ADAPTIVE EQUALIZATION FILTER AND SIGNAL PROCESSING DEVICE

Abstract

Each of a first filter to a fourth filter is an FIR filter having one tap. Each of a fifth filter and a sixth filter is an FIR filter having less than 46 taps. As a result, it is possible to obtain an adaptive equalization filter having a smaller total number of taps than an adaptive equalization filter using an FIR filter having 24 taps as each of the first to fourth filters.

Claims

1. An adaptive equalization filter comprising: a first filter extracting a horizontal polarization component from a first polarization; a second filter extracting a vertical polarization component from the first polarization; a third filter extracting a horizontal polarization component from a second polarization; a fourth filter extracting a vertical polarization component from the second polarization; a fifth filter receiving a horizontal polarization component extracted by the first filter and a horizontal polarization component extracted by the third filter, and outputting a horizontal polarization component to which waveform distortion compensation is applied; and a sixth filter receiving a vertical polarization component extracted by the second filter and a vertical polarization component extracted by the fourth filter, and outputting a vertical polarization component to which waveform distortion compensation is applied, wherein each of the first to fourth filters is a finite impulse response filter having one tap, and each of the fifth and sixth filters is a finite impulse response filter having less than 46 taps, and wherein each of filter coefficients in the first to sixth filters is updated on a basis of constant modulus algorithm.

2. The adaptive equalization filter according to claim 1, wherein each of the fifth and sixth filters is a finite impulse response filter having 24 taps.

3. (canceled)

4. A signal processing device comprising: a coherent receiver extracting each of a first polarization and a second polarization from an optical signal by performing coherent detection of the optical signal in which the first polarization and the second polarization are multiplexed using local oscillation light; and an adaptive equalization filter compensating for waveform distortion in each of the first polarization and the second polarization extracted by the coherent receiver, wherein the adaptive equalization filter includes: a first filter extracting a horizontal polarization component from the first polarization; a second filter extracting a vertical polarization component from the first polarization; a third filter extracting a horizontal polarization component from the second polarization; a fourth filter extracting a vertical polarization component from the second polarization; a fifth filter receiving a horizontal polarization component extracted by the first filter and a horizontal polarization component extracted by the third filter, and outputting a horizontal polarization component to which waveform distortion compensation is applied; and a sixth filter receiving a vertical polarization component extracted by the second filter and a vertical polarization component extracted by the fourth filter, and outputting a vertical polarization component to which waveform distortion compensation is applied, each of the first to fourth filters is a finite impulse response filter having one tap, and each of the fifth and sixth filters is a finite impulse response filter having less than 46 taps, and wherein each of filter coefficients in the first to sixth filters is updated on a basis of constant modulus algorithm.

5. The signal processing device according to claim 4, wherein each of the fifth and sixth filters is a finite impulse response filter having 24 taps.

6. (canceled)

Description

BRIEF DESCRIPTION OF DRAWINGS

[0022] FIG. 1 is a configuration diagram illustrating a signal processing device according to a first embodiment of the present invention.

[0023] FIG. 2 is a configuration diagram illustrating an adaptive equalization filter 4 according to the first embodiment of the present invention.

[0024] FIG. 3A is a configuration diagram illustrating a first filter 11, FIG. 3B is a configuration diagram illustrating a second filter 12, FIG. 3C is a configuration diagram illustrating a third filter 13, and FIG. 3D is a configuration diagram illustrating a fourth filter 14.

[0025] FIG. 4A is a configuration diagram illustrating a fifth filter 15, and FIG. 4B is a configuration diagram illustrating a sixth filter 16.

[0026] FIG. 5 is an explanatory diagram illustrating a result of simulating a relationship between BER and input power of a first polarization E.sub.H,in and a second polarization E.sub.V,in multiplexed in an optical signal.

DESCRIPTION OF EMBODIMENTS

[0027] Hereinafter, in order to describe the present invention in more detail, an embodiment for carrying out the present invention will be described with reference to the accompanying drawings.

First Embodiment

[0028] FIG. 1 is a configuration diagram illustrating a signal processing device according to a first embodiment of the present invention.

[0029] In FIG. 1, a coherent receiver 1 performs coherent detection of an optical signal in which a first polarization E.sub.H,in and a second polarization E.sub.V,in are multiplexed using local oscillation light, and thereby extracts each of the first polarization E.sub.H,in and the second polarization E.sub.V,in from the optical signal.

