JP2016208257A - Digital optical receiver and optical communication system employing the same - Google Patents

Abstract

PROBLEM TO BE SOLVED: To keep a quality of a reception signal high by digital signal processing at a reception side even in the case where constellation distortion caused by analog front end property deterioration at a transmission side is residual.SOLUTION: A reception demodulation processing unit comprises a plurality of demodulation parts and at least one or more constellation distortion correction parts. Each of the demodulation parts includes: a channel separation part which performs channel separation processing for polarization separation from a reception signal in which an X polarization signal and a Y polarization signal are mixed in a polarization multiplex QPSK system, into the X polarization signal and the Y polarization signal; and a carrier frequency offset compensation part which performs carrier frequency offset compensation processing for compensating a deviation between a carrier frequency of the reception signal and a local oscillation optical frequency of a local oscillation optical signal for the X polarization signal and the Y polarization signal. The at least one or more constellation distortion correction parts are provided on a post-stage of the channel separation part and the carrier frequency offset compensation part and performs constellation distortion correction processing for correcting the constellation distortion of an inputted signal.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a digital coherent digital optical receiver and an optical communication system using the same.

  With the explosive demand for broadband multimedia communication services in the Internet and video distribution, etc., long-distance, large-capacity and highly reliable optical fiber communication systems are being introduced.

  In an optical fiber communication system, it is important to reduce the installation cost of an optical fiber that is an optical transmission path and to improve the transmission band utilization efficiency per optical fiber. Therefore, in recent years, digital coherent optical communication technology using a digital optical transceiver has attracted attention and its importance is increasing.

  In an optical communication system, an analog optical transceiver using a modulation method such as OOK (On-Off Keying), which has been widely applied in general, has been used. Note that compared to analog optical transceivers, digital coherent optical transceivers can transmit distortion on the transmission side or on the reception side even if they include waveform distortion caused by wavelength dispersion in the transmission path or imperfections in the transceiver. Compensation can be made by performing digital signal processing (DSP) on the side.

  With such a technique, the performance of the optical communication device is improved and the cost can be reduced, and a large-capacity backbone optical communication system is becoming widespread.

  In particular, by making use of the advantages of digitalization, multilevel modulation techniques such as QPSK (Quadrature Phase Shift Keying) modulation and QAM (Quadrature Amplitude Modulation) modulation, which have been difficult to realize by analog processing, can be applied. In addition, a technique of increasing the number of wavelength multiplexed channels by narrowing the signal band using Nyquist filtering can be applied. Thereby, the transmission band utilization efficiency is improved and the transmission capacity per optical fiber can be expanded.

  On the other hand, as advanced transmission schemes such as the multi-level modulation technique and the Nyquist filtering technique are applied, the signal waveform becomes more complicated and high signal accuracy is required. For this reason, the characteristic deterioration of the analog front end part constituting the transmitter / receiver such as the transmission light modulator and the coherent detection receiver may significantly affect the transmission characteristic.

  FIG. 13 is a block diagram showing a general configuration of a digital coherent optical communication system. The cause of the characteristic deterioration of the analog front end unit described above is that the DC offset and I / Q imbalance (In-phase signal and Quadrature-phase signal generated in both the transmission analog front end 101 and the reception analog front end 102 are generated. Orthogonal deviation), nonlinear characteristics, extinction ratio occurring in the transmission analog front end, and the like. Since these characteristic deteriorations cause distortion such as deviation from the ideal constellation position and deformation, transmission characteristics are greatly deteriorated.

  FIG. 14 is a diagram showing a specific example of constellation distortion, where (a) is an ideal constellation, (b) is extinction ratio degradation, (c) is a DC offset, (d) is an I / Q imbalance, (e ) Shows the case of linearity degradation. Note that the constellation defines the arrangement of signal points indicating a combination of phase and / or amplitude of an in-phase channel (I channel) and a quadrature channel (Q channel) in a digital quadrature modulation scheme such as QPSK or 16QAM. .

  As a specific example of the analog characteristic degradation occurring on the transmission side, the imperfection of the transmission I / Q optical modulator, the drive circuit, and the bias control circuit, which are mainly combined with the MZ type optical modulator, can be exemplified. On the other hand, as a specific example of analog characteristic degradation occurring on the receiving side, incompleteness such as a 90 deg hybrid mixer that performs coherent detection and a TIA (Trans-Impedance Amplifier) can be exemplified.

  In contrast to such analog characteristic degradation, until now, it has been common to compensate for characteristic degradation that has occurred on the transmission side on the transmission side, and characteristic degradation that has occurred on the reception side on the reception side.

