WO2010131323A1

WO2010131323A1 - Polarization demultiplexing apparatus

Info

Publication number: WO2010131323A1
Application number: PCT/JP2009/058762
Authority: WO
Inventors: 剛吉田; 聖史斧原; 竜也小林
Original assignee: 三菱電機株式会社
Priority date: 2009-05-11
Filing date: 2009-05-11
Publication date: 2010-11-18

Abstract

There are included a light source part (1) that emits a light at a frequency identical to the carrier frequency of a received signal; polarization splitting parts (2-1, 2-2) that polarize the optical signals; optical interfering parts (3-1, 3-2) that output optical signals having difference and sum components of the electric fields of the polarized optical signals; balanced photon detecting parts (4-1 to 4-4) that perform photoelectric conversions of the optical signals having the sum and difference components to output the differences as electric signals; analog signal branching parts (5-1 to 5-4) each of which bifurcates the respective electric signal; real time ADC parts (61-1 to 61-4) each of which converts one of the bifurcated electric signals to a discrete value of N1-bit resolution (where N1 > 1) in a time period (T1); low-speed ADC parts (71-1 to 71-4) each of which converts the other of the bifurcated electric signals to a discrete value of N2-bit resolution (where N2 > N1) in a time period (T2) longer than the foregoing time period (T1); a matrix generating part (72) that generates a complex matrix based on the outputs from the low-speed ADC parts (71-1 to 71-4); and a matrix multiplying part (62) that generates, based on the outputs from the real time ADC parts (61-1 to 61-4), complex vectors and multiplies the complex matrix by the generated complex vectors.

Description

Polarization demultiplexer

The present invention relates to a polarization multiplexing / demultiplexing device that separates a polarization multiplexed signal.

In order to perform large-capacity communication by optical fiber communication, the bit rate per wavelength is being increased. Since optical fibers have dispersion characteristics and the transmission distance is limited in proportion to the square of the symbol rate, it is essential to reduce the symbol rate. For this reason, transmission schemes that increase the number of bits per symbol are being studied, and polarization multiplexing schemes are particularly attracting attention. For example, the following non-patent document 1 and the following non-patent document 2 disclose a technique based on a 40 Gbit / sec polarization multiplexed quadrature phase-shift keying (PDM-QPSK) system. As shown in the following Non-Patent Document 2, it has already been put into practical use. The PDM-QPSK system can transmit 40 Gbit / sec while suppressing the symbol rate to 10 G symbol / sec, and has high dispersion tolerance.

In a system that performs polarization multiplexing, a polarization multiplexed signal transmitter assigns a signal to each of the orthogonal x and y polarization components independently and transmits a multiplexed polarization signal that is multiplexed. The polarization multiplexed signal receiving unit receives the polarization multiplexed signal via the transmission path. In the transmission line, waveform distortion and noise superposition occur, and the polarization state also changes. In the polarization multiplexed signal receiving unit, the polarization multiplexed signal distributed to the wavelength unit by the digital signal processing in the electric domain is tracked while tracking the polarization state changing with time in the transmission path. Are separated into x and y polarization components before polarization multiplexing. Conventionally, optical polarization demultiplexing has been studied as shown in Patent Document 1 below, but in recent years, polarization separation by digital signal processing is becoming the mainstream of investigation.

Specifically, since the polarization change in the transmission line can be expressed by multiplying the transmission signal vector by the Jones matrix, the polarization multiplexed signal receiving unit multiplies the reception signal vector by the inverse matrix of the Jones matrix. By obtaining a transmission signal vector, polarization multiplexing / demultiplexing is possible. Actually, the transmission signal vector itself cannot be restored due to the influence of noise and waveform distortion, but it can be restored to the extent that there is no problem in code identification. In order to perform such polarization multiplexing / demultiplexing, it is necessary to use a synchronous detection method capable of directly obtaining optical electric field information as a method of light / current conversion. In addition, since the polarization state changes at the fastest in the order of microseconds (1 MHz), the inverse matrix needs to be updated following the change in the polarization state. For the calculation of the inverse matrix, CMA (Constant Modulus Algorithm) is used in a modulation / demodulation method such as QPSK which does not take multiple values in the intensity direction.

