JP6045723B2 - Level spacing for M-PAM optical systems using coherent detection - Google Patents

Description

  This application is a U.S. patent to US provisional application 61 / 769,078 filed February 25, 2013 and US provisional application 61 / 772,664 filed March 5, 2013. Claims the benefit of priority under Article 119 (e) of the Act, the contents of both are incorporated by reference in their entirety. This application is a continuation-in-part of US Patent Application No. 13 / 929,757, filed June 27, 2013, which is incorporated herein by reference in its entirety.

  The present disclosure relates generally to the field of optical communications.

  This section introduces aspects that can help facilitate a better understanding of the present invention. Accordingly, the statements in this section should be read in this light and should not be understood as admissions about what is in the prior art or what is not in the prior art.

  The rapid growth of Internet and cloud computing applications is driving an increasingly important demand for communication network capacity. Commercialization and deployment of 100 Gb / s technology using polarization division multiplexing quadrature phase shift keying (PDM-QPSK) and digital coherent detection in optical transmission networks, and higher bit rates such as 400 Gb / s and 1 Tb / s technologies There is also an urgent need to upgrade metro networks from 10 Gb / s to over 100 Gb / s in the near future. Digital coherent detection is one way to achieve high spectral efficiency and networking flexibility. However, compared to optical transmission networks, metro networks are more sensitive to cost, footprint, and power consumption. There is a significant need for a low cost optical transmission system that meets the future requirements of metro networks.

A. S. Karar, J .; C. Cartridge, et al., “Generation and Detection of a 112 Gb / s Dual Polarization Signal Using a Directly Modulated Laser and Half-Cycle16-QAM Nyquir-T20 3. A. 4, 2012 C. Xie, et al., “Colorless coherent receiving using 3 × 3 coupler hybrids and single-ended detection”, Optics Express, Vol. 20, no. 2, pp. 1164-1171, 2012

  One embodiment provides an apparatus that includes an optical transmitter configured to provide an optical signal that is amplitude modulated between M different levels. The constellation control module is configured to provide a drive signal to control the optical signal. The feedback module is configured to receive an amount of spacing between amplitude peaks of the signal constellation of the optical signal and adjust the optical signal in response to the amount of symbol spacing.

  Another embodiment provides a method of forming an optical transmitter, for example. The method includes configuring an optical transmitter that provides an optical signal that is amplitude modulated between M different levels. The method further includes configuring a constellation control module that controls the drive signal to control the optical signal. The method further includes configuring a feedback module that receives an amount of spacing between amplitude peaks of the symbol constellation of the optical signal. The feedback module is further configured to adjust the optical signal according to the amount of spacing.

  In either embodiment, the constellation control module and the feedback controller may be configured to substantially equalize the spacing between symbols of the amplitude shift keyed symbol constellation. In any embodiment, the optical transmitter may include a vertical cavity surface emitting laser (VCSEL) configured to generate an optical signal in response to the drive signal. In either embodiment, the drive signal can indicate amplitude, bias level, and amplitude peak spacing. In any embodiment, the feedback module may be configured to provide an amplitude peak spacing adjustment signal. In any embodiment, the optical transmitter can include an electroabsorption modulator configured to modulate light from the laser to M different levels in response to the drive signal. In any embodiment, the optical transmitter may include a Mach-Zehnder modulator (MZM) configured to modulate light from the laser to M different levels in response to the drive signal.

  Any embodiment may further include a coherent optical receiver. The receiver may be configured to determine the spacing between the amplitude peaks of the symbol constellation, and from there may be further configured to generate an amount of symbol spacing. In either embodiment, the two lasers may be configured to provide a polarization multiplexed signal for the first and second different polarizations of the optical signal.

  Another embodiment provides a second device, eg, an optical receiver. The apparatus includes an optical detector and a constellation characterization module. The optical detector may be configured to demodulate the received optical signal and generate a received symbol constellation therefrom. The constellation characterization module may be configured to determine spacing between amplitude peaks of the received symbol constellation and provide an amount of symbol spacing.

  The following embodiments are described in connection with the second device. In either embodiment, the optical receiver can include a local oscillator and an optical hybrid configured to determine an in-phase component and a quadrature component of the optical signal. In either embodiment, the optical receiver may include an optical 120 ° hybrid configured to determine an in-phase component and a quadrature component of the optical signal. In either embodiment, each symbol in the symbol constellation may be represented by a closed curve in the in-phase / quadrature (I / Q) plane. In such embodiments, the amplitude detector can determine the spacing between closed curves.

  A more complete understanding of the present invention can be obtained by reference to the following detailed description when read in conjunction with the accompanying drawings.

