rotated to be oriented with vertical polarization as the
light enters the second PM fiber 1258. Conversely, vertically polarized light passing from the
second PM fiber 1258 to the first PM fiber 1252 through the nonreciprocal phase shifter 1250
in the second direction has the fast birefringence axis of the quarter wave plate 1274. It
propagates through 1282 and does not undergo relative phase delay. The vertically polarized
light propagating in the second direction is rotated to be oriented with horizontal polarization as
the light enters the first PM fiber 1252. As described in more detail below, the relative phase
shift between the horizontally polarized light propagating in the first direction and the vertically
polarized light propagating in the second direction is π / 2 phase Give a bias.
[0137]
FIGS. 28 and 29 illustrate an alternative embodiment of the nonreciprocal phase shifter 1250,
wherein the first Faraday rotator 1272 is a quarter wave plate 1274 (here with the first quarter
wave plate). And a second quarter-wave plate 1294. In FIG. 28, light from the second end 1256
of the first PM fiber 1252 is collimated by the first collimating lens 1270 as before. The light is
originally in horizontal polarization. As light passes through the first quarter wave plate 1274, it
is converted to light with circular polarization. Since circularly polarized light passes through the
first Faraday rotator 1272, circularly polarized light produces a phase shift of φ. In the preferred
embodiment, the first Faraday rotator 1272 is selected to cause a phase shift of π / 4. The light
from the Faraday rotator 1272 remains circularly polarized and passes through the second
quarter wave plate 1294. The quarter wave plate converts circularly polarized light into linearly
polarized light in the direction of vertical polarization. In addition to being vertically polarized,
the light has undergone a phase shift of φ (eg π / 4).
[0138]
FIG. 29 illustrates the operation of an alternative embodiment of non-reciprocal phase shifter
1250 for light propagating in the opposite direction. In FIG. 29, vertically polarized light from the
second end 1262 of the second PM fiber 1260 is collimated by the second collimating lens 1278
and passes through the second quarter wave plate 1294. The second quarter wave plate 1294
converts the vertically polarized light into light with circular polarization. The circularly polarized
light passes through the first Faraday rotator 1272 and undergoes a phase shift as before. As the
light propagates in the opposite direction through the first Faraday rotator 1272, the light
undergoes an opposite phase shift -.phi. (E.g. -.pi./4). The light from the first Faraday rotator
05-05-2019
54
1272 then passes through the first quarter wave plate 1274 where the circularly polarized light
is converted to linearly polarized light with horizontally polarized light Be done. Thus, light
propagating in two directions undergoes a relative phase shift of 2φ (eg, π / 2) in total. This has
the same effect as the first embodiment of the non-reciprocal phase shifter 1250 illustrated in
FIGS. 26 and 27.
[0139]
The effects of the non-reciprocal phase shifter 1250 on polarization orientation and phase delay
provide the biasing effects described above and re-described in conjunction with FIG. As shown
in FIG. 24, light entering the second PM fiber 1258 with vertical polarization passes through the
common rung 712 from the first PM coupler 1264 to the second PM coupler 1266 at the second
PM coupler 1266. It is combined with the transmitted light. For reasons that will become
apparent from the following description, it is desirable for the light entering the second PM
coupler 1266 from the common rung 712 to have the same polarization as the light entering the
second PM fiber from the second PM fiber 1258. Thus, in the preferred embodiment, the
vertically polarized light of the second PM fiber 1258 is the light of the horizontally polarized
light of the common rung 712 by rotating either the second PM fiber 1258 or the common rung
712 by 90 °. In the same direction as. This is by rotating the second end 1262 of the second PM
fiber 1258 close to the second collimating lens 1278 so that vertically polarized light is in the
horizontal polarization axis of the second PM fiber 1258 It is easily achieved by entering the
second end 1262 with the polarization state oriented along. Thus, light exiting nonreciprocal
phase shifter 1250 in the vertical polarization state is applied to coupler 1266 as light in the
horizontal polarization state relative to the polarization axis of coupler 1266. Thus, the light from
non-reciprocal phase shifter 1250 has the same polarization state as the light from common rung
712.
