Components for optical network

Abstract

An optical communication network includes an optical amplifier to amplify an input signal and provide an amplified signal. The optical amplifier includes a polarization splitter- rotator to split the input signal into a first polarization path and a second polarization path through the optical amplifier, and two or more components in the first polarization path. The two or more components include a polarization controller and amplifier or a parallel in-phase amplifier with two or more amplifiers. A polarization beam combiner combines signals at an output of the first polarization path and the second polarization path as the amplified signal. A transmitter of the input signal includes a light source to output a light signal. A modulator modulates a first portion of the light signal to output a modulated signal, and a polarization beam combiner combines the modulated signal and a second portion of the light signal to output the input signal.

Description

COMPONENTS FOR OPTICAL NETWORKRELATED APPLICATION

[0001] This application claims the benefit of priority under 35 U.S.C. § 119(e) to U.S. Provisional Application No., 63 / 551,196, filed February 8, 2025, which is incorporated herein by reference in its entirety.BACKGROUND

[0002] Optical fibers are used for transmission of data signals in a variety of applications and environments. For example, transatlantic optical cables carry signals across the Atlantic Ocean and terrestrial optical fiber networks may connect cities or countries in what may be referred to as long-haul links (generally cables longer than 20 to 50 kilometers (km)). As another example, fiber optic cables convey signals within data centers in what may be referred to as short-haul links. Regardless of the application and environment, optical fibers suffer from three major limitations or impairments: attenuation (a decrease in signal strength proportional to a length of the optical fiber), chromatic dispersion (a dispersion or spread in wavelength causing overlaps with adjacent signals), and four-wave mixing (nonlinear interactions between two or three wavelengths producing two or one new wavelengths).SUMMARY

[0003] In some embodiments, an optical communication network is provided for data communication.

[0004] According to some embodiments, an optical communication network includes an optical amplifier to amplify an input signal and provide an amplified signal. The optical amplifier includes a polarization splitter-rotator to split the input signal into a first polarization path and a second polarization path through the optical amplifier and two or more components disposed in the first polarization path. The two or more components include a polarization controller and amplifier or a parallel in- phase amplifier with two or more amplifiers. The optical amplifier also includes a polarization beam combiner to combine signals at an output of the first polarization path and the second polarization path as the amplified signal.

[0005] According to optional embodiments, the two or more components include the polarization controller and the amplifier. The amplifier may be a semiconductor-based optical amplifier. In some embodiments, the amplifier may be a parallel in-phase amplifier. The parallel in-phase amplifier may include two or more semiconductor-based optical amplifiers.

[0006] According to some embodiments, the optical communication network may include an optical switch with a plurality of ports. The optical communication network may include a plurality of the optical amplifiers and each of the plurality of ports of the optical switch may be connected to one of the plurality of the optical amplifiers. A transmitter may transmit a self-homodyne signal to the optical switch, the self-homodyne signal including an unmodulated portion and a modulated portion.

[0007] According to optional embodiments, the optical communication network may include a routing layer coupled with a plurality of the optical amplifiers to form an amplified switch. A switch may include N input ports and N output ports, and one or more of the N output ports may be respectively connected to the optical amplifier.

[0008] According to optional embodiments, the optical communication network may include a chromatic dispersion detector to detect chromatic dispersion in the input signal and facilitate chromatic dispersion compensation. The chromatic dispersion may be detected based on a tone added to the input signal.

[0009] According to optional embodiments, an optical communication network may use an input signal that includes multiple wavelengths. Multiple local oscillator portions of the input signal may be at the multiple wavelengths and have multiple first polarization orientations, respectively. Multiple modulated signal portions of the input signal may be at the multiple wavelengths and have multiple second polarization orientations, respectively. The first polarization orientations and the second polarization orientations may be interleaved across the multiple wavelengths that are multiplexed into the input signal, thereby mitigating four-wave mixing as the input signal propagates along a fiber. At each of the multiple wavelengths, a respective one of the first polarization orientations may be different than a respective one of the second polarization orientations.

[0010] According to some embodiments, an optical communication network includes a light source to output a light signal, a modulator to modulate a first portion of the light signal to output a modulated signal, and a polarization beam combiner configured to combine the modulated signal and a second portion of the light signal to output a transmission signal for transmission through optical fiber.

[0011] According to optional embodiments, a tunable coupler may generate and control respective powers of the first portion of the light signal and the second portion of the light signal. The light source may be external to a photonic integrated circuit including the modulator and the light signal may be conveyed from the light source to the photonic integrated circuit in an optical fiber. The optical fiber may be a polarization maintaining optical fiber. The optical communication network may include a polarization controller on the photonic integrated circuit to separate the first portion of the light signal and the second portion of the light signal according to polarization. The optical communication network may also include a phase modulator to add a tone to the modulated signal or the second portion of the light signal to facilitate chromatic dispersion detection.

[0012] It should be understood that aspects and features made possible by this disclosure at not limited to those described above but include additional or alternate aspects and features not explicitly mentioned that would clearly be understood by a person of ordinary relevant skill based on the description above and the details provided below.BRIEF DESCRIPTION OF THE DRAWINGS

[0013] The embodiments described throughout the disclosure will be better understood with reference to the following drawings and descriptions but should not be construed as being limited by the drawings. In the figures, like-referenced numerals designate corresponding parts throughout the different views. In addition, every instance of a part, illustrated similarly to another labeled instance of the part, may not be labeled for readability.

[0014] FIG. 1 is a block diagram of an optical amplifier according to some embodiments;

[0015] FIG. 2 is a block diagram detailing aspects of an exemplary polarization- multiplexed SOA-based amplifier (PMSA) used as an optical amplifier according to some embodiments;

[0016] FIG. 3 illustrates the exemplary PMSA of FIG. 2 and indicates a spectrum corresponding to each polarization for a polarization multiplexed self-homodyne (PM-SH) signal as it propagates along the fiber;

[0017] FIG. 4 illustrates is a block diagram of an optical amplifier with an exemplary PMSA used as an optical amplifier according to some embodiments;

[0018] FIG. 5 is a block diagram of an optical amplifier with a PMSA used as an optical amplifier according to some embodiments;

[0019] FIG. 6 is a block diagram of an exemplary optical amplifier employing parallel in-phase amplifiers (PIAs) to amplify each polarization component according to some embodiments;

[0020] FIG. 7 is a block diagram of an exemplary optical amplifier used with a polarization-maintaining (PM) fiber according to some embodiments;

[0021] FIG. 8 is a block diagram of a bidirectional optical amplifier according to some embodiments;

[0022] FIG. 9 is a block diagram of a bidirectional optical amplifier according to some embodiments;

[0023] FIG. 10 is a block diagram of a bidirectional optical amplifier according to some embodiments;

[0024] FIG. 11 is a block diagram of a bidirectional optical amplifier with signals of different wavelengths in each direction according to some embodiments;

[0025] FIG. 12 is a block diagram of a bidirectional optical amplifier with signals of different wavelengths in each direction according to some embodiments;

[0026] FIG. 13 is a block diagram of a bidirectional optical amplifier with signals of different wavelengths in each direction according to some embodiments;

[0027] FIG. 14 is a block diagram of a multi-wavelength optical amplifier according to some embodiments;

[0028] FIG. 15 is a block diagram of a multi-wavelength optical amplifier according to some embodiments;

[0029] FIG. 16 is a block diagram of a multi-wavelength optical amplifier according to some embodiments;

[0030] FIG. 17 is a block diagram of a multi-wavelength optical amplifier according to some embodiments;

[0031] FIG. 18 is a block diagram of a multi-wavelength optical amplifier according to some embodiments;

[0032] FIG. 19 illustrates an exemplary frequency spectrum for a multi- wavelength Input Signal designed to mitigate four-wave mixing (FWM) effects according to some embodiments;

[0033] FIG. 20 is a block diagram of an exemplary optical amplifier according to some embodiments;

[0034] FIG. 21 is a block diagram of aspects of an exemplary polarization controller (PC) with four degrees of freedom;

[0035] FIGS. 22-24 pertain to incorporating chromatic dispersion detection and compensation in optical amplifiers according to embodiments:

[0036] FIG. 22 is a block diagram of an exemplary optical amplifier that includes a tunable chromatic dispersion compensator (CDC) according to some embodiments;

