Phase-sensitive amplification in a fiber

Information

  • Patent Application
  • 20060285197
  • Publication Number
    20060285197
  • Date Filed
    June 16, 2005
    19 years ago
  • Date Published
    December 21, 2006
    18 years ago
Abstract
A method of and device for generating an amplified optical signal directly in an optical fiber by way of phase-sensitive amplification based on one or more four-wave mixing (FWM) processes. In one embodiment, an input signal and two pump waves are applied to a highly nonlinear fiber (HNLF). The input signal is amplified in the HNLF due to energy transfer from the pump waves to the input signal via a degenerate phase-conjugation (PC) process. In another embodiment, an input signal and first and second pump waves are applied to a first HNLF to generate, via a Bragg scattering (BS) process, an idler signal corresponding to the input signal. The second pump wave is then filtered out and the first pump wave, a third pump wave, and the input and idler signals are applied to a second HNLF, where they interact via a non-degenerate PC process to produce an amplified output signal.
Description
BACKGROUND OF THE INVENTION

1. Field of the Invention


The present invention relates to optical communication equipment and, more specifically, to optical amplifiers.


2. Description of the Related Art


Optical communication systems employ optical amplifiers, e.g., to compensate for signal attenuation in optical fibers. One type of amplifier that may be used in a fiber-based optical communication system is an optical fiber amplifier (OFA). Many current optical communication systems employ erbium-doped and/or Raman fiber amplifiers. These amplifiers are examples of phase-insensitive amplifiers (PIAs). More specifically, a PIA produces signal gain that is independent of the signal phase. In contrast, a phase-sensitive amplifier (PSA) produces signal gain that does depend on the signal phase. Potential advantages of using PSAs instead of or in addition to PIAs in optical communication systems include, but are not limited to, noise reduction, the reduction of noise-induced frequency and phase fluctuations, dispersion compensation, and suppression of modulational instability. A more detailed discussion of these and other potential advantages of PSA use in optical communication systems can be found, e.g., in the following publications: (1) R. Loudon, “Theory of Noise Accumulation in Linear Optical-Amplifier Chains,” Institute of Electrical and Electronic Engineers (IEEE) Journal of Quantum Electronics, vol. 21, pp. 766-773 (1985); (2) H. P. Yuen, “Reduction of Quantum Fluctuation and Suppression of the Gordon-Haus Effect with Phase-Sensitive Linear Amplifiers,” Optics Letters, vol. 17, pp. 73-75 (1992); (3) Y. Mu and C. M. Savage, “Parametric Amplifiers in Phase-Noise-Limited Optical Communications,” Journal of the Optical Society of America B, vol. 9, pp. 65-70 (1992); (4) R. D. Li, P. Kumar, W. L. Kath, and J. N. Kutz, “Combating Dispersion with Parametric Amplifiers,” IEEE Photonics Technology Letters, vol. 5, pp. 669-672 (1993); and (5) W. Imajuku and A. Takada, “Reduction of Fiber-Nonlinearity-Enhanced Amplifier Noise by Means of Phase-Sensitive Amplifiers,” Optics Letters, vol. 22, pp. 31-33 (1997), the teachings of all of which are incorporated herein by reference. However, PSAs suitable for use in fiber-based optical communication systems are not yet sufficiently developed.


SUMMARY OF THE INVENTION

Problems in the prior art are addressed, in accordance with the principles of the present invention, by a method of and device for generating an amplified optical signal directly in an optical fiber by way of phase-sensitive amplification based on one or more four-wave mixing (FWM) processes. In one embodiment, an input signal and two pump waves are applied to a highly nonlinear fiber (HNLF). The input signal is amplified in the HNLF due to energy transfer from the pump waves to the input signal via a degenerate phase-conjugation (PC) process. The amplified signal is then separated from the pump waves by an optical filter placed at the end of the HNLF.


In another embodiment, an input signal and first and second pump waves are applied to a first HNLF to generate, via a Bragg scattering (BS) process, an idler signal corresponding to the input signal. A first optical filter placed at the end of the first HNLF then filters out the second pump wave while passing the first pump wave and the input and idler signals to a second HNLF, which also receives a third pump wave. In the second HNLF, the first and third pump waves and the input and idler signals interact via a non-degenerate PC process, which results in energy transfer from the pump waves to the input signal. The amplified input signal is then separated from the pump waves and the idler signal by a second optical filter placed at the end of the second HNLF.




BRIEF DESCRIPTION OF THE DRAWINGS

Other aspects, features, and benefits of the present invention will become more fully apparent from the following detailed description, the appended claims, and the accompanying drawings in which:



FIG. 1 graphically illustrates a simplified diagram of the frequency structure developed via a representative optical parametric amplification (OPA) process in a single-pump configuration;



FIG. 2 graphically illustrates a simplified diagram of the frequency structure developed via OPA in a double-pump configuration;



FIG. 3 graphically illustrates a scalar four-wave mixing (FWM) process according to one embodiment of the invention;



FIG. 4 graphically illustrates a vector FWM process according to another embodiment of the invention;



FIG. 5 shows an optical fiber amplifier (OFA), in which the processes illustrated in FIGS. 3 and 4 can be carried out according to one embodiment of the invention;



FIG. 6 graphically illustrates a scalar cascaded FWM process according to one embodiment of the invention;



FIG. 7 graphically illustrates a vector cascaded FWM process according to another embodiment of the invention; and



FIG. 8 shows an OFA, in which the processes illustrated in FIGS. 6 and 7 can be carried out according to one embodiment of the invention.




