JP2015109576A - Photoreceiver and optical signal receiving method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect the deviation of an interference phase of two orthogonal delay detector from an optimal value with one simple detection mechanism and to optimize a phase amount of a delay time and orthogonality of the two delay detector.SOLUTION: In the detection mechanism, a phase amount φ of an inclined axis in a signal point arrangement and a ratio R of an amplitude in a direction of the inclined axis in the signal point arrangement and an amplitude in a direction at 90° to the inclined axis are detected and a simple matrix operation created by using them is applied to a digital signal. Therefore, the digital signal is reproduced in an ideal signal point arrangement in which a phase of at least one delay detector is optimized and orthogonality of two delay detectors is optimized.

Description

The present invention relates to an optical receiver and an optical signal receiving method, and more particularly to an optical receiver and an optical signal receiving method suitable for transmitting and receiving an optical signal transmitted through an optical fiber.

  Internet traffic is increasing year by year due to the spread of large-capacity content such as videos and social networking services including Twitter and Facebook. As a result, there is a demand for an increase in capacity of an optical transmission system that performs data communication via an optical fiber. In such a background, in a long distance network such as a core / metro network connecting between cities or in cities, an optical transmission system that performs 10 Gbit / s information transmission at one wavelength is currently operated in a wide range.

  In addition, as an optical modulation / demodulation method for next-generation long-distance networks, a polarization multiplexed QPSK (Quadrature Phase Shift Keying) digital coherent method has attracted the interest of many research institutions.

  An optical receiver using digital coherent detection will be briefly described. The coherent optical receiver uses a local laser light source having the same wavelength as the input optical signal as a reference for the optical phase. The input optical signal and the local light output from the local laser light source are input to an optical 90-degree hybrid in which the two lights interfere. The optical 90-degree hybrid combines an input optical signal and local light. The optical 90-degree hybrid generates I (in-phase) component output light composed of in-phase components of local light and an optical signal, and Q (quadrature) component light composed of quadrature components of the local light and the optical signal. Both are received by the balanced optical receiver, and the received optical signal is converted into an electrical signal by the 90-degree optical hybrid. For the optical signal, each of the two A / D converters performs time sampling to produce a digitized output signal. It is assumed that the optical field of the received optical signal is r (n) exp (jθ (n)), and the local electric field is 1.

  The definition of the amplitude and phase of a digital signal will be described with reference to FIG. In the IQ plane shown in FIG. 1, there is a signal point in the first quadrant. The distance r (n) between the signal point and the origin of the IQ plane is the amplitude of the digital signal. The angle formed by the line segment connecting the signal point and the origin of the IQ plane and the I axis is the phase θ (n). Furthermore, n is a sample number. Here, j indicates an imaginary unit.

  Local light and signal light actually have random phase noise and difference frequency components. However, since they can be removed by digital signal processing performed after reception, they will be ignored. The balanced optical receiver performs local light emission and homodyne detection on the input optical signal. Each of the balanced optical receivers outputs an in-phase component and a quadrature component of an optical electric field of an optical signal based on local light. Therefore, electrical signals output from the two A / D converters are expressed by Equations 1 and 2.

I (n) = r (n) cos (θ (n)) (Formula 1)
Q (n) = r (n) sin (θ (n)) (Formula 2)
As described above, the coherent optical electric field receiver obtains all information (I and Q components) representing the optical electric field r (n) exp (jθ (n)) from the received optical signal. In addition, the digital arithmetic circuit provided in the digital coherent optical receiver is a complex electric field arithmetic circuit, and provides an inverse function such as linear degradation such as chromatic dispersion that the optical signal undergoes during transmission after data timing extraction. And counteract that effect. By using a digital coherent optical receiver, electric field information of a received signal can be obtained, and in principle, any complex multilevel signal can be received. However, the coherent receiver has a complicated receiver configuration. Further, a laser provided as a signal light source and a local light source is expensive because high performance is required.

  In contrast to the coherent detection method, the optical delayed detection method that detects the phase difference (differential phase) of successive symbols does not use local light. For this reason, the optical delay detection method eliminates the need for frequency estimation and high-speed phase estimation by digital signal processing. In these respects, the optical delay detection method is cheaper and simpler than the coherent detection method.

  Hereinafter, a reception method for demodulating the I-phase and Q-phase of an optical multilevel information signal using two optical delay detectors whose interference phases are orthogonal will be described. The orthogonal optical delay detector will be described. The first optical delay detector is set so that the delay time Td of one of the two paths is equal to the symbol time T of the received optical multilevel information signal, and the optical phase difference between both paths is zero. The second optical delay detector has a delay time Td = T in one of the two paths, and is set so that the optical phase difference between both paths is π / 2. An optical signal received using these two delay detectors and a balanced receiver can be described as follows. If the optical electric field of the received optical electric field signal is described as r (t) exp (jθ (t)), dI and dQ can be expressed by Expressions 14 and 15 from the principle of optical delay detection. Here, t is time.

dI (t)
= R (t) r (t−Td) cos (θ (t) −θ (t−Td)) (Formula 14)
dQ (t)
= R (t) r (t−Td) sin (θ (t) −θ (t−Td)) (Equation 15)
Information transmission is possible by superimposing information to be transmitted in advance on the phase difference Δθ = (θ (t) −θ (t−Td)) of consecutive symbols. However, the delay time Td fluctuates due to the temperature dependence of the delay detector, or the orthogonality is deviated, and since the extra terms are included in Equations 14 and 15, the demodulated phase An error occurs in the information Δθ. In order to detect the phase of the optical signal with high accuracy, it is necessary to accurately control the delay time of the delay detector and the relative phase amount of the two delay detectors.

  Specifically, the delay time Td varies because the optical phase difference between both paths in the first delay detector deviates from zero, and the optical phase difference between both paths in the second delay detector deviates from π / 2. Out. The amount of deviation is the same. Specifically, the orthogonality shift is caused by the relative amount of the optical phase difference between both paths in the first delay detector and the optical phase difference between both paths in the second delay detector deviating from π / 2. is there.