In order to perform such polarization multiplexing / demultiplexing, the polarization multiplexed signal receiving unit requires a high-resolution ADC (Analog to Digital Converter) that operates at high speed in synchronization with the symbol rate. This is because high-speed operation corresponding to the symbol rate of the received signal and high resolution for accurately calculating the inverse matrix are required. For example, a 40 Gbit / sec PDM-QPSK ADC shown in Non-Patent Document 3 below requires a sampling rate of 20 Gsample / sec when performing double oversampling. Furthermore, in order to perform polarization demultiplexing while sufficiently suppressing interference, the resolution of the ADC needs to be about 6 bits.

JP 2007-306248 A

However, according to the above-described conventional technique, a high-resolution ADC that operates at high speed in synchronization with the symbol rate is required on the polarization multiplexed signal receiver side. However, the development of such an ADC is technically and costly. There was a problem that it was difficult. Examples of the practical application of the ADC with the sampling rate of 20 Gsample / sec and the resolution of 6 bits shown above are limited in the industry, and are extremely difficult in terms of technology and cost. On the other hand, since the polarization change is 1000 times or more slower than the symbol rate, it is not necessary to update the matrix elements of the inverse matrix at the symbol rate class speed, and can be performed at a speed of about 10 to 100 times. is there.

The present invention has been made in view of the above, and an object of the present invention is to obtain a polarization multiplexing / demultiplexing apparatus capable of performing polarization multiplexing / demultiplexing using an ADC with reduced technical and cost difficulty. .

In order to solve the above-described problems and achieve the object, the present invention provides a polarization multiplexing method in which a received signal is polarization-demultiplexed in a receiving device that receives a polarization-multiplexed signal in a transmitting device via a transmission line. A separation device, wherein a light source means for outputting an optical signal at the same frequency as the carrier frequency of the received signal, and the received signal and the optical signal output from the light source means are each converted into two orthogonal polarization components It corresponds to the difference component of the electric field based on two polarized light signals having the same polarization direction among the polarized light separating means to be separated and output and the polarized light signals outputted from the polarized light separating means. The optical signal corresponding to the sum component is output, and the electric field of one polarization optical signal of the two polarization optical signals is held, and the electric field of the other polarization optical signal is phase-shifted. Based on the polarization optical signal, Optical interference means for outputting an optical signal corresponding to the component and an optical signal corresponding to the sum component, and the light of the sum component and the difference component generated with the polarization direction aligned and without phase shifting of the electric field Based on the signal, the difference between the values obtained by photoelectric conversion of the respective optical signals is output as an electric signal, and the sum component and the difference component of the difference component are generated by aligning the polarization direction and shifting the phase of the electric field. Based on the optical signal, the balanced photon detection means for outputting the difference between the values obtained by photoelectric conversion of the respective optical signals as an electrical signal, and the electrical signal output from the balanced photon detection means are input and branched into two. The analog signal branching means to output and one of the two branched electrical signals are input and converted to discrete values with N1 (N1> 1) bit resolution at a time period T1 oversampled with respect to the symbol rate. Then, the real time ADC means for outputting a digital signal and the other of the two branched electric signals are input and converted into discrete values with N2 (N2> N1) bit resolution in a time period T2 longer than the time period T1. A low-speed ADC means for outputting a digital signal, and a complex matrix indicating the inverse characteristic of the polarization characteristic of the transmission line is generated based on the digital signal outputted from the low-speed ADC means. Matrix generating means for updating a complex matrix; matrix multiplying means for generating a complex vector indicating a received signal vector based on the digital signal output from the real-time ADC means; and multiplying the complex vector by the complex matrix; It is characterized by providing.

The polarization multiplexing / demultiplexing device according to the present invention has an effect that polarization multiplexing / demultiplexing can be realized by using an ADC that is technically and cost-effective.

FIG. 1 is a diagram illustrating a configuration example of a polarization multiplexing transmission system. FIG. 2 is a diagram illustrating a configuration example of the polarization multiplexing / demultiplexing apparatus.