DETAILED DESCRIPTION

Reference herein to “one embodiment” or “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the invention. The appearances of the phrase “in one embodiment” in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments.


Optical Parametric Amplification


As known in the art, optical parametric amplification (OPA) is a process that produces a tunable coherent optical signal via nonlinear optical processes, in which, typically, one or two pump-wave photons are converted into two new photons with conservation of photon energy and momentum. The waves corresponding to the two new photons are usually referred to as a signal and an idler. An excellent review of the fundamentals of OPA can be found in a book by G. P. Agrawal, “Nonlinear Fiber Optics” (Third Edition), Academic Press, 2001.



FIG. 1 graphically illustrates a representative OPA process in a single-pump configuration. More specifically, in the configuration of FIG. 1, the OPA process is governed by degenerate four-wave mixing (FWM), in which two pump photons (labeled P in FIG. 1) combine to generate one “signal” photon and one “idler” photon (labeled s and i, respectively, in FIG. 1) according to the following equation:

Psi  (1)

where ωP, ωs, and ωi are the frequencies of the pump, signal, and idler photons, respectively. The nonlinear medium, in which the FWM process occurs, is characterized by a nonlinearity coefficient (γ) and a set dispersion coefficients, each of which is frequency dependent. The frequency or wavelength at which the second-order dispersion coefficient equals zero is referred to as the zero-dispersion frequency (ω0) or wavelength (λ0). The regions in which the second-order dispersion coefficient is positive and negative are referred to as the normal dispersion region and the anomalous dispersion region, respectively.


The degenerate FWM process of FIG. 1 occurs when the wavelength of the pump wave is in the anomalous dispersion region of the nonlinear optical medium (e.g., optical fiber). It is known in the art that the parametric gain depends on the intensity of the pump wave (Ip) and the interaction length in the optical fiber. Depending on the relationship between (1) the wave-vector mismatch coefficient (κ) in the optical fiber, (2) the nonlinearity coefficient, γ, and (3) the intensity of the pump wave, Ip, FWM can produce either exponential or quadratic gain. For example, if the intensity of the pump wave is chosen such that κ=−γip, then the gain is relatively high and exponential. In contrast, if κ=0, then the gain is quadratic. Other values of κ will correspond to a relatively low exponential gain.



FIG. 2 shows a simplified diagram of the frequency structure developed via OPA in a two-pump configuration. In the presence of two pump waves labeled P1 and P2 located at frequencies ω1 and ω2, respectively, and an optical communication signal (illustratively a sideband at frequency ω1−), various FWM interactions in the nonlinear optical medium produce three complementary sidebands at frequencies ω1+, ω2−, and ω2+. In general, the optical communication signal may correspond to any one of the four sidebands, with the remaining three sidebands being generated by FWM processes.


The following describes the FWM processes that lead to the frequency structure shown in FIG. 2. Assuming that the optical communication signal is at frequency ω1− and the remaining three sidebands are idler sidebands, the modulational interaction (MI) produces a first idler sideband at frequency ω1+ according to Eq. (2):

11−1+  (2)

A Bragg scattering (BS) process produces a second idler sideband at frequency ω2− according to Eq. (3):

ω1−212−  (3)

A phase-conjugation (PC) process produces a third idler sideband at frequency ω2+ according to Eq. (4):

ω121−2+  (4)

In addition, each of the three idler sidebands is coupled to the other two idler sidebands by an appropriate FWM process, i.e., MI, BS, or PC, that can be expressed by an equation analogous to one of Eqs. (2), (3), or (4).


In addition to the sidebands illustrated in FIG. 2, OPA may also generate several additional sidebands (not shown). For example, MI with P2 generates an additional sideband at frequency 2ω2−ω1+δω, where δω=ω1−ω1−. Also, a BS process with respect to P1 generates another sideband at frequency 2ω1−ω2−δω. Similarly, the sidebands at frequencies ω1+ and ω2− are each coupled to additional sidebands at frequencies 2ω2−ω1−δω and 2ω1−ω2+δω, and the sideband at frequency ω2+ is coupled to the aforementioned additional sidebands with frequencies 2ω2−ω1+δω and 2ω1−ω2−δω. However, unlike the four original sidebands shown in FIG. 2, each of which is coupled to each of the other three, none of the additional sidebands is coupled to all of the original four or all of the other three additional sidebands. Furthermore, for most values of da), the additional sidebands are driven non-resonantly. Consequently, effects of the additional sidebands on the overall OPA process are relatively small and, for all practical purposes, can be substantially ignored. Numerical simulations validate this conclusion.