[0030] The local oscillation light has the same wavelength as the optical signal in which the first polarization E.sub.H,in and the second polarization E.sub.V,in are multiplexed.

[0031] In the coherent receiver 1, since it is difficult to completely separate a signal into a horizontal polarization component and a vertical polarization component, the first polarization E.sub.H,in extracted by the coherent receiver 1 includes a vertical polarization component in addition to a horizontal polarization component.

[0032] Further, the second polarization E.sub.V,in extracted by the coherent receiver 1 includes a horizontal polarization component in addition to a vertical polarization component.

[0033] The coherent receiver 1 outputs each of the extracted first polarization E.sub.H,in and second polarization E.sub.V,in to an analog-digital converter 2.

[0034] The analog-digital converter 2 converts each of the first polarization and the second polarization output from the coherent receiver 1 from an analog signal to a digital signal, and outputs the digital signal indicating the first polarization E.sub.H,in and the digital signal indicating the second polarization E.sub.V,in to an adaptive equalizer 3.

[0035] The adaptive equalizer 3 includes an adaptive equalization filter 4 and a coefficient updater 5.

[0036] The adaptive equalization filter 4 is a digital filter including a first filter 11 to a sixth filter 16 as illustrated in FIG. 2.

[0037] The adaptive equalization filter 4 compensates for waveform distortion in the digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2 and waveform distortion in the digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2.

[0038] The adaptive equalization filter 4 outputs each of a horizontal polarization component E.sub.X,out to which waveform distortion compensation is applied and a vertical polarization component E.sub.Y,out to which waveform distortion compensation is applied.

[0039] The coefficient updater 5 is implemented by an arithmetic processing circuit, for example.

[0040] The coefficient updater 5 performs a process of updating each of first to sixth filter coefficients in the adaptive equalization filter 4 on the basis of constant modulus algorithm (CMA).

[0041] FIG. 2 is a configuration diagram illustrating the adaptive equalization filter 4 according to the first embodiment of the present invention.

[0042] In FIG. 2, the first filter 11 is a finite impulse response (FIR) filter having one tap.

[0043] The first filter 11 extracts a horizontal polarization component included in a digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2, and outputs the extracted horizontal polarization component to the fifth filter 15. In FIG. 2, the first filter 11 is expressed as 1-tap m.sub.XH.

[0044] The second filter 12 is an FIR filter having one tap.

[0045] The second filter 12 extracts a vertical polarization component included in a digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2, and outputs the extracted vertical polarization component to the sixth filter 16. In FIG. 2, the second filter 12 is expressed as 1-tap m.sub.YH.

[0046] The third filter 13 is an FIR filter having one tap.

[0047] The third filter 13 extracts a horizontal polarization component included in a digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2, and outputs the extracted horizontal polarization component to the fifth filter 15. In FIG. 2, the third filter 13 is represented as 1-tap m.sub.XV.

[0048] The fourth filter 14 is an FIR filter having one tap.

[0049] The fourth filter 14 extracts a vertical polarization component included in a digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2, and outputs the extracted vertical polarization component to the sixth filter 16. In FIG. 2, the fourth filter 14 is represented as 1-tap m.sub.YV.

[0050] The fifth filter 15 is an FIR filter having a plurality of taps. In the first embodiment, the fifth filter 15 has 24 taps.

[0051] The fifth filter 15 receives a horizontal polarization component output from the first filter 11 and a horizontal polarization component output from the third filter 13, and outputs a horizontal polarization component E.sub.X,out to which waveform distortion compensation is applied.

[0052] In FIG. 2, the fifth filter 15 is expressed as 24-tap h.sub.X.

[0053] The sixth filter 16 is an FIR filter having a plurality of taps. In the first embodiment, the sixth filter 16 has 24 taps.

[0054] The sixth filter 16 receives a vertical polarization component output from the second filter 12 and a vertical polarization component output from the fourth filter 14, and outputs a vertical polarization component E.sub.Y,out to which waveform distortion compensation is applied.

[0055] In FIG. 2, the sixth filter 16 is expressed as 24-tap h.sub.Y.

[0056] FIG. 3 is a configuration diagram illustrating an FIR filter having one tap.

[0057] FIG. 3A is a configuration diagram illustrating the first filter 11, and FIG. 3B is a configuration diagram illustrating the second filter 12.

