JP6040288B1 - Optical data transmission system - Google Patents

Abstract

[PROBLEMS] To appropriately perform pre-equalization capable of canceling waveform distortion of a signal received on a receiving side on a transmitting side. An optical data transmission system includes optical transceivers at both ends of an optical fiber. Each optical transceiver has an optical receiver and an optical transmitter. The optical receiving apparatus measures the waveform distortion of the main signal demodulated after receiving the optical signal and estimates the waveform distortion information, and the waveform distortion information estimated in the same manner as the above estimation in the other optical transceiver. Is decoded after receiving the optical signal from the other side. The optical transmission apparatus performs waveform distortion indicated by the waveform distortion information decoded by the decoding unit 63 and the multiplexing unit 42 that multiplexes the waveform distortion information estimated by the estimation unit 65 and the main signal transmitted to the other side. The pre-equalization unit 44 that pre-equalizes the multiplexed main signal so as to cancel, and the polarization is modulated with the multiplexed signal including the pre-equalized main signal and the estimated waveform distortion information. And an E / O converter that converts the optical signal. [Selection] Figure 2

Description

  The present invention provides an optical data transmission system that performs pre-equalization to suppress waveform distortion of a signal in one optical transceiver, and optically transmits the pre-equalized signal to the other optical transceiver via an optical transmission path. About.

  Currently, a transmission capacity of 100 Gbit / s channel capacity is realized in high-speed long-distance communication using a digital coherent method. The optical transport technology of 100 Gbit / s employs a polarization multiplexed QPSK (Quadrature Phase Shift Keying) method, and operates at a baud rate of about 32 GBaud. The polarization multiplexing QPSK can transmit and receive 4 bits of information per information symbol, and a communication capacity of approximately 127 Gbit / s is secured at a baud rate of 32 GBaud. Among these, 100 Gbit / s is assigned to the information payload, and 28 Gbit / s is assigned to redundant bits such as forward error correction, thereby realizing highly reliable information transmission.

  With such a transmission technology, a large capacity of the backbone network supporting the Internet network is realized, and at the same time, the transmission unit cost per bit is dramatically reduced. However, the transmission capacity flowing into the core network is increasing year by year, and it is necessary to further increase the channel capacity of the core network in the future.

  In order to increase the optical transmission capacity, there is a high possibility that technical upgrades such as multilevel and high baud rates will be performed in the future. Multi-leveling is a method of transmitting information using not only the phase of an optical signal but also the amplitude.

  In other words, the current 100 Gbit / s channel mainly uses the QPSK system in which information transmission is performed by dividing the phase of light into four and making the information bits correspond to the phases of the optical signals. On the other hand, it is possible to increase the optical transmission capacity by a method of improving the transmission capacity by performing multi-leveling that puts information not only on the phase of light but also on the amplitude.

  For example, a 16 QAM (Quadrature Amplitude Modulation) system that divides the phase of light into four and modulates with four levels of amplitude is one of the modulation system candidates in the next-generation optical transport network. On the other hand, increasing the baud rate simply means improving the information symbol transmission / reception rate. The current situation is about 32 GBaud, but by increasing this to 64 GBaud, it becomes possible to double the transmission capacity.

  However, by implementing multi-value and high baud rate, the demand for analog electronic circuits constituting the transmission / reception apparatus becomes more severe. In order to perform multi-value processing, the characteristics of the analog electronic circuit must be linear. This is a particularly demanding requirement for optical modulators and modulator drivers. For modulation up to QPSK, the analog electronics need only handle binary signals. For this reason, there is no problem even if the characteristics of the analog electronic circuit are non-linear. Rather, the characteristics may even be improved by a non-linear limiting operation.

  However, when using a multilevel signal such as 16QAM, the transfer function of the analog electronic circuit needs to be linear. Further, since the higher baud rate results in a wider signal bandwidth, the bandwidth of the analog electronic circuit becomes a restriction on the system performance unless the analog electronic circuit is also broadbanded.

  In other words, it is necessary to improve the performance of the analog electronic circuit in order to increase the multi-value and increase the baud rate, but on the other hand, an approach that compensates the imperfection of the analog electronic circuit by digital signal processing is also conceivable. For example, the nonlinearity of an analog electronic circuit that becomes a problem when using a multilevel signal is prepared by providing a linearizer in the DSP (Digital Signal Processor) on the transmitting side and using a nonlinear transfer function that compensates for the nonlinearity of the analog electronic circuit. The effect of nonlinearity can be reduced by acting on the signal.

  Further, the shortage of the frequency band of the analog electronic circuit that will become apparent when the baud rate is increased can be compensated to some extent by pre-emphasis on the transmission side. Pre-emphasis is pre-equalization, which amplifies the high frequency side of the transmission signal according to the attenuation characteristic at the high frequency inherent in the transmission line, sends it from the transmission side, and determines the frequency characteristic of the signal received at the reception side. This is an improved modulation technique.

  These compensation / reduction approaches using digital signal processing are effective when the transmission side DSP performs the compensation / reduction. This is because if the band compensation is performed on the receiving side, the effect is limited by the noise enhancement problem, and if the nonlinearity compensation is performed on the receiving side, the noise profile is changed. This is because it has an adverse effect.

  For this reason, in the prior art, pre-equalization has been performed by the transmission side DSP. There is a technique described in Non-Patent Document 1 as a communication technique using this pre-equalization. In this technology, the waveform distortion of the received signal due to the non-linearity and frequency characteristics of the analog electronic circuit is calculated on the transmission side, and the inverse distortion of the calculated waveform distortion is given to the waveform of the transmission signal for transmission, The signal waveform distortion is canceled (cancelled) during reception.

Takagi Sugihara “Present and Future of Pre-equalization Technology in High-speed Optical Communications” IEICE Technical Report OCS2011-41 (2011-7) pp.83-88.

  However, waveform distortion caused by non-linearity and frequency characteristics of analog electronic circuits can be measured using the received signal or training signal on the receiving side, but is basically measured on the transmitting side. Can not. For this reason, there is a problem that even if waveform distortion is obtained by calculation as in Non-Patent Document 1, there is an actual error, so that proper pre-equalization cannot be performed on the transmission side.

  It is also possible to measure the waveform distortion of the signal in advance on the receiving side and use this waveform distortion on the transmitting side to perform pre-equalization. If the frequency characteristics fluctuate, a waveform distortion and an error that have been obtained in advance are generated, which causes a problem that proper pre-equalization cannot be performed on the transmission side.

  The present invention has been made in view of such circumstances, and an optical data transmission system capable of appropriately performing pre-equalization capable of canceling waveform distortion of a signal received on the receiving side on the transmitting side. The issue is to provide.