Phase-Sensitive Amplification in a χ(2) Medium


A representative prior-art scheme for phase-sensitive amplification (and squeezed-state generation) is described in a book by R. Loudon, “The Quantum Theory of Light,” 3rd Ed., Oxford University Press, Oxford, 2000. More specifically, phase-sensitive amplification in a χ(2) medium can occur via degenerate parametric amplification (DPA, also referred to as frequency down-conversion). As known in the art, a χ(2) medium is a nonlinear optical medium characterized in that its nonlinear optical response is governed primarily by the second-order nonlinear optical susceptibility (χ(2), a second-rank tensor). Typically, a χ(2) medium is a non-centrosymmetric crystal, two representative examples of which are (i) a lithium iodate crystal (LiIO3) and (ii) a KTP crystal (potassium titanium oxide phosphate, KTiOPO4).


In a DPA process, a pump wave interacts with signal and idler waves of a lower frequency, where the signal and idler waves have the same (i.e., common) frequency. This interaction is governed by the frequency-matching condition

χ2=2ω1  (5)

where ω2 is the pump frequency and ω1 is the common signal and idler frequency. Amplitude changes of waves <1> and <2> are governed by Eqs. (6) and (7):

dzA1=i2{overscore (γ)}A2Ai*exp(−iβz)  (6)
dzA2=i{overscore (γ)}A12 exp(iβz)  (7)

where Aj is the amplitude of the corresponding wave, z is the coordinate along the wave propagation direction, {overscore (γ)} is the nonlinear coupling coefficient, and β=2β1−β2 is the linear wavenumber mismatch. One can choose the amplitude units in such a way that |Aj|2 is proportional to the photon flux Pj.


Suppose that wave <2> is a relatively strong pump wave and wave <1> is a relatively weak signal wave. Then, in the small-signal (undepleted-pump) approximation, A2(z)=A2(0), where A2(0) is the wave amplitude at the beginning of the medium. Let

A1(z)=B1 exp(−iβz/2)

Then, B1 obeys the following (linearized) equation:

(dz−iδ)B1=iγB1*  (8)

where δ=β/2 and γ=2{overscore (γ)}A2. Eq. (8) has a solution

B1(z)=μ(z)B1(0)+ν(z)B1(0)*  (9a)

where the transfer functions μ(z) and ν(z) are:

μ(z)=cos hz)+i(δ/κ)sin h(κz)  (9b)
ν(z)=i(γ/κ)sin hz)  (9c)

and the growth rate κis:

κ=(|γ|2−δ2)1/2

The input-output relation described by Eqs. (9) is a property of phase-sensitive amplification. In quantum optics this relation is called a squeezing transformation.


To illustrate important characteristics of phase-sensitive amplification, consider a simple case, in which δ=0, and suppose that A2 is real (measure the signal phase relative to the pump phase). Then, according to Eqs. (9), the in-phase signal quadrature (B1+B1*)/2 is amplified by the factor exp(κz), whereas the out-of-phase quadrature (B1−B1*)/2 is attenuated by the same factor (i.e., multiplied by the factor exp(−κz)). This property of phase-sensitive amplification is largely responsible for the advantages of PSAs over PIAs specified in the background section.


Phase-Sensitive Amplification in a χ(3) Medium


As known in the art, a χ(3) medium is a nonlinear optical medium characterized in that its nonlinear optical response is governed primarily by the third-order nonlinear optical susceptibility (χ(3), a third-rank tensor). Typically, a χ(3) medium is a centrosymmetric crystal or nonlinear glass. One representative example of a χ(3) medium is a Kerr medium, in which the nonlinear response has a special form of χ(3)|E|2 E (where E is the electric field) often referred to as the Kerr nonlinearity. More details on various forms of nonlinear optical interactions in χ(2) and χ(3) media can be found, e.g., in a book by R. W. Boyd, entitled “Nonlinear Optics,” Academic Press, 1992, pp. 1-37 and 57-63, the teachings of which are incorporated herein by reference.


Representative examples of prior-art phase-sensitive amplification in a χ(3) medium are described in: (i) H. P. Yuen and J. H. Shapiro, “Generation and Detection of Two-Photon Coherent States in Degenerate Four-Wave Mixing,” Opt. Letters, vol. 4, pp. 334-336 (1979); (ii) B. Yurke, “Use of Cavities in Squeezed-State-Generation,” Phys. Rev. A, vol. 29, pp. 408-410 (1984); and (iii) P. Kumar and J. H. Shapiro, “Squeezed-State Generation via Forward Degenerate Four-Wave Mixing,” Phys. Rev. A, vol. 30, 1568-1571 (1984), the teachings of all of which are incorporated herein by reference.


The Yuen and Yurke references disclose schemes that utilize a degenerate backward FWM process. More specifically, a degenerate backward FWM process uses a counter-propagating pump configuration, in which two pump waves having the same frequency travel in opposite directions. Due to the phase-matching condition, the signal and idler waves travel in opposite directions as well. Phase sensitivity (and squeezing) is obtained by the use of a semi-transparent (50/50) beam splitter (Yuen) or a partly reflecting cavity mirror (Yurke) placed outside the χ(3) medium, which splitter/mirror combines the signal and idler waves to produce a desired squeezed state.