  A method for controlling the phase difference between two orthogonal delay detectors to π / 2 is reported in Non-Patent Document 1. The principle of Non-Patent Document 1 is a configuration similar to a Costas loop, which is a phase control method adopted in a wireless communication system, in which a correlation between an I phase and a Q phase is taken and fed back to a phase control mechanism. The principle of Non-Patent Document 1 is that a part of the I-phase main signal is taken out, multiplied by a reference signal orthogonal to the multiplier, and their correlation is taken, and the correlation is based on the error information obtained as a result. The phase of the delay detector is feedback-controlled so that it becomes zero. Thereby, the phase difference is optimized to π / 2. Here, the reference signal is generated from the orthogonal Q-phase signal using an identification circuit.

  As a phase control method of the delay time of the delay detector, there is a technique described in Patent Document 1. The technique described in Patent Document 1 includes a phase adjustment unit in at least one of two paths of a delay detector, calculates an error rate of a bit pattern reproduced after receiving an optical signal, Disclosed is phase optimization by feedback-controlling the phase adjuster so as to minimize the error rate.

JP 2012-253400 A

Z. Tao, et al., “Dither-free, Accurate, and Robust Phase Offset Monitor and Control Method for Optical DQPSK Demodulator”, ECOC2007, 2007, 3-5-2, pp75-76.

  Without receiving an identification circuit as used in Non-Patent Document 1, digitizing a received signal with an AD converter, performing correlation detection in digital signal processing, and controlling the phase of two delay detectors using the result Is also possible. However, in order to detect the correlation with high sensitivity, an averaging process using a large amount of symbol data is required, and a considerable time is required for detection.

  The present invention detects a deviation from the optimum value of the interference phase of two orthogonal delay detectors with a simple detection mechanism, optimizes the phase amount of the delay time, and determines the orthogonality of the two delay detectors. The objective is to perform optimization without a heater.

  The above-described problem is input to an optical receiver including a first optical delay detector, a second optical delay detector, a first balanced receiver, and a first balanced receiver. The obtained optical electric field signal is branched into two, one optical electric field signal is input to the first optical delay detector, and the output light output from the first optical delay detector is the first balanced receiver. The other optical electric field signal is input to the second optical delay detector, and the output light output from the second optical delay detector is photoelectrically converted by the second balanced receiver. A complex optical electric field is generated from the two pairs of converted electric signals, and the difference in interference phase between the two paths of the first optical delay device or the second optical delay detector is determined from the shape of the signal point arrangement of the complex optical electric field. By an optical receiver that detects and detects a phase difference between interference phases of the first optical delay device and the second optical delay detector It can be achieved.

  In addition, the first optical delay detector without phase difference between the two paths, the second optical delay detector with phase difference between the two paths of π / 2, and the first optical delay detector were connected. A first optical intensity detector; a second optical intensity detector connected to the second optical delay detector; and a first optical signal that converts an output of the first optical intensity detector into a first digital signal. Based on the AD converter, the second AD converter that converts the output of the second light intensity detector into a second digital signal, and the first digital signal and the second digital signal, the optical complex electric field is generated. A quadrature / rotation correction unit that generates the quadrature / rotation correction unit, wherein the quadrature / rotation correction unit converts the first digital signal and the second digital signal based on the optical complex electric field. By detecting the amount of rotation and orthogonality of the elementary symbol, and applying digital signal processing to compensate for distortion of the optical complex electric field, The optical receiver for reproducing an optical complex electric field can be achieved.

  Further, with respect to the inputted optical electric field signal, the step of bifurcating, the step of inputting one of the two branched optical electric field signals to the first optical delay detector, and the other optical electric field signal of the second optical delay detection. Input to the detector, photoelectric conversion of the output light output from the first optical delay detector by the first balanced receiver, and output light output from the second optical delay detector to the first The first optical delay device or the step of photoelectric conversion by two balanced receivers, the step of generating a complex optical electric field from two pairs of photoelectrically converted electric signals, and the shape of the signal point arrangement of the complex optical electric field Detecting a difference in interference phase between the two paths of the second optical delay detector and detecting a phase difference between the interference phases of the first optical delay detector and the second optical delay detector. This can be achieved by the optical signal receiving method.

  According to the present invention, when receiving an optical electric field signal, the interference phase of two orthogonal delay detectors is one of which the interference phase of the delay detector is zero and the interference phase of the other delay detector is π / 2. In one simple detection mechanism, the deviation of the tilt axis of the signal point arrangement, the amplitude of the direction of the tilt axis of the signal point arrangement, and the direction of 90 degrees with respect to the tilt axis By detecting the ratio to the amplitude of the signal and applying a simple matrix operation to the digital signal using them, the phase amount of the delay time and the orthogonality of the two delay detectors are optimized. Can be reproduced in signal point arrangement.

Definition of the amplitude and phase of a digital signal. It is a block diagram explaining a phase preintegration type optical multilevel transceiver system. It is a figure explaining complex signal point arrangement | positioning. It is a figure explaining complex signal point arrangement | positioning when a shift | offset | difference or rotation generate | occur | produced. It is a figure explaining the arrangement | positioning of a complex signal point when a shift | offset | difference and rotation generate | occur | produce. It is a block diagram explaining a direct detection light multilevel receiver. It is a functional block diagram of an orthogonality / rotation correction part. It is a figure explaining 16QAM complex signal point arrangement | positioning. It is a functional block diagram of an inclination detection part. It is a figure explaining the optical electric field in point A, B, C of an inclination detection part. It is a figure explaining the definition of the ratio of the rhombus which a 16QAM signal point group weaves. It is a simulation result of an optical electric field. It is a simulation result of the optical electric field in an orthogonal coordinate transformation circuit. It is a block diagram explaining the other direct detection light multilevel receiver. It is a functional block diagram of another orthogonal / rotation correction unit.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings using examples.

