JP2015005805A - Optical transmission system, and digital signal processor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an optical transmission system and a digital signal processor which are arranged so that after collectively receiving optical signals phase-modulated with the same bit pattern and sent, a redundant optical signal receiver performs phase synchronization and waveform equalization among the received signals by means of digital signal processing to synthesize signals.SOLUTION: An optical transmission system comprises a redundant optical signal receiver which collectively receives identical optical signals 1 to N transmitted through optical transmission lines, and synthesizes signals. The redundant optical signal receiver includes: N optical signal receivers which perform a coherent detection on optical signals 1 to N transmitted through the optical transmission lines, perform an analog-to-digital conversion on the resultant signals, and then output digital signals respectively; and a digital signal processor which accepts the input of the digital signals output from the N optical signal receivers, and performs phase synchronization and waveform equalization on the input signals thereby to synthesize and output signals.

Description

  The present invention relates to an optical transmission system for transmitting the same optical signal from a redundant optical signal transmitter via an optical transmission line, collectively receiving the signal with a redundant optical signal receiver, and performing signal synchronization by performing phase synchronization and waveform equalization. And a digital signal processing apparatus.

  In an ultra-high-speed transmission system with a transmission rate per wavelength of 100 Gbit / s or more, a digital coherent technology combining a coherent optical communication technology and a digital signal processing technology has been widely used. For example, in the DP-QPSK (Dual Polarization-Quadrature Phase Shift Keying) method (Non-patent Document 1), which is a standard modulation / demodulation method in a 100 Gbit / s long-haul optical transmission system, 32 Gbit is used by using quaternary phase modulation. Two coherent optical signals are generated by multiplexing two signals of / s, and further multiplexed by using two polarizations to generate a coherent optical signal of 128 Gbit / s. On the receiving side, the signal coherently detected using local light of the same wavelength as the signal light is digitized using an AD (Analog to Digital) converter, and then subjected to frequency offset compensation and wavelength dispersion of the transmission path by digital signal processing. Excellent transmission characteristics are realized by performing compensation, polarization dispersion compensation, polarization signal separation, and the like.

OIF, “100G Ultra Long Haul DWDM Framework Document”. H.Y.Choi, T.Tsuritani, and I.Morita, “BER-adaptive flexible-format transmitter for elastic optical networks,” Optics Express, vol.20, no.17, pp.18652-18658, 2012. T.Tsuboi, M.Natori, and S.Mitachi, “Single-fiber optical protection ring architecture suitable for asymmetric traffic,” Global Telecommunications Conference, vol.7, pp.4044-4048, 2003. S.J.Savory, “Digital Coherent Optical Receivers: Algorithms and Subsystems,” IEEE Journal of Selected Topics in Quantum Electronics, vol.16, no.5, pp.1164-1178, 2010.

  As a further large-capacity transmission method, multi-value modulation signals such as polarization multiplexed 16QAM (Quadrature Amplitude Modulation) are being studied. However, since the reception sensitivity is lowered due to the multi-level modulation signal, the transmission distance is limited (Non-Patent Document 2).

  In addition, in a conventional optical transmission system, there is a case in which a transmission line is redundant in advance to switch from a working line to a protection line when a failure occurs in the optical network in order to improve the reliability of the optical network. (Non-patent Document 3) Since it takes about several tens of milliseconds for physical switching when a failure occurs, the optical signal is interrupted during that time.

  The present invention performs digital signal processing to perform phase synchronization and waveform equalization between received signals after a plurality of optical signals that have been phase modulated with the same bit pattern are collectively received by a redundant optical signal receiver. It is an object of the present invention to provide an optical transmission system and a digital signal processing apparatus that can synthesize signals.

  The first invention is an optical transmission system in which the same optical signals 1 to N (N is an integer of 2 or more) are transmitted via an optical transmission line, and are collectively received by a redundant optical signal receiver to perform signal synthesis. The redundant optical signal receiver performs coherent detection on the optical signals 1 to N transmitted through the optical transmission path, further converts the signals from analog to digital, and outputs digital signals, and N optical signal receivers. A digital signal processing apparatus that inputs digital signals output from the optical signal receivers, synthesizes and outputs the signals that have undergone phase synchronization and waveform equalization;

In the optical transmission system of the first invention, the optical signal receiver i (i is an integer of 1 to N) of the redundant optical signal receiver performs coherent detection of the optical signal i transmitted through the optical transmission line, and The digital signal processing apparatus is configured to output a digital signal after analog-digital conversion. The digital signal processing apparatus inputs an n-th (n is a natural number) digital signal x _{in, i} (n) output from the optical signal receiver i. adaptive equalization signal x _out, inputs the adaptive equalizer i for outputting a _i (n), the adaptive equalization signal x _out output from the adaptive equalizer _{i, i} an (n), the phase difference phi _i ( n) and the phase synchronization signal x _'out, and a phase synchronizer i for outputting a _i (n), the phase synchronization signal x is output from the phase synchronizer i' enter _{out, i} (n) is synthesized, synthesized and a signal combiner i for outputting a signal X _out (n), adaptive equalizer i the phase difference output from the phase synchronizer i phi _i (n) and the phase synchronization No. x _'out, enter the _i (n), and optionally of a phase synchronizer k (k is an integer of 1 to N) phase locked signal x is output from the' Type _{out, k} a (n), a digital signal This is a configuration for outputting an adaptive equalization signal x _{out, i} (n) obtained by performing phase synchronization and waveform equalization of x _{in, N} (n).