The Kumar reference discloses a scheme that is similar to the scheme disclosed in the Yuen reference, except that the counter-propagating pump configuration is replaced with a nominally co-propagating pump configuration. More specifically, each of the pump waves propagates at a small angle with respect to a selected axis. To satisfy the phase-matching condition, each of the signal and idler waves also propagates at that small angle with respect to that selected axis such that the plane passing through the wave vectors of the pump waves is orthogonal to the plane passing through the wave vectors of the signal and idler waves. In addition, all waves in the Kumar reference have the same frequency.


It should be noted that the χ(3) medium employed in either of the above schemes is a three-dimensional (3D) medium, i.e., a medium for which the transverse dimensions are comparable with the longitudinal dimension. In contrast, an optical fiber is a substantially one-dimensional medium because the fiber's transverse dimension (e.g., diameter) is typically much smaller than its longitudinal dimension (e.g., length).


First attempts to achieve phase-sensitive amplification in a fiber are disclosed in (1) M. Shirasaki and H. A. Haus, “Squeezing of Pulses in a Nonlinear Interferometer,” Journal of the Optical Society of America B, vol. 7, pp. 30-34 (1990) and (2) M. E. Marhic and C. H. Hsia, “Optical Amplification in a Nonlinear Interferometer,” Electronics Letters, vol. 27, pp. 210-211 (1991). More specifically, these references disclose schemes, each of which utilizes a nonlinear interferometer. The interferometer employs one or more 50/50 beam splitters to (1) direct the pump and signal waves along two interferometer arms, where the waves interact, and (2) combine the waves emerging from the arms to produce the corresponding squeezed states. Note that, in each of these interferometer schemes, (i) a single pump source is used to produce the pump waves that interact with the signal wave(s) in the interferometer arms and (ii) the signal gain increases as a quadratic function of the arm (fiber) length.


Four-Wave Mixing in a χ(3) Medium



FIG. 3 graphically illustrates a scalar FWM (inverse MI) process according to one embodiment of the invention. More specifically, FIG. 3 shows a polarization diagram, in which (i) each arrow represents a wave, with the length of the arrow schematically representing the wave's intensity, (ii) the position of the arrow along the ω-axis indicates the frequency of the wave, and (iii) the orientation of the arrow with respect to the ⊥ and ∥ axes represents the wave's polarization. In FIG. 3, waves <1> and <3> are pump waves and wave <2> is a signal wave, where all waves are (vertically) co-polarized and propagate in a fiber along the same direction.


The scalar FWM process illustrated in FIG. 3 is governed by the frequency-matching condition given by Eq. (10):

ω31=2ω2  (10)

and the amplitude equations given by Eqs. (11):

dzA1=i{overscore (γ)}(|A1|2+2|A2|2+2|A3|2A1+i{overscore (γ)}A22A3*exp(iβz)  (11a)
dzA2=i{overscore (γ)}(2|A1|2+|A2|2+2|A3|2)A2+i2{overscore (γ)}A3A1A2*exp(−iβz)  (11b)
dzA3=i{overscore (γ)}(2|A1|2+2|A2|2+|A3|2)A3+i{overscore (γ)}A1*A22 exp(iβz)  (11c)

where Aj is the amplitude of the corresponding wave, z is the wave propagation direction, {overscore (γ)} is the nonlinear coupling coefficient, and β=2β2−β3−β1 is the wavenumber mismatch. In this scalar FWM process, the signal and idler are substantially identical. The self-phase modulation (SPM) and cross-phase modulation (CPM) coefficients are not exactly equal to {overscore (γ)} and 2{overscore (γ)}, respectively. However, the deviations of these coefficients from the stated values are qualitatively unimportant and, for typical frequencies, are quantitatively insignificant. In one embodiment, ω2 is located within about 10 THz of ω0 (the zero-dispersion frequency).


Suppose that waves <1> and <3> are relatively strong pump waves and wave <2> is a relatively weak signal wave. Then, in the small-signal approximation, the pump photon-fluxes (P1,3) are constant and:

A3(z)=A3(0)exp [i{overscore (γ)}(2P1+P3)z]  (12)
A1(z)=A1(0)exp [i{overscore (γ)}(P1+2P3)z]  (13)

In a χ(3) medium, the pump waves are subject to SPM and CPM. Let

A1(0)=B1  (14a)
A2(z)=B2(z)exp [−iβz/2+i3{overscore (γ)}(P3+P1)z/2]  (14b)
A3(0)=B3  (14c)

Then, the transformed signal amplitude obeys the (linearized) equation:

dz−iδ)B2=iγB2*  (15)

where δ=β/2+{overscore (γ)}(P3+P1)/2 and γ=2{overscore (γ)}B3B1. Eq. (15) has the same form as Eq. (8). Hence, the input-output relation for the process of FIG. 3 is described by Eqs. (9), and this process provides phase-sensitive amplification in a fiber. Note that a process governed by Eqs. (9) provides signal gain that increases as an exponential function of the fiber length (see the preceding subsection). This characteristic is advantageously different from that of the above-described interferometer schemes (capable of providing only a much slower quadratic gain).