1 illustrates a prior art system that uses multiple transmitters, demultiplexers, and multiple receivers multiplexed on a single optical fiber. FIG. FIG. 2 shows a prior art system that includes a transmitter that combines the outputs of two directly modulated lasers using a polarization beam combiner (PBC) and a receiver that employs direct detection of the transmitted signal. An optical communication system 1) without using a transmitter including a laser that is directly modulated with an M-level electrical signal, eg, 4 levels, to generate an M-ASK optical signal, and 2) carrier and phase recovery FIG. 2 is a diagram illustrating an embodiment of the present disclosure including a receiver that employs an optical hybrid that demodulates an optical signal to recover data. FIG. 4 illustrates an alternative embodiment of the receiver of FIG. 3 where an optical 120 ° hybrid, such as a 3 × 3 optical coupler, provides some functionality of the optical hybrid of FIG. FIG. 5 illustrates an alternative embodiment of the transmitter of FIG. 3 in which a Mach-Zehnder modulator is employed to generate the transmitted optical signal of FIGS. 3 and 4. FIG. 5 illustrates an alternative embodiment of the transmitter of FIG. 3 in which an electroabsorption modulator is employed to generate the transmitted optical signal of FIGS. 3 and 4. FIG. 4 illustrates aspects of a received constellation, and an embodiment in which the constellation includes a continuous closed curve. FIG. 6 illustrates an embodiment of a reception constellation, wherein the constellation includes an open arc. FIG. 6 illustrates an embodiment of a reception constellation, wherein the constellation includes an open arc. FIG. 13 illustrates an experimental configuration including a shaping filter between a transmitter and a receiver that can be used to determine the characteristics presented in FIGS. 9A / B 10, 11, and 12. FIG. 9 shows the I / Q characteristics of a signal transmitted by the embodiment of FIG. 8 without a shaping filter. FIG. 9 is a diagram illustrating I / Q characteristics of a signal transmitted by the embodiment of FIG. 8 with a shaping filter. FIG. 4 is a diagram illustrating peak intensity characteristics of signals transmitted by the embodiment of FIG. 3 with and without a shaping filter, and an example response of the shaping filter. FIG. 9 illustrates the experimental performance of the embodiment of FIG. 8 for back-to-back operation (eg, negligible optical path length between transmitter and receiver). FIG. 9 illustrates the bit error rate (BER) characteristics of the embodiment of FIG. 8 as a function of output power for three span lengths, such as 320 km, 640 km, and 960 km. 7B illustrates an embodiment of a PAM transmission system configured to approximately equalize spacing between constellation symbols, such as the symbol ring of FIG. 7A. FIG. FIG. 14 illustrates an embodiment of a method for operating the system of FIG. 13 for equal symbol spacing. Experimental performance of a test system configured to implement the embodiment shown in FIG. 13 in a 4-PAM transmission system in back-to-back operation (eg, negligible optical path length between transmitter and receiver). FIG.

  The present disclosure is directed to, for example, an apparatus, system, and method for transmitting an optical communication signal.

  The embodiments presented herein describe cost effective solutions for improved optical transmission of data over metro networks, for example, typically in the range of 100 km to 1000 km. In one non-limiting example embodiment, a 100 Gb / s signal is transmitted over a 960 km standard single mode fiber (SSMF) using a 1.5 μm direct modulation VCSEL that implements the ASK modulation scheme. The embodiment can be applied to, for example, a next generation 100 Gb / s metro network. Reception of the transmitted signal is enabled by digital coherent detection. Such transmission distances and rates represent a significant improvement, such as over known methods and systems for transmitting optical data using direct modulation VCSELs, for example.

  VCSELs are widely used in short range and low data rate applications due to their relatively low cost, energy efficiency, and small footprint. Recent developments have enabled 40 Gb / s operation of single mode 1.5 μm VCSELs and a maximum transmission distance of 60 km at 10 Gb / s. A 100 Gb / s short-range link using VCSEL with direct modulation has also been recently demonstrated by 4-level pulse amplitude modulation (PAM), polarization division multiplexing (PDM), and direct detection. However, only a transmission distance of 100 m was achieved. Because of this, and other considerations, direct modulation VCSELs have traditionally not generally been considered suitable for metro networks with transmission distances from 100 km to 1000 km at high data rates.

  FIG. 1 shows a first prior art implementation that uses multiple transmitters and receivers. For example, such a system may include ten 10 Gb / s transmitters 110 and receivers 120 or four 25 Gb / s transmitters 110a-110d and receivers 120a-120d (as shown). 4 × 25 Gb / s). One deficiency of such an implementation is that it typically occupies a large bandwidth, such as being unable to accommodate 50 GHz channel spacing. Also, such systems typically cannot transmit over distances greater than about a few hundred kilometers without using optical dispersion compensation.