[0140]
The light passing through common rung 712 and the light passing through non-reciprocal phase
shifter 1250 then enter port 1124 of polarizing beam splitter (PBS) 1104. Horizontally polarized
light is output from port 1123 of PBS 1104 to fiber 740. Fiber 740 includes a delay loop 750
and is terminated at a Faraday rotating mirror (FRM) 1106. Delay loop 750 and FRM 1106
operate as described above and the reflected and delayed pulse is returned to port 1123 of PBS
1104 in vertically polarized light. The pulse is output from the port 1121 of the PBS 1104 to the
array 716 via the fiber 720 and propagates through the sensors 722 (i) of the array 716 in a
clockwise direction.
05-05-2019
55
[0141]
The pulses are output from the array 716 through the fiber 714 and the depolarizer 1110 to the
2 × 2 coupler 1220 where the clockwise propagating light and the counterclockwise
propagating light are combined. The light propagating backward also starts as horizontally
polarized light. The light is depolarized and passes through the sensor array 716. Light emerging
from sensor array 716 in vertically polarized light is reflected by PBS 1123 and discarded
through port 1122 and terminator 732. Light emerging from sensor array 716 in horizontally
polarized light passes through PBS 1123, is delayed by loop 750, and is rotated to vertically
polarized light by FRM 1106. The return light, which is in vertical polarization, is reflected by
PBS 1123 to port 1124 and thus directed to the second PM coupler 1266. A portion of the light
passes through the delay loop 1269 of the common rung 712 and a portion of the light passes
through the non-reciprocal phase shifter 1250. As described above, light entering the
nonreciprocal phase shifter 1250 with vertical polarization propagates through the fast
birefringent axis 1282 of the quarter wave plate 1274 (FIG. 27) and does not undergo relative
phase delay. Thus, two pulses of counterclockwise light propagate to the coupler 1220 where
they are combined with the clockwise propagating light pulse. The optical signals passed through
the common rung 712 and the delay loop 1269 in both directions do not undergo relative phase
shift and are combined as described above. Since the optical signal passing through the
nonreciprocal phase shifter 1250 in both directions undergoes a relative phase shift of π / 2
between the clockwise propagating signal and the counterclockwise propagating signal, as
described above, π / 2 Have a phase bias of At both outputs of coupler 1220, portions of the
two pulses of light returning from sensor array 1200 are directed to polarizer 1224 and the
remaining portions are directed to polarizer 1232. The role of the two polarizers 1224 and 1232
is to ensure that the light entering the loop has the same polarization as the light leaving the
loop, thereby ensuring reciprocity. As mentioned above, the two pulses that reach detector 1230
are in quadrature, which allows using a number of signal processing techniques known in the art
to avoid signal fading. The same applies to the detector 1226. In the embodiment of FIG. 24, the
generation of the two pulses in quadrature is the main reason for incorporating the rung
including the non-reciprocal phase shifter 1250.
[0142]
Figures 30 to 36 illustrate a further alternative embodiment of the present invention, wherein
the folded Sagnac sensor array uses polarization based biasing for multiple detectors, each
detector being independent of the bias point of the other detector Has a bias point that can be
05-05-2019
56
set. The embodiment of FIGS. 30-36 includes the sensor array 716 described above in detail. It
should be understood that other configurations of amplified sensor arrays may be used instead
of the sensor array 716 of the embodiment of FIGS. 30-36.
[0143]
In the folded Sagnac sensor array 1300 illustrated in FIG. 30, a polarizing fiber superfluorescent
source (SFS) 1310 is coupled to a polarization controller 1312 via a fiber 1314. Fiber 1314
further couples polarization controller 1312 to the first port of 2 × 2 coupler 1316. The second
port of coupler 1316 is an output port, which will be described later. The third port of coupler
1316 is coupled to non-reflective termination 1320 via fiber 1318. The fourth port of coupler
1316 is coupled to a first port 1330 of polarizing beam splitter (PBS) 1332 via a common array
input / output fiber 1334. The second port 1336 of the polarizing beam splitter 1332 is coupled
to a first horizontal polarizer 1338. The first horizontal polarizer 1338 is coupled to the second
array input / output fiber 720 of the array 716. The third port 1340 of the polarizing beam
splitter 1332 is connected to a common delay fiber 1342 formed in the delay loop 1344 and
terminated with a Faraday rotator mirror (FRM) 1346. The fourth port 1348 of the polarizing
beam splitter 1332 is coupled to the second horizontal polarizer 1350 and then to the
depolarizer 1352. Depolarizer 1352 is coupled to the first array input / output fiber 714.