As means for solving the above-mentioned problems, the invention according to claim 1 is directed to an optical transmitter for transmitting an optical signal, the polarization of which is modulated by a main signal as transmission data, to the optical transmission line, and the transmitted optical signal And an optical transceiver including an optical receiver that demodulates the main signal at both ends of the optical transmission path, and performs optical transmission, wherein the optical receiver includes an optical transceiver on one side. An estimator that measures waveform distortion of the demodulated main signal after receiving an optical signal by a transceiver and estimates waveform distortion information related to the waveform distortion, and a waveform estimated in the same manner as the above-mentioned estimation in the other optical transceiver A decoding unit that decodes distortion information after receiving an optical signal from the other optical transceiver, and the estimation unit measures nonlinear distortion of the demodulated main signal, and measures the measured nonlinear distortion. From the waveform In addition to estimating the nonlinear transfer function as information, the energy of a specific frequency band of the training signal sequence multiplexed with the main signal received from the optical transceiver on the other side is measured, and the energy amount of the measured specific frequency band is calculated. A corresponding linear transfer function is estimated, and the decoding unit calculates the linear transfer function and the nonlinear transfer function estimated by the estimation unit in the other optical transceiver, and calculates the optical signal from the other optical transceiver. Decoding after reception, the optical transmission device encodes the linear transfer function and the nonlinear transfer function estimated by the estimation unit to generate a control signal, and the main signal has energy in a specific frequency band. a sequence multiplexing unit for multiplexing the known training signal sequence but centered, the waveform distortion information estimated by the estimating unit, the other The multiplexed main signal is pre-equalized so as to cancel the waveform distortion indicated by the waveform distortion information decoded by the decoding unit, and a multiplexing unit that multiplexes the main signal to be transmitted to the optical transceiver A pre-equalization unit; and a conversion unit that modulates polarization with a multiplexed signal including the pre-equalized main signal and the estimated waveform distortion information and converts the modulated signal into an optical signal, and the generation unit includes: The linear transfer function estimated by the estimation unit is encoded to generate a control signal, and the multiplexing unit trains the generated control signal and the sequence multiplexing unit to transmit to the other optical transceiver. The pre-equalization unit uses the linear transfer function and the nonlinear transfer function decoded by the decoding unit to multiplex an optical signal received by the other optical transceiver. In order to cancel the waveform distortion, An optical data transmission system characterized by performing pre-equalization of the converted main signal .

  According to this configuration, when one optical transceiver (also referred to as one side) transmits an optical signal to the other side optical transceiver (also referred to as the other side) via the optical transmission path, optical transmission is performed from one side. The main signal in the multiplexed signal to be transmitted is pre-equalized so as to cancel the waveform distortion indicated by the waveform distortion information received from the other side with respect to the main signal included in the multiplexed signal. The waveform distortion information described above is waveform distortion information estimated by receiving an optical signal transmitted from one side and received from the other side, and thus indicates waveform distortion of the signal due to nonlinearity and frequency characteristics of the analog electronic circuit on one side.

Therefore, as described above, if pre-equalization is performed on the main signal transmitted to the other side with the waveform distortion information received from the other side on the one side, the following effects can be obtained. That is, when an optical signal transmitted from one side is received on the other side, the optical signal is pre-equalized so as to cancel the waveform distortion that originally occurs, so that it is received on the other side. Waveform distortion can be reduced or eliminated in the optical signal. As described above, according to the present invention, on the transmission side (one side optical transceiver), the pre-equalization capable of canceling the waveform distortion of the signal received on the receiving side (the other side optical transceiver) is appropriately performed. Can do. Further, as the waveform distortion information for pre-equalization, the linear transfer function and the nonlinear transfer function estimated by measuring the linear distortion and nonlinear distortion of the main signal obtained by demodulating the optical signal from the other side are used. These transfer functions represent a signal state as a function of how an optical signal is transmitted from the one side to the other side via the optical transmission path. For this reason, if these transfer functions are used for pre-equalization, pre-equalization of the main signal transmitted from one side to the other side can be performed with higher accuracy. Furthermore, when the energy of a specific frequency in the training signal sequence is measured, if the energy is reduced, it can be seen that the signal waveform at that frequency is distorted. Since the distorted signal is transmitted from the other side, waveform distortion caused by nonlinearity and frequency characteristics of the analog electronic circuit on the other side occurs. Therefore, the linear transfer function associated with the reduced energy amount is transmitted to the other side, and the other side pre-equalization unit performs pre-equalization using the linear transfer function as described above. With this pre-equalization, if the main signal is transmitted as an optical signal from the other side and received on one side, the received main signal is in a state where the linear distortion component is reduced or absent.

According to a second aspect of the present invention, in the optical receiver, the estimation unit obtains a waveform distortion by comparing a main signal received from the other optical transceiver by a provisional determination process with a known reference main signal. as linear distortion components of the distortion is minimized, the tap coefficient vector corresponding to said linear by adaptive equalization algorithm estimation, the decoding unit, as before Ki推 constant in the optical transceiver of the other side to the tap coefficient vector estimated and decoded after reception of the optical signal from the optical transceiver of the other side, in the optical transmitting apparatus, the generating unit, encoding the tap coefficient vectors that are pre Ki推 constant Generating a control signal, the multiplexing unit multiplexes the generated control signal and a main signal to be transmitted to the optical transceiver on the other side, and the pre-equalization unit includes an FIR filter. For example, to perform a pre-equalization of the multiplexed main signal, optical data according to claim 1 which comprises using the tap coefficient vector decoded by the decoder as a weighting coefficient of the FIR filter It is a transmission system.

  According to this configuration, the following operational effects can be obtained. For example, in the received main signal sequence, it is assumed that a signal (symbol) immediately before the current signal is mixed with the current signal and waveform distortion occurs. In this case, in the FIR filter, if the previous signal is made negative by a tap coefficient vector as a weighting coefficient corresponding to the linear transfer function and added to the current signal, the above-mentioned mixing is eliminated and the waveform distortion is reduced. The linear distortion component can be canceled out. Thus, if the tap coefficient vector of the FIR filter is estimated and weighted, the main signal can be pre-equalized so as to cancel the linear distortion of the main signal transmitted to the other side.

According to a third aspect of the present invention, in the optical receiver, the estimation unit obtains a waveform distortion by taking a difference between a main signal received from the other optical transceiver and a known reference main signal, and the waveform distortion as non-linear distortion component is minimized, the tap coefficient vector corresponding to the non-linear transfer function to estimate the adaptive equalization algorithm, said decoder is estimated as before Ki推 constant in the optical transceiver of the other side the tap coefficient vectors, and decoded after reception of the optical signal from the optical transceiver of the other side, in the optical transmitting apparatus, the generating unit, prior Ki推 constant tap coefficients vector coded by control signal The multiplexing unit multiplexes the generated control signal and the main signal to be transmitted to the other optical transceiver, and the pre-equalization unit includes a nonlinear FIR filter. For example, to perform a pre-equalization of the multiplexed main signal light according to claim 1 which comprises using the tap coefficient vector decoded by the decoder as a weighting coefficient of the nonlinear FIR filter A data transmission system.

  According to this configuration, the following operational effects can be obtained. For example, in the received main signal sequence, when a signal (symbol) immediately before the current signal is mixed with the current signal and waveform distortion occurs, the weighting coefficient is applied to the previous signal in the nonlinear FIR filter. The tap coefficient vector corresponding to the nonlinear transfer function is negative. If this is added to the current signal, the mixing can be eliminated and the nonlinear distortion component of the waveform distortion can be canceled. Thus, if the tap coefficient vector of the nonlinear FIR filter is estimated and weighted, the main signal can be pre-equalized so as to cancel the nonlinear distortion of the main signal transmitted to the other side.