  FIG. 2 shows a second prior art implementation that directly modulates two lasers 210a, 210b with a four level signal. The laser output is polarized and combined differently using a polarizing beam combiner (PBC) 220, eg, at right angles. The signal is received by a polarizing beam splitter (PBS) 230 that separates the two polarizations. Two direct detection receivers 240a and 240b then receive the separated polarization signals. One deficiency of such an implementation is that optical polarization tracking is typically required and is generally cumbersome. Also, for distances over tens of kilometers, optical dispersion compensation is typically required.

  A third prior art implementation (not shown) uses subcarrier modulation. Polarization division multiplexing (PDM) can be used in this scheme to reduce the bandwidth of the signal. On the receiver side, optical polarization tracking is typically required. In addition, this scheme is sensitive to polarization mode dispersion (PMD). (For example, A. S. Karar, J. C. Cartridge, et al., “Generation and Detection of a 112 Gb / s Dual Polarization Signal Using a Directly Modulated Laser and Half-C16 , Paper Th.3.A.4, 2012).

  Embodiments within the scope of the present disclosure overcome some of the disadvantages of the prior art implementations described above. The present invention finds that a light source, such as a laser, can be directly modulated with an M-level electrical signal to generate an M-Amplitude Shift Keyed (ASK) optical signal. In some embodiments described below, but not limited to, M is shown to be equal to 4. The complexity of the optical transmitter is significantly reduced compared to a coherent transmitter. Polarization division multiplexing (PDM) is used in some embodiments to reduce signal bandwidth. For example, a PDM-4ASK modulation system can implement a 100 Gb / s transmission rate using a 25 Gbaud symbol rate. On the receiver side, the received symbol stream can be detected coherently and, if necessary, digital signal processing (DSP) can provide polarization demultiplexing and chromatic dispersion compensation in the electrical domain. Since the transmitted signal is ASK modulated, carrier frequency and phase recovery are not required, significantly reducing the power consumption and complexity of the receiver. In some embodiments, the laser is a VCSEL, further reducing costs.

  FIG. 3 shows a block diagram of a system 300 in a non-limiting embodiment. System 300 includes a transmitter 310 and a receiver 320. The transmitter 310 includes two lasers 330a and 330b. Lasers 330a and 330b are not limited to any particular type of laser, and each is, for example, an edge-emitting distributed feedback (DFB), distributed Bragg reflector (DBR) or Fabry-Perot (FP) laser. Or a laser followed by an external modulator, such as a VCSEL, or an electroabsorption modulator or an interference-based modulator. Although not limited to VCSELs, embodiments that include this laser type include, for example, that VCSELs can be lower cost and have higher reliability than comparable edge emitting laser diodes. May be advantageous.

  Each of the lasers 330a, 330b is directly modulated with an M-level electrical signal, where, but not limited to, M is shown as being equal to four. The output of laser 330a has a first distinct polarization, eg horizontal (H). The polarization rotator (PR) 335 rotates the polarization of the output of the laser 330b so that the polarization rotation light has a second polarization different from the first polarization, such as vertical (V). The outputs of lasers 330a, 330b are combined in a polarization beam combiner (PBC) 340, thereby generating a PDM- (M) ASK signal 350, eg, M = 4.

  Referring to receiver 320, the polarization component of signal 350, which can be arbitrarily rotated with respect to the polarization at the output of transmitter 310, is separated by PBS 355. A local oscillator (LO) 360 generates two polarization components separated by PBS 365, such as H and V, for example. Each polarization component from PBS 355 beats with the corresponding polarization component from PBS 365 in the corresponding one of the two polarization diversity 90 ° optical hybrids 370a, 370b. Unreferenced photodetectors, such as single-ended or balanced photodetectors, low pass filters, and analog-to-digital converters (ADCs) convert the outputs of hybrids 370a, 370b into the digital electrical domain. The DSP 380 provides chromatic dispersion (CD) compensation, polarization demultiplexing, and intersymbol interference (ISI) equalization. In some embodiments, symbol identification may be performed immediately after the equalizer. In particular, carrier frequency and phase recovery are not required by the illustrated embodiment or are not used in the illustrated embodiment.