[0144]
The second port of coupler 1316 is coupled to detector subsystem 1360 via fiber 1362. In the
example of FIG. 30, detector subsystem 1360 includes a 1 × n coupler 1364 having a single
input port that receives light from the second port of coupler 1316. A first output port of 1 × n
coupler 1364 is coupled to polarization controller 1366. Polarization controller 1366 is coupled
to polarizer 1368, which is in turn coupled to first detector 1370. A second output port of 1 × n
coupler 1364 is coupled to polarization controller 1372. Polarization controller 1372 is coupled
to polarizer 1374, which is coupled to second detector 1376. Additional polarization controllers,
polarizers and detectors (not shown) can be connected to additional ports (not shown) of the 1 ×
n coupler 1364.
[0145]
The folded Sagnac sensor array 1300 of FIG. 30 operates in the following manner. Polarized SFS
05-05-2019
57
1310 provides a polarized output signal that passes through polarization controller 1312 via
fiber 1314. The polarization controller 1312 is adjustable to change the polarization to the
desired polarization state. For example, in FIG. 30, the polarization state is adjusted to give
linearly polarized light which is directed at 45 ° to the vertical and horizontal axes at the input
to polarizing beam splitter 1332. The light remains in fiber 1314 and is provided as an input to
coupler 1316. The coupler 1316 couples approximately 50% of the incoming light to the first
output fiber 1318 and is thereby discarded at the non-reflective termination 1320. Couplers
1316 couple approximately 50% of the incoming light to a common array input / output fiber
1334.
[0146]
A common array input / output fiber 1334 directs light to polarizing beam splitter 1330, which
reflects horizontally polarized light to second port 1336 and vertically polarized light to third
port 1340. Pass through. Reflected horizontally polarized light from the second port 1336
travels through the first horizontal polarizer 1338 to the second array input / output fiber 720
and propagates through the array 716 in a clockwise direction. . Light propagating clockwise
exits array 716 via depolarizer 1352 and array input / output fiber 714. As mentioned above, the
depolarizer 1352 ensures that after exiting the sensors in the array 716, the exiting light is
distributed substantially equally in the horizontal and vertical polarization modes. Next,
clockwise propagating light passes through the second horizontal polarizer 1350, which rejects a
portion of the vertically polarized light. Next, the clockwise propagating horizontally polarized
light enters the fourth port 1348 of the polarizing beam splitter 1330, is reflected to the third
port 1340, and propagates to the common delay fiber 1342. Clockwise return light passes
through the delay loop 1344 to the Faraday rotator mirror 1346 where it is reflected as
vertically polarized light. The vertically polarized light returns to the third port 1340 of the
polarizing beam splitter 1332 and is passed to the first port 1330.
[0147]
As described above, the light originally incident on the first port 1330 of the polarizing beam
splitter 1332 was oriented at about 45 ° to the horizontal and vertical polarizations. Thus, about
50% of the light corresponding to the component of the vertically polarized light passes through
the polarizing beam splitter 1332 to the third port 1340 and thus to the common delay fiber
1342. The vertically polarized light propagates through the delay loop 1344 and is reflected by
the Faraday rotator mirror 1346 as horizontally polarized light. The reflected horizontally
polarized light passes through the delay loop 1344 and returns to the third port 1340 of the
05-05-2019
58
polarizing beam splitter 1332. Because the light is horizontally polarized, it is reflected to the
fourth port 1348 of the polarizing beam splitter 1332 and thus passes through the second
horizontal polarizer 1350 via the first array input / output fiber 714 and depolarized It
propagates through the child 1352 into the array 716 and is caused to propagate therein in a
counterclockwise direction. The depolarizer 1352 is such that when counterclockwise
propagating light emerges from the array 716, the counterclockwise propagating light has
components of all polarizations such that at least a portion of the light is in horizontal
polarization. To guarantee.