[0058] FIG. 3C is a configuration diagram illustrating the third filter 13, and FIG. 31) is a configuration diagram illustrating the fourth filter 14.

[0059] In FIG. 3A, a multiplier 21a multiplies a digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2 by a filter coefficient m.sub.XH of the first filter 11 output from the coefficient updater 5, and outputs the digital signal E.sub.H,inm.sub.XH multiplied by the filter coefficient m.sub.XH to a delay element 22a.

[0060] The delay element 22a delays the digital signal E.sub.H,inm.sub.XH output from the multiplier 21a by one sampling time, and outputs the delayed digital signal E.sub.H,inm.sub.XH to the fifth filter 15 as a horizontal polarization component included in the digital signal indicating the first polarization E.sub.H,in.

[0061] In FIG. 3B, a multiplier 21b multiplies a digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2 by a filter coefficient m.sub.YH of the second filter 12 output from the coefficient updater 5, and outputs the digital signal E.sub.H,inm.sub.YH multiplied by the filter coefficient m.sub.YH to a delay element 22b.

[0062] The delay element 22b delays the digital signal E.sub.H,inm.sub.YH output from the multiplier 21b by one sampling time, and outputs the delayed digital signal E.sub.H,inm.sub.YH to the sixth filter 16 as a vertical polarization component included in the digital signal indicating the first polarization E.sub.H,in.

[0063] In FIG. 3C, a multiplier 21c multiplies a digital signal indicating the second polarization output from the analog-digital converter 2 by a filter coefficient m.sub.XV of the third filter 13 output from the coefficient updater 5, and outputs the digital signal E.sub.V,inm.sub.XV multiplied by the filter coefficient m.sub.XV to a delay element 22c.

[0064] The delay element 22c delays the digital signal E.sub.V,inm.sub.XV output from the multiplier 21c by one sampling time, and outputs the delayed digital signal E.sub.V,inm.sub.XV to the fifth filter 15 as a horizontal polarization component included in the digital signal indicating the second polarization E.sub.V,in.

[0065] In FIG. 3D, a multiplier 21d multiplies a digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2 by a filter coefficient m.sub.YV of the fourth filter 14 output from the coefficient updater 5, and outputs the digital signal E.sub.V,inm.sub.YV multiplied by the filter coefficient m.sub.YV to a delay element 22d.

[0066] The delay element 22d delays the digital signal E.sub.V,inm.sub.YV output from the multiplier 21d by one sampling time, and outputs the delayed digital signal E.sub.V,inm.sub.YV to the sixth filter 16 as a vertical polarization component included in the digital signal indicating the second polarization E.sub.V,in.

[0067] FIG. 4 is a configuration diagram illustrating an FIR filter having 24 taps.

[0068] FIG. 4A is a configuration diagram illustrating the fifth filter 15, and FIG. 4B is a configuration diagram illustrating the sixth filter 16.

[0069] In FIG. 4A, a multiplier 31-1 receives the digital signal E.sub.H,inm.sub.XH output from the delay element 22a of the first filter 11 and the digital signal E.sub.V,inm.sub.XV output from the delay element 22c of the third filter 13.

[0070] The multiplier 31-1 multiplies an addition signal (E.sub.H,inm.sub.XH+E.sub.V,inm.sub.XV) which is the sum of the input digital signals E.sub.H,inm.sub.XH and E.sub.V,inm.sub.XV by the filter coefficient h.sub.X1 output from the coefficient updater 5, and outputs the addition signal multiplied by the filter coefficient h.sub.X1.

[0071] Similarly to the multiplier 31-1, the delay element 32-1 receives the digital signal E.sub.H,inm.sub.XH output from the delay element 22a of the first filter 11 and the digital signal E.sub.V,inm.sub.XV output from the delay element 22c of the third filter 13.

[0072] The delay element 32-1 delays the addition signal (E.sub.H,inm.sub.XH+E.sub.V,inm.sub.XV) which is the sum of the input digital signals E.sub.H,inm.sub.XH and E.sub.V,inm.sub.XV by one sampling time.

[0073] A multiplier 31-n (n=2, 3, . . . , 24) multiplies the addition signal, which is delayed by one sampling time by a delay element 32-(n1), by the filter coefficient h.sub.Xn, and outputs the addition signal multiplied by the filter coefficient h.sub.Xn.