According to a fourth aspect of the present invention, in the optical receiver, the estimation unit calculates an average value of error vector groups in which the distribution of main signal symbols received from the other optical transceiver is biased by provisional determination processing, said non-linear transfer function corresponding to the calculated average value was estimated, the decoding unit, a non-linear transfer function estimated as before Ki推 constant in the optical transceiver of the other side, of the other side light decoded after reception of the optical signal from the transceiver, in the optical transmitting apparatus, the generating unit, prior to encoding the Ki推 constant nonlinear transfer function to generate a control signal, the multiplexing unit was the product The control signal and the main signal to be transmitted to the other optical transceiver are multiplexed, and the pre-equalization unit receives the non-linear transfer function decoded by the decoding unit by the other optical transceiver. So as to cancel the waveform distortion of the optical signal is an optical data transmission system according to claim 1, characterized in that the pre-equalization of the multiplexed main signal.

  According to this configuration, the pre-equalization is performed by the average value (nonlinear transfer function) of the error vector group indicating that the waveform distortion has occurred. In other words, the inverse vector that cancels the average value of the error vector group is obtained. Pre-equalization is performed on the main signal that is generated and transmitted to the other side using this inverse vector. Thus, one side decodes the nonlinear transfer function corresponding to the average value of the error vector group indicating the waveform distortion from the main signal received from the other side. This decoded non-linear transfer function corresponds to non-linear distortion due to non-linearity of the analog electronic circuit on one side. Therefore, if one side performs pre-equalization on the main signal optically transmitted to the other side with the decoded nonlinear transfer function, the main signal received on the other side has reduced or no nonlinear distortion component. It becomes a state.

  ADVANTAGE OF THE INVENTION According to this invention, the optical data transmission system which can perform appropriately the pre-equalization which can cancel the waveform distortion of the signal received at the receiving side on the transmission side can be provided.

It is a block diagram which shows the structure of the optical data transmission system which concerns on embodiment of this invention. It is a block diagram which shows the structure of the optical transceiver of the optical data transmission system of this embodiment. It is a circuit diagram which shows the specific structural example of the optical data transmission system of this embodiment. It is a circuit diagram which shows the connection structure of the optical modulator in the specific structure of the optical data transmission system of this embodiment. It is a block diagram which shows the structure of a FIR filter. It is a block diagram which shows the structure of a non-linear FIR filter. It is a block diagram which shows the structure of the optical transceiver of the modification 3 of this embodiment. (A) Frequency spectrum diagram showing that power concentrates on two or more specific frequencies, (b) Diagram showing BPSK signals -S, S on the IQ plane. It is a figure which shows the deviation of the error vector on an IQ plane, and distribution of an error vector.

Embodiments of the present invention will be described below with reference to the drawings.
<Configuration of Embodiment>
FIG. 1 is a block diagram showing a configuration of an optical data transmission system according to an embodiment of the present invention.
An optical data transmission system 10 shown in FIG. 1 includes two optical fibers 11 and 12 as optical transmission paths laid in parallel, and optical transceivers 14 and 15 connected to both ends of the optical fibers 11 and 12. Configured. In the middle of the optical fibers 11 and 12, optical amplifiers 17 and 18 are inserted as necessary. For example, when the length of the optical fibers 11 and 12 exceeds approximately 100 km, the optical amplifiers 17 and 18 are connected in the middle of the optical fibers 11 and 12 to amplify the optical signals.

  The optical transceivers 14 and 15 include an optical transmission device 20 and an optical reception device 30. The optical transmitter 20 of the optical transceiver 14 and the optical receiver 30 of the optical transceiver 15 are connected by a single optical fiber 11, and the optical transmitter 20 of the optical transceiver 15 and the optical receiver 30 of the optical transceiver 14 are connected. They are connected by a single optical fiber 12.

  The feature of this embodiment is that waveform distortion information is first estimated by measuring waveform distortion of a signal received by the optical receiver 30 of the optical transceiver 14 on one side. The measured waveform distortion is a waveform distortion caused by nonlinearity and frequency characteristics of the analog electronic circuit of the optical transceiver 15 on the other side. Next, in the optical transmission device 20 of the one-side optical transceiver 14, the estimated waveform distortion information is multiplexed as a control signal into the main signal, and this multiplexed signal is transmitted to the optical fiber 11.

  The signal transmitted via the optical fiber 11 is received by the optical receiver 30 of the optical transceiver 15 on the other side, and the control signal is separated. In the optical transmitter 20 of the optical transceiver 15 on the other side, The main signal is pre-equalized with the waveform distortion information indicated in the separated control signal. That is, the transmission signal is pre-equalized with waveform distortion (for example, reverse waveform distortion) that cancels the waveform distortion caused by the waveform distortion information.

The pre-equalization is similarly performed in the optical transceiver 14 on one side. The waveform distortion is similarly estimated in the optical transceiver 15 on the other side.
However, the main signal is a digital signal including a transmission data sequence (transmission information). The waveform distortion information includes a linear transfer function and a nonlinear transfer function, which will be described later. The optical transceiver 14 on one side is also referred to as one side, and the optical transceiver 15 on the other side is also referred to as the other side.

The optical transmission device 20 has a function of transmitting the main signal in parallel via the optical fibers 11 and 12 using parallel or orthogonal X polarization and Y polarization, and includes a transmission signal processing unit 40 and E / O (electricity / light) conversion unit 50. Note that the conversion unit described in the claims includes the E / O conversion unit 50.
The optical receiver 30 has a function of demodulating signals received from the optical fibers 11 and 12, and includes a received signal processing unit 60 and an O / E (optical / electrical) converter 70.

In each of the optical transceivers 14 and 15, the reception signal processing unit 60 and the transmission signal processing unit 40 are connected by a digital IF (interface) 80 that performs data input / output processing. For the digital IF 80, for example, an I ² C (Inter-Integrated Circuit) which is a serial bus for performing serial communication is used.

  As shown in FIG. 2, the transmission signal processing unit 40 includes an input IF 41, a control signal multiplexing unit 42, a control signal generation unit 43, a transmission side pre-equalization unit 44, and an output IF 45. Yes. The control signal multiplexing unit 42 is also referred to as a multiplexing unit 42, the control signal generation unit 43 is also referred to as a generation unit 43, and the transmission side pre-equalization unit 44 is also referred to as a pre-equalization unit 44. The input IF 41 and the output IF 45 perform signal or data input / output processing.

  The reception signal processing unit 60 includes an input IF 61, a control signal separation unit 62, a control signal decoding unit 63, a reception side equalization unit 64, a linear / nonlinear transfer function estimation unit 65, and an output IF 66. Has been. The control signal separation unit 62 is also called a separation unit 62, the control signal decoding unit 63 is also called a decoding unit 63, the reception side equalization unit 64 is also called an equalization unit 64, and the linear / nonlinear transfer function estimation unit 65 is also called an estimation unit 65. The linear / nonlinear transfer function estimation unit 65 may be provided separately for a linear transfer function estimation unit and a nonlinear transfer function estimation unit. The input IF 61 and the output IF 66 perform signal or data input / output processing.