[0148]
The counterclockwise propagating light emerges from the array 716 via the second array input /
output fiber 720 and the component of the horizontally polarized light passes through the first
horizontal polarizer 1338. Horizontal polarizers reject light of other polarization orientations.
Horizontally polarized light originating from the portion of light propagating counterclockwise
enters the second port 1336 of the polarizing beam splitter 1332 and is reflected to the first port
1330 of the polarizing beam splitter 1332 where it is It is combined with vertically polarized
light originating from the part of the light propagating around.
[0149]
The combined light propagates to the fourth port of coupler 1316 where about 50% of the
combined light is coupled via fiber 1362 to the second port of coupler 1316 and thus to detector
subsystem 1360 Ru. A 1 × n coupler 1364 splits the light into N parts. For example, in FIG. 30,
N is equal to 2 and a first portion of light is coupled to polarization controller 1366 to propagate
through polarizer 1368 to first detector 1370 and a second portion of light Are coupled to the
polarization controller 1372 to propagate through the polarizer 1374 to the second detector
1376. The orientations of the polarization controller 1366, 1372 and the polarizers 1368, 1374
are adjustable to bias the light signal incident on the first detector 1370 and the second detector
1376 in different phases. For example, the signal applied to the second detector 1376 can be
biased to be in quadrature with the signal applied to the first detector 1370 so that one signal
has minimal sensitivity If so, the other signal has the greatest sensitivity, and vice versa.
[0150]
05-05-2019
59
As mentioned above, each of the two signal portions travels the same distance through the array
716, the common delay fiber 1342 and the delay loop 1344. Thus, in the absence of an acoustic
signal or other noise induced perturbations impinging on the sensors in array 716, the two parts
are in phase and constructively interfere to produce a composite optical signal with 45 ° linear
polarization . However, light has a polarization state that is orthogonal to the original polarization
state. Thus, if the original polarization state is + 45 ° (again, without phase perturbations), then
the polarization state of the output signal is -45 °.
[0151]
When an acoustic signal is present, the clockwise propagating light and the counterclockwise
propagating light undergo a relative phase shift. As the relative phase shift increases, the
polarization states of the two interfering beams change from -45 ° linear polarization to left
circular polarization, then to + 45 ° polarization, and then to right circular polarization, with -45
° polarization. Return to The transition over these four polarization states defines a circle on the
Poincare sphere. The polarization state at the output of the polarization beam splitter 1332
corresponds to a point along this circle on the Poincare sphere, and its location on the circle is a
function of the acoustically induced nonreciprocal phase shift.
[0152]
After traveling from the output of polarizing beam splitter 1332 through common array input /
output fiber 1334 through coupler 1316 to detector subsystem 1360, the polarization state of
the composite signal is arbitrarily changed by the unknown birefringence of fiber 1334. Be done.
For each detector 1370, 1376 using a polarization controller 1366 close to the polarizer 1368
before the first detector 1370 and a polarization controller 1372 close to the polarizer 1374
before the second detector 1376 , Redirect the polarization state back to the respectively
selected polarization state. Polarization controllers 1366, 1372 are set, for example, when no
acoustic signal is applied to array 716 and thus no relative phase shift is introduced into the
counter-propagating optical signal.
[0153]
For example, to provide the first detector 1370 with a bias point of ± 90 °, the polarization
controller 1376 controls the first detector when the combined light at the output of the
05-05-2019
60
polarizing beam splitter 1332 has a left circular polarization state. 1370 is set to detect either
the maximum or minimum intensity of light. For other polarization states of the output light, the
first detector 1370 detects light having an intensity between maximum and minimum intensity.
[0154]
As a further example, the second detector 1376 can advantageously be set to different bias
points, such as 0 ° and 180 °. For this bias point, when the light at the output of polarization
beam splitter 1332 has a polarization state of −45 °, polarization controller 1372 either
detects the maximum or minimum intensity of the light. Set to detect. For other polarization
states of the output light, the second detector 1376 detects light having an intensity between
maximum and minimum intensity.
[0155]
It should be understood that the light applied to the input of polarizing beam splitter 1332 may
have polarization states other than ± 45 °. For example, if the input light originally has a left
circular polarization state, the polarization controller 1366, 1372 is set accordingly to provide
the first detector 1370 and the second detector 1376 with appropriate bias points.