[0074] A delay element 32-n (n=2, 3, . . . , 23) delays the output signal of a delay element 32-(n1) by one sampling time.

[0075] The sum of the output signals of the multipliers 31-1 to 31-24 is output as a horizontal polarization component E.sub.X,out to which waveform distortion compensation is applied.

[0076] In FIG. 4B, a multiplier 41-1 receives the digital signal E.sub.H,inm.sub.VH output from the delay element 22b of the second filter 12 and the digital signal E.sub.V,inm.sub.YV output from the delay element 22d of the fourth filter 14.

[0077] The multiplier 41-1 multiplies an addition signal (E.sub.H,inm.sub.YH+E.sub.V,inm.sub.YV) which is the sum of the input digital signals E.sub.H,inm.sub.YH and E.sub.V,inm.sub.YV by the filter coefficient h.sub.Y1 output from the coefficient updater 5, and outputs the addition signal multiplied by the filter coefficient h.sub.Y1.

[0078] Similarly to the multiplier 41-1, a delay element 42-1 receives the digital signal E.sub.H,inm.sub.YH output from the delay element 22b of the second filter 12 and the digital signal E.sub.V,inm.sub.YV output from the delay element 22d of the fourth filter 14.

[0079] The delay element 42-1 delays the addition signal (E.sub.H,inm.sub.YH+E.sub.V,inm.sub.YV) which is the sum of the input digital signals E.sub.H,inm.sub.YH and E.sub.V,inm.sub.YV by one sampling time.

[0080] A multiplier 41-n (n=2, 3, . . . , 24) multiplies the addition signal, which is delayed by one sampling time by a delay element 42-(n1), by the filter coefficient h.sub.Yn, and outputs the addition signal multiplied by the filter coefficient h.sub.Yn.

[0081] A delay element 42-n (n=2, 3, . . . , 23) delays the output signal of a delay element 42-(n1) by one sampling time.

[0082] The sum of the output signals of the multipliers 41-1 to 41-24 is output as the vertical polarization component E.sub.Y,out to which waveform distortion compensation is applied.

[0083] Next, operation will be described.

[0084] When receiving an optical signal in which the first polarization E.sub.H,in and the second polarization E.sub.V,in are multiplexed, the coherent receiver 1 performs coherent detection of the optical signal using local oscillation light, thus extracting each of the first polarization E.sub.H,in and the second polarization E.sub.V,in from the optical signal. The local oscillation light has the same wavelength as the optical signal.

[0085] The coherent receiver 1 outputs each of the extracted first polarization E.sub.H,in and second polarization E.sub.V,in to an analog-digital converter 2.

[0086] The analog-digital converter 2 converts each of the first polarization E.sub.H,in and the second polarization E.sub.V,in output from the coherent receiver 1 from an analog signal to a digital signal.

[0087] The analog-digital converter 2 outputs the digital signal indicating the converted first polarization E.sub.H,in and the digital signal indicating the converted second polarization E.sub.V,in to the adaptive equalizer 3.

[0088] The adaptive equalization filter 4 of the adaptive equalizer 3 compensates for waveform distortion in the digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2 and waveform distortion in the digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2.

[0089] The adaptive equalization filter 4 outputs each of a horizontal polarization component E.sub.X,out to which waveform distortion compensation is applied and a vertical polarization component E.sub.Y,out to which waveform distortion compensation is applied.

[0090] Hereinafter, operation of the adaptive equalizer 3 will be specifically described.

[0091] The first filter 11 in the adaptive equalization filter 4 of the adaptive equalizer 3 extracts a horizontal polarization component included in the digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2. The first filter 11 outputs the extracted horizontal polarization component to the fifth filter 15.

[0092] Specifically, the multiplier 21a of the first filter 11 multiplies the digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2 by the filter coefficient m.sub.XH output from the coefficient updater 5, and outputs the digital signal E.sub.H,inm.sub.XH multiplied by the filter coefficient m.sub.XH to the delay element 22a.

[0093] The delay element 22a of the first filter 11 delays the digital signal E.sub.H,inm.sub.XH output from the multiplier 21a by one sampling time, and outputs the delayed digital signal E.sub.H,inm.sub.XH to the fifth filter 15 as a horizontal polarization component included in the digital signal indicating the first polarization E.sub.H,in.