  FIG. 4 shows an alternative embodiment, a receiver 400, that includes a receiver 410. In this embodiment, optical 120 ° hybrids 420a, 420b, such as 3 × 3 couplers, replace the 90 ° optical hybrids 370a, 370b of system 300, for example. It has previously been shown that a properly configured 3x3 coupler can be used instead of an optical hybrid in an optical receiver. For example, C.I. Xie, et al., “Colorless coherent receiving using 3 × 3 coupler hybrids and single-ended detection”, Optics Express, Vol. 20, no. 2, pp. 1164-1171, 2012, within. In the embodiment represented by receiver 400, couplers 420a-b are expected to provide significantly lower costs compared to embodiments using optical hybrids, such as receiver 320, for example. Three single-ended detectors 430 are used for each 3 × 3 coupler, and additional signal processing may be required to obtain I and Q components.

The DSP 440 can include DSP 380 functionality and additional functionality to determine the I and Q of each received polarization channel. In summary, the optical signal field at the input of the coupler (field) E _S and LO field (field) E _L, such as single-ended detector, the output photocurrent of the detector 430 is described by equation (1).

が存在する場合、直接検波の項は、うなりの項に比べて比較的大きくなり得る。ここで、Ｍは、ＷＤＭチャネルの数であり、Ｅ_Ｓｉは、チャネルｉの光信号場（フィールド）である。Ｉ成分およびＱ成分は、以下の簡単な演算で得ることができる。 Here, φ represents the phase difference between the LO and the signal. The first term of Equation (1) is a direct detection term, and the second term is a beat term. Low local oscillator to signal power ratio (LOSPR) and / or multiple wavelength division multiplexed (WDM) channels

If is present, the direct detection term can be relatively large compared to the beat term. Here, M is the number of WDM channels, and E _Si is the optical signal field (field) of channel i. The I component and the Q component can be obtained by the following simple calculation.

_{I 1 = I 2 -0.5I 1 -0.5I} 3 = | E L || E S | cosφ (2a)

  FIG. 5 shows an alternative embodiment of the (M) ASK transmitter 510, eg, M = 4. Transmitter 510 includes a non-modulated (CW) laser source 520, Mach-Zehnder modulators (MZM) 530a, 530b, previously referenced PR 335, and previously referenced PBC 340. MZM 530a is driven by an M level signal source 540a, eg, M = 4. MZM 530b is again driven by an M level signal source 540b, eg, M = 4. Each CW optical portion received by the MZM 530a / b is thereby modulated into one of the M values, eg, M = 4. The PR 335 rotates the polarization of the modulated light from the MZM 530b with respect to the light from the MZM 530a, for example, by π / 2 radians, and these signals are recombined by the PBC 340.

  FIG. 6 shows another alternative embodiment of a (M) ASK transmitter 610, eg, M = 4, where electroabsorption modulators (EAM) 620a and 620b replace the MZM of FIG. As will be appreciated by those skilled in the optical arts, EAM can modulate the intensity of an optical signal propagating therethrough in response to an applied voltage, where the voltage modulates the band gap of the propagation medium. To do. The EAM 620a / b can thereby be used to modulate the intensity of the CW light received from the laser 520. The remaining elements of transmitter 610 can operate as described above.

  7A-7C illustrate aspects of received signal constellations 700A, 700B, and 700C in various embodiments. Each figure shows a complex IQ space, such as a plane having in-phase (horizontal axis) and quadrature (vertical axis) components of constellations 700A-700C, for example.

  Constellation 700A includes a number of closed curves 710 and symbol points 720, such as concentric rings. This constellation represents the simulated data in an embodiment after equalization for one polarization of the dual-polarized transmit signal, where each of the closed curves 710 and the symbol point 720 are: As further described, it represents a transmitted symbol. The closed curve 710 can be considered as an arc having an angular dimension of 2π. In the simulation, the line width of the transmitter laser (eg, laser 330a / b) and the line width of the LO laser (eg, laser 360) are 500 MHz and 10 MHz, respectively, but are not limited thereto. The frequency offset between the transmitter and the LO laser is 1 GHz and 2 GHz for x-polarized light and y-polarized light, respectively, but is not limited thereto.

  Constellation 700B illustrates an aspect of the received signal constellation when the concentric rings of constellation 700A are not fully closed, eg, open arc. This constellation includes an arc 730 and a symbol point 720. Each of open arc 730 and point 720 represents a transmitted symbol. Each open arc 730 has an angular dimension of less than 2π, and in this example is approximately 11π / 6. Each open arc 730 may be extrapolated along its radius to form a closed curve or ring 740 that is similar to the closed curve 710.

  Constellation 700C illustrates an aspect of the received signal constellation, where arc 750 has an angular dimension substantially less than 2π, such as, for example, approximately π / 4. Arc 750 may also be extrapolated to form a closed curve or ring 760 that is similar to closed curve 710.