[0156]
FIG. 31 illustrates an alternative configuration of a folding Sagnac sensor array 1300
'substantially similar to the folding Sagnac sensor array 1300 of FIG. In the folded Sagnac sensor
array 1300 ′ of FIG. 31, the depolarizer 1352 is located in the second array input / output fiber
720 rather than the first array input / output fiber 714. Repositioning depolarizer 1352 in fiber
720 due to the conflicting structure of sensor array 716 does not change the overall operation of
folding Sagnac sensor array 1300 'relative to the operation of folding Sagnac sensor array 1300.
The operation of the folding Sagnac sensor array 1300 'is similar to the operation of the folding
Sagnac sensor array 1300 and will not be described in detail here.
[0157]
05-05-2019
61
FIG. 32 illustrates a further alternative embodiment of a folding Sagnac acoustic sensor array
1400 similar to the folding Sagnac sensor array 1300 of FIG. 30, the same elements being
numbered accordingly. Unlike the folded Sagnac sensor array 1300, the folded Sagnac sensor
array 1400 replaces the 2 × 2 coupler 1316 with a polarization independent optical circulator
1410. The optical circulator performs the same function as the 2 × 2 coupler 1316, but in the
folding Sagnac sensor array 1300, about 50% of the input light is lost when the input light is
split by the coupler 1316 and the output light is split by the coupler 1316 About 50% of the
output light is lost. In example 1400, substantially all input light passes from polarized SFS 1310
through circulator 1410 to polarizing beam splitter 1332 and substantially all output light from
polarizing beam splitter 1332 through circulator 1410 to a detector It reaches subsystem 1360.
[0158]
FIG. 33 illustrates an alternative configuration of a folded Sagnac sensor array 1400
'substantially similar to the folded Sagnac sensor array 1400 of FIG. In the folded Sagnac sensor
array 1400 ′ of FIG. 33, the depolarizer 1352 is located in the second array input / output fiber
720 rather than the first array input / output fiber 714. Repositioning depolarizer 1352 in fiber
720 due to the conflicting structure of sensor array 716 does not change the overall operation of
embodiment 1400 'with respect to the operation of folded Sagnac sensor array 1400. Thus, the
operation of the folded Sagnac sensor array 1400 'will not be described in detail here.
[0159]
FIG. 34 illustrates a further alternative embodiment of a folded Sagnac sensor array 1600
according to the invention, which is coupled to the array 716 in the same manner as described
above in connection with FIGS. 30 to 33. An input / output subsystem 1610 is included.
[0160]
In FIG. 34, polarization source 1620 provides linearly polarized input light along the axis of
polarization maintaining fiber 1622.
The polarization maintaining fiber 1622 is rotated such that the polarization axis is oriented at ±
45 ° with respect to the vertical polarization axis of the input / output system 1610. Light from
fiber 1622 is coupled to input / output subsystem 1610 via a first collimating lens 1630. The
first collimating lens 1630 directs light towards a first port 1634 of a first polarizing beam
05-05-2019
62
splitter (PBS) 1632. The polarization beam splitter also has a second port 1636, a third port
1638 and a fourth port 1640. The second port 1636 directs a portion of the input light towards
the first 45 ° Faraday rotator (45 ° FR) 1642. The third port 1638 directs a portion of the
input light towards the second 45 ° Faraday rotator 1644. As described below, the fourth port
1640 directs selected portions of the output light to the detection subsystem 1650.
[0161]
Light passing through the first Faraday rotator 1642 is collimated by the second collimating lens
1660 and coupled into the array input / output fiber 720, thus propagating to the sensor portion
of the array 716 and clocking therein. Propagating around.
[0162]
Light passing through the second Faraday rotator 1644 passes through the half-wave (λ / 2)
plate 1662.
Half-wave plate 1662 has first and second birefringence axes (not shown). One of the
birefringence axes extends at an angle of 22.5 ° to the vertical polarization axis of the incoming
light and is directed at −22.5 ° to the 45 ° polarization of light traveling from the source
towards it (Ie, an axis exists between the light's normal and the polarization). The purpose of this
orientation is described below. Light passing through half-wave plate 1662 enters first port 1672
of second polarization beam splitter 1670. The polarization beam splitter also has a second port
1674, a third port 1676 and a fourth port 1678. As described below, the second port 1674 is not
coupled to a further element. Light output from the third port 1676 is directed to the third
collimating lens 1680. Light output from the fourth port 1678 is directed to the fourth
collimating lens 1682.