[0037] FIG. 23 is a block diagram of aspects of an exemplary chromatic dispersion (CD) detector that may be included in a PMSA according to some embodiments;

[0038] FIG. 24 is a block diagram showing aspects of a transmitter according to some embodiments;

[0039] FIG. 25 is a block diagram of an exemplary amplified optical according to exemplary embodiments;

[0040] FIG. 26. illustrates exemplary optical links within a network that integrate an optical amplifier according to embodiments;

[0041] FIG. 27 is a block diagram of an all-optical network connecting N nodes using an NxN amplified optical switch according to some embodiments;

[0042] FIG. 28 shows aspects of an exemplary network using the exemplary amplified optical switch shown in FIG. 27;

[0043] FIG. 29 illustrates aspects of an exemplary network that includes five clusters connected to each other;

[0044] FIG. 30 illustrates connections to clusters used as uplinks and downlinks in order to form superclusters according to some embodiments;

[0045] FIG. 31 illustrates an exemplary network comprising interconnected super clusters according to some embodiments;

[0046] FIG. 32 shows an exemplary cluster in which the nodes are not interconnected using a local switch, as illustrated in FIG. 27;

[0047] FIG. 33 shows an exemplary amplified optical switch consisting of uplinks and downlinks according to embodiments;

[0048] FIG. 34 shows aspects of an exemplary network with N clusters connected to N clusters using a layer of N NXN amplified optical switches according to embodiments;

[0049] FIG. 35 shows aspects of an exemplary network according to some embodiments;

[0050] FIG. 36 illustrates an exemplary amplified optical switch according to some embodiments;

[0051] FIG. 37 shows a modification of the embodiment shown in FIG. 36;

[0052] FIG. 38 shows a broadband switch in which the received polarizations are routed in tandem to the desired output port according to some embodiments;

[0053] FIG. 39 depicts a broadband switch in which the routing layer is polarization- diverse according to some embodiments;

[0054] FIG. 40 is a block diagram of an exemplary embodiment of a single wavelength polarization multiplexed intensity modulated self-homodyne (PM-SH-IM) transmitter according to some embodiments;

[0055] FIG. 41 is a block diagram of an exemplary PM-SH-IM transmitter that incorporates a multi-wavelength laser and ring-resonator modulators according to some embodiments;

[0056] FIG. 42 is an exemplary multi-channel transmitter according to some embodiments;

[0057] FIG. 43 is an exemplary multi-channel transmitter in which the modulation is performed for a multi-wavelength signal of one of the polarizations following multiplexers according to some embodiments;

[0058] FIG. 44 is an exemplary multi-channel transmitter according to some embodiments in which the lasers are external to a photonic integrated circuit according to some embodiments;

[0059] FIG. 45 is an exemplary multi-channel transmitter according to some embodiments;

[0060] FIG. 46 is an exemplary multi-channel transmitter according to some embodiments;

[0061] FIG. 47 is a block diagram of an exemplary PM-SH-IM transmitter similar to the one shown in FIG. 44;

[0062] FIG. 48 is a block diagram of an exemplary PM-SH-IM transmitter similar to the one shown in FIG. 45;

[0063] FIG. 49 is a block diagram of an exemplary intensity-modulated direct- detection (IMDD) transmitter using external light sources according to some embodiments;

[0064] FIG. 50 is an exemplary PM-SH-IM receiver according to some embodiments; and

[0065] FIG. 51 illustrates an exemplary PM-SH-IM receiver according to some embodiments.DETAILED DESCRIPTION

[0066] As previously noted, optical fibers suffer from attenuation, chromatic dispersion, and four- wave mixing in various environments for various applications. In long- haul applications, a solution to address the impairments generally includes Erbium-doped fiber amplifiers (EDFAs), which led to the use of C-band wavelength range (1530 nm to 1565 nm), to address attenuation, digital signal processing (DSP) to address chromatic dispersion, and a combination of low transmission power and operation at high chromatic dispersion wavelengths to address four-wave mixing.

[0067] Data centers, which house servers, data storage, and network equipment, represent a fast- growing application for short-haul optical fiber links. Intra-data center links typically operate in the O-band range (1260 nm to 1360 nm) to obviate the need for chromatic dispersion compensation. In this exemplary application and other short-haul applications, routing of signals (e.g., from a signal source through multiple optical switches within a data center to a destination) can result in signal loss and distortion due to fiber impairments. The fact that channel spacing may be compact and wavelengths are close to the zero-dispersion point can cause four-wave mixing to have an increased impact on the transmission. Four-wave mixing in data center links, for example, can cause crosstalk and signal degradation, particularly in densely packed wavelength-division multiplexing (WDM) systems.

[0068] Electronic packet switches have typically been used for routing signals within data centers. However, this approach requires optical-electrical-optical (OEO) conversion at every switching point (referred to as a hop) so that the incoming optical signal can be switched electronically but sent to the next hop or destination via another optical fiber. In addition to requiring OEO conversion, electronic packet switches can introduce latency, which reduces the utilization rates of computing resources such as graphical processing units (GPUs) and tensor processing units (TPUs). The under-utilization may, in turn, necessitate a larger number of resources to achieve the same computational output, causing an increase hardware and operational costs.

[0069] Optical switches can reduce cost and power requirements, as compared to electronic packet switches, because the OEO conversion is not needed. However, optical switches can attenuate an incoming signal and simply amplifying the switch output (e.g., with Praseodymium-doped fiber amplifiers (PDFAs) akin to EDFAs used in long-haul applications) can lead to gain-bandwidth constraints resulting in a narrower range of usablewavelengths. That is, optical switches inevitably add insertion loss and, therefore, may require optical links with sufficient unallocated link budget (ULB). ULB is the measure of optical loss a link can tolerate while maintaining performance within a specified bit-error rate (BER). And, addressing the loss through amplification via PDFAs, for example, can increase cost, weight, and power consumption, in addition to limiting bandwidth. Semiconductorbased optical amplifiers (SOAs), as an alternative to PDFAs, can present their own challenges, because they introduce nonlinear distortion and pattern effects. In addition, the performance of SOAs is polarization dependent.

[0070] The inventors have recognized that optical switches and, more generally, a fully optical communication network can offer significant cost and power savings in optical communication applications such as in data centers. The inventors have also recognized that prior challenges posed by optical switches can be addressed through homodyne detection at the receivers. Homodyne detection refers to detection using a local oscillator (EO) (unmodulated signal) at the same frequency / wavelength as the modulated, data-containing, signal. According to some embodiments, self-homodyne transceivers may be used. In this case, the unmodulated signal generated by a single light source is split and used as both an EO and the basis for the modulated signal that is processed at a receiver by using the LO.

[0071] According to embodiments pertaining to a self-homodyne system, the short- haul links each convey two signals (e.g., LO and modulated signal). The inventors recognized that the self-homodyne approach addresses prior challenges to amplified optical switches, because the nonlinearity and distortion introduced by SOAs, which have made prior amplification methods impractical, can be addressed by amplifying the two signals independently. The modulated signal power is typically much lower than the SOA gain saturation power, and therefore does not suffer significantly from nonlinear distortion. Conversely, the higher power unmodulated signal, usually a continuous wave, avoids nonlinear distortion penalties from SOA because the generated harmonics fall outside of the signal band.

[0072] Amplified optical switches according to some embodiments include a polarization-multiplexed amplifier or, more particularly, a polarization-multiplexed SOA- based amplifier (PMSA). A PMSA facilitates individual configuration of polarization alignment for each transmitted wavelength in the optical link such that polarization- sensitive impairments (e.g., four- wave mixing) may be mitigated. The PMSA amplifies polarization- multiplexed signals (e.g., LO and modulated signals) and can also be thought of and referredto as a polarization-demultiplexed SOA-based amplifier since the two signals (e.g., LO and modulated signal) are first separated (demultiplexed) and amplified separately.