  In the reception signal processing unit 60, the separation unit 62 is an electrical signal obtained by photoelectrically converting the optical signal transmitted from the other optical transceiver 15 (FIG. 1) and received on one side by the O / E conversion unit 70. The main signal is separated into a control signal and a main signal, and the control signal is output to the decoding unit 63 and the main signal is output to the equalization unit 64 and the estimation unit 65.

  The equalization unit 64 restores transmission information on the transmission side by performing equalization processing on the main signal, and outputs this via the output IF 66.

  The estimation unit 65 estimates (first estimation) a linear transfer function and a nonlinear transfer function as waveform distortion information on the other side from the main signal as described later. The main signal includes waveform distortion caused by nonlinearity and frequency characteristics of the analog electronic circuit of the optical transceiver 15 on the other side, and this waveform distortion is composed of linear distortion and nonlinear distortion. Therefore, the estimation unit 65 measures the linear distortion of the received signal, estimates a linear transfer function that gives the linear distortion to the signal waveform from the measured linear distortion data (first estimation), and also nonlinearizes the received signal. The distortion is measured, and the nonlinear transfer function given the nonlinear distortion is estimated (first estimation) from the measured nonlinear distortion data. The estimated linear transfer function and nonlinear transfer function on the other side are output to the generation unit 43.

  The decoding unit 63 decodes the control signal transmitted from the other side separated by the separation unit 62, acquires the linear transfer function and the nonlinear transfer function of the optical transceiver 14 on one side, and predicts these transfer functions. To the conversion unit 44. These transfer functions are obtained when the optical transceiver 15 on the other side receives the main signal when an optical signal is transmitted from one side to the other side, and the estimator 65 on the other side receives the main signal from the waveform distortion contained in the main signal. Similarly, a linear transfer function and a non-linear transfer function on one side estimated (first estimation). That is, these transfer functions relate to waveform distortion caused by nonlinearity and frequency characteristics of the analog electronic circuit of the optical transceiver 15 on one side.

  In the transmission signal processing unit 40, the generation unit 43 generates a control signal by encoding the other-side linear transfer function and nonlinear transfer function estimated by the estimation unit 65, and outputs the control signal to the multiplexing unit 42. The control signal is notified from one side to the optical transceiver 15 on the other side, and used for the pre-equalization processing in the optical transmitter 20 on the other side (FIG. 1).

  The multiplexing unit 42 multiplexes the control signal generated by the generation unit 43 and the main signal input from the input IF 41, and outputs the multiplexed signal to the pre-equalization unit 44.

  The pre-equalization unit 44 outputs the multiplexed signal from the multiplexing unit 42 to the E / O conversion unit 50 via the output IF 45, and the linear transfer function and the nonlinear transfer function on one side decoded by the decoding unit 63. To perform the pre-equalization process described later. That is, the pre-equalization unit 44 performs linear pre-equalization on the main signal included in the multiplexed signal transmitted from the optical transceiver 14 on the one side using the linear transfer function on the one side and nonlinear transmission on the one side. Perform nonlinear pre-equalization with functions. These pre-equalizations are performed by controlling the amplitude and phase of the main signal to be transmitted so as to cancel the waveform distortion of the optical signal (main signal) reaching the optical transceiver 15 on the other side from the optical transceiver 14 on the one side. It is. In other words, pre-equalization is performed on the main signal with waveform distortion (for example, reverse waveform distortion) that cancels the waveform distortion of the optical signal that has arrived from the optical transceiver 14 on one side to the optical transceiver 15 on the other side.

  That is, the pre-equalization unit 44 performs pre-equalization processing on the main signal so as to cancel the waveform distortion originally generated when the main signal is received by the optical transceiver 15 on the other side. In the received optical signal, waveform distortion is reduced or absent.

<Specific configuration and operation of optical transceiver>
FIG. 3 shows an example in which the optical transceivers 14 and 15 that perform the pre-equalization processing described above are specifically configured by devices. The optical transceiver 14 is also referred to as a first optical transceiver 14, and the optical transceiver 15 is also referred to as a second optical transceiver 15.
As shown in the optical transmission device 20 of FIG. 3, the transmission signal processing unit 40 includes a DSP (Digital Signal Processor) 40a, and the E / O conversion unit 50 includes DACs (Digital to Analog Converters) 47a to 47d, A polarization multiplexing QPSK modulator module 50a (also referred to as module 50a) and a transmission LD (Laser Diode) 50b are provided.

The module 50a includes optical modulators 52x and 52y, drivers 53a to 53d, and a polarization beam combiner 54.
In the optical modulators 52x and 52y, as shown as a representative of the optical modulator 52x in FIG. 4, an optical waveguide 55 is formed on an optical material substrate K such as LiNbO3 (lithium niobate), and one optical waveguide. Electrodes Ea and Eb are mounted and fixed on the branched optical waveguides 55a and 55b obtained by bifurcating 55. Further, the branched optical waveguides 55a and 55b are coupled to the single optical waveguide 55 on the output side of the optical transmission signal to constitute the optical modulators 52x and 52y.

  As shown in FIG. 4, the pre-equalization unit 44 (see FIG. 2) configured by the DSP 40a includes a transmission-side pre-equalization unit 44c and a transmission-side pre-equalization unit 44s. The pre-equalization unit 44c outputs a digital pre-equalization signal for providing a cos component (quadrature phase component) in X polarization. The pre-equalization unit 44s outputs a digital pre-equalization signal for providing a sin component (in-phase component).

The output side of the pre-equalization unit 44c is connected to the electrode Ea on the branch optical waveguide 55a via the output IF 45c, the DAC 47a, and the driver 53a.
The output side of the pre-equalization unit 44s is connected to the electrode Eb on the branch optical waveguide 55b via the output IF 45s, the DAC 47b, and the driver 53b.

The laser light output side of the transmission LD 50b (FIG. 3) is connected to the optical waveguide 55 on the input side of the optical modulator 52x. A polarization beam combiner 54 (FIG. 3) is connected to the optical waveguide 55 on the output side. The output side of the polarization beam combiner 54 is connected to the optical fiber 11 (FIG. 3).
Similar to the X-polarized light modulator 52x, a Y-polarized light modulator 52y is also configured.

  Next, referring back to FIG. 3, as shown in the optical receiver 30 of the optical transceiver 15, the O / E converter 70 includes an optical reception front-end module 70 a (also referred to as a module 70 a) and a local oscillation LD (Laser Diode). The reception signal processing unit 60 is configured by a DSP 60a.

The module 70a includes a polarization separator 71, 90-degree hybrid PDs (Photo Diodes) 72x and 72y, and ADCs (Analog to Digital Converters) 73a to 73d.
The input side of the polarization separator 71 is connected to the optical fiber 11. The output side branched to the X polarization side and the Y polarization side of the polarization separator 71 is connected to the input side of the 90-degree hybrid PD 72x with the output side of the X polarization paired with the laser light output side of the local oscillation LD 70b. The output side of the Y polarization is paired with the laser light output side of the local oscillation LD 70b and connected to the input side of the 90-degree hybrid PD 72y.
The output sides of the 90-degree hybrid PDs 72x and 72y are connected to the DSP 60a via ADCs 73a to 73d.