[0163]
Light passing through the fourth collimating lens 1682 is coupled to the first array input /
output fiber 714 and passes through the depolarizer 1352 to reach the sensor portion of the
array 716 and propagates therein in a counterclockwise direction. Do.
[0164]
05-05-2019
63
Light passing through the third collimating lens 1680 is focused at the end of the common delay
fiber 1342 and propagates through the delay loop 1344 to the Faraday rotator mirror 1346 and
back through the delay loop 1344 to the collimating lens 1680 .
Thus, the reflected light is directed back to the third port 1676 of the second polarizing beam
splitter 1670.
[0165]
As described above, light from the fourth port 1640 of the first polarizing beam splitter 1632
enters the detection subsystem 1650. The detection subsystem 1650 includes a first beam
splitter 1690, a second beam splitter 1692, a first birefringent element 1694, a second
birefringent element 1696, a first detector 1698, a second detector 1700, A first polarizer 1702
and a second polarizer 1704 are included. The first fraction of light from the fourth port 1640 is
reflected by the first beam splitter 1690 and passes through the first birefringent element 1694
and the first polarizer 1702 to reach the first detector 1698. The remainder of the light from the
fourth port 1640 passes through the first beam splitter 1690 and is incident on the second beam
splitter 1692. Here, a second fraction of light is reflected by the second beam splitter 1692
through the second birefringent element 1696 and the second polarizer 1704 to the second
detector 1700. The remaining part of the light passes through the second beam splitter 1692 to
a further element (not shown). If only two detectors are provided, the ratio of the first binding is
preferably 50% and the second ratio is preferably 100%, so that both detectors 1698, 1700 have
approximately the same amount Receive the light of If a third detector (not shown) is included,
the first fraction is preferably about 33% and the second fraction is preferably about 50%. Even
receives about 33% of the original light. The third detector then receives the remaining 33%.
[0166]
The folded Sagnac sensor array 1600 of FIG. 34 operates in the following manner. As mentioned
above, light incident on the first lens 1630 is directed at 45 ° to the vertical and horizontal
polarization axes. Thus, light passing through lens 1630 and entering first port 1634 of first
polarization beam splitter 1632 has a component of horizontal polarization and a component of
vertical polarization. The horizontal component is reflected by the polarizing beam splitter 1632
to the second port 1636 and the vertical component passes through the polarizing beam splitter
1632 to the third port 1638.
05-05-2019
64
[0167]
The horizontal component from the second port 1636 passes through the first Faraday rotator
1642 and the polarization state is a linear polarization of 45 degrees of light emerging from the
first Faraday rotator 1642 and incident on the second lens 1660 To have a state, it is rotated by
45 ° in a first direction (eg clockwise). Light propagates through the second lens 1660, into the
second array input / output fiber 720, and propagates through the array 716 in a clockwise
direction. Light may encounter changes in polarization within the array 716. Thus, as described
above, light exiting the array 716 through the first array input / output fiber 714 passes through
the depolarizer 1352. This ensures that at least a portion of the light is in horizontal and vertical
polarization states.
[0168]
Clockwise propagating light from the first array input / output fiber 714 enters the input /
output subsystem 1610 via the fourth lens 1682 and is incident on the second polarizing beam
splitter 1670. The vertical component of the light passes through the second polarizing beam
splitter 1670 and is output from the second port 1674 and discarded. The component of the
horizontally polarized light is reflected to the third port 1676 of the second polarizing beam
splitter 1670 and passes through the third lens 1680 to the common delay fiber 1342 so that
the light propagates through the delay loop 1344 And is reflected by the Faraday rotation mirror
1346 to a vertically polarized state and back through the delay loop 1344 and the common delay
fiber 1342 to the third lens 1680. The reflected light in the vertical polarization state passes
from the third port 1676 to the first port 1672 of the second polarizing beam splitter 1670,
passes through the half-wave plate 1662 to the second Faraday rotator 1644, The third port
1638 of the polarizing beam splitter 1632 is reached. The half-wave plate 1662 is oriented such
that one of its birefringence axes is at 22.5 ° to the vertical polarization axis, so that the vertical
light incident on the half-wave plate 1662 is reflected near the birefringence axis The
polarization state of the light emerging from the half wave plate 1662 is directed at 45 ° to the
vertical and horizontal axes. The second Faraday rotator 1644 further rotates the polarization
state by 45 ° so that light emerging from the second Faraday rotator 1644 and incident on the
third port 1638 of the first polarization beam splitter 1632 has horizontal polarization state.