  Referring to FIG. 2, phase pre-integration is performed on the transmission side to generate an optical multilevel signal, and the optical multilevel signal is received by combining the delay detection method and the direct detection method on the reception side. An integrated optical multilevel transceiver system will be described. The optical electric field receiver of FIG. 2 can detect the optical phase by using optical delayed detection without using local light. For this reason, the optical electric field receiver can be realized simply and inexpensively. Furthermore, since the phase pre-integration is used, the optical electric field receiver can restore the differential phase after delay detection to the phase of the optical electric field. For this reason, the transceiver can perform chromatic dispersion compensation by digital signal processing after reception.

  In the following, the configuration of the optical transceiver system when the interference phase of the delay detector is the optimum value will be described with reference to FIG. Next, a complex signal point arrangement (constellation) when a 16QAM signal is generated and received using this transceiver will be described with reference to FIG.

  In FIG. 2, the phase preintegration type optical multilevel transceiver system includes a phase preintegration type optical electric field transmitter 201, an optical transmission line 212, and an optical electric field receiver 213. The phase pre-integration type optical electric field transmitter 201 includes a laser light source 202, an optical electric field modulator 203, a complex multilevel signal generation circuit 204, a phase pre-integration unit 205, a sampling rate conversion circuit 206, and a pre-equalization circuit. 207, a DA converter 208, and a drive circuit 210 are included. The optical electric field receiver 213 includes a delay detector 214, a delay detector 215, a light intensity detector 216, a BPD 217, an AD converter 218, a sampling rate conversion circuit 221, an arctangent calculation circuit 222, and a square root. A circuit 223 and an orthogonal coordinate conversion circuit 224 are included.

  In the phase pre-integration type optical electric field transmitter 201, the unmodulated laser light output from the laser light source 202 is modulated by the optical electric field modulator 203 into an optical electric field signal subjected to a desired electric field modulation, The data is output to the transmission line 212.

  For the information signal to be transmitted, the complex multi-level signal generation circuit 204 receives an input to the digital information input terminal as a parallel binary high-speed digital electric signal sequence. The complex multi-level signal generation circuit 204 converts the input signal into a complex multi-level information signal. The converted signal is a digital electrical multilevel signal expressed as (i, q) on a two-dimensional IQ plane, and its real part i and imaginary part q are output at every time interval T (= symbol time). The For the converted signal, the phase pre-integration unit 205 receives an input. The phase preintegration unit 205 digitally integrates only the phase components of the input signal at a time interval T, and converts it into a phase preintegration complex multilevel information signal. When the complex multilevel information signal (i, q) is input, the phase preintegration unit 205 converts it into polar coordinates (r, θ) on the complex plane. The polar coordinates (r, θ) can be expressed as Equation 3. FIG. 3A shows the signal point arrangement of the 16QAM optical electric field at this time.

Ei (n) = i (n) + jq (n)
= R (n) exp (jθ (n)) (Formula 3)
The phase preintegration signal to be output at this time can be expressed by Equation 4 using polar coordinates.

Eo (n) = i ′ (n) + jq ′ (n)
= R (n) exp (jφ (n))
= R (n) exp (jΣθ (n)) (Formula 4)
At this time, φ (n) is the phase angle of the output signal, and Σθ (n) is a value obtained by cumulatively adding past phase angles θ (1) to θ (n). The output signal is again converted into the orthogonal coordinate system, and then output as a phase preintegrated complex multilevel information signal. FIG. 3B shows the signal point arrangement of the 16QAM optical electric field at this time. This phase pre-integrated complex multilevel information signal is input to sampling rate conversion circuits 206-1 and 206-2, and the sampling points are complemented so that the sampling rate is 2 samples / symbol or more. Thus, the Nyquist theorem is satisfied and complete electric field equalization processing is possible. Thereafter, an inverse function of deterioration generated in the optical transmission line 212 by the pre-equalization circuit 207 is applied to the phase pre-integrated complex multi-level information signal, and then separated into a real part i ″ and an imaginary part q ″. The two DA converters 208 and 209 convert the separated signals into high-speed analog signals, respectively. These two analog signals are amplified by the drive circuit 210 and the drive circuit 211 and then input to the two modulation terminals I and Q of the optical electric field modulator 203. As a result, an optical electric field signal having the pre-equalized phase integration signal (i ″, q ″) in the in-phase component I and the quadrature component Q of the optical electric field is generated. The optical electric field of the optical electric field signal is (i ″ (n) + jq ″ (n)) expjω (n). Here, ω (n) is the optical angular frequency of the laser light source. That is, the optical electric field signal is equivalent to (i ″ (n), q ″ (n)) by an equivalent low-frequency approximation with the optical frequency component removed.

  The optical electric field signal is transmitted through the optical fiber transmission line 212. The optical fiber transmission line 212 gives transmission degradation due to chromatic dispersion to an optical signal. The non-coherent optical electric field receiver 213 receives the received optical electric field signal.

  The transmission deterioration in the optical fiber transmission line 212 is mutually canceled with the inverse function applied in advance by the pre-equalization circuit 207. For this reason, the optical electric field of the received signal is equal to the phase preintegrated complex multilevel information signal. Even when the inverse function applied in advance and the transfer function of the transmission path deterioration do not completely cancel each other, it is possible to completely cancel with the transfer function applied again on the receiving side.

  The optical electric field receiver 213 branches the received optical electric field signal and inputs it to the two optical delay detectors 214 and 215 and the optical intensity detector 216. The first optical delay detector 214 is set so that the delay time Td of one of the two paths is approximately equal to the symbol time T of the received optical multilevel information signal, and the optical phase difference between both paths is zero. Has been. The second optical delay detector 215 has a delay time Td = T in one of the two paths, and is set so that the optical phase difference between both paths is π / 2. The delay detectors 214 and 215 output the output light to the balanced photodetectors 216 and 217. The balanced photodetectors 216 and 217 convert the output light into an electrical signal. The AD converter 218 converts the converted electrical signal into a digital signal. The sampling rate conversion circuits 221-1 and 221-2 convert to 1 sample / symbol, and become dI (n) and dQ (n).