The operation of the optical transceivers 14 and 15 including the optical transmission device 20 and the optical reception device 30 configured as described above will be described.
In the optical transmitter 20 of the first optical transceiver 14, the DSP 40a performs the same process as the process in the transmission signal processing unit 40 described above. That is, the main signal in the multiplexed signal of the digital main signal and the digital control signal is pre-equalized by the processing of the DSP 40a, and the digital multiplexed signal after this pre-equalization is converted into an analog signal by the DACs 47a to 47d. The The converted analog multiplexed signal is amplified by the drivers 53a to 53d and then applied to the electrodes Ea and Eb of the optical modulators 52x and 52y.

Since the laser light of the transmission LD 50b is incident on each optical waveguide 55 of the optical modulator 52x on the X polarization side, the amplitude of this laser light depends on the multiplexed signal applied to the electrodes Ea and Eb. And an X-polarized optical signal whose phase is modulated (also referred to as modulated light). Each of the X-polarized modulated light has a cos component and a sin component.
The Y polarization side is similarly modulated. Further, the X polarization and the Y polarization are in a parallel or orthogonal state.

  In the above modulation, the main signal in the multiplexed signal is pre-equalized, so that the main signal component in the modulated light of X polarization and Y polarization is pre-equalized. The optical signals modulated by the X polarization and the Y polarization are combined by the polarization combiner 54 and transmitted to the optical fiber 11.

  The optical signal transmitted through the optical fiber 11 is received by the optical receiver 30 of the second optical transceiver 15, and the polarization separator 71 performs polarization separation on the received optical signal in the optical region. It is separated into two orthogonal polarizations (X polarization and Y polarization) and is incident on each of the 90-degree hybrid PDs 72x and 72y. At this time, the laser light from the local oscillation LD 70b is also incident on the 90-degree hybrid PDs 72x and 72y.

  The 90-degree hybrid PD 72x, 72y performs photoelectric conversion, and separates the optical electric field of the X-polarized and Y-polarized signal light coherently received using the laser light from the local oscillation LD 70b into components orthogonal to each other. And convert it into an analog electrical signal. The converted cos and sin component signals of the X polarization are converted into digital signals by the ADCs 73a and 73b, and the cos and sin component signals of the Y polarization are converted into digital signals by the ADCs 73c and 73d. Input to the DSP 60a.

  The DSP 60a performs the same processing as the received signal processing unit 60 described above from the input digital signal, estimates the linear transfer function and the nonlinear transfer function on the first optical transceiver 14 side (first side), and The control signal transmitted by the optical transceiver 14 is decoded to obtain a linear transfer function and a nonlinear transfer function on the second optical transceiver 15 side (second side). The linear transfer function and nonlinear transfer function on the first side, and the linear transfer function and nonlinear transfer function on the second side are output to the optical transmitter 20 on the second side.

  In the second-side optical transmitter 20, the transmission signal processing unit 40 configured by a DSP (not shown) generates the first-side control signal from the first-side linear transfer function and the nonlinear transfer function, as described above. The control signal is multiplexed with the main signal. Further, the pre-equalization processing described above is performed on the main signal transmitted from the second side to the first side by the linear transfer function and the nonlinear transfer function on the second side. Since the main signal transmitted as the optical signal is subjected to pre-equalization processing so as to cancel the waveform distortion originally generated when it is received by the first optical transceiver 14, the first optical transceiver 14 In the optical signal received at, the waveform distortion is reduced or absent.

<Effect of embodiment>
As described above, the optical data transmission system 10 according to this embodiment includes the optical transmission device 20 that transmits an optical signal whose polarization is modulated by the main signal as transmission data to the optical fiber 11 or 12, and the optical transmission device 20 In this system, an optical transceiver including an optical receiver 30 that receives an optical signal and demodulates a main signal is provided at both ends of the optical fibers 11 and 12 to perform optical transmission.

  The feature of this embodiment is that the optical receiver 30 measures the waveform distortion of the main signal demodulated after receiving the optical signal by the optical transceiver 14 on one side, and estimates the waveform distortion information related to this waveform distortion. 65 and a decoding unit 63 that decodes the waveform distortion information estimated in the same manner as the above estimation in the optical transceiver 15 on the other side after receiving the optical signal from the optical transceiver 15 on the other side.

  Further, the optical transmission device 20 is configured to multiplex the waveform distortion information estimated by the estimation unit 65 and the main signal to be transmitted to the optical transceiver 15 on the other side, and the waveform distortion decoded by the decoding unit 63. A pre-equalization unit 44 for pre-equalization of the multiplexed main signal so as to cancel the waveform distortion indicated by the information, and a multiplexed signal including the pre-equalized main signal and the estimated waveform distortion information And an E / O converter 50 that modulates the polarization and converts it into an optical signal.

  According to this configuration, when one optical transceiver 14 (also referred to as one side) transmits an optical signal to the other side optical transceiver 15 (also referred to as the other side) via the optical fiber 11, light is transmitted from one side. The main signal in the multiplexed signal to be transmitted is canceled so as to cancel the waveform distortion indicated by the waveform distortion information received from the other side via the optical fiber 12 with respect to the main signal included in the multiplexed signal to be transmitted. Perform pre-equalization. The waveform distortion information described above is waveform distortion information estimated by receiving an optical signal transmitted from one side and received from the other side, and thus indicates waveform distortion of the signal due to nonlinearity and frequency characteristics of the analog electronic circuit on one side.

  From this, as described above, if pre-equalization is performed on the one side on the main signal transmitted to the other side, the following effects can be obtained. That is, when an optical signal transmitted from one side is received on the other side, the optical signal is pre-equalized so as to cancel the waveform distortion that originally occurs, so that it is received on the other side. Waveform distortion can be reduced or eliminated in the optical signal. As described above, according to the present invention, pre-equalization that can cancel the waveform distortion of the signal received at the receiving side (the optical transceiver 15 on the other side) is properly performed on the transmitting side (the optical transceiver 14 on the one side). It can be carried out.

  Further, the estimation unit 65 measures linear distortion and nonlinear distortion of the main signal demodulated after receiving the optical signal on one side, and linear transmission as the above-described waveform distortion information from the measured linear distortion and nonlinear distortion. A function and a nonlinear transfer function are estimated (first estimation). The decoding unit 63 converts the linear transfer function and the nonlinear transfer function as waveform distortion information estimated in the same manner as the above estimation (first estimation) in the other optical transceiver 15 into the optical signal from the other optical transceiver. Decrypt after receiving.

Furthermore, the optical transmission device 20 includes a generation unit 43 that encodes the linear transfer function and the nonlinear transfer function that are first estimated by the estimation unit 65 to generate a control signal.
The multiplexing unit 42 multiplexes the generated control signal and the main signal to be transmitted to the optical transceiver 15 on the other side, and the pre-equalization unit 44 uses the linear transfer function and the nonlinear transfer function decoded by the decoding unit 63. Is used to pre-equalize the multiplexed main signal so as to cancel the waveform distortion of the optical signal received by the optical transceiver 15 on the other side.