In the optical transmission system of the first invention, the optical signal receiver i (i is an integer of 1 to N) of the redundant optical signal receiver performs polarization separation on the optical signal i transmitted through the optical transmission line. The digital signal processing apparatus is configured to output a digital signal by performing coherent detection and further analog-to-digital conversion. The digital signal processing apparatus outputs an n-th (n is a natural number) digital signal x that is output from the optical signal receiver i. adaptive equalizers xi and yi that input _{in, i} (n) and y _{in, i} (n) and output adaptive equalization signals x _{out, i} (n) and y _{out, i} (n), and adaptive The adaptive equalization signals x _{out, i} (n) and y _{out, i} (n) output from the equalizers xi and yi are input, the phase differences φ _xi (n) and φ _yi (n) and the phase synchronization signal x _{'out, i} (n) and y' _out, and a phase synchronizer xi and yi outputs a _i (n), the phase output from the phase synchronizer xi and yi Period signal x _{'out, i} (n) and y' _out, enter the _i (n), and a respective combined signal X _out (n) and Y _out (n) signal combiner xi and yi outputs a, The adaptive equalizers xi and yi are connected to the phase differences φ _xi (n) and φ _yi (n) output from the phase synchronizers xi and yi and the phase synchronization signals x ′ _{out, i} (n) and y ′ _{out, i.} (n) is inputted, and phase synchronization signals x ′ _{out, k} (n) and y ′ _{out, k} (n) outputted from arbitrary phase synchronizers xk and yk (k is an integer of 1 to N) are further inputted. The input equalized signals x _{out, i} (n) and y _{out, i} (n) are obtained by performing phase synchronization and waveform equalization of the digital signals x _{in, N} (n) and y _{in, N} (n). It is the structure which outputs.

According to a second aspect of the present invention, in the digital signal processing apparatus of the redundant optical signal receiver of the optical transmission system according to the first aspect of the invention, the nth output (n is an integer from 1 to N) output from the optical signal receiver i. A natural number) digital signal x _{in, i} (n) and an adaptive equalization signal x _{out, i} (n), and an adaptive equalization signal output from the adaptive equalizer i a phase synchronizer _i that inputs x _{out, i} (n) and outputs a phase difference φ _i (n) and a phase synchronization signal x ′ _{out, i} (n), and a phase synchronization signal output from the phase synchronizer i x ' _{out, i} (n) is inputted and synthesized, and a signal synthesizer i that outputs a synthesized signal X _out (n) is provided, and the adaptive equalizer i outputs a phase difference output from the phase synchronizer i. phi _i (n) and the phase synchronization signal x _'out, enter the _i (n), and optionally of a phase synchronizer k (k is an integer of 1 to n) phase locked signal x is output from the' _{out, k} ( n) Enter a digital signal _in, the phase synchronization and waveform equalization adaptive equalization signal x _out of _performing, configured to output a _i (n) of _N (n).

In the digital signal processing apparatus according to the second aspect of the invention, the adaptive equalizer i includes M−1 delay units (M is an integer of 2 or more) delaying the nth digital signal x _{in, N} (n) sequentially. , M tap coefficients h _i (= [h _i (1), h _i (2),..., H _i (M)]) and the nth digital signal x _{in, N} (n) and M−1. The digital signals multiplied by M tap coefficients h _i are added to the digital signals x _{in, i} (n−1),..., X _{in, i} (n−M + 1) delayed by the delay units. (transposed matrix of x ^T _{in, i} (n) is x _{in, i} (n)) adaptive equalization signal _{x out, i (n) =} h i x T in, i (n)
From an adder that outputs a reference signal, a phase synchronization signal x ′ _{out, k} (n) output from an arbitrary phase synchronizer k, a reference signal generator that outputs a reference signal d _k (n), and an adder The output adaptive equalization signal x _{out, i} (n), the phase difference φ _i (n) and phase synchronization signal x ′ _{out, i} (n) output from the phase synchronizer i, and the reference signal generator enter the reference signal d _k which is output (n), to update the tap coefficients h _i, and the tap coefficient calculating section for outputting the error signal epsilon _i a (n), the error signal epsilon _i output from the tap coefficient calculating section a convergence determination unit that performs convergence determination by inputting (n), the convergence determination unit performs convergence determination using the error signal ε _i (n) output from the tap coefficient calculation unit, and tap coefficients after convergence The calculation unit is configured to update the tap coefficient h _i using the reference signal d _k (n) output from the reference signal generation unit.

According to a third aspect of the present invention, there is provided a digital signal processing apparatus for a redundant optical signal receiver in the optical transmission system according to the first aspect of the present invention. The nth (n is a natural number) digital signal x _{in, i} (n) and y _{in, i} (n) are input, and adaptive equalization signals x _{out, i} (n) and y _{out, i} (n) are output. The adaptive equalizers xi and yi and the adaptive equalization signals x _{out, i} (n) and y _{out, i} (n) output from the adaptive equalizers xi and yi are input, and the phase difference φ _xi (n) And φ _yi (n) and phase synchronizers xi and yi that output phase synchronization signals x ′ _{out, i} (n) and y ′ _{out, i} (n), and phase synchronization output from the phase synchronizers xi and yi Signal synthesizers xi and yi for inputting signals x ′ _{out, i} (n) and y ′ _{out, i} (n) and outputting synthesized signals X _out (n) and Y _out (n), respectively, are adaptive Equalizers xi and y , Type phase difference phi _xi output from the phase synchronizer xi and yi (n) and φ _yi (n) phase synchronization signal x _{'out, i} (n) and y' _{out, i} a (n), Further, phase synchronization signals x ′ _{out, k} (n) and y ′ _{out, k} (n) output from arbitrary phase synchronizers xk and yk (k is an integer of 1 to N) are input, and digital signal x _{in , N} (n) and y _{in, N} (n) are subjected to phase synchronization and waveform equalization to output adaptive equalization signals x _{out, i} (n) and y _{out, i} (n).