[0073] According to some embodiments, the two signals (e.g., LO and phase modulated signal) transmitted in a self-homodyne system are separated in polarization when introduced in the fiber (although the polarizations undergo rotation in the fiber). The transmitted signal is referred to as a polarization multiplexed self-homodyne coherent (PM- SH-C) signal in this case. The PMSA separates the LO and modulated signals, for example, by demultiplexing the polarizations of the PM-SH-C signal to amplify the demultiplexed polarizations independently. The two polarizations of a PM-SH-C signal (e.g., the phase modulated signal and the unmodulated signal (LO)), exhibit distinct characteristics that mitigate the SOA penalties as noted above. Specifically, a polarization controller (PC) may be used to separate the two signals (e.g., LO and modulated signal) for individual amplification. The amplified signals may be recombined for transmission on the next hop. The phase modulated signal can be further divided into two parts with orthogonal phases (i.e., a first modulated signal in a first quadrature and a second modulated signal in a second quadrature) when quadrature phase shift keying, rather than binary phase shift keying, is used.

[0074] According to other embodiments, the two signals may result from intensity modulation of one of the signals in an intensity modulated self-homodyne (PM-SH-IM) system. In this case, the two signals (e.g., LO and intensity modulated signal) may again be separated by polarization and individually amplified. A PC, different to the one used with a PM-SH system, may be used to separate the two signals for individual amplification.

[0075] A self-homodyne transmitter may include a tunable coupler that facilitates controlling the power split in the light signal that results in the LO and modulated signals. When the light source and modulator are on different integrated circuits, a polarization maintaining fiber may be used to obtain the light signal to be split and result in the LO and modulated signals. In other embodiments with the light source and modulator on different integrated circuits, a polarization controller may be used with the modulator to ensure that the light signal obtained and split results in LO and modulated signals that are separable in polarization. In additional or alternate embodiments, a parallel in-phase amplifier may be used with or instead of a polarization controller to separately amplify the signals.

[0076] Amplified optical switches, according to some embodiments and based on the self-homodyne architecture, can address all the major impairments of the optical fiber (e.g., attenuation, chromatic dispersion, four-wave mixing) between hops so that the transmitting and receiving components (e.g., transceivers) need not accommodate components (in terms ofspace or power) to address these impairments. According to some embodiments four-wave mixing may be diminished by aligning the LO and modulated signals in different polarizations. According to some embodiments chromatic dispersion may be addressed by a chromatic dispersion compensator included in the amplified optical switch (e.g., as part of the amplifier).

[0077] FIG. 1 is a block diagram of an optical amplifier 100 according to some embodiments. The exemplary optical amplifier 100 includes a polarization-multiplexed (demultiplexed) SOA-based amplifier (PMSA) 105. The PMSA 105 generally includes a polarization splitter-rotator (PSR) 110, polarization controller (PC) 120, two semiconductor optical amplifiers (SOAs) 130, and a polarization beam combiner (PBC) 140. The input signal may include a local oscillator (LO) signal and a modulated signal, which may have different polarizations and, more particularly, orthogonal polarizations relative to each other. The PSR 110 may split the orthogonal polarizations and rotate one of the polarizations to generate polarization components 115a and 115b. The PC 120 may separate (demultiplex) a first component signal 125a (e.g., LO signal) and second component signal 125b (e.g., modulated signal) for individual amplification by an SOA 130. The amplified signals 135a and 135b may be combined by the PBC 140 for output as the output signal. The components of the PMSA 105 are further discussed with reference to FIGS. 2 and 3.

[0078] FIG. 2 is a block diagram detailing aspects of an exemplary PMSA 105, used as an optical amplifier 100 according to some embodiments. The exemplary PMSA 105 may be implemented fully or partially as a photonic integrated circuit (PIC) or using individual components, for example using bulk optics or micro-optic components. As indicated in FIG. 2, a signal is transmitted over optical fiber and, following a routing layer, for example, an Input Signal is received at the input of the PMSA 105. The initial state of polarization (SOP), transmitted SOP 21, is shown as a point on the Poincare sphere. Due to fiber asymmetries and environmental factors, the transmitted SOP 21 of the signal undergoes time-varying changes, such that the instantaneous Input Signal has an arbitrary input signal SOP 22.

[0079] The PSR 110 splits the two orthogonal polarizations of the Input Signal and rotates one polarization such that both signals propagate along the PIC waveguides 205, in the exemplary case of a PIC implementation and in case the waveguides support only one polarization. A PC 120 transforms the input signal SOP 22 (with transformations 25, 26 indicated) such that the PC output SOP 23 lies on a circle 24 corresponding to the value of the first Stokes Parameter (S 1 = Ex2 - Ey2) such that the two polarizations (Ex and Ey) of the Transmitted Signal are demultiplexed into different output waveguides 215.

[0080] In the exemplary embodiment, a PC 120 with three degrees of freedom (DOF) is depicted. Additional DOF may be incorporated to enable reset-free operation for arbitrary input and output SOP. While several embodiments herein are illustrated with a three-DOF PC 120, additional DOF may be included as needed. The DOFs are a tunable coupler 27a, a phase shifter 28, and a second tunable coupler 27b (generally referred to as tunable coupler 27). The tunable couplers 27 are made up of 2x2 couplers 225 and electro-optic phase shifters 230 (denoted by ^), such as thermo-optic phase shifters, plasma-dispersion phase shifters, and phase shifters based on the Pockels or Kerr effects. For each of the three phase shifters 230 indicated in FIG. 2, it should be understood that there is a complementary phase shifter 230 on the other path, which is omitted for simplicity, operating in conjunction with the illustrated phase shifter 230 to achieve the polarization control. A small percentage of the signals at the output of the PC 120 are tapped optically and measured with low-frequency receivers 240 (tap & RX). The low frequency receiver 240 may include a reverse-biased photodiode and a transimpedance amplifier, which outputs a voltage proportional to the measured power. The outputs of the RXs may be converted to digital signals using analog to digital converters (ADCs). The digital signals are read by a controller (microcontroller, indicated by uC) 210, as shown in FIG. 2.

[0081] The Transmitted Signal may include one or multiple low-frequency tones which can be filtered within the controller 210 to obtain their amplitude. The controller 210 may utilize the measured power of the tone(s) as an error signal in closed-loop control. Alternatively, the total power of the signals measured by the low frequency receivers 240 of the PC 120 may be used as the error signals. The controller 210 outputs analog voltage signals using internal or external digital to analog converters. The controller 210 voltage outputs drive the PC phase shifters 230 directly or drive electrical amplifiers 220, which in turn drive the phase shifters 230. One or both of the demultiplexed signals (125a, 125b) are then separately amplified using SOAs 130. The demultiplexed signals may have different power levels, spectral characteristics, linearity requirements and / or noise considerations, and therefore it may be advantageous to have separate individually controlled SOAs 130 for each polarization. For example, the optical link may employ self-homodyne detection (PM-SH).

[0082] In a self-homodyne detection (PM-SH) link, a modulated signal and an unmodulated optical signal, also referred to as a local oscillator (LO), are polarization- multiplexed. The modulated signal may have lower power than the unmodulated signal, and therefore the SOA pattern effects and nonlinear distortions due to gain saturation may be minimized. Conversely, the LO may have higher power, which can saturate the SOA 130 atlower gain, but it can be more tolerant to nonlinear distortions and pattern effects because it maintains a constant power level.

[0083] FIG. 3 illustrates the exemplary PMSA 105 of FIG. 2 and indicates a spectrum corresponding to each polarization for an PM-SH signal as it propagates along the fiber, at the output of the PC 120, and after amplification by the SOAs 130. The spectrum for the Input Signal and for the Output Signal indicates the orthogonal polarization of the two components of the Input Signal and the Output Signal. After amplification by the SOAs 130, the two polarizations are combined using the PBC 140 to generate the Output Signal transmitted to a Receiver 2620 (FIG. 26), a switch, or another optical amplification stage. In a long-haul system, the exemplary amplifier of FIGS. 2 and 3 may be used as a range extender.

[0084] FIG. 4 illustrates an exemplary PMSA 105, used as an optical amplifier 100 according to some embodiments, that incorporates attenuators 410 (indicated as att) to compensate for polarization-dependent loss (PDL) at the input and output introduced by the PSR 110, PBC 140, or other components. The attenuators 410 may be passive, such as dummy crossings, or active, such as variable optical attenuators (VOAs). The VOAs may be controlled using closed-loop control implemented with the controller 210 by tapping off a portion of the signal before or after the PC 120 using a tap & RX (240) block. Tunable filters 420 (indicated as TF) may be used to filter out undesired wavelengths such as those generated due to nonlinearities of the SOAs 130.