1 Light source unit 2-1, 2-2 Polarization separation unit 3-1, 3-2 Optical interference unit 4-1, 4-2, 4-3, 4-4 Balanced photon detection unit 5-1, 5- 2, 5-3, 5-4 Analog signal branching unit 6 Real time processing unit 7 Low speed processing unit 61-1, 61-2, 61-3, 61-4 Real time ADC unit 62 Matrix multiplication unit 71-1, 71 -2, 71-3, 71-4 Low-speed ADC unit 72 Matrix generation unit 100 Polarization multiplexed

signal transmission unit

101, 102 Wavelength multiplexing unit 103 Polarization multiplexing unit 200 Transmission path unit 300 Polarization multiplexed signal receiving unit 301 Wavelength multiplexing separation Part 302 Polarization demultiplexing part

Hereinafter, embodiments of a polarization multiplexing / demultiplexing device according to the present invention will be described in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

Embodiment.
First, the polarization multiplexing transmission system including the polarization multiplexing / demultiplexing device according to the present embodiment will be briefly described. FIG. 1 is a diagram illustrating a configuration example of a polarization multiplexing transmission system. The polarization multiplexing transmission system includes a polarization multiplexing signal transmission unit 100, a transmission path unit 200, and a polarization multiplexing signal reception unit 300. The polarization multiplexed signal transmission unit 100 assigns a signal to each of the orthogonal x and y polarizations independently, and transmits a wavelength multiplexed / polarized multiplexed signal obtained by multiplexing the signals. The polarization multiplexing signal transmission unit 100 includes

wavelength multiplexing units

101 and 102 and a polarization multiplexing unit 103. The wavelength multiplexing unit 101 performs wavelength multiplexing on the x polarization signal. The wavelength multiplexing unit 102 multiplexes the wavelength of the y polarization signal. The polarization multiplexing unit 103 polarization multiplexes the wavelength-multiplexed signal.

The transmission path unit 200 is, for example, a single-mode fiber (SMF), an erbium-doped fiber amplifier (EDFA), a band-pass optical filter (OBPF), and reconfigurable. It consists of an optical add / drop multiplexer (ROADM: Reconfigurable Optical Add-Drop Multiplexer), which causes waveform distortion and noise superposition, and changes the polarization state. The change in polarization state reaches the order of microseconds (1 MHz) at the fastest speed.

The polarization multiplexed signal receiving unit 300 performs polarization separation by digital signal processing in the electrical domain while tracking the polarization state that changes with time in the transmission path unit 200. The polarization multiplexing signal receiving unit 300 includes a wavelength multiplexing / separating unit 301 and a polarization multiplexing / separating unit 302. The wavelength demultiplexing unit 301 distributes the wavelength multiplexed / polarized multiplexed signal in units of wavelengths. The polarization multiplexing / separating unit 302 separates the polarization multiplexed signal distributed in wavelength units into x and y polarization components before being polarization multiplexed.

Next, the operation of the polarization multiplexing transmission system will be described. The wavelength multiplexing unit 101 performs wavelength multiplexing on x-polarized signals of all wavelengths, and the wavelength multiplexing unit 102 performs wavelength multiplexing on y-polarized signals of all wavelengths. Thereafter, the polarization multiplexing unit 103 performs polarization multiplexing on the x-polarization wavelength multiplexed signal and the y-polarization wavelength multiplexed signal and then transmits them. In this case, it is desirable to generate the x polarization signal and the y polarization signal having the same wavelength based on the same light source. This is because when the x-polarization signal and the y-polarization signal are generated based on different light sources, the influence of the phase noise of the light source is also different, and the polarization multiplexing / demultiplexing performance in the polarization multiplexing signal receiving unit 300 is deteriorated.

In the polarization multiplexed signal receiving unit 300, the wavelength multiplexing / separating unit 301 distributes the wavelength multiplexed / polarized multiplexed signal in units of wavelengths. The polarization multiplexing / separating unit 302 separates the polarization multiplexed signal distributed in wavelength units into x and y polarization components before polarization multiplexing. Here, the polarization change in the transmission line unit 200 is represented by the Jones matrix T (t) (2 × in the transmission signal vector “E _T (t) = [E _x (t), E _y (t)] ^T ”. It is expressed by “T (t) E _T (t)” obtained by multiplying two elements by a complex number. Therefore, the received signal vector “E _R (t) = [E _X (t), E _Y (t)] ^T ” is multiplied by the inverse matrix T ⁻¹ (t) of the matrix T (t) to obtain “E ′ _T (T) = T ⁻¹ (t) E _R (t) = E _T (t) ”, the transmission signal vector E _T (t) can be obtained, and polarization multiplexing / demultiplexing is possible. CMA is used for calculation of the inverse matrix T ⁻¹ (t) as in the conventional case. The Jones matrix T (t) is represented by the following formula (1).