[0094] The second filter 12 in the adaptive equalization filter 4 of the adaptive equalizer 3 extracts a vertical polarization component included in the digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2. The second filter 12 outputs the extracted vertical polarization component to the sixth filter 16.

[0095] Specifically, the multiplier 21b of the second filter 12 multiplies a digital signal indicating the first polarization E.sub.H,in output from the analog-digital converter 2 by the filter coefficient m.sub.YH output from the coefficient updater 5, and outputs the digital signal E.sub.H,inm.sub.YH multiplied by the filter coefficient m.sub.YH to the delay element 22b.

[0096] The delay element 22b of the second filter 12 delays the digital signal E.sub.H,inm.sub.YH output from the multiplier 21b by one sampling time, and outputs the delayed digital signal E.sub.H,inm.sub.YH to the sixth filter 16 as a vertical polarization component included in the digital signal indicating the first polarization E.sub.H,in.

[0097] The third filter 13 in the adaptive equalization filter 4 of the adaptive equalizer 3 extracts a horizontal polarization component included in the digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2. The third filter 13 outputs the extracted horizontal polarization component to the fifth filter 15.

[0098] Specifically, the multiplier 21c of the third filter 13 multiplies the digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2 by the filter coefficient m.sub.XV output from the coefficient updater 5, and outputs the digital signal E.sub.V,inm.sub.XV multiplied by the filter coefficient m.sub.XV to the delay element 22c.

[0099] The delay element 22c of the third filter 13 delays the digital signal E.sub.V,inm.sub.XV output from the multiplier 21c by one sampling time, and outputs the delayed digital signal E.sub.V,inm.sub.XV to the fifth filter 15 as a horizontal polarization component included in the digital signal indicating the second polarization E.sub.V,in.

[0100] The fourth filter 14 in the adaptive equalization filter 4 of the adaptive equalizer 3 extracts a vertical polarization component included in the digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2. The fourth filter 14 outputs the extracted vertical polarization component to the sixth filter 16.

[0101] Specifically, the multiplier 21d of the fourth filter 14 multiplies the digital signal indicating the second polarization E.sub.V,in output from the analog-digital converter 2 by the filter coefficient m.sub.YV output from the coefficient updater 5, and outputs the digital signal E.sub.V,inm.sub.YV multiplied by the filter coefficient m.sub.YV to the delay element 22d.

[0102] The delay element 22d of the fourth filter 14 delays the digital signal E.sub.V,inm.sub.YV output from the multiplier 21d by one sampling time, and outputs the delayed digital signal E.sub.V,inm.sub.YV to the sixth filter 16 as a vertical polarization component included in the digital signal indicating the second polarization E.sub.V,in.

[0103] The fifth filter 15 receives a horizontal polarization component output from the first filter 11 and a horizontal polarization component output from the third filter 13, and outputs a horizontal polarization component E.sub.X,out to which waveform distortion compensation is applied.

[0104] The fifth filter 15 compensates for waveform distortion in the digital signal indicating the first polarization E.sub.H,in, and outputs the horizontal polarization component E.sub.X,out to which waveform distortion compensation is applied.

[0105] Hereinafter, operation of the fifth filter 15 will be specifically described.

[0106] When (E.sub.H,inm.sub.XH+E.sub.V,inm.sub.XV), which is an addition signal of the output signal of the first filter 11 and the output signal of the third filter 13, is input to the multiplier 31-1 of the fifth filter 15, the multiplier 31-1 multiplies the input addition signal by the filter coefficient h.sub.X1 output from the coefficient updater 5, and outputs the addition signal multiplied by the filter coefficient h.sub.X1.

[0107] The delay element 32-1 of the fifth filter 15 delays the addition signal of the output signal of the first filter 11 and the output signal of the third filter 13 by one sampling time.

[0108] The multiplier 31-n (n=2, 3, . . . , 24) of the fifth filter 15 multiplies the addition signal, which is delayed by one sampling time by the delay element 32-(n1), by the filter coefficient h.sub.Xn output from the coefficient updater 5, and outputs the addition signal multiplied by the filter coefficient h.sub.Xn.

[0109] The delay element 32-n (n=2, 3, . . . , 23) of the fifth filter 15 delays the output signal of the delay element 32-(n1) by one sampling time.

[0110] The sum of the output signals of the multipliers 31-1 to 31-24 is output from the fifth filter 15 as the horizontal polarization component E.sub.X,out to which waveform distortion compensation is applied.