[0085] FIG. 5 is a schematic of a PMSA 105, used as an optical amplifier 100 according to some embodiments, that incorporates a Parallel In-Phase Amplifier (PIA) 510 instead of a single SOA 130 for the second component signal 125b (e.g., modulated signal). The PIA 510 incorporates tunable couplers 27 and N SOAs 130 within nested Mach-Zehnder Interferometers (MZIs) 530. As indicated, the tunable couplers 27 and MZIs 530 include phase shifters 230. The incoming second component signal 125b is split N times, such that the input power at each SOA 130 of the PIA 510 is the total signal power of the second component signal 125b divided by N. This division of the input power enables increasing the total saturation input power by using the N SOAs 130 compared with using the single SOA 130. The tunable couplers are biased to precisely control the number of splits (N) experienced by the second component signal 125b, which in turn determines the number (N) of parallel SOAs 130 that are utilized. The MZI phase shifters 230, which may be implemented using thermo-optic phase shifters, are biased such that the amplified signals output from each of the N SOAs 130 constructively interfere to maximize the signal power at the output of the PIA 510 that is directed to the PBC 140. The control for both the biasing of the tunable couplers27 and the MZI phase shifters 230 are implemented by low frequency closed loop control loops employing the measured powers at the receivers (RX) 520 that are fed back to the controller 210. The PIA 510 can be used in one of the polarizations, as shown in the exemplary case of FIG. 5, or both polarizations, as shown in FIG. 6.

[0086] FIG. 6 is a block diagram of an exemplary optical amplifier 100 employing PIAs 510 to amplify each polarization component 115a, 115b according to some embodiments. Using a PIA 510 instead of a single SOA 130 may reduce pattern effects and nonlinear distortions by limiting the power input to each parallel SOA 130 that is part of the PIA 510. This can prevent the unwanted intermixing of two orthogonal polarizations, even in the absence of a PC 120. Consequently, PIAs 510 may offer the potential to bypass the need for a PC 120, as shown in FIG. 6. This creates a setup that is agnostic to the modulation format used in the optical link. That is, an intensity modulated signal on a single polarization, with or without an unmodulated signal separated in polarization (i.e., not necessarily selfhomodyne) may be amplified using a PMSA 105 with PIAs 510. This is further discussed with reference to FIGS. 46 and 49.

[0087] FIG. 7 is a block diagram of an exemplary optical amplifier 100 used with a polarization-maintaining (PM) fiber 710, according to some embodiments. In the exemplary embodiment of FIG. 7, the PMSA 105 may be used without a PC 120 despite having a single SOA 130 in each amplification path, rather than a PIA 510 as in FIG. 6. This is facilitated by using the PM fiber 710. The PM fiber 710 allows polarizations of the Transmitted Signal to be passively demultiplexed by the PSR 110.

[0088] FIGS. 8-19 illustrate extensions of the embodiments discussed with reference to FIGS. 1-7. FIGS. 8-13 illustrate bidirectional optical amplifiers 100 and FIGS. 14-19 illustrate multi-wavelength optical amplifiers 100 according to various embodiments. The embodiments illustrated in FIGS. 11-19 contemplate wavelength multiplexed signals in addition to polarization multiplexed signals.

[0089] FIG. 8 is a block diagram of a bidirectional optical amplifier 100 according to some embodiments. As illustrated, the bidirectional optical amplifier 100 of FIG. 8 includes two PMSAs 105a, 105b that are both similar to the one shown in FIG. 3. In the exemplary embodiment of FIG. 8, circulators 810a, 810b facilitate bidirectional operation such that an Input Signal from the left, according to the arrangement in FIG. 8, is directed to the PMSA 105a by the circulator 810a and an Input Signal from the right is directed to the PMSA 105b by the circulator 810b.

[0090] FIG. 9 is a block diagram of a bidirectional optical amplifier 100 according to some embodiments. Like the embodiment illustrated in FIG. 8, the embodiment illustrated in FIG. 9 includes circulators 810a, 810b to facilitate bidirectional operation. Unlike the embodiment of FIG. 8, the optical amplifier 100 in the embodiment of FIG. 9 includes an exemplary PMSA 105 for only one direction (right to left) of the Input Signal. When the Input Signal arrives at the circulator 810a, according to the exemplary arrangement shown in FIG. 9, it is directed to the PSR 110 and the two polarization components 115a, 115b output by the PSR 110 are respectively carried via two waveguides 910 to the PBC 140. Each of the waveguides 910 may only support one polarization, for example.

[0091] FIG. 10 is a block diagram of a bidirectional optical amplifier 100 according to some embodiments. Like the optical amplifiers 100 according to the embodiments illustrated in FIGS. 8 and 9, the optical amplifier 100 according to the embodiment illustrated in FIG. 10 includes circulators 810a, 810b to facilitate bidirectional operation. Like the embodiment of EIG. 9, the embodiment of EIG. 10 includes an exemplary PMSA 105 for only one direction (right to left) of the Input Signal. When the Input Signal arrives at the circulator 810a, according to the exemplary arrangement shown in PIG. 9, it is directed through a waveguide 1010 to the circulator 810b for output. The waveguide 1010 may support both polarizations, for example.

[0092] In both the PIG. 9 and PIG. 10 embodiments, amplification may only be needed in one direction of the signal. Although that amplification is shown for an Input Signal traveling from the circulator 810b to 810a, the illustration is not intended to limit similar optical amplifiers 100 that have the PMSA 105 arranged to amplify the Input Signal entering at circulator 810a.

[0093] FIGS. 11-13 illustrate bidirectional optical amplifiers 100 used with signals of different wavelengths in each direction. Although two wavelengths (XI and 2) are indicated in the figures for explanatory purposes, the exemplary embodiments may also be used with more than two wavelengths. For example, signals with two wavelengths (e.g., XI and 2) may be propagated in one direction and signals with two different wavelengths (e.g., X3 and 4) may be propagated in the opposite direction. Accordingly, a multiplexer / demultiplexer 1110a, 1110b may be used at each end of the optical amplifier 100 instead of the circulators 810a, 810b. In the examples illustrated in FIGS. 11-13, the Input Signal reaching the multiplexer / demultiplexer 1110a may have a wavelength component of XL In this direction, the multiplexer / demultiplexer 1110a acts as a demultiplexer separating signal with the wavelength component of I, which may be some or all of the Input Signal. At the other end,the multiplexer / demultiplexer 1110b acts as a multiplexer. On the other hand, the Input Signal reaching the multiplexer / demultiplexer 1110b may have a wavelength component of X2. In this direction, the multiplexer / demultiplexer 1110b acts as a demultiplexer separating signal with the wavelength component of X2, which may be some or all of the Input Signal. At the other end, the multiplexer / demultiplexer 1110a acts as a multiplexer.

[0094] FIG. 11 is a block diagram of a bidirectional optical amplifier 100 according to some embodiments. As illustrated, FIG. 11 includes two PMSAs 105a, 105b that are similar to the one shown in FIG. 3. When an Input Signal reaches the multiplexer / demultiplexer 1110a, a signal with a wavelength of I, which may represent a portion or all of the Input Signal, is directed to the PMSA 105a. As previously discussed, the PMSA 105a uses the PSR 110 to separate orthogonal polarizations and rotate one of the polarities. The output of the PBC 140 of the PMSA 105a is output via the multiplexer / demultiplexer 1110b as the Output Signal. Similarly, when an Input Signal reaches the multiplexer / demultiplexer 1110b, a signal with a wavelength of X2, which may represent a portion or all of the Input Signal, is directed to the PMSA 105b. The output of the PBC 140 of the PMSA 105b is output via the multiplexer / demultiplexer 1110a as the Output Signal.

[0095] FIG. 12 is a block diagram of a bidirectional optical amplifier 100 according to some embodiments. Like the embodiment illustrated in FIG. 9, only one direction of the bidirectional optical amplifier 100 of FIG.12 includes a PMSA 105 while the other direction includes waveguides 910. Like the embodiment illustrated in FIG. 11, each end includes a multiplexer / demultiplexer 1110a, 1110b rather than a circulator.