FIG. 4 graphically illustrates a vector FWM process according to another embodiment of the invention. FIG. 4 is a polarization diagram analogous to that shown in FIG. 3. In FIG. 4, waves <1> and <4> are cross-polarized pump waves, and waves <2> and <3> are signal and idler waves. All waves propagate in a fiber along the same direction. When the pump waves are cross-polarized (a distinguishing feature of vector FWM), e.g., as shown in FIG. 4, then the signal wave is aligned with one pump wave and the idler is aligned with the other pump wave. Note that the process illustrated in FIG. 4 is a degenerate process, where ω32.


The vector FWM process of FIG. 4 is governed by the frequency-matching condition given by Eq. (16):

ω41=2ω2  (16)

and the following amplitude equations:

dzA1=i{overscore (γ)}(|A1|2+2|A2|2+ε|A3|2+ε|A4|2)A1+i{overscore (γ)}εA22A3A4*exp(iβz)  (17a)
dzA2=i{overscore (γ)}(2|A1|2+|A2|2+ε|A3|2+ε|A4|2)A2+i{overscore (γ)}εA3*A4A1 exp(−iβz)  (17b)
dzA3=i{overscore (γ)}(ε|A1|2+ε|A2|2+|A3|2+2|A4|2)A3+i{overscore (γ)}εA4A1A2*exp(−iβz)  (17c)
dzA4=i{overscore (γ)}(ε|A1|2+ε|A2|2+2|A3|2+|A4|2)A4+i{overscore (γ)}εA1*A2A3 exp(iβz)  (17d)

where Aj is the amplitude of the corresponding wave, z is the wave propagation direction, β=β23−β4−β1 is the wavenumber mismatch, {overscore (γ)} is the nonlinear coupling coefficient for co-polarized waves, and ε is the ratio of the coupling coefficients for cross-polarized and co-polarized waves. For (polarization-maintaining) fibers with constant dispersion, ε=⅔; and for (non-polarization-maintaining) fibers with random dispersion, ε=1.


Suppose that waves <4> and <1> are relatively strong pump waves, and waves <2> and <3> are the polarization components of a relatively weak signal. Then, in the small-signal approximation, the pump photon-fluxes (P1,4) are constant and:

A4(z)=A4(0)exp [i{overscore (γ)}(εP1+P4)z]  (18)
A1(z)=A1(0)exp [i{overscore (γ)}(P1+εP4)z]  (19)

As in the scalar process described above, the pump waves are subject to SPM and CPM. Let

A1(0)=B1  (20a)
A2(z)=B2(z)exp [−iβz/2+i{overscore (γ)}3P1z/2+i{overscore (γ)}(ε−½)P4z]  (20b)
A3(z)=B3(z)exp [−iβz/2+i{overscore (γ)}(ε−½)P1z+i{overscore (γ)}3P4z/2]  (20c)
A4(0)=B4  (20d)

Then, the transformed signal components obey the following (linearized) equations:

(dz−iδ)B2=iγB3*  (21)
(dz+iδ)B3*=−iγ*B2  (22)

where δ=β/2+{overscore (γ)}(P4+P1)/2 and γ={overscore (γ)}εB4B1. It follows from Eqs. (21)-(22) that:

B2(z)=μ(z)B2(0)+ν(z)B3*(0)  (23)
B3*(z)=ν*(z)B2(0)+μ*(z)B3*(0)  (24)

where μ and ν are defined in Eqs. (9b)-(9c). In quantum optics, the input-output relation defined by Eqs. (23) and (24) is called a two-mode squeezing transformation. If there is no input idler (i.e., B3(0)=0), then the output idler is proportional to the complex conjugate of the input signal, and the FWM process illustrated in FIG. 4 is a phase conjugation (PC) process. If the input signal is split evenly between the two orthogonal polarizations (i.e., B3(0)=B2(0)), then Eq. (23) reduces to Eq. (9a). Hence, the degenerate vector FWM of FIG. 4 produces phase-sensitive amplification characterized by an exponential gain function (i.e., with the signal gain increasing exponentially with the fiber length).



FIG. 5 shows an optical fiber amplifier (OFA) 500, in which the processes illustrated in FIGS. 3 and 4 can be carried out according to one embodiment of the invention. OFA 500 is coupled between two sections 502 and 502′ of an optical fiber and has couplers 504a-b configured to combine an optical communication signal from section 502 with pump waves generated by pump-wave sources (e.g., lasers) 506a and 506b. For example, in reference to FIG. 3, pump-wave source 506a can generate pump wave <1> while pump-wave source 506b generates pump wave <3>. In reference to FIG. 4, pump-wave sources 506a and 506b generate pump waves <1> and <4>, respectively. The two pump waves are combined together using coupler 504a, and the resulting wave is then combined with the optical communication signal using coupler 504b. Couplers 504a-b incorporate polarization filters as appropriate and/or necessary to provide a desired polarization configuration, e.g., as illustrated in FIGS. 3 and 4. In another embodiment, a three-way coupler may be used to combine the optical communication signal with the pump waves. In general, as long as the two pump waves and the optical communication signal are coupled into a highly nonlinear fiber (HNLF) 508, they can be combined in any combination/sequence. Depending on the implementation of OFA 500, the pump waves may be continuous-wave (CW) or pulsed optical signals. The optical communication signal interacts with the pump waves in HNLF 508, e.g., as described above in the context of FIGS. 3 and 4, and is amplified due to this interaction. A filter 510 placed at the end of HNLF 508 separates the amplified optical communication signal from the pump waves for further transmission in the communication system via section 502′.