  The AD converter 220 and the sampling rate conversion circuit 221-3 convert the electrical signal output from the light intensity detector 216 into a digital signal P (n). For the digital signals dI (n) and dQ (n), the arc tangent calculation circuit 222 performs an inverse tangent operation of two arguments using dI (n) as an X component and dQ (n) as a Y component, and the digital signal dI (n ), And the phase angle of dQ (n) is calculated. If the optical electric field of the received optical electric field signal is described as r (n) exp (jθ (n)), dI and dQ can be expressed by equations (5) and (6).

dI (n) = r (n) r (n−1) cos (Δφ (n)) (Formula 5)
dQ (n) = r (n) r (n-1) sin (Δφ (n)) (Expression 6)
Here, Δφ (n) is a phase difference (φ (n) −φ (n−1)) of the received n-th optical electric field symbol from the immediately preceding symbol. Since di and dQ are the sine component and cosine component of Δφ (n), respectively, the arc tangent calculation circuit 222 performs an arc tangent calculation in four quadrants to calculate Δφ (n). Here, since the phase pre-integration is performed on the transmission side, the phase angle of the received optical electric field signal can be expressed as in Expression 7.

φ (n) = Σθ (n) (Expression 7)
The output signal of the arctangent calculation circuit 222 can be expressed by Equation 8. In Equation 8, the phase component θ (n) of the original complex multilevel information signal can be extracted. FIG. 3C shows the signal point arrangement of the 16QAM optical electric field at this time.

Δφ (n) = Σθ (n) −Σθ (n−1)
= Θ (n) (Formula 8)
For the output signal P (n) of the light intensity detector 216, the square root circuit 223 obtains the electric field amplitude expressed by Equation 9 as an output.

r (n) = sqrt (P (n)) (Formula 9)
Based on the obtained amplitude component r (n) and phase component θ (n), the orthogonal coordinate transformation circuit 224 reproduces the original digital electric multilevel signal represented by Expression 10 from the reproduction complex information output terminal. . FIG. 3D shows the signal point arrangement of the 16QAM optical electric field at this time.

(I, q) = r (n) exp (jθ (n)) (Equation 10)
The complex signal arrangement will be described with reference to FIG. In FIG. 3, FIG. 3 (a) shows a 16QAM complex signal point arrangement before phase pre-integration. FIG. 3B shows a 16QAM complex signal point arrangement after phase pre-integration. FIG. 3C shows a 16QAM complex signal point arrangement after delay detection. FIG. 3D shows a 16QAM complex signal point arrangement after the orthogonal coordinate conversion circuit.

In the phase pre-accumulation transmitter / receiver, a case where a shift such as the following “1” and “2” occurs in the two delay detectors used will be described.
“1”: the optical phase difference between the two paths in the first delay detector deviates from zero, and the optical phase difference between both paths in the second delay detector deviates from π / 2. Is the same amount δ including the sign “2”: The relative amount of the optical phase difference between both paths in the first delay detector and the optical phase difference between both paths in the second delay detector is π / When deviating from 2 to (π / 2 + 2α) If they occur independently, dI and dQ shown in Equations 5 and 6 are converted by matrix operation as shown in Equations 11 and 12, and dI ′ and dQ 'become.

このとき図３（ｃ）に示した遅延検波後の１６ＱＡＭ複素信号点配置は、”１”の場合は図４（ａ）に、”２”の場合は図４（ｂ）（理想的な信号点配置が正方形の形状をなしている場合、ひし形のように、また円形の場合は楕円のように）になる。また、”１””２”が一度に起きた場合は図５（ａ）のようになり、その信号点配置から算出できる位相角と強度検出部で検出された光振幅を用いて複素光電界を再生すると図５（ｂ）のようになる。図５（ｂ）は、全体が回転していることと、第１象限と第３象限の位相方向の間隔が狭く、第２象限と第４象限の位相方向の間隔が広いことが特徴である。このように遅延検波器に”１””２”のような事象が起きた場合、元の光電界（ｉ、ｑ）を再生できない。

At this time, the 16QAM complex signal point arrangement after delayed detection shown in FIG. 3C is “1” in FIG. 4A, and in “2” in FIG. 4B (ideal signal). If the point arrangement has a square shape, it will be like a diamond, and if it is circular, it will be an ellipse). Further, when “1” and “2” occur at the same time, as shown in FIG. 5A, the complex optical electric field is obtained using the phase angle that can be calculated from the signal point arrangement and the light amplitude detected by the intensity detector. Is reproduced as shown in FIG. FIG. 5 (b) is characterized in that the whole is rotating, the interval in the phase direction between the first quadrant and the third quadrant is narrow, and the interval in the phase direction between the second quadrant and the fourth quadrant is wide. . Thus, when an event such as “1” or “2” occurs in the delay detector, the original optical electric field (i, q) cannot be reproduced.

  The complex signal arrangement will be described with reference to FIGS. 4 and 5. In FIG. 4, FIG. 4 (a) shows a 16QAM complex signal point arrangement after the delay detector (in the case of “1”). FIG. 4B shows a 16QAM complex signal point arrangement after phase pre-integration (in the case of “2”). FIG. 5A shows a 16QAM complex signal point arrangement after the delay detector (in the case of “1” and “2”). FIG. 5B shows a 16QAM complex signal point arrangement after the orthogonal coordinate conversion circuit (in the case of “1” and “2”).

  In response to this phenomenon, the first embodiment detects and corrects the phase shift and the orthogonal shift of the delay detector by the following method.

  The configuration of the optical electric field receiver will be described with reference to FIG. In FIG. 6, an optical electric field receiver 600 includes a delay detector 602, a delay detector 603, a light intensity detector 604, an AD converter 608, a sampling rate conversion circuit 609, an amplitude normalization circuit 610, An arctangent calculation circuit 612, a square root circuit 613, an orthogonal coordinate conversion circuit 614, and an orthogonal / rotation correction unit 620 are configured.