  As described above, the linear transfer function and the nonlinear transfer function estimated by measuring and estimating the linear distortion and nonlinear distortion of the main signal obtained by demodulating the optical signal from the other side are used as the waveform distortion information for pre-equalization. These transfer functions represent a signal state as a function of how an optical signal is transmitted from one side to the other side via an optical fiber. For this reason, if these transfer functions are used for pre-equalization, pre-equalization of the main signal transmitted from one side to the other side can be performed with higher accuracy.

<Modification 1 of Embodiment>
Next, Modification 1 of the present embodiment will be described. However, in Modification 1, the estimation unit 65 does not measure linear distortion as in the above-described embodiment. The same applies to Modifications 2 to 4 described later.

The feature of the first modification is that the pre-equalization unit 44 shown in FIG. 2 is configured to include an FIR (Finite Impulse Response) filter 90 shown in FIG.
Furthermore, the estimation unit 65 shown in FIG. 2 compares the main signal received from the optical transceiver 15 (FIG. 1) on the other side by the provisional determination process with a reference main signal stored in a storage unit (not shown), thereby generating waveform distortion. Ask for. Further, the estimation unit 65 uses a decision-oriented LMS (Least Mean Square) algorithm as an adaptive equalization algorithm so that the obtained linear distortion component of the waveform distortion is minimized so that the tap coefficient vector ( (Also referred to as a tap coefficient) is estimated (second estimation). This tap coefficient is treated as the linear transfer function described above.

  Then, the generation unit 43 generates a control signal by encoding the estimated (second estimation) tap coefficient, and outputs the control signal to the multiplexing unit 42. The multiplexing unit 42 multiplexes the main signal and the control signal, and the multiplexed signal is converted into an optical signal and transmitted to the optical transceiver 15 on the other side.

  Further, an optical signal in which the tap coefficient corresponding to the linear transfer function estimated (second estimation) in the same manner as described above in the optical transceiver 15 (FIG. 1) on the other side is multiplexed with the main signal and optically converted is sent to one side. It is received by the side optical transceiver 14. Further, the decoding unit 63 decodes the tap coefficient from the received signal. The decoded tap coefficient is given as a weighting coefficient to the FIR filter 90 to perform pre-equalization.

  Here, the FIR filter 90 shown in FIG. 5 will be described. The FIR filter 90 includes a plurality of delay circuits 91a to 91c connected in series, a plurality of multipliers 92a to 92d, and an adder 93. A multiplier 92a is connected to the signal input side of the delay circuit 91a, a multiplier 92b is provided between the delay circuits 91a and 91b, a multiplier 92c is provided between the delay circuits 91b and 91c, and a multiplier is provided on the signal output side of the delay circuit 91c. A device 92d is connected. The output sides of the multipliers 92 a to 92 d are connected to the adder 93.

In the delay circuits 91 a - 91 c, by the input signal sequence x _k go by a predetermined time delay, and can detect back past signal. Then, the individual weighting factors in the multipliers 92a~92d * a0, * a1, * a2, a * a2, individually weighted by multiplying the input signal sequence _{x k} and the delay signals of each delay circuit 91a~91c These results are added by an adder 93 and output as an output signal sequence y _k . One output signal sequence y _k is output every time one input signal sequence x _k is input.

…式（１）

但し、ａ_ｍはタップ係数（重み付け係数）に対応している。 The FIR filter 90, the input-output relationship between the input signal sequence _{x k} and the output signal sequence _{y k} is represented by the following formula (1).

... Formula (1)

However, a _m corresponds to the tap coefficient (weighting coefficient).

In this FIR filter 90, the estimated tap coefficients are used for the multipliers 92a to 92d as the weighting coefficients * a0 to * a2.
Here, in the FIR filter 90, in order to compensate for the waveform distortion of the transmission side optical transceiver, there is a just good weighting coefficient that can optimally remove the distortion of the signal. This coefficient will be described.

  For example, when the waveform of the received signal is distorted, the components of the main signal symbols before and after are slightly mixed on the time axis. As an example of this, if the signal (symbol) immediately before the current signal is mixed with the current signal in the main signal sequence, the FIR filter 90 taps the signal immediately before the current signal with a linear transfer function. If the coefficient (weighting coefficient) is set to minus and added to the current signal, the linear distortion component in the waveform distortion can be canceled without mixing. Thus, if the tap coefficient of the FIR filter 90 is estimated and weighted, the main signal is pre-equalized so as to cancel the linear distortion component of the main signal transmitted to the optical transceiver 15 on the other side. be able to.

<Modification 2 of Embodiment>
Next, a second modification of the present embodiment will be described. The feature of the modified example 2 is that the pre-equalization unit 44 is configured to include the nonlinear FIR filter 100 shown in FIG.
Further, the estimation unit 65 shown in FIG. 2 obtains the waveform distortion by taking the difference between the main signal received from the optical transceiver 15 (FIG. 1) on the other side and the reference main signal stored in the storage unit (not shown). . Further, the estimation unit 65 estimates (third estimation) the tap coefficient used in the nonlinear FIR filter 100 by the decision-oriented LMS algorithm so that the nonlinear distortion component of the obtained waveform distortion is minimized. This tap coefficient is treated as the above-described nonlinear transfer function.

  Then, the generation unit 43 generates a control signal by encoding the estimated (third estimation) tap coefficient, and outputs the control signal to the multiplexing unit 42. The multiplexing unit 42 multiplexes the main signal and the control signal, and the multiplexed signal is converted into an optical signal and transmitted to the optical transceiver 15 on the other side.

  Further, in the optical transceiver 15 on the other side (FIG. 1), the tap coefficient corresponding to the nonlinear transfer function estimated (third estimation) in the same manner as described above is multiplexed with the main signal, and the optical signal is optically converted. Reception is performed by the optical transceiver 14 on one side. Further, the decoding unit 63 decodes the tap coefficient from the received signal. The decoded tap coefficient is given to the nonlinear FIR filter 100 as a weighting coefficient to perform pre-equalization.

  Here, the non-linear FIR filter 100 shown in FIG. 6 will be described. The nonlinear FIR filter 100 includes a plurality of power circuits 101a to 101b, a plurality of multipliers 102a to 102c, and an adder 103. However, the power circuit 101a is a square circuit 101a, and the power circuit 101b is a cube circuit 101b.

The multiplier 102a, the square circuit 101a, and the cube circuit 101b are configured to receive the input signal series _xk in parallel. A multiplier 102b is connected to the output side of the square circuit 101a, and a multiplier 102c is connected to the output side of the cube circuit 101b. The output sides of the multipliers 102 a to 102 c are connected to the adder 103.

To the multiplier 102a, the input signal sequence _{x k} is inputted as it is. The squaring circuit 101a squared input signal sequence _{x k,} the non-linear signal input to the multiplier 102b, the third power circuit 101b is the cube of the input signal sequence _{x k,} enter the nonlinear signal to the multiplier 102c To do. Next, each of the multipliers 102 a to 102 c performs weighting by multiplying the individual weighting coefficients * b 0, * b 1, * b 2 by the input signal sequence x _k and each nonlinear signal individually, and these results are added by the adder 103. The result is added and output as an output signal series _yz . The output signal sequence y _z, the input signal sequence x _k are output one per input one.