In the digital signal processing apparatus of the third invention, the adaptive equalizer xi sequentially delays the nth digital signal x _{in, N} (n) 2 (M−1) (M is an integer of 2 or more). Delay unit x and M tap coefficients h _{xx, i} (= [h _{xx, i} (1), h _{xx, i} (2),..., H _{xx, i} (M)]) and h _{xy, i} (= [H _{xy, i} (1), h _{xy, i} (2),..., H _{xy, i} (M)]) and the nth digital signal x _{in, N} (n) and y _{in, N} ( n) and digital signals x _{in, i} (n-1),..., x _{in, i} (n-M + 1) and y _{in, i} (n-1) delayed by M-1 delay units x, respectively. ),..., Y _{in, i} (n−M + 1) are added with digital signals multiplied by M tap coefficients h _{xx, i} and h _{xy, i} , respectively, and the adaptive equalized signal x _{out, i} ( n) = h _{xx, i} x ' ^T _{in, i} (n) + h _{xy, i} y' ^T _{in, i} (n)
And a reference signal generation unit x that outputs a reference signal d _xk (n), and an adder x that outputs the reference signal d ′ and a phase synchronization signal x ′ _{out, k} (n) output from an arbitrary phase synchronizer xk Adaptive equalization signal x _{out, i} (n) output from the device x, phase difference φ _xi (n) and phase synchronization signal x ′ _{out, i} (n) output from the phase synchronizer xi, reference signal A tap coefficient calculator x that receives the reference signal d _xk (n) output from the generator x, updates the tap coefficients h _{xx, i} and h _{xy, i} , and outputs an error signal ε _xi (n); A convergence determination unit x that receives the error signal ε _xi (n) output from the tap coefficient calculation unit x and performs convergence determination. The convergence determination unit x includes the error signal ε output from the tap coefficient calculation unit x. _xi (n) is used to determine convergence. After convergence, the tap coefficient calculation unit x uses the reference signal d _xk (n) output from the reference signal generation unit x to perform tap coefficient h _{xx, The} adaptive equalizer yi is configured to update _i and h _{xy, i} , and the adaptive equalizer yi sequentially delays the nth digital signal y _{in, N} (n) (M is an integer of 2 or more). ) Delay unit y and M tap coefficients h _{yx, i} (= [h _{yx, i} (1), h _{yx, i} (2),..., H _{yx, i} (M)]) and h _{yy , i} (= [h _{yy, i} (1), h _{yy, i} (2),..., h _{yy, i} (M)]) and the nth digital signal x _{in, N} (n) and y _{in, N} (n) and digital signals x _{in, i} (n-1),..., X _{in, i} (n-M + 1) and y _{in, i} (n −1),..., Y _{in, i} (n−M + 1) are added with digital signals obtained by multiplying M tap coefficients h _{yx, i} and h _{yy, i} , respectively, and the adaptive equalization signal y _{out, i} (n) = h _{yx, i} x ′ ^T _{in, i} (n) + h _{yy, i} y ′ ^T _{in, i} (n)
And a reference signal generation unit y for inputting a phase synchronization signal y ′ _{out, k} (n) output from an arbitrary phase synchronizer yk and outputting a reference signal d _yk (n), and addition The adaptive equalization signal y _{out, i} (n) output from the unit y, the phase difference φ _yi (n) and the phase synchronization signal y ′ _{out, i} (n) output from the phase synchronizer y _, and the reference signal A tap coefficient calculator y that inputs the reference signal d _yk (n) output from the generator y, updates the tap coefficients h _{yx, i} and h _{yy, i} , and outputs an error signal ε _yi (n); A convergence determination unit y that receives the error signal ε _yi (n) output from the tap coefficient calculation unit y and performs convergence determination, and the convergence determination unit y includes the error signal ε output from the tap coefficient calculation unit y. performs the convergence determination using _yi (n), the tap coefficient calculation unit y after convergence, the tap coefficients using the reference signal d _yk output from the reference signal generator y (n) h _{yx, i} And h _yy, is configured to update the _i.

  The present invention performs digital signal processing to perform phase synchronization and waveform equalization between received signals after a plurality of optical signals that have been phase modulated with the same bit pattern are collectively received by a redundant optical signal receiver. Signal synthesis. As a result, transmission quality can be improved to facilitate large-capacity transmission, and optical signal disconnection when an optical network failure occurs can be prevented to improve stability.

It is a figure which shows the structure of Example 1 of the optical transmission system of this invention. It is a figure which shows the structure of Example 2 of the optical transmission system of this invention. It is a figure which shows the structural example of the adaptive equalizer 23-i of Example 1,2. It is a figure which shows the structural example of the digital signal processing apparatus 22 in Example 3 of the optical transmission system of this invention. It is a figure which shows the structural example of the adaptive equalizer 23x-i of Example 3. FIG. It is a figure which shows the structural example of the adaptive equalizer 23y-i of Example 3. FIG. It is a figure which shows the constellation map after digital signal processing of DP-QPSK signal. It is a figure which shows the constellation map after the digital signal processing of the DP-QPSK signal at the time of cut | disconnecting the optical transmission line 3. FIG.

FIG. 1 shows a configuration of a first embodiment of an optical transmission system according to the present invention.
In FIG. 1, a redundant optical signal transmitter 10A receives the same electrical signal, converts it to the same optical signal 1 to N (N is an integer of 2 or more), and outputs it in parallel to the optical transmission line. 11-1 to 11-N.

  The optical transmission path for transmitting the same optical signals 1 to N is, for example, a configuration using N single-core optical fibers, or L × K (= N) parallel using K multi-core fibers with L cores. It may be configured to perform transmission. Alternatively, the optical signals 1 to N may be wavelength-multiplexed and transmitted using a one-core fiber or a multi-core fiber.

  The redundant optical signal receiver 20 coherently detects the optical signals 1 to N transmitted through the optical transmission path, and performs AD conversion to output a digital signal, respectively. The digital signal processing device 22 is configured to receive digital signals output from the optical signal receivers 21-1 to 21-N and perform signal synchronization by performing phase synchronization and waveform equalization.

The digital signal processing device 22 which is a feature of the present invention includes an adaptive equalizer 23-1 to 23 -N, a phase synchronizer 24-1 to 24 -N, and a signal synthesizer 25. The adaptive equalizers 23-1 to 23-N respectively receive the nth digital signals x _{in, 1} (n) to x _{in, N} (n) output from the optical signal receivers 21-1 to 21-N. The adaptive equalization signals x _{out, 1} (n) to x _{out, N} (n) are output. The phase synchronizers 24-1 to 24-N receive the adaptive equalization signals x _{out, 1} (n) to x _{out, N} (n) output from the adaptive equalizers 23-1 to 23-N, respectively. , Phase differences φ ₁ (n) to φ _N (n) and phase synchronization signals x ′ _{out, 1} (n) to x ′ _{out, N} (n) are output. The signal synthesizer 25 inputs the phase synchronization signals x ′ _{out, 1} (n) to x ′ _{out, N} (n) output from the phase synchronizers 24-1 to 24-N, and equal gain synthesis or maximum ratio The combined signal X _out (n) is output. Further, the phase differences φ ₁ (n) to φ _N (n) output from the phase synchronizers 24-1 to 24-N and the phase synchronization signals x ′ _{out, 1} (n) to x ′ _{out, N} (n) And a phase synchronization signal x ′ _{out, k} (n) output from an arbitrary phase synchronizer 24-k (k = 1 to N) as a reference is an adaptive equalizer 23-1 to 23-. N is fed back to each.