Have to have. Thus, light entering the third port 1638 is reflected to the fourth port 1640 and
enters the detection subsystem 1650 in a horizontal polarization state.
[0169]
05-05-2019
65
As noted above, the vertical component of the input light incident on the first port 1634 of the
first polarizing beam splitter 1632 passes to the third port 1638. The polarization state of the
light is rotated by 45 ° by the second Faraday rotator 1644 to a 45 ° polarization state with
respect to the vertical and horizontal polarization axes. At this time, the polarization state of the
light is reflected around the birefringence axis of the half wave plate 1662 so that the
polarization state of the light emerging from the half wave plate is again oriented in the vertical
direction. Those skilled in the art will recognize that due to the non-reciprocal action of the
second Faraday rotator 1644, vertically polarized light passing through the second Faraday
rotator 1644 from left to right and then through the half-wave plate 1646 is first 45 ° polarized
It will be appreciated that it is rotated to the state and then reflected back to the vertical
polarization state. In contrast, vertically polarized light passing from right to left is first reflected
by the half wave plate 1646 to a 45 ° polarization state and then rotated to a horizontal
polarization state by a second Faraday rotator 1644.
[0170]
Vertically polarized light from half-wave plate 1662 enters first port 1672 of second polarizing
beam splitter 1670 and passes through third port 1676 to reach third lens 1680. The vertically
polarized light passes through common delay fiber 1342, through delay loop 1344 to Faraday
rotator mirror 1346, and is reflected as horizontally polarized light back through delay loop
1344 and common delay fiber 1342. The horizontally polarized light passes through the third
lens 1680 to the third port 1676 of the polarizing beam splitter 1670. Horizontally polarized
light is reflected to the fourth port 1678 and passes through the fourth lens 1682 to the first
array input / output fiber 714 and through the depolarizer 1352 in the counterclockwise
direction the array 716 Propagate through.
[0171]
The counterclockwise propagating light emerges from the array 716 via the second array input /
output fiber 720 and passes through the second lens 1660 to the first Faraday rotator 1642. The
first Faraday rotator 1642 rotates the polarization state of light by 45 °. Because the light is
effectively depolarized by the depolarizer 1352, the light reaching the second port 1634 of the
first polarizing beam splitter 1632 through the first Faraday rotator 1642 is horizontally and
vertically polarized. Contains light having the following components. The horizontally polarized
component of light is reflected to the first port 1634 and output through the first lens 1630 to
05-05-2019
66
the input fiber 1622. It is advantageous to include an isolator (not shown) to absorb the light.
[0172]
The vertically polarized component of the counterclockwise propagating light entering the
second port 1636 of the first polarizing beam splitter 1632 passes through the fourth port 1640
and is combined with the horizontally polarized component of the clockwise propagating light. As
discussed above in connection with FIG. 30, if the counter-propagating light does not undergo a
relative phase shift, the light is combined as linearly polarized light with a 45 ° polarization
state. As described further above, the relative phase shift changes the polarization state.
[0173]
Select light incident on detector 1698, 1900 by introducing a relative phase shift to light of two
different polarizations (eg, horizontal and vertical polarization, + 45 ° and -45 ° polarization or
left circular and right circular polarization) In order to bias, it includes birefringent elements
1694, 1696. The birefringent element may advantageously comprise a linear or circular wave
plate (e.g. a quarter wave plate, a half wave plate, a Faraday rotator etc).