Cascaded Four-Wave Mixing Processes in a χ(3) Medium


The degenerate FWM processes described in the preceding subsection provide phase-sensitive amplification in a fiber because the frequency degeneracies allow the signal amplitudes to interact with their complex-conjugates (see Eqs. (9)). More specifically, in the degenerate PC process of FIG. 3, the idler and signal waves have the same frequency, ω2. Similarly, in the degenerate PC process of FIG. 4, the idler and signal waves have the same frequency, ω23. Non-degenerate PC processes that are analogous to the degenerate PC processes shown in FIGS. 3 and 4, can provide phase-sensitive amplification in a fiber if the idler amplitude is nonzero and the idler wave has a specific phase with respect to the signal wave. However, such an idler wave is usually absent. Embodiments of the cascaded FWM schemes disclosed below address this problem by providing such an idler wave, e.g., by using Bragg scattering (BS) as an additional FWM process. More specifically, BS generates an idler wave that is a frequency-shifted, but non-conjugated, image of the signal wave. Once the idler wave has been generated by BS, a non-degenerate PC process analogous to that of FIG. 4 can be used to provide phase-sensitive amplification.



FIG. 6 graphically illustrates a scalar cascaded FWM process according to one embodiment of the invention. FIG. 6 is a polarization diagram analogous to those shown in FIGS. 3 and 4. In FIG. 6, waves <1>, <3>, and <5> are pump waves, wave <2> is a signal wave, and wave <4> is an idler wave. All waves are (vertically) co-polarized and propagate in a fiber along the same direction. Wave <1> is on for both the BS and PC processes. However, during BS, wave <3> is on and wave <5> is off; and, during PC, wave <3> is off and wave <5> is on.



FIG. 7 graphically illustrates a vector cascaded FWM process according to another embodiment of the invention. FIG. 7 is analogous to FIG. 6. In FIG. 7, waves <1>, <3>, and <5> are pump waves, wave <2> is a signal wave, and wave <4> is an idler wave. Waves <1> and <2> are vertically polarized and waves <3>, <4>, and <5> are horizontally polarized. All waves propagate in a fiber along the same direction. Wave <1> is on for both the BS and PC processes; during BS, wave <3> is on and wave <5> is off; and, during PC, wave <3> is off and wave <5> is on. Note that the following description applies to both FIGS. 6 and 7.


Consider the BS process. As already explained above, BS is governed by the frequency-matching condition:

ω2341  (25)

and the amplitude equations analogous to Eqs. (17). Note that, unlike for the FWM process described in the preceding subsection, for the BS process, waves <1> and <3> are the pump waves (rather than waves <1> and <4> above), wave <2> is the signal wave, and wave <4> is the idler wave (rather than wave <3> above). Because power flows from the signal to the idler, the photon flux of the output idler wave cannot exceed the photon flux of the input signal wave. As a result, BS is intrinsically stable (i.e., does not provide gain).


Suppose that waves <1> and <3> are strong pumps, wave <2> is a weak signal, and wave <4> is a weak idler. Then, in the small-signal approximation, the pump photon-fluxes (P1,3) are constant and:

A1(z)=A1(0)exp [i{overscore (γ)}(P1+εP3)z]  (26)
A3(z)=A3(0)exp [i{overscore (γ)}(εP1+P3)z]  (27)

As in the PC process described above, the pump waves are subject to SPM and CPM. Let

A1(0)=B1  (28a)
A2(z)=B2(z)exp [−iβz/2+i{overscore (γ)}3P1z/2+i{overscore (γ)}(ε+½)P3z]  (28b)
A3(0)=B3  (28c)
A4(z)=B4(z)exp [iβz/2+i{overscore (γ)}(ε+½)P1z+i{overscore (γ)}3P3z/2]  (28d)

Then, the signal and idler amplitudes obey the following (linearized) equations:

(dz−iδ)B2=iγB4  (29)
(dz+iδ)B4=iγ*B2  (30)

where δ=β/2+{overscore (γ)}(P1−P3)/2 and γ={overscore (γ)}εB1B3*. It follows from Eqs. (29) and (30) that