  For example, Japanese Patent Application Laid-Open No. 2014-146915 discloses a transmission pre-compensation technique that compensates for static analog characteristic degradation (I / Q imbalance) occurring in a transmission-side optical modulator in advance by transmission-side digital signal processing. Has been. Japanese Patent Application Laid-Open No. 2014-165913 discloses a technique for similarly controlling a transmission optical modulator with high accuracy to suppress degradation of analog characteristics caused by bias deviation. Furthermore, Japanese Patent Application Laid-Open No. 2013-528995 discloses a technique for compensating analog characteristics generated by a 90 deg hybrid mixer on the reception side by digital signal processing on the reception side.

JP 2014-146915 A JP 2014-165913 A JP 2013-52895 A

  However, in the compensation techniques according to Patent Documents 1 to 3 described above, when the signal waveform becomes more complicated as advanced transmission methods such as multi-level modulation and Nyquist transmission proceed for further increase in capacity in the future. There is a problem that it is difficult to completely compensate for the distortion of the constellation caused by the transmission / reception analog characteristics. In particular, the pre-compensation technique and the bias control technique on the transmission side have a large part depending on the complexity of the waveform, so that it is difficult to perform highly accurate compensation.

  In addition, the analog characteristic deterioration caused by the transmission optical modulator is gradually but dynamically changed, so that there is a problem that the characteristic cannot be completely compensated and constellation distortion remains.

  This residual constellation distortion on the transmission side can be ignored in the QPSK system, which is a currently popular modulation system, with little effect on the transmission characteristics, but in a higher multilevel system such as the 8QAM and 16QAM systems. There is a problem that transmission characteristic deterioration becomes remarkable and cannot be ignored.

  For example, let us consider a case in which residual constellation distortion on the transmission side is to be removed with higher accuracy using the technology according to Japanese Patent Application Laid-Open Nos. 2014-146915 and 2014-165913. In this case, it is necessary to perform feedback control with higher accuracy for the transmission digital signal section and the optical modulator bias circuit. As a result, a mechanism such as a high-precision real-time waveform monitor of the same scale as that of the receiving side must be introduced on the transmitting side, resulting in a significant increase in circuit scale and cost.

  Japanese Patent Laid-Open No. 2013-528995 only compensates for constellation distortion generated in the 90-deg hybrid mixer on the reception side, and cannot compensate for residual constellation distortion generated on the transmission side.

  Therefore, the main object of the present invention is to maintain the quality of the received signal by the digital signal processing on the receiving side even when the constellation distortion due to the deterioration of the analog front end characteristic on the transmitting side remains. A digital optical receiver and an optical communication system using the same are provided.

  In order to achieve the above object, a digital optical receiver according to the present invention includes a coherent detection receiver that performs detection by mixing a received signal and a local light signal, and an ADC that converts the detected received analog signal into a digital signal. And a reception demodulation processing unit for demodulating the digital reception signal, wherein the reception demodulation processing unit includes a plurality of demodulation units, at least one constellation distortion correction unit, and , And the demodulator performs channel separation to separate the polarization signal into the X polarization signal and the Y polarization signal from the reception signal in which the X polarization signal and the Y polarization signal are mixed in the polarization multiplexing QPSK system. A carrier that compensates for a deviation between a carrier frequency of a received signal and a local light emission frequency of a local light signal with respect to a channel separation unit that performs processing and an X polarization signal and a Y polarization signal A carrier frequency offset compensation unit that performs wave number offset compensation processing, and at least one constellation distortion correction unit is provided at the subsequent stage of the channel separation unit and the carrier frequency offset compensation unit, and the constellation of the input signal is provided. A constellation distortion correction process is performed to correct the constellation distortion.

  According to the present invention, even when constellation distortion due to analog front-end characteristic degradation on the transmission side remains, the quality of the reception signal is suitably maintained by the digital signal processing on the reception side, and the system performance is maintained. Can do.