[0111] The sixth filter 16 inputs a vertical polarization component output from the second filter 12 and a vertical polarization component output from the fourth filter 14, and outputs a vertical polarization component E.sub.Y,out to which waveform distortion compensation is applied.

[0112] The sixth filter 16 compensates for waveform distortion in the digital signal indicating the second polarization E.sub.V,in, and outputs the vertical polarization component E.sub.Y,out to which waveform distortion compensation is applied.

[0113] Hereinafter, operation of the sixth filter 16 will be specifically described.

[0114] When (E.sub.H,inm.sub.YH+E.sub.V,inm.sub.YV), which is an addition signal of the output signal of the second filter 12 and the output signal of the fourth filter 14, is input to the multiplier 41-1 of the sixth filter 16, the multiplier 41-1 multiplies the input addition signal by the filter coefficient h.sub.Y1 output from the coefficient updater 5, and outputs the addition signal multiplied by the filter coefficient h.sub.Y1.

[0115] The delay element 42-1 of the sixth filter 16 delays the addition signal of the output signal of the second filter 12 and the output signal of the fourth filter 14 by one sampling time.

[0116] The multiplier 41-n (n=2, 3, . . . , 24) of the sixth filter 16 multiplies the addition signal, which is delayed by one sampling time by the delay element 42-(n1), by the filter coefficient h.sub.Yn output from the coefficient updater 5, and outputs the addition signal multiplied by the filter coefficient h.sub.Yn.

[0117] The delay element 42-n (n=2, 3, . . . , 23) of the sixth filter 16 delays the output signal of the delay element 42-(n1) by one sampling time.

[0118] The sum of the output signals of the multipliers 41-1 to 41-24 is output from the sixth filter 16 as the vertical polarization component E.sub.Y,out to which waveform distortion compensation is applied.

[0119] The coefficient updater 5 of the adaptive equalizer 3 updates each of the filter coefficients in the first filter 11 to the sixth filter 16 on the basis of CMA.

[0120] Hereinafter, the filter coefficient update process by the coefficient updater 5 will be specifically described.

[0121] First, the coefficient updater 5 updates the filter coefficient m.sub.XH in the first filter 11, the filter coefficient m.sub.YH in the second filter 12, the filter coefficient m.sub.XV in the third filter 13, and the filter coefficient m.sub.yv in the fourth filter 14 in accordance with the following formulas (1) to (4).

m.sub.XHm.sub.XH+h.sub.X*e.sub.X,outE.sub.X,out.sub.H,in*(1)

m.sub.YHm.sub.YH+h.sub.Y*e.sub.Y,outE.sub.Y,out.sub.H,in*(2)

m.sub.XVm.sub.XV+h.sub.X*e.sub.X,outE.sub.X,out.sub.V,in*(3)

m.sub.YVm.sub.YV+h.sub.Y*e.sub.Y,outE.sub.Y,out.sub.V,in*(4)

h.sub.X+h.sub.X1+h.sub.X2+ . . . +h.sub.24(5)

h.sub.Y=h.sub.Y1+h.sub.Y2+ . . . +h.sub.Y24(6)

[0122] In formulas (1) to (4),

[0123] represents an update step size. The update step size is determined by, for example, a laser line width of local oscillation light, and is set to a time such as 0.5 seconds.

[0124] e.sub.X,out represents a square amplitude error, and is expressed as e.sub.Y,out=(1|E.sub.X,out|).sup.2.

[0125] e.sub.Y,out represents a square amplitude error, and is expressed as e.sub.Y,out=(1|E.sub.Y,out|).sup.2.

[0126] * is a symbol indicating complex conjugate.

[0127] E bar.sub.H,in represents an average value of the first polarization E.sub.H,in in a time corresponding to the update step size .

[0128] E bar.sub.V,in represents an average value of the second polarization E.sub.V,in in a time corresponding to the update step size .

[0129] In the text of this specification, the symbol - cannot be added above the letter E in the electronic patent application, and therefore it is expressed as E bar.