In the non-linear FIR filter 100, estimated tap coefficients are used for the multipliers 102a to 102d as the weighting coefficients * b0 to * b2.
Here, in the non-linear FIR filter 100, in order to compensate for the waveform distortion of the transmission side optical transceiver, there is a just good weighting coefficient that can optimally remove the distortion of the signal. This coefficient will be described.

  For example, when the waveform of the received signal is distorted, the components of the main signal symbols before and after are slightly mixed on the time axis. As an example of this, if the signal (symbol) immediately before the current signal is mixed with the current signal in the main signal sequence, the previous signal corresponds to the nonlinear transfer function in the nonlinear FIR filter 100. The tap coefficient (weighting coefficient) is negative. If this is added to the current signal, the mixing can be eliminated and the nonlinear distortion component of the waveform distortion can be canceled. Thus, if the tap coefficient of the nonlinear FIR filter 100 is estimated and weighted, the main signal can be pre-equalized so as to cancel the nonlinear distortion of the main signal transmitted to the optical transceiver 15 on the other side. .

<Modification 3 of embodiment>
Next, Modification 3 of the present embodiment will be described. In Modification 3, as shown in FIG. 7, the transmission signal processing unit 40 includes a training signal sequence multiplexing unit (also referred to as a sequence multiplexing unit) 46, and the estimation unit 65 of the reception signal processing unit 60 includes a lookup table ( 65t).

The sequence multiplexing unit 46 multiplexes a known training signal sequence in which energy is concentrated in a specific frequency band, which will be described later, on the main signal and outputs the multiplexed signal to the multiplexing unit 42.
The storage table 65t stores the above-described linear transfer function in advance associated with the energy amount of the specific frequency band measured by the estimation unit 65 as described later. That is, a large and small linear transfer function corresponding to the energy amount is associated in advance with the measured energy amount.

  However, when energy is concentrated in the specific frequency band described above, in other words, energy is concentrated in a plurality of specific frequencies, the frequency f is plotted on the horizontal axis and the power Pw is plotted on the vertical axis, as shown in FIG. When taken, it means that the power component shown by the bar graph protrudes at the specific frequencies f1 and f2 on the f-axis. The power component is the peak frequency.

  More specifically, for example, −S, S, −S, S,... Of a BPSK (Binary Phase Shift Keying) signal is converted into a time waveform (= signal on the time axis of I, Q as shown in FIG. 8B. When the time waveforms S and -S are subjected to FFT (Fast Fourier Transform) and viewed in the frequency domain, the power components are obtained. This power component (peak frequency) is expressed as “the energy concentrates in a specific frequency band”.

  The feature of Modification 3 is that the estimation unit 65 of the optical transceiver 14 on one side shown in FIG. 7 separates the training signal sequence multiplexed with the main signal received from the optical transceiver 15 (FIG. 1) on the other side, The energy of a specific frequency band of the separated training signal sequence is measured (fourth estimation). Further, the estimation unit 65 searches and estimates the linear transfer function associated with the measured energy amount of the specific frequency band from the storage table 65 t and outputs the linear transfer function to the generation unit 43.

  Then, the generation unit 43 generates a control signal by encoding the estimated (fourth estimation) linear transfer function, and outputs the control signal to the multiplexing unit 42. The multiplexing unit 42 multiplexes the main signal on which the training signal sequence is multiplexed by the sequence multiplexing unit 46 and the control signal. The multiplexed signal is converted into an optical signal and transmitted to the optical transceiver 15 on the other side. I did it.

  Further, the optical transceiver 14 on the other side receives the optical signal obtained by multiplexing the optical signal converted by the linear transfer function estimated (fourth estimation) in the same manner as described above into the main signal. . Further, the decoding unit 63 decodes the linear transfer function from the received signal. The pre-equalization unit 44 performs pre-equalization with the decoded linear transfer function.

  Thus, when the energy of a specific frequency is measured, if the energy is reduced, it can be seen that the signal waveform at that frequency is distorted. Since this distorted signal is transmitted from the optical transceiver 15 on the other side, waveform distortion caused by nonlinearity and frequency characteristics of the analog electronic circuit on the other side occurs. Therefore, the linear transfer function associated with the decreased energy amount is transmitted to the other side, and the pre-equalization unit 44 on the other side performs pre-equalization using the linear transfer function as described above. By this pre-equalization, if the main signal is transmitted as an optical signal from the other side and received by the optical transceiver 14 on one side, the received main signal is in a state where the linear distortion component is reduced or absent.

<Modification 4 of embodiment>
Next, Modification 4 of the present embodiment will be described.
The feature of the modification 4 is that the estimation unit 65 shown in FIG. 2 first calculates an average value of error vectors (described later) of symbols of the main signal received from the optical transceiver 15 (FIG. 1) on the other side by the provisional determination process. To do.

  Here, for example, in the first quadrant on the IQ plane shown in FIG. 9, when the signal point g2 of the symbol received as described above is shifted from the original signal point g1, the arrow V1 from g1 to g2 is changed. It is called an error vector. When this error vector V1 is concentrated and distributed in a certain range (distribution is biased), for example, as indicated by a dashed ellipse 110 in the first quadrant, the main signal received at the broken ellipse 110 portion. It can be seen that the waveform distortion occurs. Therefore, the estimation unit 65 calculates an average value (referred to as an error vector average value) of the error vectors V1 group in which the distribution of the received signal symbols is biased.

  Further, the estimation unit 65 shown in FIG. 2 stores a storage table (not shown) in which a nonlinear transfer function corresponding to the error vector average value is associated in advance with the calculated error vector average value. 65t). The estimation unit 65 retrieves the nonlinear transfer function associated with the calculated error vector average value from the storage table, estimates (fifth estimation), and outputs it to the generation unit 43.

  The generation unit 43 encodes the estimated nonlinear transfer function (fifth estimation) to generate a control signal, and outputs the control signal to the multiplexing unit 42. The multiplexing unit 42 multiplexes the main signal and the control signal, and the multiplexed signal is converted into an optical signal and transmitted to the optical transceiver 15 on the other side.

  Also, in the optical transceiver 15 on the other side, the optical signal on which the nonlinear transfer function estimated (fifth estimation) is multiplexed into the main signal and optically converted in the same manner as described above is received by the optical transceiver 14 on the one side. . Further, the decoding unit 63 decodes the nonlinear transfer function from the received signal. The pre-equalization unit 44 performs pre-equalization on the main signal transmitted to the other side using the decoded linear transfer function.

  That is, the pre-equalization is performed with a nonlinear transfer function corresponding to the error vector average value indicating that waveform distortion has occurred. In other words, an inverse vector that cancels the error vector average value is generated, and this inverse vector is generated. Thus, pre-equalization is performed on the transmission main signal.

  Thus, the one-side optical transceiver 14 decodes the nonlinear transfer function corresponding to the error vector average value indicating the waveform distortion from the main signal received from the other-side optical transceiver 15. This decoded non-linear transfer function corresponds to non-linear distortion due to non-linearity of the analog electronic circuit on one side. Therefore, if one side performs pre-equalization on the main signal optically transmitted to the other side with the decoded nonlinear transfer function, the main signal received on the other side has reduced or no nonlinear distortion component. It becomes a state.