Next, the polarization multiplexing / demultiplexing device according to the present embodiment will be described. The polarization multiplexing / demultiplexing device shows the internal configuration of the polarization multiplexing / demultiplexing unit 302 in FIG.

FIG. 2 is a diagram illustrating a configuration example of a polarization multiplexing / demultiplexing device. The polarization multiplexing / demultiplexing device includes a light source unit 1, polarization separation units 2-1, 2-2, optical interference units 3-1, 3-2, balanced photon detection units 4-1, 4-2, 4-3, 4-4, analog signal branching units 5-1, 5-2, 5-3 and 5-4, a real time processing unit 6, and a low speed processing unit 7.

The light source unit 1 is a light source that emits CW (Continuous Wave) light. The polarization separation unit 2-1 separates the received light into X and Y polarization components. The polarization separation unit 2-2 separates the CW from the light source unit 1 into X and Y polarization components. The optical interference units 3-1 and 3-2 obtain a difference component and a sum component of the two input lights based on the two input lights, and further, shift the electric field of one of the input lights by 90 degrees to obtain the difference component and the sum component. Ask for. The balanced photon detectors 4-1 to 4-4 output a difference electric signal obtained by photoelectric conversion based on the two received input lights. The analog signal branching units 5-1 to 5-4 branch the input electric signal into two and output them to the real time processing unit 6 and the low speed processing unit 7.

The real-time processing unit 6 outputs four signals that have been polarization multiplexed and demultiplexed based on the four input signals and the inverse matrix input from the low-speed processing unit 7. The real time processing unit 6 includes real time ADC (Analog to Digital Converter) units 61-1, 61-2, 61-3, 61-4, and a matrix multiplication unit 62. The real-time ADC units 61-1 to 61-4 perform analog-digital (AD) conversion on the input analog signal and output the digital signal to the matrix multiplication unit 62. The matrix multiplication unit 62 outputs four signals that have been polarization multiplexed and demultiplexed based on the four digital signals and the inverse matrix from the low-speed processing unit 7.

The low speed processing unit 7 calculates an inverse matrix of a matrix representing the polarization state of the transmission path based on the four input signals. The low speed processing unit 7 includes low speed ADC units 71-1, 71-2, 71-3, 71-4, and a matrix generation unit 72. The low-speed ADC units 71-1 to 71-4 perform AD conversion on the input analog signal and output the digital signal to the matrix generation unit 72. The matrix generation unit 72 calculates an inverse matrix based on the four digital signals and outputs the inverse matrix to the matrix multiplication unit 62.

Next, the operation of the polarization multiplexing / demultiplexing device will be described. The polarization multiplexing / demultiplexing apparatus first receives E _S (t) of an optical signal that is a polarization multiplexed signal. As an example, the carrier frequency of the received light E _S (t) is 1550.2 nm, the bit rate is 43 Gbit / sec, and the modulation method is PDM-QPSK. At this time, the symbol rate is 11 Gbit / sec. The light source unit 1 continues to emit CW light having a wavelength of 1550.2 nm that is the same as the carrier frequency of the received light E _S (t) as the local oscillation light E _LO (t).

The polarization separation unit 2-1 separates the received light E _S (t) into two orthogonal X and Y polarization components E _{S, X} (t), E _{S, Y} (t) and outputs them. The polarization separation unit 2-2 separates the local oscillation light E _LO (t) into two orthogonal X and Y polarization components E _{LO, X} (t) and E _{LO, Y} (t) and outputs them. To do.