  It is believed that the constellation arc angular dimensions, such as closed curve 710 and arcs 730 and 750, are determined at least in part by the line width of laser 330. A laser 330 having a smaller line width produces a constellation arc having a smaller angular dimension, while a laser 330 having a larger line width is expected to produce a constellation arc having a larger angular dimension. Is done. For example, when the line width of the laser 330 is sufficiently large, the constellation includes a closed curve as in FIG. 7A. In the very small linewidth limit, the constellation can contain points, for example arcs with very small angular dimensions. Constellation 700B and 700C show an example between these extremes, where the arc is open.

  The angular position of one of the symbol arcs in constellations 700B and 700C can be indefinite with respect to the other angular positions of the symbol arc. This may occur, for example, when there is no restriction on the polarization rotation of the transmitted light that is decomposed into the respective symbol arcs. Each closed curve 710 may be referred to as a “symbol ring”. Also, by extrapolating each arc onto a closed curve, such as one of rings 740 or 760, each arc 730, even if the arc has an angular dimension of less than 2π, eg, open. 750 may be referred to as a symbol ring.

  In this specification and claims, the term “concentric” applies to two or more symbol rings, closed curves or arcs, where one symbol ring, closed curve or arc is another symbol ring, closed curve. Or it means to be placed in an arc. When the first arc extrapolates a closed curve having a smaller radius than the closed curve that the second arc extrapolates, the first arc is disposed within the second arc. Thus, each of constellations 700A-700C includes three concentric symbol rings, as well as symbols located approximately at the origin. For the two symbol rings, the symbol ring with the smaller radius is referred to herein as the lower order symbol ring, and the symbol ring with the larger radius is referred to herein as the higher order symbol ring. In contrast to conventional M-QAM or M-PSK constellations represented by points in the complex plane, symbols are now represented by the entire ring, ie the meaning of the symbols in these constellations is It is independent of the light field value on the ring.

  In contrast to some common meanings of “concentric”, as used herein, the term includes embodiments in which the symbol ring shares the origin, but concentric symbol rings are strictly There is no need to share the origin. Also, the symbol ring 710 is approximately circular, but embodiments include non-circular symbol rings, such as, but not limited to, a closed circle such as an ellipse. In addition, in some embodiments, one symbol ring can have a small radius such that the symbol ring is effectively a symbol point around the origin of the other symbol ring. Such symbol points, such as, for example, symbol point 720, can be considered concentric symbol rings when included within one or more other symbol rings. Thus, constellation 700 represents four received symbols.

  FIG. 8 illustrates an example experimental configuration, system 800, of another embodiment. System 800 includes laser 810, digital-to-analog converter (DAC) 820, polarization multiplexer 830, amplifier 840, shaping filter 850, optical path 855, amplifier LO 860, coherent receiver 870, and digital sampling oscilloscope. 880 and offline processing 890. Herein, when the length of the optical path 855 is negligible, the embodiment may be referred to as “back-to-back”. In other embodiments, the length of the optical path 855 can be on the order of hundreds of kilometers (km). In a non-limiting example, the driver 820 provides 35 Gbaud 3 level amplitude direct modulation of the laser 810 to achieve, for example, 52.86 Gb / s, and the polarization multiplexer 830 causes 105.70 Gb / s in the same wavelength channel. Emulates two VCSELs to transmit. In various embodiments, the optical filter 850 is configured to reduce the intensity of a true subset of a plurality of concentric symbol rings. This aspect is further described below.

  If the laser 810 is a VCSEL, it can have a large line width, for example> 500 MHz. However, this has little impact on system performance and no carrier frequency and phase recovery is required in process 890, further reducing the power consumption and complexity of the coherent receiver.

  Without limitation, FIGS. 9A and 9B present experimentally determined symbol constellations that are measured using system 800. The experimental embodiment included the operation and back-to-back configuration of the 105.70 Gb / s PDM-3ASK signal without (9A) and with (9B) the presence of the shaping filter 850. These figures show the I / Q constellation of the signal output by the amplifier 840a. Referring to FIG. 3, the shaping filter 850 can be physically implemented, for example, by an optical element, or can be implemented by a DSP at the transmitter 310 or receiver 320. Embodiments may be described generally with reference to the system 800 including the filter 850 without details, recognizing that all functional aspects of the filter 850 may be implemented by a DSP or optical element. In the illustrated embodiment, the filter 850 was implemented by, for example, without limitation, a 0.67 nm filter manufactured by JDS Uniphase Corporation, Milpitas, CA, USA.