  According to the present embodiment and the first to fourth modifications described above, when optical signals are transmitted and received by the optical transceivers 14 and 15 on both sides via the optical fibers 11 and 12, the signal quality is improved and the bit error rate is improved. Can be achieved.

  In general, the signal quality decreases as the transmission rate of the optical signal increases, so that the energy that can be used per symbol decreases as the speed increases. For this reason, if the frequency characteristic of the optical transmission device is up to 32 GBaud, the signal quality can be maintained, but if it becomes 64 GBaud, the signal is distorted and the signal quality cannot be maintained.

  However, in the present invention, since the signal quality can be improved as described above, the throughput of optical data transmission, which is limited by the performance of the analog electronic circuits of the optical transceivers 14 and 15, such as bandwidth and linearity, is greatly increased. It is possible to improve. For example, the baud rate that has been about 32 GBaud in the past can be improved to, for example, 64 GBaud. As a result, a communication capacity of 254 Gbit / s, which is twice that of the prior art, can be ensured. Furthermore, it is possible to use the 16QAM system with respect to the conventional QPSK system, and the optical data transmission throughput can be greatly improved.

DESCRIPTION OF SYMBOLS 10 Optical data transmission system 11,12 Optical fiber 14,15 Optical transceiver 20 Optical transmitter 30 Optical receiver 40 Transmission signal processing part 42 Control signal multiplexing part (multiplexing part)
43 Control signal generator (generator)
44 Transmitting side pre-equalization unit (pre-equalization unit)
46 Training signal sequence multiplexing unit (sequence multiplexing unit)
60 reception signal processing unit 62 control signal separation unit 63 control signal decoding unit (decoding unit)
64 Receiver-side equalization unit 65 Linear / nonlinear transfer function estimation unit (estimation unit)
65t lookup table 50 E / O (electrical / optical) converter 70 O / E (optical / electrical) converter

Claims (4)

An optical transceiver comprising: an optical transmitter that transmits an optical signal whose polarization is modulated by a main signal as transmission data to an optical transmission line; and an optical receiver that receives the transmitted optical signal and demodulates the main signal Is an optical data transmission system that performs optical transmission by providing both ends of an optical transmission line,
The optical receiver is
An estimation unit that measures waveform distortion of the demodulated main signal after receiving an optical signal by an optical transceiver on one side, and estimates waveform distortion information related to the waveform distortion;
A decoding unit that decodes the waveform distortion information estimated in the same manner as in the estimation in the optical transceiver on the other side after receiving the optical signal from the optical transceiver on the other side;
With
The estimation unit measures nonlinear distortion of the demodulated main signal, estimates a nonlinear transfer function as the waveform distortion information from the measured nonlinear distortion, and receives the main signal received from the other optical transceiver Measure the energy of a specific frequency band of the training signal sequence multiplexed in the above, estimate a linear transfer function corresponding to the amount of energy of the measured specific frequency band,
The decoding unit decodes the linear transfer function and the nonlinear transfer function estimated by the estimation unit in the other side optical transceiver after receiving an optical signal from the other side optical transceiver,
The optical transmitter is
A generation unit that generates a control signal by encoding the linear transfer function and the nonlinear transfer function estimated by the estimation unit;
A sequence multiplexing unit that multiplexes a known training signal sequence in which energy is concentrated in a specific frequency band on the main signal;
A multiplexing unit that multiplexes the waveform distortion information estimated by the estimation unit and a main signal to be transmitted to the optical transceiver on the other side;
A pre-equalization unit that performs pre-equalization of the multiplexed main signal so as to cancel the waveform distortion indicated by the waveform distortion information decoded by the decoding unit;
A conversion unit that modulates polarization with a multiplexed signal including the pre-equalized main signal and the estimated waveform distortion information to convert the signal into an optical signal;
Equipped with a,
The generation unit generates a control signal by encoding the linear transfer function estimated by the estimation unit,
The multiplexing unit multiplexes the generated control signal and a main signal on which a training signal sequence is multiplexed in the sequence multiplexing unit that is transmitted to the optical transceiver on the other side,
The pre-equalization unit uses the linear transfer function and the nonlinear transfer function decoded by the decoding unit to cancel the waveform distortion of the optical signal received by the optical transceiver on the other side. An optical data transmission system for performing pre-equalization of signals . In the optical receiver,
The estimation unit obtains waveform distortion by comparing the main signal received from the optical transceiver on the other side with a known reference main signal by provisional determination processing, and adapts the linear distortion component of the waveform distortion to be minimum. the tap coefficient vector corresponding to said linear and estimated by algorithm,
Said decoding unit, the tap coefficient vector estimated as before Ki推 constant in the optical transceiver of the other side, and decoded after reception of the optical signal from the optical transceiver of the other side,
In the optical transmitter,
The generating unit, before encoding the Ki推 constant tap coefficients vector to generate a control signal,
The multiplexing unit multiplexes the generated control signal and the main signal to be transmitted to the optical transceiver on the other side,
The pre-equalization unit includes an FIR filter, and uses the tap coefficient vector decoded by the decoding unit as a weighting coefficient of the FIR filter so as to perform pre-equalization of the multiplexed main signal. The optical data transmission system according to claim 1 . In the optical receiver,
The estimation unit obtains a waveform distortion by taking a difference between a main signal received from the other optical transceiver and a known reference main signal, and an adaptive equalization algorithm so that a nonlinear distortion component of the waveform distortion is minimized. the tap coefficient vector corresponding to the non-linear transfer function to estimate the,
Said decoding unit, the tap coefficient vector estimated as before Ki推 constant in the optical transceiver of the other side, and decoded after reception of the optical signal from the optical transceiver of the other side,
In the optical transmitter,
The generating unit, before encoding the Ki推 constant tap coefficients vector to generate a control signal,
The multiplexing unit multiplexes the generated control signal and the main signal to be transmitted to the optical transceiver on the other side,
The pre-equalization unit includes a non-linear FIR filter, and uses the tap coefficient vector decoded by the decoding unit as a weighting coefficient of the non-linear FIR filter so as to pre-equalize the multiplexed main signal. The optical data transmission system according to claim 1 . In the optical receiver,
The estimating unit, the provisional decision process by calculating an average value of the error vector group distribution is biased symbols of the main signal received from the optical transceiver of the other side, the non-linear transfer function corresponding to the calculated average value estimated constant,
Said decoding unit, a non-linear transfer function estimated as before Ki推 constant in the optical transceiver of the other side, and decoded after reception of the optical signal from the optical transceiver of the other side,
In the optical transmitter,
The generating unit, before encoding the Ki推 constant nonlinear transfer function to generate a control signal,
The multiplexing unit multiplexes the generated control signal and the main signal to be transmitted to the optical transceiver on the other side,
The pre-equalization unit uses the nonlinear transfer function decoded by the decoding unit to cancel the waveform distortion of the optical signal received by the optical transceiver on the other side. The optical data transmission system according to claim 1 , wherein equalization is performed.

Images