The optical interfering unit 3-1 makes difference components “E _Xd (t) = E _{S, X with} _{respect to} two input lights E _{S, X} (t) and E _{LO, X} (t) having the same polarization. (T) −E _{LO, X} (t) ”and the sum component“ E _Xc (t) = E _{S, X} (t) + E _{LO, X} (t) ”are output. Also, the electric field of one of the two input lights is phase-shifted by 90 degrees, and the difference component “E _Xdj (t) = E _{S, X} (t) −jE _{LO, X} (t)” and the sum component “E _Xcj (t) = _ES _{, X} (t) + jE _{LO, X} (t) "is output. The optical interfering unit 3-2 _applies the difference component “E _Yd (t) = E _{S, Y to the} two input lights E _{S, Y} (t) and E _{LO, Y} (t) having the same polarization. (T) −E _{LO, Y} (t) ”and the sum component“ E _Yc (t) = E _{S, Y} (t) + E _{LO, Y} (t) ”are output. Further, the electric field of one of the two input lights is phase-shifted by 90 degrees, and the difference component “E _Ydj (t) = E _{S, Y} (t) −jE _{LO, Y} (t)” and the sum component “E _Ycj (t) = _ES _{, Y} (t) + jE _{LO, Y} (t) "is output.

The balanced photon detector 4-1 receives E _Xd (t) and E _Xc (t) as inputs and receives the balance. When the sensitivity of each photon detector is equal to R [A / W], “Re {E _X (t)} ∝R {| E _Xd (t) | ² − | E _Xc (t) | ² } "Is output. The balanced photon detector 4-2 receives E _Xdj (t) and E _Xcj (t) as inputs and receives the balance. When the sensitivity of each photon detector is equal to R [A / W], “Im {E _X (t)} ∝R {| E _Xdj (t) | ² − | E _Xcj (t) | ² } "Is output. The balanced photon detector 4-3 receives E _Yd (t) and E _Yc (t) as inputs and receives the balance. When the sensitivity of each photon detector is equal to R [A / W], “Re {E _Y (t)} ∝R {| E _Yd (t) | ² − | E _Yc (t) | ² } "Is output. The balanced photon detection unit 4-4 receives E _Ydj (t) and E _Ycj (t) as inputs and performs balanced reception. When the sensitivity of each photon detector is equal to R [A / W], “Im {E _Y (t)} ∝R {| E _Ydj (t) | ² − | E _Ycj (t) | ² } "Is output.

Analog signal branching unit 5-1, an electrical signal Re {E _X (t)} input from the balanced photon detector 4-1 2 branches, outputs one to the real time ADC 61-1, and the other Output to the low-speed ADC unit 71-1. Analog signal branching unit 5-2, an electrical signal Im {E _X (t)} input from the balanced photon detector 4-2 2 branches, outputs one to the real time ADC unit 61-2, the other Output to the low-speed ADC unit 71-2. The analog signal branching unit 5-3 branches the electrical signal Re {E _Y (t)} input from the balanced photon detection unit 4-3 into two, outputs one to the real-time ADC unit 61-3, and outputs the other Output to the low-speed ADC 71-3. The analog signal branching unit 5-4 branches the electrical signal Im {E _Y (t)} input from the balanced photon detection unit 4-4 into two, and outputs one to the real time ADC unit 61-4. Output to the low-speed ADC 71-4. As a method for the analog signal branching units 5-1 to 5-4 to branch the electric signal into two, there is a method of branching into two with equal amplitude, but the method is not limited to this.

The low-speed processing unit 7 AD-converts four pairs of 11 G symbol / sec analog signals input from the analog signal branching units 5-1 to 5-4, and performs CMA (Constant Modulus Argorithm) based on the AD-converted digital signals. used to calculate the matrix T ₀ ^-1 (t) and continues to update the T ₀ ^-1 (t) at a rate of 110 MHz.

Slow ADC unit 71-1, the analog signal Re {E _X (t)} of the input 11Gsymbol / sec, and as an example, the sampling rate 110Msample / sec, and the AD conversion resolution 6bit, and outputs to the matrix generation unit 72. Slow ADC unit 71-2, the analog signal Im {E _X (t)} of the input 11Gsymbol / sec, and as an example, the sampling rate 110Msample / sec, and the AD conversion resolution 6bit, and outputs to the matrix generation unit 72. For example, the low-speed ADC 71-3 performs AD conversion on the input 11 G symbol / sec analog signal Re {E _Y (t)} at a sampling rate of 110 Msample / sec and a resolution of 6 bits, and outputs the result to the matrix generator 72. For example, the low-speed ADC unit 71-4 converts the input analog signal Im {E _Y (t)} of 11 G symbol / sec at a sampling rate of 110 Msample / sec and a resolution of 6 bits, and outputs it to the matrix generation unit 72.