  Filter 850 can operate to reduce the strength of a true subset of concentric symbol rings. For example, looking at FIGS. 9A and 9B, it can be seen that the intensity of the low order symbol ring 910 (lowest order symbol ring) placed around the origin of the IQ plane is reduced by the presence of the filter 850, Thereby, the contrast between the symbol rings of the set is improved. Further comparing the unfiltered (FIG. 9A) and filtered (FIG. 9B) characteristics, filtering results in significant contrast enhancement of the constellation concentric symbol rings 920 and 930, improving the performance of the detected signal It is expected. In another aspect, the filter 850 suppresses lower levels of amplitude and increases the amplitude difference between different signals in the I / Q space. As further shown below, the presence of filter 850 is expected to significantly improve system performance in some embodiments by increasing signal contrast.

  The effect of the presence of the filter 850 is further illustrated by FIG. 10, which is filtered (1020) and unfiltered (1010) using the aforementioned 0.67 nm filter by the amplifier 840a in the embodiment of FIG. It shows the intensity as a function of the wavelength of the output signal. Overlapping filter response 1030 illustrates the operation of filter 850 to reduce the intensity of the optical signal at an upper wavelength and a lower wavelength of about 1526 nm. The reduction has the effect of reducing the strength of the true subset of symbol rings, such as the lowest order symbol ring described above. The filtered characteristics also show a significant reduction in off-peak optical noise floor, consistent with the improved constellation characteristics shown in FIG. 9B.

  For directly modulated lasers, such as lasers 330a and 330b, high intensity symbols are typically shifted to blue compared to low intensity symbols. In another aspect of the operation of filter 850, when filter 850 is matched to a signal wavelength in the manner shown in the exemplary embodiment of FIG. 10, the signal portion that is shifted to red (eg, a low intensity symbol) is , Subject to greater attenuation than signal portions that are shifted to blue (eg, high intensity symbols). For example, in one aspect of the illustrated embodiment, the spectrum of the signal is placed at a wavelength with a negative slope where the filter response increases, for example, with increasing wavelength. In another aspect, the peak filter response is located at about 1526.1 nm, while the peak signal strength, or signal center frequency, is located at about 1526.7 nm. In other words, the filter response can be shifted about 0.5 nm relative to the signal center wavelength in the short wavelength direction, eg, in the blue direction. In another view, the wavelength difference Δλ between the peak filter response and the signal center wavelength is about 0.5 nm. Viewed another way, this relationship between the peak filter response wavelength and the signal center wavelength results in a conversion from frequency modulation (FM) to amplitude modulation (AM). This transformation is expected to increase the eye opening for the signal and thus increase the performance of the system. In various embodiments, Δλ is such that the entire signal spectrum is located in the region of the filter response having a negative slope, eg, the wavelength of the entire signal spectrum is greater than the center wavelength of the filter. Selected. For example, in FIG. 10, the signal spectrum 1010 is substantially located at a wavelength greater than 1526 nm, and thus coincides with the portion of the filter response 1030 that has a negative slope.

Considering now FIG. 11, the bit error rate (BER) of the system 800 in back-to-back operation is shown against the optical signal-to-noise ratio (OSNR). This characteristic indicates that in this particular embodiment, there is an error floor at a BER of about 2.0 × 10 ⁻³ . A 7% overhead hard decision forward error correction (FEC) code can result in a net bit rate of about 98.80 Gb / s, and a substantially error-free operation can be expected with an OSNR of more than about 26 dB. In another aspect, a 20% overhead hard decision FEC code is used, and for a net bit rate of about 88.10 Gb / s, a substantially error free operation is expected with an OSNR of greater than 20.3 dB.

  FIG. 12 shows the BER characteristics as a function of the output power of the signal to the optical path 850 for three span lengths of 320 km, 640 km, and 960 km at a bit rate of 105.7 Gb / s. Each characteristic has a minimum BER for output power from about 2 dBm to about 3 dBm. With 7% FEC overhead, a transmission distance of 320 km can be achieved, while with 20% FEC overhead, a transmission distance of 960 km can be achieved. This transmission capability for ASK signals has not been known so far and is an unexpectedly beneficial aspect of the described embodiments. For example, the maximum transmission distance of 4-level PAM by the above-mentioned PDM and direct detection was about 100 m. Thus, the result shown in FIG. 12 is an improvement of at least about 1000 times over previous transmission distance limits and may even represent an improvement of at least about 10,000 times. Such a result clearly provides the potential benefits for cost-effective optical transmission in metro networks.

  Instead, FIG. 13 illustrates an embodiment labeled 1300 of an example optical transmission system that includes an optical transmitter 1305 and an optical receiver 1310. The transmitter 1305 is configured to transmit the modulated optical signal 1315 to the receiver 1310 via the optical path. The optical path is not limited to any particular type, but in some embodiments can include an optical fiber. As described for lasers 330a and 330b, laser 1330 is configured to provide an optical signal that is amplitude modulated between M different levels. The modulation produces a signal constellation as illustrated by constellations 700A, 700B, and 700C. Constellation control module 1340 is configured to control the amplitude and / or bias of the laser drive signal of laser 1330, such as providing amplitude modulation of optical signal 1315 to generate a signal constellation, for example.