[0096] As a result of the exemplary arrangement in FIG. 12, when an Input Signal that reaches the multiplexer / demultiplexer 1110a, a signal with a wavelength of I, which may represent a portion or all of the Input Signal, is directed to the PSR 110 and each of the resulting polarization components 115a, 115b is directed to the PBC 140 via a respective waveguide 910. The output of the PBC 140 is then output via the multiplexer / demultiplexer 1110b as the Output Signal. In the opposite direction, when an Input Signal reaches the multiplexer / demultiplexer 1110b, a signal with a wavelength of X2, which may represent a portion or all of the Input Signal, is directed to the PMSA 105. The output of the PBC 140 of the PMSA 105 is output via the multiplexer / demultiplexer 1110a as the Output Signal.

[0097] FIG. 13 is a block diagram of a bidirectional optical amplifier 100 according to some embodiments. Like the embodiment illustrated in FIG. 10, only one direction of the bidirectional optical amplifier 100 of FIG.13 includes a PMSA 105 while the other directionincludes waveguide 1010. Like the embodiments illustrated in FIGS. 11 and 12, each end includes a multiplexer / demultiplexer 1110a, 1110b rather than a circulator.

[0098] As a result of the exemplary arrangement in FIG. 13, when an Input Signal that reaches the multiplexer / demultiplexer 1110a, a signal with a wavelength of I, which may represent a portion or all of the Input Signal, is directed to through the waveguide 1010 to the multiplexer / demultiplexer 1110b as the Output Signal. In the opposite direction, when an Input Signal reaches the multiplexer / demultiplexer 1110b, a signal with a wavelength of X2, which may represent a portion or all of the Input Signal, is directed to the PMSA 105. The output of the PBC 140 of the PMSA 105 is output via the multiplexer / demultiplexer 1110a as the Output Signal.

[0099] FIGS. 14-16 illustrate exemplary optical amplifiers 100 used with multiwavelength signals. That is, unlike the embodiments of FIGS. 11-13, which pertain to bidirectional optical amplifiers 100 for use with a signal of a particular wavelength (XI or X2) in each direction, the embodiments of FIGS. 14-16 pertain to unidirectional optical amplifiers 100 that may be used with an Input Signal that is multiplexed both in polarization and wavelength. In the exemplary embodiments, wavelengths of I, X2, X3, and X4 are used for explanatory purposes. However, the embodiments discussed with reference to FIGS. 14-16 may be used with multi- wavelength signals with two or more wavelengths.

[0100] FIG. 14 is a block diagram of an optical amplifier 100 according to some embodiments. The PSR 110 separates a multi-wavelength Input Signal into polarization components 115a and 115b. Each of the polarization components 115a, 115b is demultiplexed by a demultiplexer 1410. Each signal (of wavelength XI, X2, X3, or X4) output by the demultiplexer 1410 is attenuated by an attenuator 410 and provided to a PC 120, paired with a signal of the same wavelength and the opposite polarization. For explanatory purposes, dashed lines are used to indicate the inputs to one of the PCs 120 for the two polarizations associated with wavelength XL The first component signal 125a and the second component signal 125b output by each PC 120 are amplified by corresponding SOAs 130. The amplified signals are provided to one of the multiplexers 1420 based on their polarization. The two polarizations associated with wavelength XI are again indicated with dashed lines for explanatory purposes. The signals of the different wavelengths (XI, X2, X3, and X4) but the same polarization are multiplexed by one of the multiplexers 1420. Outputs of the two multiplexers 1420 are provided to the PBC 140 for output as the Output Signal.

[0101] FIG. 15 is a block diagram of an optical amplifier 100 according to some embodiments. While the exemplary embodiment of FIG. 14 separates the polarizationcomponents 115a, 115b using the PSR 110 prior to demultiplexing the wavelength ( I, X2, X3, and X4) for each of the polarization components 115a, 115b, the embodiment of FIG. 15 includes a single demultiplexer 1410 that demultiplexes the Input Signal. As a result, a different PSR 110 is used to separate each of the four resulting signals (each of wavelength I, X2, X3, or X4) by polarization. The polarization components 115a, 115b associated with wavelength XI are indicated with dashed lines. Each of the polarization components 115a, 115b, output by each of the PSRs 110 is attenuated by an attenuator 410 and processed by a PC 120. Attenuated signals of the same wavelength but opposite polarization are provided to the same PC 120, as highlighted, for example, by the dashed lines for the signals associated with wavelength XI.

[0102] The two outputs (corresponding to the two polarizations) of each of the PCs 120 are amplified by respective SOAs 130. The amplified signals are attenuated by attenuator 410 before being combined with an attenuated signal of the same wavelength but opposite polarization at a PBC 140. The two signals associated with wavelength XI (with different polarizations) are again indicated with dashed lines for explanatory purposes. The output of each PBC 140, which is at one of the wavelengths, is multiplexed with the outputs of the other PBCs 140, which are at other wavelengths, by multiplexer 1420 to provide the Output Signal.

[0103] FIG. 16 is a block diagram of an optical amplifier 100 according to some embodiments. Like the embodiment shown in FIG. 14, the embodiment in FIG. 16 includes a PSR 110 that first separates the Input Signal by polarization. Demultiplexers 1410 then separate each of the polarization components 115a, 115b based on wavelength (XI, X2, X3, or X4). In the embodiment of FIG. 16, the outputs of the demultiplexers 1410 are matched according to wavelength and provided in pairs to a PC 120. The output of each demultiplexer 1410 at wavelength XI is indicated with a dashed line for explanatory purposes. At the output of each PC 120, each of the signals is provided to a different multiplexer 1420 based on polarization. For example, all the outputs of all the PCs 120 with a first polarization are provided to one of the multiplexers 1420 while all of outputs of all the PCs 120 with a second polarization are provided to the other of the multiplexers 1420, as shown. The outputs of the PC 120 associated with wavelength XI are indicated with dashed lines for explanatory purposes. The output of each of the multiplexers 1420 is amplified with an SOA 130. The amplified multiplexed signals associated with each polarization are then combined by the PBC 140 as the Output Signal.

[0104] FIG. 17 is a block diagram of an optical amplifier 100 according to some embodiments. FIG. 17 includes gain clamping lasers (X5 lasers) 1710 to improve linearity of the SOAs 130. While the gain clamping lasers could be of the same wavelength as one of the components of the Input Signal, the exemplary gain clamping lasers (X5 lasers) 1710 have a wavelength X5 different from the wavelengths (XI, X2, X3, and X4) of the Input Signal. The power of the X5 lasers 1710 may be fixed or, as illustrated, may be controlled by the controller 210. The controller 210 may set the power of each X5 laser 1710 using a lookup table or using closed loop control based on monitoring the power of the signals prior to amplification by the SOAs 130. According to the embodiment shown in FIG. 17, each of the X5 lasers 1710 is coupled into an SOA 130. The outputs of the SOAs 130 are provided to one of two multiplexers 1420 and the multiplexer 1420 outputs are combined by the PBC 140, as discussed with reference to FIG. 14.

[0105] FIG. 18 is a block diagram of an optical amplifier 100 according to some embodiments. FIG. 18 shows a variation of the embodiment of FIG. 17. The outputs of the PCs 120 pertaining to the second polarization are handled in the same way as the outputs of the PCs 120 in FIG. 14, for example. That is, the outputs of the PCs 120 pertaining to the second polarization are amplified by SOAs 130, as indicated by the dashed line for the output pertaining to wavelength I. The amplified signals are then multiplexed by one of the multiplexers 1420.

[0106] However, the outputs of the PCs 120 pertaining to the first polarization are not amplified until they are first multiplexed by the other of the multiplexers 1420. This is indicated for example, for the output of the PC 120 pertaining to the first polarization and wavelength XI by the bold line. The output of the multiplexer 1420 is then amplified by an SOA 130 that is coupled with the output of a clamping laser, X5 laser 1710. According to alternate embodiments, the outputs of the PCs 120 pertaining to the second polarization may also be multiplexed prior to amplification with a single SOA 130, which may also be coupled with the output of a clamping laser.