The matrix generation unit 72 uses the four digital signals (X / Y polarization, I / Q axis) input from the low-speed ADC units 71-1 to 71-4 to generate a low-speed received signal vector “E _RS (t) = [E _X (t), E _Y (t)] ^T ”, an inverse matrix T ⁻¹ (t) is calculated by CMA, and output to the matrix multiplier 62. The matrix generation unit 72 continues to update the inverse matrix T ⁻¹ (t) at a period corresponding to the sampling rate of the low-speed ADC units 71-1 to 71-4.

The real-time processing unit 6 AD-converts 11 Gsymbol / sec analog signal pairs input from the analog signal branching units 5-1 to 5-4, respectively, and converts the AD converted digital signal and the matrix T input from the low-speed processing unit 7 _A matrix operation is performed based on ₀ ⁻¹ (t), and 4 pairs of signals subjected to polarization demultiplexing are output.

Real-time ADC unit 61-1, the analog signal Re of inputted _{11Gsymbol / sec {E X (t} )}, as an example, the sampling rate 22Gsample / sec, and the AD conversion resolution 4bit, digital signal matrix generator 72 Output to. Real-time ADC unit 61-2, the analog signal Im of the inputted _{11Gsymbol / sec {E X (t} )}, as an example, the sampling rate 22Gsample / sec, and the AD conversion resolution 4bit, digital signal matrix generator 72 Output to. For example, the real-time ADC unit 61-3 AD-converts the input 11 G symbol / sec analog signal Re {E _Y (t)} at a sampling rate of 22 G sample / sec and a resolution of 4 bits, and converts the digital signal into a matrix generation unit 72. Output to. For example, the real-time ADC unit 61-4 AD-converts the input 11 G symbol / sec analog signal Im {E _Y (t)} at a sampling rate of 22 G sample / sec and a resolution of 4 bits, and converts the digital signal into a matrix generation unit 72. Output to.

The symbol rate was oversampled twice in order to set the sampling rate to 22 Gsample / sec, but it is generally about 2 to 4 times. In addition, the sampling rate in the real-time ADC units 61-1 to 61-4 is about 100 to 1000 times the sampling rate of the low-speed ADC units 71-1 to 71-4 (the low-speed ADC units 71-1 to 71-71). -4 has a cycle for outputting a digital signal 100 to 1000 times longer than the cycle of the real-time ADC units 61-1 to 61-4).

The matrix calculation unit 62 calculates the real-time received signal vector “E _RR (t) = from the four digital signals (X / Y polarization, I / Q axis) input from the real-time ADC units 61-1 to 61-4. [E _X (t), E _Y (t)] _T ”and the product“ E ′ _T0 (t) = T ₀ ⁻¹ ”with the inverse matrix T ₀ ⁻¹ (t) input from the matrix generation unit 72. (T) E _RR (t) ”. Thus, the four elements of the restored transmission vector E ′ _T0 (t), Re {E ′ _x0 (t)}, Im {E ′ _x0 (t)}, Re {E ′ _y0 (t)}, Im {E ' _y0 (t)} is output as a polarization-demultiplexed signal.

It should be noted that the frequency / phase estimation of the carrier wave may be performed after the matrix calculation unit 62, and adaptive equalization of chromatic dispersion may be performed, but this is not particularly shown.

In addition, it is common for wavelength dispersion equalization to be performed by a FIR (Finite Impulse Response) filter between the balanced photon detection units 4-1 to 4-4 and the analog signal branching units 5-1 to 5-4. Not shown.