1 is a block diagram of a digital optical receiver in a digital coherent optical communication system according to a first embodiment. It is a block diagram of a specific digital optical receiver. It is a block diagram of the constellation distortion correction part which performs DC offset compensation processing. It is a block diagram of the constellation distortion correction | amendment part which performs an I / Q imbalance compensation process. It is a block diagram of the constellation distortion correction | amendment part which performs an extinction ratio compensation process. It is a block diagram of the constellation distortion correction | amendment part which performs the extinction ratio compensation process using LUT. It is a block diagram of the constellation distortion correction | amendment part which performs a nonlinearity compensation process. It is a block diagram of the constellation distortion correction | amendment part which performs the nonlinearity compensation process using LUT. FIG. 2 is a configuration diagram of an MZ type I channel optical modulator constituting an I / Q optical modulator, where (a) is a block diagram of the MZ type I channel optical modulator, and (b) is an upper phase modulator and a lower phase modulator; The figure which shows the output signal of a modulator, (c) is a figure which shows the output signal of the optical modulator for MZ type | mold I channels. It is the figure which illustrated the straight line approximation characteristic of the LUT part provided with LUT. FIG. 5 is a diagram in which a plurality of carrier frequency offset compensators for performing coarse adjustment / fine adjustment functions are provided. It is a block diagram of the digital optical receiver concerning 2nd Embodiment. 1 is a block diagram of a general polarization multiplexing type digital coherent optical communication system applied to a description of related technology. FIG. It is a figure which shows the specific example of a constellation distortion, (a) is ideal constellation, (b) is extinction ratio degradation, (c) is DC offset, (d) is I / Q imbalance, (e) is linearity It is a figure which shows the case of deterioration.

<First Embodiment>
A first embodiment of the present invention will be described. FIG. 1 is a block diagram of a digital optical receiver 2 in the digital coherent optical communication system according to the present embodiment.

  The digital optical receiver 2 according to the present embodiment includes a reception analog front end unit 10 and a reception demodulation processing unit 20.

  The reception analog front end unit 10 includes a coherent detection receiver 11 that performs detection by mixing a reception signal and a signal by local light (local light signal), and an ADC (Analog−) that converts the detected reception analog signal into a digital signal. to-Digital Converter) 12.

  The reception demodulation processing unit 20 includes a first to m-th demodulation unit 21 (21a to 21m), a channel separation unit 22, a first to n-th constellation distortion correction unit 23 (23a, 23b, ..., 23n), a carrier frequency. The digital signal received from the reception analog front end unit 10 is demodulated including the offset compensation unit 24.

  At least one constellation distortion correction unit 23 is provided in the subsequent stage of the channel separation unit 22 and the carrier frequency offset compensation unit 24. That is, one or more constellation distortion correction units 23 are provided in the subsequent stage (output side) of the channel separation unit 22 and / or one or more constellations in the subsequent stage (output side) of the carrier frequency offset compensation unit 24. A distortion correction unit 23 is provided.

  At this time, the number of demodulation units 21 and constellation distortion correction units 23 is not limited, and each demodulation unit 21 and constellation distortion correction unit 23 does not exclude performing different processes. Further, each demodulation processing procedure including the processing of the channel separation unit 22 and the carrier frequency offset compensation unit 24 is not limited. For example, the demodulation processing may be performed in the order of the frequency offset compensation unit 24 and the channel separation unit 22.

  Hereinafter, in order to simplify the description, a digital optical receiving apparatus in which the configuration according to the present embodiment is applied to the general polarization multiplexing digital coherent optical communication system shown in FIG. 13 will be described as an example. FIG. 2 is a block diagram of the digital optical receiver 2. The first demodulator 21a performs chromatic dispersion compensation processing, and the second demodulator 21b includes three-stage constellation distortion correction units 23a to 23c as an example of performing carrier phase compensation processing. Details of each process will be described.

  The first demodulator 21a performs chromatic dispersion compensation processing for compensating for the chromatic dispersion effect generated in the optical fiber that is the transmission path.

  The channel separation unit 22 converts the X polarization signal in the polarization multiplexed signal and the Y polarization signal orthogonal to the X polarization signal into a polarized wave from the X polarization signal to the Y polarization signal. A channel separation process is performed. The channel separation unit 22 performs adaptive equalization processing by performing MIMO (Multi-input Multi-output) processing using a FIR filter of several tens of taps. At this time, compensation for polarization mode dispersion and the like is performed at the same time. Done.

  The carrier frequency offset compensator 24 compensates for the deviation between the carrier frequency of the received signal and the frequency of the local light signal (local light frequency) for the X polarization signal and the Y polarization signal subjected to polarization separation. As a result, the carrier frequency offset is removed on average, and the rotation of the constellation stops in time.

  The second demodulator 21b performs carrier phase compensation processing on the input signal, so that the constellation is corrected to a desired position as shown in the ideal constellation diagram of FIG.