[0107] FIG. 19 illustrates an exemplary frequency spectrum for a multi-wavelength Input Signal designed to mitigate four-wave mixing (FWM) effects according to some embodiments. That is, the LO signal and the modulated signal in the Input Signal each have components at different wavelengths, each of the LO signal components differ in polarization from each of the modulated signal components, and four-wave mixing is mitigated by interleaving the LO and modulated signal components, at the different polarizations, across the different wavelengths that are multiplexed. In the example illustrated at FIG. 19, LOsignal components have polarizations of Y, X, Y, Y at wavelengths 1330 nanometers (nm), 1310 nm, 1290 nm, and 1270 nm, respectively. Modulated signal components have polarizations of X, Y, X, X at wavelengths 1330 nm, 1310 nm, 1290 nm, and 1270 nm, respectively. That is, at each wavelength, the polarization of the LO component differs from that of the modulated signal component. The different polarizations are interleaved before the different wavelengths are multiplexed to generate the Input Signal. As a result, the Input Signal is split according to polarization (by the PSR 110) and demultiplexed (by demultiplexers 1410) prior to amplification, as shown in FIG. 18.

[0108] An exemplary configuration may employ self-homodyne coherent detection (PM-SH-C) for all wavelengths. FWM is a polarization- sensitive phenomenon that generates intermodulation products from optical signals aligned to the same polarization in an optical fiber, with its severity heightened by low dispersion. According to some embodiments, the LO and modulated signals are aligned in specific polarizations to diminish the radio frequency (RF) power of FWM products detected at the Receiver 2620. In the example shown in FIG. 19, the polarization states of wavelengths near the low-dispersion zone, specifically at 1310 nm, are arranged in an interleaved manner to reduce the RF power of FWM products at the Receiver 2620. The exemplary PMSA 105 according to some embodiments enables individual configuration of the polarization alignment for each transmitted wavelength directly within the optical link, thereby substantially enhancing the reach and performance of optical links.

[0109] FIG. 20 is a block diagram of an exemplary optical amplifier 100 according to some embodiments. The embodiment of FIG. 20 may be a special case of an optical amplifier 100 that immediately precedes a Receiver 2620 of a routed signal. The Input Signal may be a self-homodyne intensity modulated (PM-SH-IM) signal, which is a single quadrature signal. Unlike the embodiment of FIG. 14, for example, in the embodiment of FIG. 20, after amplification by the SOAs 130, the two signals output by each of the PCs 120 may be aligned in phase at the output of a 2x2 coupler 2010 and then directed to multiplexer 1420.

[0110] There is only one output polarization at the output of the multiplexer. Thus, as illustrated in FIG. 20, a PBC 140 is not required. The exemplary PCs 120 required by the embodiment are further discussed with reference to FIG. 21 and have four DOFs since the relative phase at the output of each PC 120 needs to be such that the signals interfere constructively before the multiplexer 1420. The phase is optimized by using a control loop in which the controller 210 uses the other outputs of the 2x2 couplers 2010, obtained by thereceivers 2020. Specifically, the measured power is used as the error signal in closed loop control. For example, for a PM-SH-IM signal, the phase is set such that the average RF signal power at the receiver 2020 is minimized. By using the exemplary embodiment of the PMSA 105 shown in FIG. 20 at the last hop before a Receiver 2620 of the routed signal, the Receiver 2620 of the routed signal need not include a PC 120, unlike the receivers 2620 discussed with reference to FIGS. 50 and 51, for example.

[0111] In terms of the amplification, because the system is still a self-homodyne system, the PMSA architecture shown in FIG. 20 can separate and independently amplify the LO and intensity modulated signals, separable according to polarization, to avoid the deleterious effects of prior optical amplifiers, while enabling the use of a traditional intensity- modulated direct-detection (IMDD) receiver (i.e., a receiver without a PC 120) as the Receiver 2620 of the routed signal. The modulated signal is intensity modulated, and therefore its phase is not time-varying. Thus, the LO and modulated signal phases can be aligned at the output of the 2x2 coupler 2010 following amplification by the SOAs 130 and before the wavelength multiplexing by the multiplexer 1420. A traditional IMDD receiver, which is typically polarization independent, does not need to do any phase alignment and can convert the optical signal to an electrical signal by directing the full signal (both received polarizations) to a photodiode.

[0112] FIG. 21 is a block diagram of aspects of an exemplary PC 120 with four DOFs. While a phase shifter 230 is shown only on one of the two paths at each of the four stages, it should be understood that there is a complementary phase shifter in the opposite direction on each of the other paths. As previously noted with reference to FIG. 2, this applies to each of the PCs 120, which are known devices.

[0113] FIGS. 22-24 pertain to incorporating chromatic dispersion detection and compensation in optical amplifiers 100 according to some embodiments. FIG. 22 is a block diagram of an exemplary optical amplifier 100 that includes a tunable chromatic dispersion compensator (CDC) 2210 according to some embodiments. The hybrid 2220 may separate quadrature signals (I and Q). The receiver 2230 may detect low frequency tones such as those used for signal and LO phase locking (carrier recovery) and / or high frequency tones such as an additional tone added to one of the quadrature. The additional tone may be in- band or out of band. Based on the CDC 2210, the optical amplifier 100 may address the chromatic dispersion impairment of the optical fiber bringing the Input Signal. Generally, a high frequency tone may be added to the LO or signal at a transmitter of the Input Signal and may be filtered out at the Receiver 2620 using a low-pass filter. At an optical amplifier 100implementing chromatic dispersion compensation in the PMSA 105, as in the exemplary embodiment of FIG. 22, the high frequency tone may be used to detect chromatic dispersion, as detailed with reference to FIG. 23, and the detected chromatic dispersion may be addressed with a known compensator.

[0114] FIG. 23 is a block diagram of aspects of an exemplary chromatic dispersion (CD) detector 2300 that may be included in a PMSA 105 according to some embodiments. A hybrid 2200 may be used to separate the two quadrature signals (I and Q). The high- frequency tone may be down-converted by mixing with an equal frequency tone 2310, as shown. The ratio of the powers of the low-frequency and high-frequency tones in the two different quadratures may then be used to control the CDC 2210 in an optical amplifier 100 such as in FIG. 22. That is, if the high frequency tone was added to Q, as in the exemplary case, quantifying the portion at I provides an error signal corresponding with chromatic dispersion. The power ratio for a known transmission wavelength may also be used to calculate the optical fiber transmission length. The optical fiber transmission length may, in turn, be used to compensate for chromatic dispersion in other wavelength channels. Thus, in a multi-wavelength system, the CD detector 2300 need only operate on one of the wavelengths, and the chromatic dispersion of the other wavelengths may be compensated in open loop, for example by using a lookup table of CD compensation settings as a function of fiber length.

[0115] FIG. 24 is a block diagram showing aspects of a transmitter 200 according to some embodiments. The transmitter 200 incorporates an additional tone (high-frequency tone 2310) at one of the quadratures (at I in the exemplary case of FIG. 24) for detecting chromatic dispersion at the PMSA 105 (of FIG. 22, for example) or at a Receiver 2620. The high-frequency tone 2310 may be in-band or out of band (e.g., 100 gigahertz for a 50 gigabaud signal). A tunable coupler 27, indicated in the dashed box, controls the power split to the modulated and unmodulated paths. The additional tone incorporated by the transmitter 200 may also be incorporated into a Transmitter 2610 (FIG. 26) discussed with reference to various figures.

[0116] FIGS. 25-35 pertain to aspects of an optical amplifier 100 and an all-optical network 10 (FIG. 26) based on a PMSA 105 according to some embodiments such as the exemplary embodiments discussed with reference to FIGS. 1-24. With a PMSA 105 or another optical amplifier 100 between each hop, an all-optical network 10 without the need for electronic packet switches that require optical-electrical-optical (OEO) conversion and also without the deleterious effects of amplifiers in prior approaches is possible. An opticalamplifier 100 (e.g., PMSA 105) may address fiber impairments of attenuation, chromatic dispersion, and four-wave mixing. Thus, a Receiver 2620 that ultimately receives the routed signals may not require additional components to address signal effects introduced by the impairments of the cable during routing.