Further, a matrix generation / calculation unit having the functions of both the matrix calculation unit 62 and the matrix generation unit 72 may be connected in series in a subsequent stage of the matrix calculation unit 62. In this case, in the first stage (first stage) matrix generation / operation unit, Re {E ′ _x0 (t)}, Im {E ′ _x0 (t)}, Re {E ′ _y0 output from the matrix operation unit 62. (T)}, Im {E ′ _y0 (t)}, the signal vector “E ′ _T0 (t) = [E ′ _x0 (t), E ′ _y0 (t)] ^T ” is reformed, An inverse matrix T ₁ ⁻¹ (t) is generated by CMA. Thereafter, the product “E ′ _T1 (t) = T ₁ ⁻¹ (t) E ′ _T0 (t)” is taken, and the four elements Re {E ′ _x1 (t)} of E ′ _T1 (t), Im {E ' _x1 (t)}, Re { _E'y1 (t)}, Im { _E'y1 (t)} are output. When generalized, the k-th (k is a natural number) matrix generation / operation unit outputs Re {E ′ _xk−1 (t)} and Im {E ′ _xk−1 (t) output from the matrix operation unit 62. }, Re {E ′ _yk−1 (t)}, Im {E ′ _yk−1 (t)}, the signal vector “E ′ _Tk (t) = [E ′ _xk−1 (t), E ' _yk-1 (t)] ^T "is recreated and the inverse matrix T _k ^-1 (t) is generated by CMA. Thereafter, the product “E ′ _Tk (t) = T _k ⁻¹ (t) E ′ _Tk−1 (t)” is taken, and the four elements Re {E ′ _xk (t)} of E ′ _Tk (t), Im _{{E 'xk (t)}} , Re {E' yk (t)}, and outputs the _{Im {E 'yk (t)} }.

As described above, in this embodiment, the polarization multiplexing / demultiplexing apparatus uses a high-speed, low-resolution ADC for a signal to be polarization-demultiplexed, and uses the inverse matrix of the polarization change of the transmission line. Polarization demultiplexing was performed using a low-speed, high-resolution ADC for the signal required for the determination. As a result, since high-speed and high-resolution ADC is not used, polarization multiplexing / demultiplexing with reduced technical and cost difficulty is possible.

As described above, the polarization multiplexing / demultiplexing device according to the present invention is useful for communication using optical fibers, and is particularly suitable for high-capacity communication using optical fibers.

Claims

A polarization multiplexing / demultiplexing device that performs polarization multiplexing / demultiplexing on a received signal in a receiving device that receives a polarization multiplexed signal by a transmission device via a transmission line,
Light source means for outputting an optical signal at the same frequency as the carrier frequency of the received signal;
Polarization separation means for separating and outputting the received signal and the optical signal output from the light source means, respectively, into two orthogonal polarization components;
Based on two polarized optical signals having the same polarization direction among the polarized optical signals output from the polarization separating means, an optical signal corresponding to the difference component of the electric field and an optical signal corresponding to the sum component In addition, based on two polarized optical signals that hold the electric field of one of the two polarized optical signals and phase-shift the electric field of the other polarized optical signal, An optical interference means for outputting an optical signal corresponding to the difference component and an optical signal corresponding to the sum component;
Based on the optical signals of the sum component and the difference component that are generated with the polarization directions aligned and without shifting the phase of the electric field, the difference between the values obtained by photoelectric conversion of the respective optical signals is output as an electrical signal, Further, based on the optical signals of the sum component and the difference component generated by aligning the polarization directions and phase-shifting the electric field, a difference between values obtained by photoelectrically converting the optical signals is output as an electric signal. Balanced photon detection means;
An analog signal branching unit that inputs the electrical signal output from the balanced photon detection unit, splits and outputs the electrical signal, and
Real-time ADC means for taking one of the two-branched electrical signals as an input, converting it into a discrete value with N1 (N1> 1) bit resolution at a time period T1 oversampled with respect to the symbol rate, and outputting a digital signal;
Low-speed ADC means for inputting the other of the two-branched electrical signals and converting it to a discrete value with N2 (N2> N1) bit resolution in a time period T2 longer than the time period T1 and outputting a digital signal;
Based on the digital signal output from the low-speed ADC unit, a matrix generating unit that generates a complex matrix indicating an inverse characteristic of the polarization characteristic of the transmission path, and that updates the complex matrix at a time period T2,
Matrix multiplication means for generating a complex vector indicating a received signal vector based on the digital signal output from the real-time ADC means, and multiplying the complex vector by the complex matrix;
A polarization multiplexing / demultiplexing device.
The polarization multiplexing / demultiplexing device according to claim 1, wherein the magnitude of the phase shift is 90 degrees.
2. The polarization multiplexing / demultiplexing device according to claim 1, wherein the oversampling is performed at 2 to 4 times a symbol rate.
3. The polarization multiplexing / demultiplexing device according to claim 2, wherein the oversampling is performed at 2 to 4 times a symbol rate.
The polarization multiplexing / demultiplexing device according to any one of claims 1 to 4, wherein the length of the time period T2 is 100 to 1000 times the time period T1.