  Each constellation distortion correction unit 23 performs constellation distortion correction processing (DC offset compensation processing, I / Q imbalance compensation processing, extinction ratio compensation processing, LUT (LUT)) on the input X polarization signal and Y polarization signal. At least one of extinction ratio compensation processing using a Look-Up Table, nonlinear compensation processing, and nonlinear compensation processing using an LUT is performed, thereby correcting constellation distortion of signals in the I channel and the Q channel ( Constellation distortion correction processing).

  3 to 8 are diagrams illustrating a specific configuration of the constellation distortion correction unit 23. 3 shows a case where DC offset compensation processing is performed, FIG. 4 shows a case where I / Q imbalance compensation processing is performed, FIG. 5 shows a case where extinction ratio compensation processing is performed, and FIG. 6 shows a case where extinction ratio compensation processing using an LUT is performed. 7 shows a case where nonlinear compensation processing is performed, and FIG. 8 shows a case where nonlinear compensation processing using an LUT is performed.

  The constellation distortion correction unit 23 shown in FIG. 3 subtracts the detected DC component from the input signal, and a DC component detection unit 30 (30a, 30b) that detects the DC component of each signal (for example, the average value of the input signal). And a subtractor 31 (31a, 31b) for removing the DC component.

  This makes it possible to correct the constellation deviation from the origin, and to correct constellation residual distortion due to imperfect bias control and extinction ratio in the optical modulator as shown in FIG. It becomes possible.

  The constellation distortion correcting unit 23 shown in FIG. 4 multiplies the input I channel signal by A and B, and multiplies the input Q channel signal by C and D (a first amplifier 33a to a fourth amplifier 33d). Further, an adder 34 (a first adder 34a and a second adder 34b) for adding a part of the amplified I channel signal and Q channel signal to the other channel is included.

  That is, the constellation distortion correction unit 23 first and second sets the I / Q signal of X / Y polarization to a value obtained by branching the I channel signal into two and multiplying each by a preset value. Two amplifiers 33a and 33b, third and fourth amplifiers 33c and 33d, each of which is divided into two by multiplying a Q channel signal by a preset value, and a first amplifier 33a branched from an I channel signal. The first adder 34a for adding and outputting the output of the third amplifier 33c branched from the output and the Q channel signal, and the output of the second amplifier 33b branched from the I channel signal and the second of the Q channel signal. And a second adder 34b that adds and outputs the outputs of the four amplifiers 33d.

In other words, the constellation distortion correction unit 23 shown in FIG.

The arithmetic processing represented by

  Note that the magnifications A, B, C, and D are preset values, and are determined according to the magnitude of the I / Q imbalance. Essentially, the I channel signal and the Q channel signal should be orthogonal to each other.

However, for example, when the angle of the I channel is deviated by θ _I and the angle of the Q channel is deviated from orthogonal by θ _Q due to imperfect bias of the transmission optical modulator or the like, it was affected by I / Q imbalance. I channel signal I and Q channel signal Q, the ideal I / Q-channel signals _I ideal _{/ Q ideal} using equation (2)

Can be expressed by multiplying the matrix shown in FIG.

This distorts the constellation into a rhombus. However, the constellation distorted in this way can be easily compensated by multiplying the inverse matrix of equation (2). To simplify the explanation, consider a case where an I / Q imbalance of θ _I = −θ and θ _Q = θ is generated. At this time, the inverse matrix of Expression (2) is expressed by Expression (3).

It is expressed.

Therefore, the I / Q imbalance can be compensated by setting the matrix represented by the magnifications A, B, C, and D to Equation (3) obtained above. That is, the I channel signal and the Q channel signal affected by the I / Q imbalance represented by Expression (2) are substituted into Expression (1), and the matrix defined by Expression (3) is further expressed by Expression (1). ) Of A, B, C, and D.
As a result, the equation (4)

As shown in FIG. 4, the influence of the I / Q imbalance is compensated, and an ideal I / Q channel signal (I _Ideal / Q _Ideal ) can be restored.

  As described above, the constellation distortion correction unit 23 shown in FIG. 4 performs the I / Q imbalance (In-phase signal and Quadrature due to imperfect bias control of the optical modulator as shown in FIG. The constellation residual distortion due to -phase signal orthogonal deviation) can be corrected.

  The constellation distortion correction unit 23 shown in FIG. 5 is a signal processing unit 36 (first signal processing unit) that performs processing using known transfer functions FI (x) and FQ (x) on the I channel signal and the Q channel signal. 36a, a second signal processing unit 36b), and an adder 37 (first adder 37a, second adder 37b) for adding to the other channel.