  The transmitter 1305 also includes a feedback module 1350. The feedback module 1350 is configured to receive the amount of symbol spacing of the signal constellation of the optical signal 1315 and adjust the control module 1340 to adjust the drive signal of the laser 1330 according to the amount of symbol spacing. The In the illustrated embodiment, the feedback module 1350 provides a first electrical signal to the control module 1340 and a second electrical signal to a digital-to-analog converter (DAC) 1360. An optional amplifier 1370 can adjust the output of the DAC 1360 to an appropriate level. The first signal can be a bias adjustment and / or amplitude signal that can be directed to the control module 1340, for example, to change the DC bias and / or signal amplitude applied to the drive signal of the laser 1330. . The second signal can be a peak amplitude spacing adjustment signal that can be directed to the DAC 1360 to change, for example, one or more peak amplitude intervals between constellation symbols. The bias / amplitude adjustment signal and the symbol spacing adjustment signal are applied dynamically, for example on the time scale of the individual symbols carried by the optical signal 1315 so that the spacing between the amplitude peaks of the signal constellation can be controlled. Can be done.

  In some embodiments, the spacing between constellation amplitude peaks is determined at receiver 1310 by constellation characterization module 1335. Module 1335 may determine the spacing between the amplitude peaks of at least two received constellations, eg, between rings of constellation 700A. Referring to FIG. 7A, for example, module 1335 can determine the spacing 730 between adjacent pairs of amplitude peaks of closed curve 710. For embodiments in which the symbol constellation includes symbol arcs as illustrated in FIG. 7C, module 1335 may determine a spacing 730 between extrapolated closed curves, such as ring 760. Module 1335 can communicate amplitude peak spacing information to feedback module 1350 from which feedback module 1350 can change the amplitude and / or DC bias of the laser control signal via control module 1340, and / or Alternatively, a symbol spacing change via the DAC 1360 can be commanded. In such embodiments, symbol spacing can include the effects of any transmission medium through which signal 1315 propagates, eg, optical fiber.

  The spacing between amplitude peaks, such as between constellation symbols, is preferably substantially equal between adjacent symbols. In various embodiments, “substantially equal” means that the spacing difference between peak amplitudes in the symbol constellation is less than about 10%. In some cases, it may be preferred that the spacing difference is about 5% or less. In still other cases, it may be preferred that the spacing difference is about 1% or less. Also, as illustrated in the constellation of FIG. 9B, in some cases, the constellation symbol or ring may deviate substantially from circular symmetry. In such cases, module 1335 determines any suitable amount of spacing between the peak amplitudes of adjacent symbol pairs, eg, average spacing, or the best of constellation symbols, eg, ellipses. Spacing between conforming geometric models can be determined.

  Note that the embodiment of FIG. 13 is described as a non-limiting example of controlling the laser 1330 to generate a signal 1315 in response to the output of the drive module 1340. In some other embodiments, signal 1315 may be generated by transmitter 510 (FIG. 5) or transmitter 610 (FIG. 6), for example. In such an embodiment, the control module 1340 can control the electroabsorption modulator or MZM as appropriate for the particular embodiment.

  FIG. 14 illustrates a method 1400 for adjusting the amplitude and / or DC bias of the signal 1315 to at least partially equalize differences in amplitude between symbols of a received constellation, eg, a pulse amplitude modulation constellation. An embodiment is presented. The method 1400 may be performed by the control module 1340 and the DAC 1360, for example. In the following, the method 1400 will be described with reference to the functional entity shown in FIG. 13, but not limited to the embodiment shown. Further, those skilled in the relevant arts will appreciate that the method 1400 shown may include additional steps or may include different steps that provide the desired amplitude characteristics while remaining within the scope of the present disclosure. Will.

In step 1410, the laser control signal is set such that, for example, by control module 1340 and DAC 1360, the laser is driven with equally spaced diving signals and an initial bias. By "equally spaced diving signal" it is meant that the laser drive signal has M different levels, each approximately equidistant from its adjacent level. Thus, for example, to generate constellation 700A, the laser drive signal may have four levels that are approximately integer multiples of the initial level V _O , such as V _O , 2V _O , 3V _O , and 4V _O. . The initial bias can be any value, for example about zero volts DC. Spacing can be calculated between two adjacent symbol rings, for example (2V _O −V _O ) = V _O.