[0130] Next, the coefficient updater 5 updates the filter coefficient h.sub.Xn (n=1, 2, . . . , 24) in the fifth filter 15 and the filter coefficient h.sub.Yn (n=1, 2, . . . , 24) in the sixth filter 16 in accordance with the following formulas (7) and (8).

h.sub.Xnh.sub.Xn+e.sub.X,outE.sub.X,out(m.sub.XHE.sub.H,in+m.sub.XVE.sub.V,in)*(7)

h.sub.Ynh.sub.Yn+e.sub.Y,outE.sub.Y,out(m.sub.YHE.sub.H,in+m.sub.YVE.sub.V,in)*(8)

[0131] In the first embodiment, the adaptive equalization filter 4 includes the first filter 11 to the fourth filter 14 each having one tap, and the fifth filter 15 and the sixth filter 16 each having 24 taps. Therefore, the total number of taps of the adaptive equalization filter 4 is 52.

[0132] The adaptive equalization filter described in Patent Literature 1 includes first to fourth filters each having 24 taps. Therefore, the total number of taps of the adaptive equalization filter described in Patent Literature 1 is 96.

[0133] Therefore, the adaptive equalization filter 4 according to the first embodiment has a smaller total number of taps than the adaptive equalization filter described in Patent Literature 1.

[0134] In the first embodiment, an example in which each of the fifth filter 15 and the sixth filter 16 has 24 taps is illustrated. However, even in another configuration, if each of the fifth filter 15 and the sixth filter 16 has less than 46 taps, the total number of taps is reduced as compared with the adaptive equalization filter described in Patent Literature 1.

[0135] Here, FIG. 5 is an explanatory diagram illustrating a result of simulating a relationship between BER and input power of the first polarization E.sub.H,in and the second polarization E.sub.V,in multiplexed in an optical signal.

[0136] In the simulation in FIG. 5, assuming that coherent technology is applied to access optical communication, a DP-QPSK signal with a transmission rate of 100 Gb/s is transmitted, and minimum receiving sensitivity of the DP-QPSK signal after 80 km transmission is measured.

[0137] In FIG. 5, the horizontal axis indicates optical power (input power) input to the coherent receiver 1 per channel, and the vertical axis indicates BER before error correction.

[0138] In FIG. 5, a forward error correction (FEC) limit value is the BER of 2.010.sup.2.

[0139] In the signal processing device of Patent Literature 1 including the first to fourth filters each having 24 taps, the minimum reception sensitivity satisfying the FEC limit value is about 34 dBm.

[0140] In the signal processing device including the adaptive equalization filter 4 according to the first embodiment, the minimum reception sensitivity satisfying the FEC limit value is about 33.9 dBm.

[0141] Therefore, although the adaptive equalization filter 4 according to the first embodiment has a smaller total number of taps than the adaptive equalization filter described in Patent Literature 1, the difference in the minimum reception sensitivity satisfying the FEC limit value between them is 0.1 dBm (=|3433.9|), and it is possible for the adaptive equalization filter 4 according to the first embodiment to obtains similar reception performance to the adaptive equalization filter described in Patent Literature 1.

[0142] As apparent from the above, according to the first embodiment, each of the first filter 11 to the fourth filter 14 is an FIR filter having one tap, and each of the fifth filter 15 and the sixth filter 16 is an FIR filter having less than 46 taps. Therefore, it is possible to obtain an adaptive equalization filter having a smaller total number of taps than an adaptive equalization filter using an FIR filter having 24 taps as each of the first to fourth filters.

[0143] Therefore, according to the first embodiment, it is possible to obtain a signal processing device capable of reducing power consumption as compared with a signal processing device having an adaptive equalization filter using an FIR filter having 24 taps as each of the first to fourth filters.

[0144] Note that any component in the embodiment can be modified, or any component in the embodiment can be omitted within the scope of the present invention.

INDUSTRIAL APPLICABILITY

[0145] The present invention is suitable for an adaptive equalization filter and a signal processing device for outputting a polarization to which waveform distortion compensation is applied.

REFERENCE SIGNS LIST

[0146] 1: Coherent receiver, 2: Analog-digital converter, 3: Adaptive equalizer, 4: Adaptive equalization filter, 5: Coefficient updater, 11: First filter, 12: Second filter, 13: Third filter, 14: Fourth filter, 15: Fifth filter, 16: Sixth filter. 21a, 21b, 21c, 21d: Multiplier, 22a, 22b, 22c, 22d: Delay element, 31-1 to 31-24: Multiplier, 32-1 to 32-23: Delay element, 41-1 to 41-24: Multiplier, 42-1 to 42-23: Delay element.