  The optical electric field receiver 600 outputs the optical electric field signal propagated through the transmission line to the two optical delay detectors 602 and 603 and the optical intensity detector 604 in two branches. The balanced photodetectors 606 and 607 convert the output light from the delay detectors 602 and 603 into electric signals, respectively. Thereafter, the AD converters 608-1 and 608-2 convert the converted electrical signals into digital signals. Sampling rate conversion circuits 609-1 and 609-2 convert the digital signal into one sample / symbol. The outputs of the sampling rate conversion circuits 609-1 and 609-2 are dI (n) and dQ (n). For dI (n) and dQ (n), the amplitude normalization circuit 610 normalizes the respective amplitudes so as to have the same value (specifically, 256). This is because when the transimpedance amplifier in the balanced receiver is automatically gain-controlled, the gains on the I side and the Q side are different, and the amplitudes of dI (n) and dQ (n) are different. This is to correct this. Normalization also increases the amplitude by 1 / √2 in the case of a square signal point arrangement such as 16QAM. This is because when the digital signal is processed, the signal is digitized (in the case of quantization bit 8, the amplitude is expressed in the range of 256). This is because it is necessary to reduce the amplitude in advance so as not to occur.

  For dI (n) and dQ (n) output from the normalization circuit 610, the orthogonal / rotation correction unit 620 detects the phase shift amount from the distorted signal point arrangement as shown in FIG. And dI ′ (n) and dQ ′ (n) are output. The quadrature / rotation correction unit 620 detects the phase shift amount and the orthogonal shift amount of the delay detector from the complex optical electric field generated by the input dI (n) and dQ (n). The orthogonal / rotation correction unit 620 further performs digital signal processing (matrix operation) described below on the complex optical electric field to compensate for the distortion of the optical electric field and regenerate the original optical electric field. By using the orthogonal / rotation correction unit 620, the delay detector can eliminate the need for a phase adjustment circuit using a heater.

  The configuration of the orthogonal / rotation correction unit 620 will be described with reference to FIG. In FIG. 7, the orthogonal / rotation correction unit 620 includes a dI-dQ complex optical electric field generation unit 621, an inclination rotation calculation unit 622, an orthogonality correction calculation unit 623, a fixed amount rotation calculation unit 624, and an inclination detection unit 625. And an orthogonality deviation detection unit 626. The orthogonality deviation detection unit 626 includes an inclination rotation unit 6261 and an orthogonality deviation calculation unit 6262.

  When dI and dQ are inputted, the dI-dQ complex optical field generator 621 generates a complex optical field as shown in FIG. 8A (in the case of 16QAM) by calculating dI (n) + jdQ (n). To do. When there is no phase shift or orthogonal shift in the delay detector, the tilt axis of the complex optical electric field shown in FIG. 8A (shown by an arrow) is rotated 45 degrees counterclockwise from the I axis. Indicates the direction. Therefore, the tilt detector 625 detects the direction of the tilt axis when the optical electric field is distorted.

  Specifically, the tilt detection unit 625 can detect the tilt axis by diverting the phase estimation technique used in the coherent detection method to the delay detection method.

  The 16QAM complex signal point arrangement will be described with reference to FIG. 8A is a 16QAM complex signal point arrangement of the output of the dI-dQ complex optical electric field generation unit 621. FIG. FIG. 8B shows the 16QAM complex signal point arrangement of the output of the tilt rotation unit 6261. FIG. 8C shows the 16QAM complex signal point arrangement of the output of the orthogonality correction calculation unit 623. FIG. 8D shows the 16QAM complex signal point arrangement of the output of the fixed amount rotation calculation unit 624.

  With reference to FIG. 9, the configuration of the tilt detection unit will be described. In FIG. 9, the inclination detection unit 625 includes a maximum amplitude signal point extraction unit 901, a restored optical electric field m-th power unit 902, an averaging unit 903, an m-fold declination calculation unit 904, and a declination calculation unit 905. , Including.

  When the 16QAM dI-dQ optical electric field is input to the tilt detection unit 625, the maximum amplitude signal point extraction unit 901 has four corners (45 degrees, 135 degrees, 225 degrees, 315 as phase angles in an ideal optical electric field). It is narrowed down to signal points corresponding to degrees (maximum amplitude) (FIG. 10A). Since the maximum signal point extraction unit 901 has many signal point arrangements of the 16QAM optical electric field, the maximum signal point extraction unit 901 also has the amplitude information P (n) of the optical electric field output from the light intensity detection unit provided in the direct detection optical multilevel receiver. Then, focus on the signal points at the four corners.

  The complex optical electric field m-th power unit 902 raises the complex optical electric field r (n) exp (jθ (n)) to the m-th power (m = 4 because it is narrowed down to signal points at the four corners) (FIG. 10B). For the optical electric field calculated to the m-th power, the averaging unit 903 averages and combines them into one point. The m-fold deflection angle calculation unit 904 calculates a deflection angle (4 <φ>) from the signal points determined by the averaging unit 903 as one point. The deflection angle calculation unit 905 determines the tilt amount <φ> by reducing the deflection angle to ¼.

  Returning to FIG. 7, the information (tilt amount) of the tilt axis detected by the tilt detection unit 625 is input to the tilt rotation unit 6261. The tilt rotation unit 6261 rotates the dI-dQ optical electric field input to the orthogonal / rotation correction unit 620 by the tilt amount φ, and the optical field in which the tilt axis coincides with the I axis (the complex photoelectric shown in FIG. 8B). ). The tilt rotation unit 6261 outputs the generated signal to the orthogonality deviation calculation unit 6262. The complex optical electric field shown in FIG. 8B is ideally square due to a shift in orthogonality, and has a rhombus-like shape. Therefore, the ratio of the two diagonals of the rhombus as shown in FIG. 8B (R = t / s) is calculated, and the inverse is applied to make the rhombus an ideal square.