  That is, the constellation distortion correction unit 23 branches the I channel signal for each I / Q signal of X / Y polarization, and performs signal processing using a preset transfer function. The second signal processing unit 36b that branches the Q channel signal and performs signal processing using a preset transfer function, and the output of the second signal processing unit 36b branched from the I channel signal and the Q channel are added. And a second adder 37b that adds and outputs the Q channel signal and the output of the first signal processing unit 36a branched from the I channel.

  At this time, the transfer functions FI (x) and FQ (x) are selected so as to have characteristics opposite to the constellation residual strain due to the extinction ratio deterioration. Here, the bow-shaped constellation distortion due to the extinction ratio degradation shown in FIG. 14B is caused by imperfection of the interferometer constituting the MZ type optical modulator, and the magnitude of the I channel signal is large. This means that it leaks in an arcuate shape on the Q channel side. Similarly, it means that the Q channel signal leaks in an arcuate shape on the I channel side according to its magnitude.

  FIG. 9 illustrates the configuration of the MZ type I-channel optical modulator 50 constituting the I / Q optical modulator. The MZ type optical modulator 50 includes an upper phase modulator 51a (51) and a lower phase modulator 51b (51). Since the Q channel has the same configuration, in the following description, the I channel optical modulator 50 will be described as an example. 9A is a block diagram of the MZ type I channel optical modulator 50, and FIG. 9B shows output signals of the upper phase modulator 51a (51) and the lower phase modulator 51b (51). FIG. 9C is a diagram showing an output signal of the MZ type I channel optical modulator 50.

An I _in optical signal exp (jωt) (j is an imaginary unit, ω is an optical signal frequency) input to the MZ type optical modulator 50 is branched into two optical signals to be divided into an upper phase modulator 51a and a lower phase. The signal is input to the modulator 51b. The electric field strength of the optical signal input to the upper phase modulator 51a is given by A ₊ exp (jωt), and the electric field strength of the optical signal input to the lower phase modulator 51b is given by A ₋ exp (jωt).

At this time, the amount of phase rotation applied in the upper phase modulator 51a is exp (jπV / 2V _π), the amount of phase rotation is applied in the lower phase modulator 51b is given by exp (-jπV / 2V _π). Here, V is a driving voltage for driving the upper phase modulator 51a and the lower phase modulator 51b, the V _[pi is the applied voltage phase rotation amount becomes [pi.

In the upper phase modulator 51a and the lower phase modulator 51b, the optical signals E ₊ and E ₋ to which phase modulation is applied are combined to become E _out (= E ₊ + E ₋₎ , and the MZ type optical modulator. 50.

First, the case where A ₊ ≠ A ₋ will be described. FIG. 9B is a graph showing E ₊ and E ₋ when A ₊ ≠ A ₋ . In FIG. 9 (b), the solid line _{E + = A + exp (iπV} / (2V π)) dotted line _E - ₌ A - represents _{exp (-iπV / (2V π)} ). Further, FIG. 9C is a graph of E _out = E ₊ + E ₋ .

As can be seen from FIGS. 9A to 9C, when A ₊ ≠ A ₋ , Q components that should cancel each other are generated when ideal A ₊ = A ₋ , and phase error Err occurs. To do. It is known from a general theoretical calculation that this phase error Err is inversely proportional (Err∝ER ⁻¹ ) to an extinction ratio (ER) = (A ₊ + A ₋ ) / (A ₊ −A ₋ ). Yes. In other words, the amount that leaks from the I channel to the Q channel according to the extinction ratio is the phase error Err itself, and can be expressed by a transfer function. In this way, signals that leak into each other's channels can be represented by a transfer function.

  Therefore, if the inverse characteristic is set as FI (x) and subtracted from the Q channel signal as shown in FIG. 5, the component leaking from the I channel to the Q channel can be canceled out. The same applies to the Q channel.

  In this way, in the configuration of FIG. 5, by setting FI (x) and FQ (x), which have characteristics opposite to the constellation residual distortion, according to the extinction ratios of the optical modulators of the I / Q channels, respectively. The mutual leakage signals can be canceled out. This makes it possible to correct the bow-shaped constellation residual distortion due to the extinction ratio deterioration of the transmission light modulator as shown in FIG.

  Further, FI (x) and FQ (x) do not need to have a highly accurate inverse characteristic that completely cancels the constellation residual distortion, and may be approximate functions, for example. This makes it possible to simplify the arithmetic processing while allowing some errors.

  Further, as shown in FIG. 6, the LUT unit 55 (first LUT unit 55a, second LUT unit 55b) including an LUT may be used instead of FI (x) and FQ (x).