The phase synchronizer 24-i (i = 1 to N) estimates a component out of phase synchronization caused by a frequency offset at the time of coherent detection in the optical signal receiver 21-i as a phase difference φ _i (n). Output. The phase difference φ _i (n) can be estimated by, for example, a calculation algorithm described in Non-Patent Document 4. In the phase synchronizer 24-i, the estimated phase difference φ _i (n) is used to correct the phase difference φ _i (n) from the adaptive equalization signal x _{out, i} (n), and the phase synchronization signal x ′. _{out, i} (n) = x _{out, i} (n) exp (-jφ _i (n)) is output (j is an imaginary unit). Then, the phase difference φ _i (n) and the phase synchronization signal x ′ _{out, i} (n) output from the phase synchronizer 24-i are input to the adaptive equalizer 23-i, and an arbitrary phase synchronizer 24 By inputting the phase synchronization signal x ′ _{out, k} (n) output from −k to the adaptive equalizer 23-i, phase synchronization and waveform equalization are performed. The configuration and operation of the adaptive equalizer 23-i will be described later.

FIG. 2 shows the configuration of Embodiment 2 of the optical transmission system of the present invention.
In FIG. 2, the redundant optical signal transmitter 10 </ b> B includes an optical signal transmitter 11 that inputs an electrical signal, converts the electrical signal into an optical signal, and outputs the optical signal. And an optical signal branching device 12 that outputs to the optical transmission line. The configuration of the redundant optical signal receiver 20 is the same as that of the first embodiment shown in FIG.

In the configuration of the digital signal processing device 22 shown in FIG. 1 and FIG. 2, in order to compensate for signal degradation caused by chromatic dispersion occurring in the optical transmission line, a wavelength is placed before the adaptive equalizers 23-1 to 23-N. a dispersion compensator disposed conversion, a digital signal x _{in, 1} (n) ~x _{in the} digital signal x and _n (n) is the chromatic dispersion compensation _{'in, 1 (n) ~x} ' in, the _n (n) Then, it may be configured to input to the adaptive equalizers 23-1 to 23-N, respectively. Further, an arbitrary compensator such as compensation for waveform distortion due to a nonlinear optical effect can be incorporated into the digital signal processing device 22.

FIG. 3 shows a configuration example of the adaptive equalizer 23-i of the first and second embodiments (i = 1 to N).
3, the adaptive equalizer 23-i is, M-1 or a delay unit 31 (M is an integer of 2 or more), M number of tap coefficients _{h i (= [h i (} 1), h i ( 2,..., H _i (M)]) 32, an adder 33, a tap coefficient calculation unit 34, a convergence determination unit 35, and a reference signal generation unit 36.

The adder 33 includes the digital signal x _{in, i} (n) and the digital signals x _{in, i} (n−1),..., X _{in, i} (n−M + 1) delayed by the M−1 delay units 31. to) by adding the digital signals obtained by multiplying each of M tap coefficients h _i 32, the adaptive equalization signal _{x out, i (n) =} h i x T in, i (n) (x T in, i ( n) is x _{in, i} (n) transpose matrix)
Is output.

The reference signal generator 36 uses the phase synchronization signal x ′ _{out, k} (n) output from the arbitrary arbitrary phase synchronizer 24-k as a reference signal d _k according to a predetermined tap coefficient update algorithm. Output (n).

The tap coefficient calculation unit 34 includes the adaptive equalization signal x _{out, i} (n) output from the adder 33, the phase difference φ _i (n) output from the phase synchronizer 24-i, and the phase synchronization signal x ′. _The tap coefficient h _i is updated using _{out, i} (n) and the reference signal d _k (n) output from the reference signal generator 36, and the error signal ε _i (n) is output.

The convergence determination unit 35 performs convergence determination using the error signal ε _i (n) output from the tap coefficient calculation unit 34, and after convergence, the tap coefficient calculation unit 34 outputs a reference output from the reference signal generation unit 36. The tap coefficient h _i is updated using the signal d _k (n).

  A tap coefficient updating method will be described by taking a QPSK signal as an example. In the tap coefficient calculation unit 34, the tap coefficient before convergence can be updated by the following equation using, for example, a CMA (Constant Modulus Algorithm) which is a blind algorithm.

μはステップサイズパラメータ、ｘ^* _in,i(n)はｘ_in,i(n) の複素共役を示す。

μ represents a step size parameter, and x ^* _{in, i} (n) represents a complex conjugate of x _{in, i} (n).

  After convergence, for example, by using a decision-directed least mean square (DD-LMS) algorithm, which is a decision feedback algorithm, the tap coefficients are updated by the following equation to synchronize the phase between signals and perform waveform equalization. Can do.

At this time, the reference signal d _k (n) is generated by the reference signal generation unit 36 using the phase synchronization signal x ′ _{out, k} (n) output from the arbitrary phase synchronizer 24-k. In the case of a QPSK signal, the reference signal d _k (n) is determined by the following equation.

Further, after convergence, the phase of the adaptive equalization signal x _{out, 1} (n) to x _{out, N} (n) is synchronized by updating the tap coefficient by the following formula using CMA which is a blind algorithm. Waveform equalization can also be performed.

At this time, since the phase synchronization signal x ′ _{out, k} (n) is used as the reference signal d _k (n), the reference signal generator 36 uses the phase synchronization signal x ′ output from the arbitrary phase synchronizer 24-k. _{out, k} (n) is output as is.