  The optical electric field at points A, B, and C in FIG. 9 will be described with reference to FIG. 10A shows an optical electric field at point A. FIG. FIG. 10B shows an optical electric field at point B. FIG. 10C shows an optical electric field at the point C.

  With reference to FIG. 11, the definition regarding a rhombus is demonstrated. In FIG. 11, FIG. 11 (a) is a diagram for explaining the definition of the diagonal length of the rhombus. FIG. 11B illustrates the definition of the length between the centroids of the upper and lower triangles with respect to the I axis. FIG. 11C illustrates the definition of the length between the centroids of the left and right triangles with respect to the Q axis.

  In FIG. 11A, the I-axis direction distance of the rhombus apex is s, the Q-axis direction distance is t, and the rhombus ratio R is t / s. In FIG. 11B, the length between the centroids of the upper and lower triangles with respect to the I axis is defined as t ′. In FIG. 11C, the length between the centroids of the left and right triangles with respect to the Q axis is defined as s ′.

  A method for calculating the ratio R of the two diagonals of the rhombus in the orthogonality deviation calculating unit 6262 will be described. First, as shown in FIGS. 11 (b) and 11 (c), the rhombus shown in FIG. 11 (a) is divided into upper and lower triangles on the I axis, and the rhombus is divided into left and right triangles on the Q axis. The orthogonality deviation calculation unit 6262 calculates the length (t ′) between the centroids of the upper and lower triangles shown in FIG. Further, the orthogonality deviation calculating unit 6262 calculates the length (s ′) between the centroids of the left and right triangles shown in FIG. The orthogonality deviation calculating unit 6262 calculates the ratio of the lengths of the original rhombuses to calculate the ratio R = t / s = t ′ / s ′ of the original rhombus.

  As a method of calculating the center of gravity of the triangle, the center of gravity can be obtained by adding the optical electric fields of the signal points included in the triangle and dividing by the number of signal points. Other methods for calculating the ratio of the rhombus include a method of comparing the maximum and minimum widths in the I-axis direction and the maximum and minimum widths in the Q-axis direction, an n-order moment in the I-axis direction, and an n-order moment in the Q-axis direction. There is a way to compare moments. These can be easily calculated because the tilt axis coincides with the I axis.

  In FIG. 7, the tilt rotation calculation unit 622, the orthogonality correction calculation unit 623, and the fixed amount rotation calculation unit 624 are the rhombuses calculated by the tilt amount φ detected by the tilt detection unit 625 and the orthogonality deviation calculation unit 6262. The optical field (dI, dQ) distorted by the phase shift and the orthogonality shift of the delay detector is corrected using the ratio R. When the corrected optical electric field is (dI ′, dQ ′), the arithmetic expression of the orthogonal / rotation correction unit 620 is expressed by Expression 13. The third item on the right side of Equation 13 is a rotation matrix that matches the inclination axis (inclination amount φ) of FIG. 8B with the I axis. The second item is a matrix that transforms the rhombus shown in FIG. 8C into a square. The first item is a rotation matrix for rotating the tilt axis shown in FIG. 8D in an ideal phase direction (45 degrees in this case).

図６に戻って、直交・回転補正部６２０において、遅延検波部の位相ずれと直交性のずれの検出および補正処理を経て出力された（ｄＩ'，ｄＱ'）について、逆正接演算回路６１２は、元の複素多値情報信号の位相成分θ（ｎ）を算出する。この位相成分θ（ｎ）と、光強度検出器６０４、ＡＤ変換器６０８−３、サンプリング速度変換回路６０９−３、平方根回路６１３を経て得られた元の複素多値情報信号の振幅成分ｒ（ｎ）とについて、直交座標変換回路６１４は、極座標から直交座標に変換する。この結果、光電界受信器６００は、複素多値情報信号（ｉ，ｑ）を再生できる。

Returning to FIG. 6, with respect to (dI ′, dQ ′) output after the detection and correction processing of the phase shift and the orthogonality shift of the delay detection unit in the quadrature / rotation correction unit 620, the arctangent calculation circuit 612 The phase component θ (n) of the original complex multilevel information signal is calculated. This phase component θ (n) and the amplitude component r () of the original complex multilevel information signal obtained through the light intensity detector 604, AD converter 608-3, sampling rate conversion circuit 609-3, and square root circuit 613 are obtained. n), the orthogonal coordinate conversion circuit 614 converts polar coordinates into orthogonal coordinates. As a result, the optical electric field receiver 600 can reproduce the complex multilevel information signal (i, q).

  The number of signal points of the digital signal used when detecting the interference phase information of the delay detector is variable, and the signal point is calculated based on the calculation result by calculating the variance or standard deviation of the interference phase information. It may be controlled. The determinant of Expression 13 is changed with a period depending on the number of signal points.

  With reference to FIG. 12 and FIG. 13, simulation results obtained by detecting and correcting the phase shift and orthogonality shift of the delay detection unit using the technique of the first embodiment will be described. In FIG. 12, FIG. 12A shows an optical electric field in the dI-dQ complex electric field generation unit. FIG. 12B shows an optical electric field in the tilt rotation calculation unit. FIG. 12C shows an optical electric field in the orthogonality correction calculation unit. FIG. 12D shows an optical electric field in the fixed amount rotation calculation unit. In FIG. 13, FIG. 13A shows an optical electric field in the orthogonal coordinate conversion circuit at the time of non-correction. FIG. 13B shows an optical electric field in the orthogonal coordinate conversion circuit at the time of correction.