  That is, the constellation distortion correction unit 23 branches the I channel signal for each I / Q signal of X / Y polarization, and corrects the channel signal by a preset LUT, A first addition is performed by adding the output of the second LUT unit 55b that branches the Q channel signal and corrects the channel signal using a preset LUT and the output of the I channel signal and the second LUT unit 55b branched from the Q channel. 37a, and a second adder 3b that adds and outputs the Q channel signal and the output of the first LUT unit 55a branched from the I channel.

  In this case, for example, as shown in FIG. 10, the calculation amount can be greatly reduced by approximating FI (x) or FQ (x) by linear approximation defined by a multi-coordinate LUT.

  The constellation distortion correction unit 23 shown in FIG. 7 performs nonlinear compensation processing for compensating nonlinear characteristics generated in the transmission optical modulator and its drive circuit.

  The constellation distortion correction unit 23 includes a non-linear transfer function unit 38 (38a, 38b) that performs processing by a known non-linear transfer function NLI (x), NLQ (x) on the I channel signal and the Q channel signal. Yes.

  At this time, the nonlinear transfer functions NLI (x) and NLQ (x) are selected so as to have characteristics opposite to the residual strain due to the nonlinear characteristics. Here, an example of the nonlinear characteristic is the extinction characteristic of the transmission light modulator.

  As shown in FIG. 9A, the MZ type optical modulator 50 operates in principle by interfering with the optical signals phase-modulated by the upper and lower two optical phase modulators 51.

For this reason, it is known that the output light E _out does not operate linearly with respect to the drive voltage of the optical modulator and has sin characteristics of a trigonometric function. Therefore, in general, a QAM signal riding on an equidistant lattice can be obtained by driving the optical modulator after correcting by the arcsin (= sin ⁻¹ ) characteristic which is the reverse characteristic.

  However, due to imperfect bias of the optical modulator, there is a difference between the correction by arcsin and the sin characteristic of the optical modulator, which cannot be completely canceled and the linearity is impaired. That is, this non-linear characteristic is caused by a difference between the arcsin characteristic and the sin characteristic, and can be calculated theoretically. Therefore, nonlinearity can be compensated by setting the inverse characteristic to NLI (x) and NLQ (x).

  Another example of the non-linear characteristic is that generated in a drive circuit (linear amplifier) that drives the optical modulator. In a transistor differential amplifier constituting a general linear amplifier, input / output transfer characteristics have characteristics approximated by a tanh characteristic of a hyperbolic function.

For this reason, in a region where the input voltage is small, it can be regarded as a linear operation, but the linearity is lost as the input voltage increases. Therefore, similarly to the nonlinearity due to the optical modulator, the nonlinearity of the drive circuit can be compensated by setting the arctanh (= tanh ⁻¹ ) characteristic to NLI (x) and NLQ (x).

  It is also possible to simultaneously compensate for both characteristics by setting the transfer function obtained by multiplying the inverse characteristics of the optical modulator and the drive circuit to NLI (x) and NLQ (x).

  As described above, it is possible to correct the constellation residual distortion at unequal intervals due to the nonlinear characteristics as shown in FIG.

  The constellation distortion correction unit 23 shown in FIG. 8 includes an LUT correction unit 39 (39a, 39b) having a preset LUT (Look-Up Table).

  The LUT correction unit 39 corrects the I channel signal and the Q channel signal according to the LUT. At this time, the LUT is appropriately selected so as to have a characteristic opposite to the characteristic of the constellation residual strain. As a result, various constellation residual distortions can be corrected.

  In the constellation distortion correction shown in FIG. 7 described above, the transfer functions NLI (x) and NLQ (x) are set and corrected. In order to calculate these transfer functions with high accuracy, The circuit scale tends to increase.

  Therefore, there is a method using a certain degree of approximation function, but there is an advantage that the amount of calculation can be greatly reduced by performing correction using the LUT shown in FIG.

  Furthermore, the correction using the LUT is effective even when it cannot be approximated by a transfer function because it can be flexibly handled by simply setting the LUT appropriately for all transfer characteristics.

  Each constellation distortion correction unit 23 described above does not need to be configured by only a single process, and a plurality of correction processes are combined as long as the signal is optimally corrected according to the type of residual distortion. May be combined.

  With the above configuration, the I-channel signal and the Q-channel signal of each polarization are separated into the orthogonal coordinate system, and the demodulation process is completed.

  Note that the I / Q imbalance compensation technique proposed in Patent Document 3 is performed immediately after the ADC that converts the received analog signal into a digital signal. Therefore, constellation distortion caused by deterioration of analog front end characteristics such as a coherent detection receiver or ADC on the receiving side is compensated.