  As described above, the case of the QPSK signal has been described as an example of updating the tap coefficient. However, the digital signal processing apparatus of the present invention can be applied to a signal modulated by an arbitrary modulation method. Further, although CMA is used before convergence and DD-LMS algorithm or CMA is used after convergence, any adaptive equalization algorithm can be used.

  FIG. 4 shows a configuration example of the digital signal processing device 22 in the third embodiment of the optical transmission system of the present invention.

  A redundant optical signal transmitter (not shown) in the optical transmission system of the present embodiment outputs the same optical signal modulated and polarization multiplexed by the 128 Gpbs DP-QPSK system to the optical transmission lines 1 to 3. Each optical transmission line is composed of four sections with one section between the optical amplifiers, and the loss of the optical fiber in one section is 29 dB. In addition, in the relationship with Example 1 and Example 2, the structural example equivalent to the case where N = 3 and k = 2 is shown.

The optical signal receivers 21-1 to 21-3 of the redundant optical signal receivers separate the polarization-multiplexed optical signals via the optical transmission lines 1 to 3 into x-polarized waves and y-polarized waves, respectively. Coherent detection, AD conversion, and the n-th x-polarized digital signal x _{in, 1} (n) to x _{in, 3} (n) and y-polarized digital signal y _{in, 1} (n) to y _{in , 3} (n) is output. An x-polarized signal is indicated by a solid line, and a y-polarized signal is indicated by a broken line. The digital signal processing device 22 includes n-th x polarization digital signals x _{in, 1} (n) to x _{in, 3} (n) and y polarization output from the optical signal receivers 21-1 to 21-3. Digital signals y _{in, 1} (n) to y _{in, 3} (n) are inputted, and the signals are synthesized by phase synchronization and waveform equalization. In FIG. 4, the notation (n) indicating the nth digital signal is omitted.

  The digital signal processing device 22 characterized by the present embodiment includes a chromatic dispersion compensator 26x-1 to 26x-3 corresponding to x polarization, an adaptive equalizer 23x-1 to 23x-3, and a phase synchronizer 24x-1. 24x-3, signal synthesizer 25x, chromatic dispersion compensators 26y-1 to 26y-3 corresponding to y polarization, adaptive equalizers 23y-1 to 23y-3, phase synchronizers 24y-1 to 24y- 3. Consists of a signal synthesizer 25y.

The chromatic dispersion compensators 26x-1 to 26x-3 corresponding to the x polarization input the digital signal xin _{, 1} (n) to xin _{, 3} (n) of the nth x polarization, and the chromatic dispersion is obtained. Compensated chromatic dispersion compensation signals x ′ _{in, 1} (n) to x ′ _{in, 3} (n) are output. The chromatic dispersion compensators 26y-1 to 26y-3 corresponding to the y polarization input digital signals y _{in, 1} (n) to y _{in, 3} (n) of the nth y polarization _{, and} perform chromatic dispersion. Compensated chromatic dispersion compensation signals y ′ _{in, 1} (n) to y ′ _{in, 3} (n) are output.

The adaptive equalizers 23x-1 to 23x-3 corresponding to x-polarizations are chromatic dispersion compensation signals x'in _{, 1} (n) to x'in _{, 3} (n) and y'in _{, 1} (n) to y ′ _{in, 3} (n) is input, and adaptive equalization signals x _{out, 1} (n) to x _{out, 3} (n) are output. The phase synchronizers 24x-1 to 24x-3 receive the adaptive equalization signals xout _{, 1} (n) to xout _{, 3} (n), respectively, and the phase differences φ _x1 (n) to φ _x3 (n) and The phase synchronization signals x ′ _{out, 1} (n) to x ′ _{out, 3} (n) are output. The signal synthesizer 25x receives the phase synchronization signals x ′ _{out, 1} (n) to x ′ _{out, 3} (n), and outputs the synthesized signal X _out (n) by performing equal gain synthesis or maximum ratio synthesis. Further, the phase differences φ _x1 (n) to φ _x3 (n) output from the phase synchronizers 24x-1 to 24x-3 and the phase synchronization signals x ′ _{out, 1} (n) to x ′ _{out, 3} (n) The phase synchronization signal x ′ _{out, 2} (n) output from the reference phase synchronizer 24-2 is fed back to the adaptive equalizers 23x-1 to 23x-3.

The adaptive equalizers 23y-1 to 23y-3 corresponding to the y-polarizations are chromatic dispersion compensation signals x'in _{, 1} (n) to x'in _{, 3} (n) and y'in _{, 1} (n) to y ′ _{in, 3} (n) is input, and adaptive equalization signals y _{out, 1} (n) to y _{out, 3} (n) are output. The phase synchronizers 24y-1 to 24y-3 receive the adaptive equalization signals y _{out, 1} (n) to y _{out, 3} (n), respectively, and the phase differences φ _y1 (n) to φ _y3 (n) and The phase synchronization signal y ′ _{out, 1} (n) to y ′ _{out, 3} (n) is output. The signal synthesizer 25y receives the phase synchronization signals y ′ _{out, 1} (n) to y ′ _{out, 3} (n) _{, and} performs equal gain synthesis or maximum ratio synthesis to output a synthesized signal Y _out (n). Further, the phase differences φ _y1 (n) to φ _y3 (n) and the phase synchronization signals y ′ _{out, 1} (n) to y ′ _{out, 3} (n) output from the phase synchronizers 24y-1 to 24y-3. The phase synchronization signal y ′ _{out, 2} (n) output from the reference phase synchronizer 24-2 is fed back to the adaptive equalizers 23y-1 to 23y-3.

  As described above, in the digital signal processing device 22 according to the third embodiment, the adaptive equalizers 23x-1 to 23x-3 and 23y-1 are used to compensate for signal degradation caused by chromatic dispersion occurring in the optical transmission lines 1 to 3. To 23y-3, the chromatic dispersion compensators 26x-1 to 26x-3 and 26y-1 to 26y-3 are arranged, but the other phase synchronizers 24x-1 to 24x-3 and 24y-1 are arranged. To 24y-3 and the signal synthesizers 25x and 25y have the same functions as those of the first and second embodiments.