In FIG. 12, the optical signal used in the simulation is 16QAM. As a feature of the 16QAM signal, the phase angle is divided into 16 on the complex plane, and the signal points are arranged so that the phase angle of all the signal points is an integral multiple of 360/16 = 22.5 degrees. This signal is subjected to phase pre-integration on the transmission side to generate a phase pre-integrated optical signal. The generated optical signal has an optical SN ratio of 25 dB and is input to the optical multilevel receiver. FIG. 12A shows a signal point arrangement formed by dI and dQ of an input optical signal when input to a delay detector having a phase shift of 5 degrees and an orthogonality shift of 15 degrees. If the optical electric field is reproduced from the amplitude of the optical electric field obtained by the intensity detector, the distorted signal point arrangement shown in FIG. 13A is obtained. On the other hand, the signal point arrangement when the orthogonal / rotation correction technique is used will be described with reference to FIG. 7. In the dI-dQ complex electric field generation unit 625, the complex optical electric field shown in FIG. Information on the amount of tilt detected by the unit 625 is transmitted to the tilt rotation calculation unit 622 so that the tilt axis coincides with the I axis in FIG. 12B, and the orthogonality correction calculation unit 623 detects it by the orthogonality deviation detection unit 626. 12C is corrected to a square using the corrected orthogonality correction amount, and an ideal signal is obtained by rotating the fixed amount rotation calculation unit 624 by the fixed amount of FIG. 12D. Get a point arrangement.
As a result, by combining the obtained dI ′ and dQ ′ and the amplitude information P of the optical electric field obtained by the intensity detector into the optical electric field, the signal point arrangement without distortion of FIG. 13B is obtained. It is done.

  According to this embodiment, when an optical electric field signal is received, the interference phase of two orthogonal delay detectors is one of which is zero, and the other delay detector is π /. 2 is an optimum state, but in a simple detection mechanism, the deviation in the direction of the tilt axis of the signal point arrangement, the direction of the tilt axis of the signal point arrangement, and 90 degrees with respect to the tilt axis. By detecting the ratio to the amplitude in the direction and applying simple matrix operations to the digital signal using them, the phase amount of the delay time is optimized and the orthogonality of the two delay detectors is optimized It is possible to reproduce in a typical signal point arrangement.

  According to this embodiment, it is possible to eliminate a phase adjustment heater in the receiver, so that reduction of power consumption can be expected, and between the interference phases of the quadrature delay detector in which two delay detectors are integrated. Specification relaxation for the difference is also possible.

  The decisive difference between this embodiment and the orthogonality detection technique using the coherent detection method is that the signal point depends on the frequency difference, phase difference, and polarization state of continuous light output from the two light sources in the coherent detection method. Whereas the arrangement fluctuates at a high speed, this technology uses a delayed detection method, so there is no frequency difference or polarization state difference between the two interfering lights. It is mentioned that arrangement does not change. Therefore, the orthogonality detection technique using the coherent detection method cannot be applied to the optical signal after transmission.

  With reference to FIG. 14, the optical electric field receiver of Example 2 is demonstrated. In FIG. 14, the difference from the optical electric field receiver of the first embodiment described in FIG. 6 will be described. The optical electric field receiver 600A of FIG. 14 includes an orthogonal / rotation correction unit 620A instead of the orthogonal / rotation correction unit 620. The optical electric field receiver 600A further includes a residual dispersion compensation unit 615, a data decoding unit 616, and a frame detection unit 617.

  When chromatic dispersion remains in the optical electric field signal input to the optical electric field receiver 600A, the complex multilevel optical electric field (i ′ ″, q ′ ″) output from the orthogonal coordinate conversion circuit 614 includes A residual chromatic dispersion transfer function is applied. The residual dispersion compensator 615 applies an inverse function of the transfer function of the residual dispersion based on the complex multilevel optical electric field (i ′ ″, q ″ ′) output from the orthogonal coordinate conversion circuit 614 to restore the original A complex multilevel signal (i, q) is obtained. The residual dispersion compensator 615 converts the input signal into the frequency domain once. The residual dispersion compensator 615 applies a transfer function to remove residual dispersion. After that, the residual dispersion may be removed in the time domain as it is using the digital correction filter composed of a plurality of taps.

  The data decoding unit 616 converts the complex multilevel signal (i, q) into a binary digital electric signal sequence. The converted binary digital electric signal sequence is configured into a frame conforming to the Ethernet (registered trademark) transmission standard. The frame detection unit 617 detects whether or not the frame configuration is expected. The frame detection unit 617 determines whether the input digital signal sequence is correctly decoded. When the frame detection unit 617 detects that the frame is incorrect, the information is transmitted to the orthogonal / rotation correction unit 620A as frame detection information.

  With reference to FIG. 15, the orthogonal / rotation correction unit according to the second embodiment will be described. In FIG. 15, the orthogonal / rotation correction unit 620 </ b> A further includes a fixed amount changing unit 627 in comparison with the orthogonal / rotation correction unit 620 of FIG. 7. With respect to the transmitted frame detection information, the fixed amount changing unit 627 changes the fixed value of the rotation amount used in the fixed amount rotation calculating unit 624 when notified that the frame is incorrect. As a method of changing the fixed value, if the initially set value is 45 degrees, it is changed to an integral multiple thereof, that is, 90 degrees, 135 degrees,..., And continues until the frame detection information is canceled. This is an initial pull-in when the interference phase of the delay detector is significantly shifted, such as when the phase preintegration type optical multilevel transceiver 600A used in the second embodiment is started in the optical transmission apparatus. It becomes effective. The application of the second embodiment is not limited to the phase preintegration type optical multilevel transceiver 600.