[0174]
FIG. 35 illustrates an example of a folded Sagnac acoustic sensor array 1750 similar to the folded
Sagnac acoustic sensor array 1600 of FIG. 34, where like elements are identified with the same
numbers as in FIG. Unlike the embodiment of FIG. 34, folded Sagnac acoustic sensor array 1750
includes non-polarized light source 1720 instead of polarized light source 1620. In order to use a
non-polarizing light source 1720, the folding Sagnac acoustic sensor array 1750 includes a 45 °
polarizer 1730 between the first collimating lens 1630 and the first polarizing beam splitter
1632. The 45 ° polarizer 1730 directs light incident on the first port 1634 of the first
polarizing beam splitter 1632 to 45 °, thus having substantially equal horizontal and vertical
polarization components. Thus, the folding Sagnac acoustic sensor array 1750 of FIG. 35
operates in substantially the same manner as the folding Sagnac acoustic sensor array 1600 of
FIG. 34, and the operation of the folding Sagnac acoustic sensor array 1750 will not be described
in detail here.
05-05-2019
67
[0175]
FIG. 36 illustrates a further embodiment of a folding Sagnac acoustic sensor array 1800 similar
to the folding Sagnac acoustic sensor arrays 1600 and 1750 of FIGS. 34 and 35, respectively, the
same elements being identified with the same numbers as in FIGS. Ru. Unlike the embodiments of
FIGS. 34 and 35, in the folded Sagnac acoustic sensor array 1800, the light signal through the
polarizers 1702 and 1704 is not directed to the detectors 1698 and 1700. Rather, the folded
Sagnac acoustic sensor array 1800 includes a collimating lens 1810 positioned proximate to the
polarizer 1702 and a collimating lens 1812 positioned proximate to the polarizer 1704.
Collimating lens 1810 directs light from polarizer 1702 to a first end 1822 of fiber 1820. The
fiber 1820 has a second end 1824 proximate to the first detector 1698 such that light entering
the fiber 1820 from the collimating lens 1810 is incident on the first detector 1698. Similarly,
collimating lens 1812 directs light from polarizer 1702 to a first end 1832 of fiber 1830. The
fiber 1830 has a second end 1834 proximate to the second detector 1700 such that light
entering the fiber 1830 from the collimating lens 1812 is incident on the second detector 1700.
By including collimating lenses 1810 and 1812 and fibers 1820 and 1830, the fiber can carry
light by a distance to detectors 1698 and 1700 so that it is remote from the detection electronics
(not shown). A detector may be provided at the place.
[0176]
In FIGS. 34, 35 and 36, the depolarizer 1352 is input as the first array without significantly
affecting the operating characteristics of the folded Sagnac acoustic sensor array 1600, the
folded Sagnac acoustic sensor array 1750 or the folded Sagnac sensor array 1800. Note that it is
possible to relocate from the / output fiber 714 to the second array input / output fiber 720.
[0177]
It should further be noted that the above embodiments have been described in the context of
superfluorescent light sources.
One skilled in the art will appreciate that advantageously other light sources (eg laser light
sources) may also be used.
[0178]
05-05-2019
68
Although the above description of the array according to the invention has dealt with underwater
acoustic sensing, it is understood that the invention can be used to sense any measured quantity
that can be made to generate non-reciprocal phase modulation in the fiber. I want to be For
example, replacing the hydrophone with an alternative sensing device that responds to different
measured quantities, the array will detect that measured quantity in the same manner as it
detects acoustic waves. The arrays of the invention can advantageously be used to sense
vibrations, intrusions, shocks, chemicals, temperatures, liquid levels and strains. The array of the
invention is also used to combine a number of different sensors provided at the same or different
locations (eg, for detection of various defects at various points along the hull or building shell)
obtain. Other exemplary applications include detection and tracking of vehicles on a freeway or
planes on a runway for traffic monitoring and control.
[0179]
Although described above in connection with particular embodiments of the present invention, it
should be understood that the description of the embodiments is illustrative of the present
invention and is not intended to be limiting. Various modifications and applications may occur to
those skilled in the art without departing from the true spirit and scope of the invention as
defined in the appended claims.
[0180]
Acoustic sensor 100, delay loop 102, hydrophone 104, source 110, 3 × 3 coupler 112.
05-05-2019
69
Документ
Категория
Без категории
Просмотров
0
Размер файла
118 Кб
Теги
jp2011232347
1/--страниц
Пожаловаться на содержимое документа