FIG. 5 illustrates a configuration example of the adaptive equalizer 23x-i according to the third embodiment (i = 1, 2, 3).
In FIG. 5, the adaptive equalizer 23x-i includes 2 (M-1) delay units 31 and M tap coefficients h _{xx, i} (= [h _{xx, i} (1), h _{xx, i} (2), ..., h _{xx, i} (M)]) and M tap coefficients h _{xy, i} (= [h _{xy, i} (1), h _{xy, i} (2), ..., h _{xy, i} (M)]) 32, an adder 33, a tap coefficient calculation unit 34, a convergence determination unit 35, and a reference signal generation unit 36.

The adder 33 receives the chromatic dispersion compensation signals x ′ _{in, i} (n) and y ′ _{in, i} (n) and the chromatic dispersion compensation signals x ′ _{in, i} (delayed by M−1 delay units 31). n-1), ..., x'in _{, i} (n-M + 1) and y'in _{, i} (n-1), ..., y'in _{, i} (n-M + 1), respectively. The chromatic dispersion compensation signals multiplied by the tap coefficients h _{xx, i} and h _{xy, i} 32 respectively are added, and the adaptive equalization signal x _{out, i} (n) = h _{xx, i} x ′ ^T _{in, i} (n) + H _{xy, i} y ' ^T _{in, i} (n)
Is output.

The reference signal generator 36 uses the phase synchronization signal x ′ _{out, 2} (n) output from the arbitrary arbitrary phase synchronizer 24-2 as a reference signal d ₂ according to the tap coefficient update algorithm. Output (n).

The tap coefficient calculation unit 34 includes the adaptive equalization signal x _{out, i} (n) output from the adder 33, the phase difference φ _xi (n) output from the phase synchronizer 24-i, and the phase synchronization signal x ′. _The tap coefficients h _{xx, i} and h _{xy, i} are updated using _{out, i} (n) and the reference signal d _x2 (n) output from the reference signal generator 36 _, and the error signal ε _xi (n) is updated. Output.

The convergence determination unit 35 performs convergence determination using the error signal ε _xi (n) output from the tap coefficient calculation unit 34, and the reference is output from the reference signal generation unit 36 in the tap coefficient calculation unit 34 after the convergence. The tap coefficients h _{xx, i} and h _{xy, i} are updated using the signal d _xk (n).

FIG. 6 illustrates a configuration example of the adaptive equalizer 23y-i according to the third embodiment (i = 1, 2, 3).
In FIG. 6, the adaptive equalizer 23y-i includes 2 (M-1) delay devices 31 and M tap coefficients h _{yx, i} (= [h _{yx, i} (1), h _{yx, i} (2), ..., h _{yx, i} (M)]) and M tap coefficients h _{yy, i} (= [h _{yy, i} (1), h _{yy, i} (2), ..., h _{yy, i} (M)]) 32, an adder 33, a tap coefficient calculation unit 34, a convergence determination unit 35, and a reference signal generation unit 36. Each unit has the same function as each unit of the adaptive equalizer 23x-i illustrated in FIG.

5 and 6, for example, as in the first and second embodiments, DD− using CMA before convergence and reference signals d _x2 (n) and d _y2 (n) after convergence. By switching to the LMS algorithm and updating, it is possible to perform waveform equalization in synchronization with the phase between signals.

In the tap coefficient calculation unit 34, the tap coefficient update formula by the CMA before convergence is as follows.

Further, the tap coefficient update formula by the DD-LMS algorithm after convergence using the reference signals d _x2 (n) and d _y2 (n) is as follows.

Here, the reference signals d _x2 (n) and d _y2 (n) are output from the phase synchronization signals x ′ _{out, 2} (n) output from the phase synchronizers 24 x−2 and 24 y−2 in the reference signal generator 34. And y ′ _{out, 2} (n) are generated in the same manner as in Examples 1 and 2.

  As in the first and second embodiments, an arbitrary adaptive equalization algorithm can be used as the tap coefficient update algorithm before and after convergence.

  FIG. 7 shows a constellation map after digital signal processing of the DP-QPSK signal. The transmission characteristics when the signals after transmission through the optical transmission lines 1 to 3 were received by normal digital coherent were 8.45 dB, 8.68 dB, and 8.90 dB as Q values, respectively.

  Next, digital signal processing was performed with the configuration described in Example 3. As shown in FIG. 7, since the phase between the signals is not synchronized before convergence, the combined signal of the optical transmission lines 1 to 3 does not have the QPSK phase point arrangement, but the reference signal is used after convergence. When switching to the DD-LMS algorithm, the QPSK phase point arrangement is obtained because the waveforms can be equalized and synthesized in synchronism with the phase between the signals. The transmission characteristic at this time was 12.72 dB in terms of Q value, confirming an improvement in transmission characteristic of 3.82 dB.

  FIG. 8 shows a constellation map after digital signal processing of the DP-QPSK signal when the optical transmission line 3 is disconnected. This assumes that a failure has occurred in the optical transmission line.

  In the digital signal processing in the optical transmission system of the present invention, since the tap coefficient is updated using the reference signal, as shown in FIG. 8, noise components generated when the optical transmission line is interrupted can be suppressed. When the optical transmission line 3 was disconnected, the transmission characteristic of the combined signal of the optical transmission lines 1 to 3 was 11.37 dB in terms of Q value. Even when a failure occurs in the optical transmission line, the transmission characteristics are improved as compared with the single transmission. For example, the transmission distance is determined by the transmission characteristic obtained by combining two optical transmission lines, and three transmission lines are normally used. In this way, even if a failure occurs in any of the routes, transmission can be continued without disconnecting the signal.