  However, even with this method, since the compensation is performed on the signal input to the demodulator 21a and the like, the residual distortion of the constellation applied by the optical modulator on the transmission side cannot be corrected.

  This is because the signal output from the ADC 12 is a signal before channel separation, and the residual distortion applied to each channel of the transmission signal is also in a complicated state. For this reason, the residual distortion of the constellation cannot be corrected correctly.

  On the other hand, in the digital optical receiver 2 according to the present embodiment, the constellation residual distortion applied to the transmission signal is also separated by the channel separation unit 22, so that the constellation distortion correction unit 23 applies to each polarization signal. Residual distortion can be appropriately corrected.

  Further, since the I / Q signal is also separated in the signal after the carrier phase compensation in the carrier carrier frequency offset compensating unit 24 and the second demodulating unit 21b, it is possible to appropriately correct each I / Q signal. Become.

  The constellation correction process may be performed collectively after the second demodulator 21b that has performed all the demodulation processes. However, if the constellation distortion that can be appropriately corrected is corrected as early as possible in the demodulation process, the demodulation process can be performed with higher accuracy. That is, it is better to perform the demodulation process in a state where residual distortion is removed as much as possible, rather than performing the demodulation process in a state including constellation distortion. For this reason, in this embodiment, as shown in FIG. 1, a plurality of constellation distortion correction units 23 are arranged between the demodulation processing units 21.

  In FIG. 1, a carrier frequency offset compensator 24 is arranged after the channel separator 22. However, it is not limited to such a configuration. For example, as shown in FIG. 11, carrier frequency offset compensation units 25 that perform a coarse adjustment / fine adjustment function may be provided at a plurality of positions. FIG. 11 is a diagram illustrating a case where the carrier frequency offset compensator 25 is provided in front of the channel separator 22.

  Also in this case, since it is preferable to perform the constellation distortion correction process after the carrier frequency offset compensation unit 25, the constellation distortion correction 23d is provided between the carrier frequency offset compensation unit 25 and the first demodulation unit 21a. As a result, it is possible to suppress carrier leakage caused by imperfect bias control of the transmission light modulator, extinction ratio, or the like.

  As described above, the constellation residual distortion generated on the transmission side can be compensated for in the demodulation processing on the reception side. As a result, the quality of the received signal can be suitably maintained and the system performance can be maintained without using an optimum control technique using a large-scale real-time waveform monitor on the transmission side.

Second Embodiment
Next, a second embodiment of the present invention will be described. In addition, about the same structure as 1st Embodiment, description is abbreviate | omitted suitably using the same code | symbol.

  As shown in FIG. 12, the digital optical receiver 2 according to the present embodiment has a plurality of constellation distortion detectors 41 (41a, 41b, 41c) in the configuration of the digital optical receiver 2 shown in the first embodiment. The constellation distortion correction control unit 42 is added.

  Note that the constellation distortion detection unit may be arranged at any place where the detection can be performed most efficiently, and may be arranged at an appropriate place according to the type of constellation residual distortion to be detected.

  The constellation distortion detection unit 41 detects the degree of distortion of the constellation of the signal input to the constellation distortion correction unit 23.

  The detected distortion degree of the constellation is input to the constellation distortion correction control unit 42. The constellation distortion correction control unit 42 controls the correction amount in each constellation distortion correction unit 23 based on the degree of distortion of the constellation. Also, if the degree of distortion of the constellation is sufficiently small, it is possible to bypass the unnecessary constellation correction unit and suppress power consumption.

  As a result, even when dynamic constellation residual distortion occurs, the constellation residual distortion is automatically compensated, and the received signal constellation can be suitably maintained and system performance can be maintained.

  As described above, the constellation residual distortion generated on the transmission side can be reduced on the reception side even when a transmission method using a multi-value modulation signal such as QAM or a more complicated transmission waveform such as Nyquist filtering is applied. It is possible to compensate in the demodulation process. As a result, the quality of the received signal can be suitably maintained and the system performance can be maintained without using an optimal control technique using a large-scale real-time waveform monitor on the transmission side.

  In addition, since the performance required for the optical modulator and the analog front-end device can be relaxed, the yield of used parts can be improved, and a digital optical transmitter can be provided at low cost.