B2(z)=μ(z)B2(0)+ν(z)B4(0)  (31)
B4(z)=−ν(z)*B2(0)+μ*(z)B4(0)  (32)

where the transfer functions are

μ(z)=cos(kz)+i(δ/k)sin(kz)  (33)
ν(z)=i(γ/k)sin(kz)  (34)

and the wavenumber k=(|γ|22)1/2. The transfer functions satisfy the auxiliary equation |μ|2+|ν|2=1. In quantum optics, the input-output relation defined by Eqs. (31)-(34) is called a beam-splitter transformation. For a representative case, in which δ=0, k=π/2 and γ is real, B2(z)=B2(0)/21/2 and B4(z)=iB2(0)/21/2. Thus, the output idler wave is a phase-shifted, but non-conjugated image of the signal wave. (Note that the phase shift results from photon-flux conservation.) Because the signal and idler frequencies are distinct, their relative phases can be different in the BS and PC processes (note that, because k4≠k2, propagation effects can change the relative phase naturally).


Now consider the PC process in which waves <5> and <1> are pump waves, wave <2> is a signal wave, and wave <4> is an idler wave. This process is governed by the frequency-matching condition

ω5124  (35)

and the amplitude equations given by Eqs. (17), in which the subscripts 3 and 4 are replaced by the subscripts 4 and 5, respectively. The effects of SPM and CPM on the pump waves are described by Eqs. (18) and (19). The signal and idler amplitudes obey the following (linearized) equations:

(dz−iδ)B2=iγB4*  (36)
(dz+iδ)B4*=−iγ*B2  (37)

where δ=β/2+{overscore (γ)}(P5+P1)/2, γ={overscore (γ)}εB5B1, and β=β24−β5−β1. It follows from Eqs. (36) and (37) that

B2(z″)=μ(z″−z′)B2(z′)+ν(z″−z′)B4*(z′)  (38)
B4*(z″)=ν*(z″−z′)B2(z′)+μ*(z″−z′)B4*(z′)  (39)

where μ and ν are defined by Eqs. (9b) and (9c), respectively. Because B2(z′) is proportional to B2(0) and B4*(z′) is proportional to B2*(0), cascaded BS and PC provide PSA in a fiber. By controlling the phase of pump wave <5>, one controls the relative phase of the pump, signal, and idler waves and, hence, the orientation of the squeezing axis. For the representative case described above after Eq. (34), in which |B4(z′)|=|B2(z′)|, Eq. (38) reduces to Eq. (9a). It should be noted that the choice of pump frequencies is not unique. That is, different combinations of pump frequencies can provide phase-sensitive amplification by cascaded BS and PC. Furthermore, BS can also be used to combine a signal wave with the frequency-shifted, and conjugated, idler wave produced by prior PC.



FIG. 8 shows an OFA 800, in which the processes illustrated in FIGS. 6 and 7 can be carried out according to one embodiment of the invention. OFA 800 is similar to OFA 500 (FIG. 5) in that it is coupled between two sections 802 and 802′ of an optical fiber. However, one difference between OFAs 500 and 800 is that the latter OFA has two cascaded stages. The first stage in OFA 800 includes pump-wave sources (e.g., lasers) 806a-b, couplers 804a-b, an HNLF 808a, and a filter 810a. The output of the first stage is fed into the second stage that includes a pump-wave source 806c, a coupler 804c, an HNLF 808b, and a filter 810b. In one embodiment, couplers 804a-c incorporate polarization filters as appropriate and/or necessary to provide a desired polarization configuration, e.g., as illustrated in FIGS. 6 and 7.


Coupler 804a combines the pump waves generated by pump-wave sources 806a and 806b, and coupler 804b combines the resulting pump wave the optical communication signal received from section 802. The combined pump and signal waves interact in HNLF 808a via a BS process to generate an idler wave, e.g., as described above. Filter 810a placed at the end of HNLF 808a then filters out one of the pump waves (e.g., the pump wave generated by pump-wave source 806b) while passing the other pump wave (e.g., the pump wave generated by pump-wave source 806a) and the signal and idler waves to the second stage.


Coupler 804c adds to the waves received from the first stage a pump wave generated by pump-wave source 806c. These pump, signal, and idler waves interact in HNLF 808b via a PC process to generate an amplified wave, e.g., as described above. Filter 810b placed at the end of HNLF 808b then separates the amplified wave from the other waves present in HNLF 808b for further transmission in the communication system via section 802′. In general, since both the signal wave and the idler wave are amplified in HNLF 808b, either one of them, or both, can be used as an amplified output of OFA 800. In reference to FIGS. 6-7, pump sources 806a-c and filter 810a can be configured as follows: pump sources 806a-c generate pump waves <1>, <3>, and <5>, respectively, and filter 810a blocks pump wave <3>.


While this invention has been described with reference to illustrative embodiments, this description is not intended to be construed in a limiting sense. For example, in different OFA implementations, optical fibers having different properties (e.g., polarization maintaining or regular fibers) may be used. Polarization sensitive filters or polarizers can be appropriately inserted at different locations of the OFA to provide a desired polarization configuration. The pump waves may be CW or pulsed. Two, three, or more pump waves may be used. Idler waves may be filtered out or used in the communication system, e.g., as protection signals in a 1+1 protection scheme. Various modifications of the described embodiments, as well as other embodiments of the invention, which are apparent to persons skilled in the art to which the invention pertains are deemed to lie within the principle and scope of the invention as expressed in the following claims.