  DESCRIPTION OF SYMBOLS 201 ... Phase preintegration type optical electric field transmitter, 202 ... Laser light source, 203 ... Optical electric field modulator, 204 ... Complex multi-value signal generation circuit, 205 ... Phase preintegration unit, 206 ... Sampling speed conversion circuit, 207 ... Prediction , 208, 209 ... DA converter, 210 ... drive circuit, 212 ... optical transmission path, 213,600 ... optical electric field receiver, 214, 215, 602, 603 ... delay detector, 216,604 ... light intensity detection 217, 606, 607 ... balanced receiver, 218, 220, 608 ... AD converter, 221, 609 ... sampling rate conversion circuit, 222, 612 ... arctangent calculation circuit, 223, 613 ... square root circuit, 224, 614 ... Cartesian coordinate transformation circuit, 610 ... Amplitude normalization circuit, 615 ... Residual dispersion compensation section, 616 ... Data decoding section, 617 ... Frame detection section, 620 ... Cross / rotation correction unit, 621... DI-dQ complex optical electric field generation unit, 622... Tilt rotation calculation unit, 623... Orthogonality correction calculation unit, 624... Fixed amount rotation calculation unit, 625. Deviation detection unit, 627... Fixed amount change unit, 901... Maximum amplitude signal point extraction unit, 902... Complex optical electric field m squared unit, 903... Averaging unit, 904. , 6261... Tilt rotation unit, 6262... Orthogonality deviation calculation unit.

Claims (10)

In an optical receiver comprising a first optical delay detector, a second optical delay detector, a first balanced receiver, and a first balanced receiver,
The input optical electric field signal is branched into two, one optical electric field signal is input to the first optical delay detector, and the output light output from the first optical delay detector is the first optical delay signal. Photoelectric conversion with a balanced receiver
The other optical electric field signal is input to the second optical delay detector, and the output light output from the second optical delay detector is photoelectrically converted by the second balanced receiver,
A complex optical electric field is generated from two pairs of photoelectrically converted electric signals,
A difference in interference phase between two paths of the first optical delay device or the second optical delay detector is detected from the shape of the signal point arrangement of the complex optical electric field, and the first optical delay device and the second optical delay device are detected. An optical receiver for detecting a phase difference between interference phases of two optical delay detectors. The optical receiver according to claim 1,
Furthermore, a first AD converter that converts the output signal of the first balanced receiver into a first digital signal, and an output signal of the second balanced receiver is converted into a second digital signal. A second AD converter, a tilt / orthogonality shift detection unit, and a digital signal processing unit;
The tilt / orthogonality shift detection unit generates a digital complex optical field from the first digital signal and the second digital signal, and based on the shape of the signal point arrangement of the digital reconstruction optical field, A phase difference between the interference phases of the first optical delay device and the second optical delay detector is detected by detecting a difference in interference phase between the two paths of the first optical delay device or the second optical delay detector. Detect
The digital signal processing unit is configured such that the signal point arrangement shape of the digital complex optical electric field is zero, the difference in interference phase between the two paths of the first delay detector is zero, and the interference position of the two paths of the second delay detector is An optical receiver characterized by converting into a signal point arrangement having a phase difference of π / 2. The optical receiver according to claim 2,
The inclination / orthogonality deviation detection unit is
The amount of rotation of the signal point arrangement when the difference in interference phase between the two paths of one delay detector and the difference in interference phase between the two paths of the other delay detector are shifted by the same phase;
The phase difference between the interference phases of the two delay detectors is deviated from π / 2, and the shape of the signal point arrangement is detected at a ratio when the specific rotation direction is expanded or contracted with respect to the axis, and
The optical receiver according to claim 1, wherein the specific rotation direction coincides with a rotation direction according to the rotation amount. The optical receiver according to claim 2,
Furthermore, an amplitude normalization unit is provided,
The amplitude normalization unit includes:
When the standard deviation of the amplitude of the first digital signal and the second digital signal does not match, the amplitude is adjusted so that the standard deviation becomes the same amount,
Further, the optical receiver is set so that the amplitude does not exceed a signal range processed by the digital signal processing unit. The optical receiver according to claim 3, wherein
The digital signal processing unit performs a rotation matrix calculation on the rotation amount of the first digital signal and the second digital signal in a reverse direction, and performs an expansion / contraction operation on the rotated digital signal at an inverse of the ratio. An optical receiver characterized in that a rotation matrix is calculated by a fixed amount for a stretched digital signal. The optical receiver according to claim 5, wherein
In addition, a frame detection unit that detects whether the decoded data sequence has a frame structure,
The fixed amount is changed according to specific conditions,
The optical condition is that the specific condition is a case where the frame detection unit detects that the decoded data string is not the frame structure. The optical receiver according to claim 5, wherein
The fixed amount that is changed when the specific condition is satisfied is a multiple of a predetermined value, and is continuously changed until the specific condition is canceled. The optical receiver according to claim 2,
The signal number of the digital signal used when detecting the information of the interference phase of the delay detector in the tilt / orthogonality deviation detector is variable,
The number of signal points is controlled based on a calculation result obtained by calculating a dispersion or standard deviation of information on the interference phase. A first optical delay detector having no phase difference between the two paths, a second optical delay detector having a phase difference of π / 2 between the two paths, and a first optical delay detector connected to the first optical delay detector. A first light intensity detector, a second light intensity detector connected to the second optical delay detector, and a first digital signal that converts the output of the first light intensity detector into a first digital signal. Based on the AD converter, the second AD converter that converts the output of the second light intensity detector into a second digital signal, the first digital signal, and the second digital signal An optical receiver configured to include an orthogonal / rotation correction unit that generates an optical complex electric field,
The orthogonality / rotation correction unit detects a rotation amount and orthogonality deviation amount of a complex symbol between the first digital signal and the second digital signal based on the optical complex electric field, and performs digital signal processing. An optical receiver that compensates for distortion of the optical complex electric field and reproduces the original optical complex electric field. A step of bifurcating the input optical electric field signal;
Inputting one of the two branched optical electric field signals to the first optical delay detector;
Inputting the other optical electric field signal to the second optical delay detector;
Photoelectrically converting the output light output from the first optical delay detector with a first balanced receiver;
Photoelectrically converting the output light output from the second optical delay detector with a second balanced receiver;
Generating a complex optical electric field from two pairs of photoelectrically converted electrical signals;
Detecting a difference in interference phase between two paths of the first optical delay device or the second optical delay detector from the shape of the signal point arrangement of the complex optical electric field;
Detecting a phase difference between interference phases of the first optical delay device and the second optical delay detector.

Images