[0117] FIG. 25 is a block diagram of an exemplary amplified optical switch 2500 with an 8 x 8 switch 2510 and 8 Amplifiers, which may be PMSAs 105 according to exemplary embodiments discussed herein. The amplified optical switch 2500 is not limited to 8 input ports and 8 output ports and is more generally regarded as an N x N amplified optical switch 2500. Optical switches 2500 may facilitate a higher number of input and output ports, as compared with electronic packet switches. This would result in a need for fewer total switches for an application, such as in a data center.

[0118] FIG. 26. illustrates exemplary optical links within a network 10 that integrate optical amplifier 100. These configurations facilitate data transmission over specified distances between Transmitters 2610 and Receivers 2620 via optical fiber. The links may incorporate amplified optical switches 2500, according to the exemplary embodiments discussed herein, at intermediate junctures. Given that optical switches contribute to insertion loss by attenuating the signal, the use of optical amplifiers 100 may ensure sufficient signal power reaches the receiver 2620 (FIG. 26), particularly in configurations with multiple optical switches 2510 in a single link. The optical amplifiers 100 can be separate components added after the optical switches 2510 or can be incorporated as amplified optical switches 2500. The descriptions of aspects of networks 10 pertain to both configurations.

[0119] FIG. 27 is a block diagram of an all-optical network 10 connecting N nodes 2710 using an NxN amplified optical switch 2500 according to some embodiments. The nodes 2710 may represent compute resources such as graphics processing units (GPUs), tensor processing units (TPUs), or other switches, such as top-of-rack or end-of-row switches. Each of the optical amplifiers 100 (e.g., PMSAs 105) after the switch 2510 is used to increase the signal power in the optical links connecting any two nodes 2710. The optical switch 2510 internally routes optical signals and can connect any input (left side) to any output (right side). While 8x8 ports are shown as an example, an optical switch 2510 according to some embodiments can have NxN ports.

[0120] FIG. 28 shows aspects of an exemplary network 10 using the exemplary amplified optical switch 2500 shown in FIG. 27, with four nodes 2710 connected, and the other half of the inputs and outputs of the switch being used for uplinks and downlinks. This configuration may represent a local segment of a larger network 10 and is referred to as acluster 2800. Uplinks may be the connections that carry data from a lower-tier switch, such as an access switch, to a higher-tier switch, such as an aggregation or core switch, in the datacenter hierarchy. Downlinks may refer to the connections that carry data from a higher- tier switch to a lower-tier switch. It should be understood that ports utilized for node connections and uplinks and downlinks is not limited to this particular example. For example, one quarter of the inputs and outputs can be utilized for node connections. With the amplified optical switch 2500 as a building block for a cluster 2800, any number of variations and configurations are possible for a network 10, as shown by additional exemplary configurations in FIGs. 29-35.

[0121] FIG. 29 illustrates aspects of an exemplary network 10 that includes five clusters 2800 connected to each other. This shows an example of a network of N / 2 + 1 clusters 2800 connected to each other, each consisting of an NxN amplified optical switches 2500, each of which connects N / 2 nodes 2710, for a total of N2 / 4 + N / 2 nodes 2710. In an example of a 128x128 port switch 2510, 65 clusters 2800 can be interconnected, each consisting of 64 nodes 2710, for a total of 4160 nodes 2710. FIG. 30 demonstrates how connections to other clusters 2800 can instead be used as uplinks and downlinks in order to form superclusters 3000. FIG. 31 illustrates an exemplary network 10 comprising super clusters 300 interconnected to hierarchically scale this network topology according to some embodiments.

[0122] FIG. 32 shows an exemplary cluster 2800 in which the nodes 2710 are not interconnected using a local switch, as illustrated in FIG. 27, for example. Instead, all the optical switch outputs are used as uplinks. FIG. 33 shows an exemplary amplified optical switch 2500, which consists of uplinks and downlinks. FIG. 34 shows aspects of an exemplary network 10 with N clusters 2800 connected to N clusters 2800 using a layer of N NXN amplified optical switches 2500, like the exemplary amplified optical switch 2500 in FIG. 33, to produce a total of N2xN2ports. FIG. 35 shows aspects of an exemplary network 10 according to some embodiments. By cascading amplified optical switches 2500 as shown in FIG. 35, it is possible to create an all-optical N2xN2switch. This can be scaled hierarchically to produce an NMxNMswitch, where M corresponds to the number of optical switch hops divided by three.

[0123] FIGS. 36-39 illustrate exemplary amplified optical switches 2500 according to some embodiments. The exemplary amplified optical switches 2500 in FIGS. 36-39 are for use with multi-wavelength signals and are extensions of embodiments discussed, for example, with reference to FIG. 14. Each of the exemplary amplified optical switches 2500is illustrated as a 4x4 port switch, with each port supporting four multiplexed wavelengths or wavelength channels. More generally, amplified optical switches 2500 may be NXN port switches supporting any number of multiplexed wavelengths. The routing layer 3610 indicated in each of FIGS. 36-39 may be implemented using multiple methods, including MZI switches, arrayed waveguide gratings (AWGs), ring resonators, micro -electromechanic al systems (MEMS), piezoelectric actuators, and liquid crystals. After the routing layer 3610, the signals are amplified using previously described methods for the PMSA 105.

[0124] FIG. 36 is a wavelength selective embodiment similar to the one shown in FIG. 14. A PSR 110 separates the polarizations of an incoming signal at each of the input ports. After wavelength demultiplexing by the demultiplexer 1410, both polarizations of each received wavelength are routed to the PC 120 corresponding to a desired output port by the routing layer 3610. After demultiplexing the polarizations, each polarization is amplified separately using an SOA 130. The wavelengths are then multiplexed by multiplexer 1420, and the output polarizations are combined using a PBC 140.

[0125] FIG. 37 shows a modification of the embodiment shown in FIG. 36. Specifically, a passive wavelength selective routing layer 3610 is shown, whereby routing from one port to another is done by transmitting over the designated wavelength. FIG. 38 shows a broadband switch in which the received polarizations are routed in tandem to the desired output port. That is, the routing layer 3610 precedes the demultiplexing such that multi-wavelength signals are routed by the routing layer 3610. FIG. 39 depicts a broadband switch in which the routing layer is polarization-diverse. That is, the routing layer 3610 precedes the PSR 110, as well as the demultiplexers 1410, such that muti-wavelength signals of both polarizations are routed by the routing layer 3610.

[0126] FIGS. 40-49 illustrate exemplary embodiments of Transmitters 2610 that may route signals in a network 10 employing amplified optical switches 100 according to embodiments like those discussed herein. According to some embodiments, optical transmitters 2610 may incorporate a tunable coupler 27, a polarization maintaining fiber 710, or a PC 120. PMSAs 105 are compatible with various modulation formats, including intensity-modulated direct detection (IMDD) and coherent modulation. PMSAs 105, according to some embodiments, may be used with self-homodyne detection formats, such as polarization-multiplexed self-homodyne coherent detection (PM-SH-C) and polarization- multiplexed self-homodyne intensity modulation (PM-SH-IM), which enable the separate amplification of the modulated and unmodulated optical fields that make up an input signal. Exemplary self-homodyne coherent detection systems are described, for example, in U.S.Patent No. 11,811,499. In a PM-SH-IM transmitter, a laser output is split into two signals of unequal or equal power, based on the tunable coupler 27, and one of the signals is then intensity modulated using a Mach-Zehnder modulator (MZM), an electro-absorption modulation (EAM), a ring resonator modulator (RRM), or another type of electro-optic modulator. The modulated and unmodulated signals are then transmitted over parallel channels, for example on orthogonal polarizations.

[0127] FIG. 40 is a block diagram of an exemplary embodiment of a single wavelength PM-SH-IM transmitter 2610 in which an intensity modulator 4010, such as an MZM or an EAM is utilized. Output from a laser 4020 is split using a tunable coupler 27, which controls the polarization state of the two resulting optical signals to be different. One of the optical signals resulting from the tunable coupler 27 is modulated with the intensity modulator 4010 before the modulated and unmodulated signals of different polarizations are combined by the PBC 140 and output as the Transmitted Signal. The tunable coupler 27 may set the portion of the laser power that is directed to the intensity modulator 4010 and thereby control the power of the modulated signal, with the remaining portion of the laser power being the (unmodulated) local oscillator. FIG. 41 is a block diagram of an exemplary PM- SH-IM transmitter 2610 that incorporates a multi-wavelength laser 4020 and ring-resonator modulators 4110 as the intensity modulator 4010.