Although the steps in the following method claims, if any, are recited in a particular sequence with corresponding labeling, unless the claim recitations otherwise imply a particular sequence for implementing some or all of those steps, those steps are not necessarily intended to be limited to being implemented in that particular sequence.

Claims
  • 1. A method of generating an optical signal, comprising: (a) applying an input signal and first and second pump waves to a first fiber, wherein the first pump wave has a first wavelength and the second pump wave has a second wavelength different from the first wavelength; and (b) generating, in the first fiber, an output signal corresponding to the input signal by way of phase-sensitive amplification, wherein energy is transferred from the first and second pump waves to the output signal.
  • 2. The invention of claim 1, wherein, for step (b), the generation of the output signal is based on a phase-conjugation (PC) process, which couples the first pump wave, the second pump wave, and the input signal.
  • 3. The invention of claim 1, wherein: for step (b), signal gain increases as an exponential function of fiber length; and the first and second pump waves are co-propagating waves.
  • 4. The invention of claim 1, wherein polarizations of the first and second pump waves are substantially parallel.
  • 5. The invention of claim 1, wherein the first and second pump waves are substantially orthogonally polarized with respect to one another.
  • 6. The invention of claim 1, further comprising separating the output signal from the first and second pump waves using an optical filter.
  • 7. The invention of claim 1, further comprising: applying the input signal, the first pump wave, and a third pump wave to a second fiber, wherein the third pump wave has a third wavelength different from the first and second wavelengths; and generating, in the second fiber, an idler signal corresponding to the input signal by way of a Bragg scattering (BS) process, which couples the first pump wave, the third pump wave, the input signal, and the idler signal.
  • 8. The invention of claim 7, further comprising separating the first pump wave, the input signal, and the idler signal from the third pump wave using an optical filter, wherein the separated first pump wave, the input signal, and the idler signal are applied to the first fiber.
  • 9. The invention of claim 8, wherein, for step (b), the generation of the output signal is based on a phase-conjugation (PC) process, which couples the first pump wave, the second pump wave, the input signal, and the idler signal.
  • 10. The invention of claim 7, wherein polarizations of the first, second, and third pump waves are substantially parallel.
  • 11. The invention of claim 7, wherein: the first and second pump waves are substantially orthogonally polarized with respect to one another; polarizations of the second and third pump waves are substantially parallel; polarization of the input signal is substantially parallel to the polarization of the first pump wave; and polarization of the idler signal is substantially parallel to the polarization of the third pump wave.
  • 12. The invention of claim 7, wherein the input signal and the idler signal are substantially orthogonally polarized with respect to one another.
  • 13. The invention of claim 7, wherein the output signal comprises an amplified idler signal.
  • 14. A device, comprising: a first fiber; first means for applying an input signal, a first pump wave, and a second pump wave to the first fiber, wherein: the first pump wave has a first wavelength and the second pump wave has a second wavelength different from the first wavelength; and the first fiber is adapted to generate an output signal corresponding to the input signal by way of phase-sensitive amplification, wherein energy is transferred from the first and second pump waves to the output signal.
  • 15. The invention of claim 14, wherein the first means for applying comprises: a first coupler adapted to combine the first pump wave and the second pump wave; and a second coupler adapted to combine (i) the combined first and second pump waves and (ii) the input signal.
  • 16. The invention of claim 14, further comprising a filter coupled to an end of the first fiber and adapted to pass the output signal and substantially block the first and second pump waves.
  • 17. The invention of claim 14, further comprising: a second fiber; second means for applying the input signal, the first pump wave, and a third pump wave to the second fiber, wherein: the third pump wave has a third wavelength different from the first and second wavelengths; and the second fiber is adapted to generate an idler signal corresponding to the input signal by way of a Bragg scattering (BS) process, which couples the first pump wave, the third pump wave, the input signal, and the idler signal; and a first filter coupled between the first and second fibers and adapted to pass the first pump wave, the input signal, and the idler signal and substantially block the third pump wave, wherein: the passed first pump wave, input signal, and idler signal are combined with the second pump wave and applied to the first fiber; and the first fiber is adapted to generate the output signal by way of a phase-conjugation (PC) process, which couples the first pump wave, the second pump wave, the input signal, and the idler signal.
  • 18. The invention of claim 17, wherein: the second means for applying comprises: a first coupler adapted to combine the first pump wave and the third pump wave; and a second coupler adapted to combine (i) the combined first and third pump waves and (ii) the input signal; and the first means for applying comprises a third coupler adapted to combine (i) the separated first pump wave, input signal, and idler signal and (ii) the second pump wave.
  • 19. The invention of claim 17, further comprising means for generating the first pump wave, the second pump wave, and the third pump wave.
  • 20. The invention of claim 17, further comprising a second filter coupled to an end of the first fiber and adapted to pass the output signal and substantially block the first and second pump waves.