2 Digital Optical Receiver 10 Reception Analog Front End Unit 11 Coherent Detection Receiver 20 Reception Demodulation Processing Unit 21 (21a-21m) Demodulation Unit 22 Channel Separation Unit 23 (23a-23n) Constellation Distortion Correction Unit 24, 25 Carrier Frequency Offset Compensator 30 (30a, 30b) DC component detector 31 (31a, 31b) Subtractor 33 (33a, 33b, 33c, 33d) Amplifier 34 (34a, 34b), 37 (37a, 37b) Adder 36 (36a, 36b) Transfer function unit 38 (38a, 38b) Nonlinear transfer function unit 41 (41a, 41b, 41c) Constellation distortion detection unit 42 Constellation distortion correction control unit 51 Optical phase modulator 51a Upper phase modulator 51b Lower phase modulation vessel

Claims (9)

A coherent detection receiver that performs detection by mixing a received signal and a local light signal, an ADC that converts the detected received analog signal into a digital signal, and a reception demodulation processing unit for demodulating the digital received signal A digital optical receiver comprising:
The reception demodulation processing unit
A plurality of demodulation units;
And at least one constellation distortion correction unit, and
The demodulator
A channel separation unit for performing a channel separation process for separating the X-polarized signal and the Y-polarized signal from the received signal obtained by mixing the X-polarized signal and the Y-polarized signal in a polarization multiplexing QPSK system; ,
A carrier frequency offset compensation unit that performs a carrier frequency offset compensation process for compensating a deviation between a carrier frequency of the reception signal and a local light emission frequency of the local light signal, with respect to the X polarization signal and the Y polarization signal; , Including
The constellation distortion correction unit is provided at least one after the channel separation unit and the carrier frequency offset compensation unit, and performs constellation distortion correction processing for correcting the constellation distortion of the input signal. A digital optical receiver characterized. The digital optical receiver according to claim 1,
The constellation distortion correction unit is
A direct current component detector that detects a direct current component of an input signal for each I / Q signal of X / Y polarization;
A subtractor for subtracting the detected DC component from the input signal;
A digital optical receiver characterized in that the DC component is removed. The digital optical receiver according to claim 1,
The constellation distortion correction unit is configured for each I / Q signal of X / Y polarization.
First and second amplifiers that branch the I channel signal into two and multiply each by a preset value, and a third amplifier that splits the Q channel signal into two and multiply each by a preset value A fourth amplifier;
A first adder for adding and outputting the output of the first amplifier branched from the I channel signal and the output of the third amplifier branched from the Q channel signal; and the second amplifier branched from the I channel signal. And a second adder for adding and outputting the output of the fourth amplifier branched from the Q channel signal;
A digital optical receiver characterized by performing I / Q imbalance compensation. The digital optical receiver according to claim 1,
The constellation distortion correction unit is configured for each I / Q signal of X / Y polarization.
A first signal processing unit for branching an I channel signal and performing signal processing with a preset transfer function; a second signal processing unit for branching a Q channel signal and performing signal processing with a preset transfer function;
A first adder for adding and outputting an I channel signal and an output of the second signal processing unit branched from the Q channel; and an output of the first signal processing unit branched from the Q channel signal and the I channel A second adder for adding and outputting,
And a constellation residual distortion caused by an extinction ratio generated in the light modulation means. The digital optical receiver according to claim 1,
The constellation distortion correction unit is configured for each I / Q signal of X / Y polarization.
A first LUT unit that branches the I channel signal and corrects the channel signal using a preset LUT (and a second LUT unit that branches the Q channel signal and corrects the channel signal using the preset LUT;
A first adder that adds and outputs the I channel signal and the output of the second LUT unit branched from the Q channel, and adds the Q channel signal and the output of the first LUT unit branched from the I channel. A second adder for outputting;
And a constellation residual distortion caused by an extinction ratio generated in the light modulation means. The digital optical receiver according to claim 1,
The constellation distortion correction unit is configured for each I / Q signal of X / Y polarization.
A signal processing unit that performs signal processing using a preset transfer function;
A digital optical receiver characterized by compensating for non-linear characteristics generated in a transmission light modulation means and a drive circuit of the transmission light modulation means. The digital optical receiver according to claim 1,
The constellation distortion correction unit is configured for each I / Q signal of X / Y polarization.
A digital optical receiver characterized in that a channel signal is corrected by an LUT. The digital optical receiver according to any one of claims 1 to 7,
At least one constellation distortion detector that detects the degree of distortion of the constellation of the input signal;
A constellation distortion correction control unit that controls a compensation amount of the constellation distortion correction unit according to a detection result by the constellation distortion detection unit;
A digital optical receiver characterized by adaptively compensating for dynamic constellation residual distortion.   An optical communication system using the digital optical receiver according to any one of claims 1 to 8.

Images