  In step 1420, for example after detection by receiver 1320, module 1335 may determine whether the amplitude level of the signal constellation is approximately equidistant. If the amplitudes are not determined to be approximately equidistant, then the method proceeds to step 1430 where the laser control signal amplitude and / or DC bias is adjusted and / or the DAC level spacing is adjusted. Such adjustment sends one or more signals that characterize the difference between the received constellation and the desired amplitude characteristic, such as uniform spacing between symbol rings, to the feedback module 1350 by module 1335. Can be included. The feedback module 1350 can then operate to provide a change in the laser control signal as described above. The method 1400 then returns to step 1420 to test again the symbol amplitude of the received constellation. If the symbol constellation is responsive to changes in laser control signal amplitude and / or DC bias and DAC level spacing in the desired manner, module 1335 and the feedback module progressively equalize the difference in constellation symbol spacing. Can operate to. Instead, if the symbol constellation responds such that the difference in constellation symbol spacing is greater, the module 1335 and the feedback module 1350 may use the difference in constellation symbol spacing in the manner of the laser control signal amplitude and It can operate to change the DC bias.

  When the spacing between constellation symbols reaches a certain threshold, such as a level that results in a reduction in the BER of the symbol stream received by receiver 1320, method 1400 may proceed to an end state 1440. Of course, the module 1335 and the feedback module 1350 continue to monitor the received symbol constellation when needed to maintain the desired level of BER and reduce the amplitude and / or DC bias and level spacing of the laser control signal. Can act to change.

  FIG. 15 shows the BER as a function of OSNR for two cases of a test system that transmits a 32 Gbaud 4-PAM signal generated by EAM. For the first BER characteristic 1510, the laser, such as laser 1330, was controlled to equalize the intensity of the constellation symbols, such as the ring of constellation 700A. This case is similar to the BER characteristic shown in FIG. For the second BER characteristic 1520, the laser was controlled to equalize the spacing between constellation symbols. A comparison between the BER characteristics for these two cases shows a factor of a 10-fold improvement in BER at 30 dB OSNR, significantly improving transmission fidelity and / or allowing longer transmission range There is expected. This result, which improves further over the already very good results discussed in FIG. 11, also provides a further unexpected and beneficial aspect of the described embodiment, which has not been known so far. A further improvement provided by the embodiment of FIGS. 13 and 14 is that symbol spacing equalization in PAM transmission communications is outside the scope of conventional optical technology and represents a significant advance in the state of the art in metro transmission systems. Clearly shows.

  While embodiments of the present invention have been shown in the accompanying drawings and described in the detailed description, the present invention is not limited to the disclosed embodiments and is described below. It should be understood that numerous reconfigurations, modifications, and substitutions are possible without departing from the invention as described and defined by the following claims.

Claims (10)

An optical transmitter configured to provide an optical signal that is amplitude modulated between M different levels;
A constellation control module configured to provide a drive signal to control the optical signal;
A feedback module configured to receive an amount of spacing between amplitude peaks of the symbol constellation of the optical signal and to adjust the constellation control module to adjust the optical signal according to the amount of spacing. ,apparatus.   The apparatus of claim 1, wherein the optical transmitter comprises a vertical cavity surface emitting laser (VCSEL) configured to generate an optical signal in response to the drive signal.   The apparatus of claim 1, wherein the constellation control module and the feedback controller are configured to substantially equalize spacing between symbols of an amplitude shift keyed symbol constellation.   The apparatus of claim 1, wherein the drive signal is indicative of amplitude, bias level, and amplitude peak spacing. An optical receiver configured to demodulate the received optical signal and generate a symbol constellation therefrom;
An apparatus comprising: a constellation characterization module configured to determine spacing between amplitude peaks of a symbol constellation and to provide an amount of spacing.   6. The apparatus of claim 5, wherein the optical receiver includes an optical hybrid configured to determine an in-phase component and a quadrature component of the optical signal, and a local oscillator.   6. The apparatus of claim 5, wherein each symbol of the symbol constellation is represented by a closed curve in an in-phase / quadrature (I / Q) plane, and the amplitude detector determines the spacing between the closed curves. Configuring an optical transmitter to provide an optical signal that is amplitude modulated between M different levels;
Configuring a constellation control module for controlling the drive signal to control the optical signal;
Configuring a feedback module that receives an amount of spacing between amplitude peaks of a symbol constellation of the optical signal and adjusts the constellation control module to adjust the optical signal according to the amount of spacing. Method.   9. The method of claim 8, wherein the constellation control module and the feedback controller are configured to substantially equalize the spacing between symbols of the amplitude shift keyed symbol constellation.   The method of claim 8, wherein the feedback module is configured to receive an amount of spacing from a coherent optical receiver.

Images