10A, 10B Redundant optical signal transmitter 11, 11-1 to 11-N Optical signal transmitter 12 Optical signal splitter 20 Redundant optical signal receiver 21-1 to 21-N Optical signal receiver 22 Digital signal processing device 23- 1-23-N, 23x-1 to 23x-3, 23y-1 to 23y-3 Adaptive equalizer 24-1 to 24-N, 24x-1 to 24x-3, 24y-1 to 24y-3 Phase synchronization vessels 25,25x, 25y signal combiner 26x-1~26x-3 chromatic dispersion compensator 26y-1~26y-3 chromatic dispersion compensator 31 delayer 32 tap coefficients h _i
33 Adder 34 Tap Coefficient Calculation Unit 35 Convergence Determination Unit 36 Reference Signal Generation Unit

Claims (7)

In an optical transmission system in which the same optical signals 1 to N (N is an integer of 2 or more) are transmitted through an optical transmission line, and received by a redundant optical signal receiver in a lump for signal synthesis,
The redundant optical signal receiver is:
N optical signal receivers that coherently detect the optical signals 1 to N transmitted through the optical transmission line, and further perform analog-digital conversion to output a digital signal;
And a digital signal processing device for inputting the digital signals respectively output from the N optical signal receivers and synthesizing and outputting the signals subjected to phase synchronization and waveform equalization. Transmission system. The optical transmission system according to claim 1,
The optical signal receiver i (i is an integer from 1 to N) of the redundant optical signal receiver performs coherent detection on the optical signal i transmitted through the optical transmission path, and further converts the signal into analog to digital form to obtain a digital signal. Is output, and
The digital signal processing device includes:
An adaptive equalizer _i that inputs an n-th (n is a natural number) digital signal x _{in, i} (n) output from the optical signal receiver i and outputs an adaptive equalization signal x _{out, i} (n). When,
Phase synchronization that inputs the adaptive equalization signal x _{out, i} (n) output from the adaptive equalizer i and outputs a phase difference φ _i (n) and a phase synchronization signal x ′ _{out, i} (n) Vessel i,
A signal synthesizer i that inputs and synthesizes the phase synchronization signal x ′ _{out, i} (n) output from the phase synchronizer i and outputs a synthesized signal X _out (n);
The adaptive equalizer i receives the phase difference φ _i (n) and the phase synchronization signal x ′ _{out, i} (n) output from the phase synchronizer _i , and an arbitrary phase synchronizer k (k Is a phase synchronization signal x ′ _{out, k} (n) output from 1 to N), and the digital signal x _{in, N} (n) is subjected to phase synchronization and waveform equalization. An optical transmission system characterized in that the signal x _{out, i} (n) is output. The optical transmission system according to claim 1,
The optical signal receiver i (i is an integer from 1 to N) of the redundant optical signal receiver performs polarization separation on the optical signal i transmitted through the optical transmission path, and performs coherent detection, respectively. It is configured to output a digital signal after analog-digital conversion.
The digital signal processing device includes:
The polarization-separated nth (n is a natural number) digital signal x _{in, i} (n) and y _{in, i} (n) output from the optical signal receiver i are input, and the adaptive equalization signal x _out, adaptive equalizers xi and yi that output _i (n) and y _{out, i} (n);
The adaptive equalization signals x _{out, i} (n) and y _{out, i} (n) output from the adaptive equalizers xi and yi are input, and phase differences φ _xi (n) and φ _yi (n) and Phase synchronizers xi and yi for outputting phase synchronization signals x ′ _{out, i} (n) and y ′ _{out, i} (n);
The phase synchronization signals x ′ _{out, i} (n) and y ′ _{out, i} (n) output from the phase synchronizers xi and yi are input, and the combined signals X _out (n) and Y _out (n) are input, respectively. And signal synthesizers xi and yi for outputting
The adaptive equalizers xi and yi are connected to the phase differences φ _xi (n) and φ _yi (n) output from the phase synchronizers xi and yi and the phase synchronization signals x ′ _{out, i} (n) and y ′ _{out. , i} (n) and phase synchronization signals x ′ _{out, k} (n) and y ′ _{out, k} (n) output from arbitrary phase synchronizers xk and yk (k is an integer of 1 to N). ) And the adaptive equalized signals x _{out, i} (n) and y _{out, i obtained} by performing phase synchronization and waveform equalization of the digital signals x _{in, N} (n) and y _{in, N} (n) An optical transmission system characterized in that (n) is output. In the digital signal processing apparatus of the redundant optical signal receiver of the optical transmission system according to claim 1,
An n-th (n is a natural number) digital signal x _{in, i} (n) output from the optical signal receiver i (i is an integer from 1 to N) is input, and an adaptive equalization signal x _{out, i} (n ) To output an adaptive equalizer i;
The phase synchronization that inputs the adaptive equalization signal x _{out, i} (n) output from the adaptive equalizer i and outputs the phase difference φ _i (n) and the phase synchronization signal x ′ _{out, i} (n) Vessel i,
A signal synthesizer i that inputs and synthesizes the phase synchronization signal x ′ _{out, i} (n) output from the phase synchronizer i and outputs a synthesized signal X _out (n);
The adaptive equalizer i receives the phase difference φ _i (n) and the phase synchronization signal x ′ _{out, i} (n) output from the phase synchronizer _i , and an arbitrary phase synchronizer k (k Is a phase synchronization signal x ′ _{out, k} (n) output from 1 to N), and the digital signal x _{in, N} (n) is subjected to phase synchronization and waveform equalization. A digital signal processing device characterized by being configured to output a signal x _{out, i} (n). The digital signal processing device according to claim 4.
The adaptive equalizer i is
M-1 delay units (M is an integer of 2 or more) for sequentially delaying the nth digital signal x _{in, N} (n);
M tap coefficients h _i (= [h _i (1), h _i (2),..., H _i (M)]),
The nth digital signal x _{in, N} (n) and the digital signals x _{in, i} (n−1),..., X _{in, i} (n−M + 1) delayed by the M−1 delay units. to, the M tap coefficients h _i by adding the digital signals obtained by multiplying each adaptive equalization signal _{x out, i (n) =} h i x T in, i (n) (x T in, i (n ) Is the transposed matrix of x _{in, i} (n))
An adder that outputs