[0128] FIG. 42 is an exemplary multi-channel transmitter 2610, where each channel n (indicated by An) includes one of the multiple wavelengths. In the illustrated example, n = 4. In this architecture, each laser 4020 is associated with one of the n wavelengths and is split using a tunable coupler 27. One of the signals associated with each channel (one of the polarizations) is modulated before wavelength multiplexers 1420 each combine the wavelengths associated with one of the polarizations. FIG. 43 is an exemplary multi-channel transmitter 2610 in which the modulation is performed for the multi-wavelength signal of one of the polarizations following the multiplexers 1420.

[0129] FIG. 44 is an exemplary multi-channel transmitter 2610 according to some embodiments in which the lasers 4020 are external to a photonic integrated circuit (PIC) 4400 that includes the intensity modulator 4010 and provides the Transmitted Signal. FIG. 44 illustrates a polarization-multiplexed self-homodyne system with intensity modulation of the modulated signal for intensity-modulated direct detection (IMDD). A PM fiber 710 may be used at the input of the PIC 4400 so that the polarizations are maintained for the multiwavelength signals of different polarizations received at the PSR 110 of the PIC 440.

[0130] FIG. 45 is an exemplary multi-channel transmitter 2610 according to some embodiments. Like the transmitter 2610 of FIG. 44, the transmitter 2610 shown in FIG. 45 is a PM-SH-IM transmitter with the lasers 4020 external to the PIC 4400. A PM fiber 710 is not shown between the light source (e.g., an integrated circuit including the lasers 4020) and the PIC 4400. Thus, polarizations in the generated light may be rotated before they reach the PSR 110 of the PIC 4400. Thus, unlike the embodiment of FIG. 44, which does use a PM fiber 710, the embodiment of FIG. 45 includes PCs 120 following the demodulators 1410.

[0131] FIG. 46 is an exemplary multi-channel transmitter 2610 according to some embodiments. Unlike the embodiments discussed with reference to FIGS. 44 and 45, the transmitter 2610 of FIG. 46 does not transmit the unmodulated (LO) polarization as part of the Transmitted Signal. As shown in FIG. 46, the second output from each of the PCs 120 is not provided to a second multiplexer 1420. Thus, the resulting Transmitted Signal may be amplified with a PMSA incorporating a parallel in-phase amplifier (PIA 510), such as the one shown in FIG. 5, for example.

[0132] FIG. 47 is a block diagram of an exemplary PM-SH-IM transmitter 2610 similar to the one shown in FIG. 44. Unlike the embodiment shown in FIG. 44, the embodiment shown in FIG. 47 does not include multiplexers 1420 at the outputs of the tunable couplers 27. Thus, instead of one PM fiber 710, the embodiment of FIG. 47 includes a separate PM fiber 710 for each of the two signals resulting from each of the four lasers 4020.

[0133] FIG. 48 is a block diagram of an exemplary PM-SH-IM transmitter 2610 similar to the one shown in FIG. 45. However, like the embodiment of FIG. 47, the embodiment of FIG. 48 lacks multiplexer(s) 1420 associated with the lasers 4020. As a result, each of the light signals output by each of the lasers 4020 is provided over a singlemode fiber, in the embodiment of FIG. 48, to the PIC 4400.

[0134] FIG. 49 is a block diagram of an exemplary intensity-modulated direct- detection (IMDD) transmitter 2610 using external light sources (e.g., lasers 4020 external to the PIC 4400). As discussed with reference to FIG. 46, the Transmitted Signal output by the transmitter 2610 shown in FIG. 49 may be amplified with a PMSA incorporating a PIA 510 on at least one path, such as the one shown in FIG. 5.

[0135] FIG. 50 is an exemplary PM-SH-IM receiver 2620 according to some embodiments with ring resonators 5010 used to demultiplex the received wavelengths. After demultiplexing, reset-free PCs 120, like the one described with reference to FIGS. 20 and 21, may be used for each received wavelength to combine the LO and modulated signals in phaseprior to a receiver 5020 including a single or differential photodiodes and a transimpedance amplifier.

[0136] FIG. 51 illustrates an exemplary PM-SH-IM receiver 2620 according to some embodiments. According to the exemplary multi-channel embodiment, each An corresponds to an individual channel n. After the PSR 110 and demultiplexers 1410, PCs 120 are used to combine the two received polarizations in phase. Subsequently, ring resonators 5010 are used to detect each individual received wavelength. In both the embodiments shown in FIGS. 50 and 51, if a PM fiber 710 is used, and there is no polarization rotation along the optical link, PCs 120 with three DOFs, such as the one described with reference to FIG. 2, can be used. 1

Claims

CLAIMS1. An optical communication network, comprising: an optical amplifier configured to amplify an input signal and provide an amplified signal, the optical amplifier comprising: a polarization splitter-rotator configured to split the input signal into a first polarization path and a second polarization path through the optical amplifier; two or more components disposed in the first polarization path, wherein the two or more components include a polarization controller and amplifier or a parallel in- phase amplifier with two or more amplifiers; and a polarization beam combiner configured to combine signals at an output of the first polarization path and the second polarization path as the amplified signal.

2. The optical communication network of claim 1, wherein the two or more components include the polarization controller and the amplifier.

3. The optical communication network of claim 2, wherein the amplifier is a semiconductor-based optical amplifier.

4. The optical communication network of claim 2, wherein the amplifier is a parallel in-phase amplifier.

5. The optical communication network of claim 4, wherein the parallel in-phase amplifier includes two or more semiconductor-based optical amplifiers.

6. The optical communication network of claim 1, further comprising an optical switch with a plurality of ports.

7. The optical communication network of claim 6, wherein the optical communication network includes a plurality of the optical amplifiers and each of the plurality of ports of the optical switch is connected to one of the plurality of the optical amplifiers.

8. The optical communication network of claim 6, further comprising a transmitter configured to transmit a self-homodyne signal to the optical switch, the self-homodyne signal including an unmodulated portion and a modulated portion.

9. The optical communication network of claim 1, further comprising a routing layer coupled with a plurality of the optical amplifiers to form an amplified switch.

10. The optical communication network of claim 1, further comprising a switch with N input ports and N output ports, wherein one or more of the N output ports is respectively connected to the optical amplifier.

11. The optical communication network of claim 1, further comprising a chromatic dispersion detector configured to detect chromatic dispersion in the input signal and facilitate chromatic dispersion compensation, wherein the chromatic dispersion is detected based on a tone added to the input signal.

12. The optical communication network of claim 1, wherein the input signal includes multiple wavelengths, multiple local oscillator portions of the input signal are at the multiple wavelengths and have multiple first polarization orientations, respectively, multiple modulated signal portions of the input signal are at the multiple wavelengths and have multiple second polarization orientations, respectively, and the first polarization orientations and the second polarization orientations are interleaved across the multiple wavelengths that are multiplexed into the input signal, thereby mitigating four-wave mixing as the input signal propagates along a fiber.

13. The optical communication network of claim 12, wherein, at each of the multiple wavelengths, a respective one of the first polarization orientations is different than a respective one of the second polarization orientations.

14. A transmitter in an optical communication network comprising: a light source configured to output a light signal; a modulator configured to modulate a first portion of the light signal to output a modulated signal; anda polarization beam combiner configured to combine the modulated signal and a second portion of the light signal to output a transmission signal for transmission through optical fiber.

15. The transmitter according to claim 14, further comprising a tunable coupler configured to generate and control respective powers of the first portion of the light signal and the second portion of the light signal.

16. The transmitter according to claim 14, wherein the light source is external to a photonic integrated circuit including the modulator and the light signal is conveyed from the light source to the photonic integrated circuit in an optical fiber.

17. The transmitter according to claim 16, wherein the optical fiber is a polarization maintaining optical fiber.

18. The transmitter according to claim 16, further comprising a polarization controller on the photonic integrated circuit configured to separate the first portion of the light signal and the second portion of the light signal according to polarization.

19. The transmitter according to claim 14, further comprising a phase modulator configured to add a tone to the modulated signal or the second portion of the light signal to facilitate chromatic dispersion detection.

Images