A reference signal generation unit that inputs the phase synchronization signal x ′ _{out, k} (n) output from the arbitrary phase synchronizer k and outputs a reference signal d _k (n);
The adaptive equalization signal x _{out, i} (n) output from the adder, the phase difference φ _i (n) and the phase synchronization signal x ′ _{out, i} (n) output from the phase synchronizer i. A tap coefficient calculator that inputs the reference signal d _k (n) output from the reference signal generator, updates the tap coefficient h _i , and outputs an error signal ε _i (n);
A convergence determination unit that performs the convergence determination by inputting the error signal ε _i (n) output from the tap coefficient calculation unit, and
The convergence determination unit performs a convergence determination using the error signal ε _i (n) output from the tap coefficient calculation unit, and after convergence, the tap coefficient calculation unit is a reference output from the reference signal generation unit A digital signal processing apparatus, wherein the tap coefficient h _i is updated using a signal d _k (n). In the digital signal processing apparatus of the redundant optical signal receiver of the optical transmission system according to claim 1,
Inputs the n-th (n is a natural number) digital signal x _{in, i} (n) and y _{in, i} (n) separated from polarization output from the optical signal receiver i (i is an integer of 1 to N). And adaptive equalizers xi and yi that output adaptive equalization signals x _{out, i} (n) and y _{out, i} (n),
The adaptive equalization signals x _{out, i} (n) and y _{out, i} (n) output from the adaptive equalizers xi and yi are input, and phase differences φ _xi (n) and φ _yi (n) and Phase synchronizers xi and yi for outputting phase synchronization signals x ′ _{out, i} (n) and y ′ _{out, i} (n);
The phase synchronization signals x ′ _{out, i} (n) and y ′ _{out, i} (n) output from the phase synchronizers xi and yi are input, and the combined signals X _out (n) and Y _out (n) are input, respectively. And signal synthesizers xi and yi for outputting
The adaptive equalizers xi and yi are connected to the phase differences φ _xi (n) and φ _yi (n) output from the phase synchronizers xi and yi and the phase synchronization signals x ′ _{out, i} (n) and y ′ _{out. , i} (n) and phase synchronization signals x ′ _{out, k} (n) and y ′ _{out, k} (n) output from arbitrary phase synchronizers xk and yk (k is an integer of 1 to N). ) And the adaptive equalized signals x _{out, i} (n) and y _{out, i obtained} by performing phase synchronization and waveform equalization of the digital signals x _{in, N} (n) and y _{in, N} (n) A digital signal processing device characterized in that it is configured to output (n). The digital signal processing apparatus according to claim 6.
The adaptive equalizer xi is
2 (M-1) delay units x (M is an integer of 2 or more) for sequentially delaying the n-th digital signal x _{in, N} (n);
M tap coefficients h _{xx, i} (= [h _{xx, i} (1), h _{xx, i} (2),..., H _{xx, i} (M)]) and h _{xy, i} (= [h _{xy , i} (1), h _{xy, i} (2),…, h _{xy, i} (M)]),
The nth digital signals x _{in, N} (n) and y _{in, N} (n) and the digital signals x _{in, i} (n−1),... Delayed by the M−1 delay units x, respectively. x _{in, i} (n−M + 1) and y _{in, i} (n−1),..., y _{in, i} (n−M + 1), the M tap coefficients h _{xx, i} and h _{xy , i} are added to each other and the adaptive equalization signal x _{out, i} (n) = h _{xx, i} x ′ ^T _{in, i} (n) + h _{xy, i} y ′ ^T _{in, i} (n)
(T represents a transposed matrix)
An adder x that outputs
A reference signal generation unit x that inputs the phase synchronization signal x ′ _{out, k} (n) output from the arbitrary phase synchronizer xk and outputs a reference signal d _xk (n);
The adaptive equalization signal x _{out, i} (n) output from the adder x, the phase difference φ _xi (n) output from the phase synchronizer xi, and the phase synchronization signal x ′ _{out, i} (n ) And the reference signal d _xk (n) output from the reference signal generation unit x, the tap coefficients h _{xx, i} and h _{xy, i} are updated, and an error signal ε _xi (n) is output. A tap coefficient calculation unit x,
A convergence determination unit x that inputs the error signal ε _xi (n) output from the tap coefficient calculation unit x and performs a convergence determination;
The convergence determination unit x performs convergence determination using the error signal ε _xi (n) output from the tap coefficient calculation unit x, and after convergence, the tap coefficient calculation unit x receives from the reference signal generation unit x. The tap coefficients h _{xx, i} and h _{xy, i} are updated using the output reference signal d _xk (n),
The adaptive equalizer yi is
2 (M-1) delay units y (M is an integer of 2 or more) for sequentially delaying the nth digital signal y _{in, N} (n);
M tap coefficients h _{yx, i} (= [h _{yx, i} (1), h _{yx, i} (2),..., H _{yx, i} (M)]) and h _{yy, i} (= [h _{yy , i} (1), h _{yy, i} (2),…, h _{yy, i} (M)]),
The nth digital signals x _{in, N} (n) and y _{in, N} (n) and the digital signals x _{in, i} (n−1),... Delayed by the M−1 delay units y, respectively. x _{in, i} (n-M + 1) and y _{in, i} (n-1),..., y _{in, i} (n-M + 1), the M tap coefficients h _{yx, i} and h _{yy , i} are added to each other, and an adaptive equalization signal y _{out, i} (n) = h _{yx, i} x ′ ^T _{in, i} (n) + h _{yy, i} y ′ ^T _{in, i} (n)
An adder y that outputs
A reference signal generation unit y that inputs the phase synchronization signal y ′ _{out, k} (n) output from the arbitrary phase synchronizer yk and outputs a reference signal d _yk (n);
The adaptive equalization signal y _{out, i} (n) output from the adder y, the phase difference φ _yi (n) output from the phase synchronizer y, and the phase synchronization signal y ′ _{out, i} (n ) And the reference signal d _yk (n) output from the reference signal generation unit y, the tap coefficients h _{yx, i} and h _{yy, i} are updated, and an error signal ε _yi (n) is output. Tap coefficient calculator y to
A convergence determination unit y that inputs the error signal ε _yi (n) output from the tap coefficient calculation unit y and performs a convergence determination;
The convergence determination unit y performs convergence determination using the error signal ε _yi (n) output from the tap coefficient calculation unit y. After the convergence, the tap coefficient calculation unit y receives the reference signal generation unit y from the reference signal generation unit y. A digital signal processing device _, wherein the tap coefficients h _{yx, i} and h _{yy, i} are updated using the output reference signal d _yk (n).

Images