Signal processing circuits, optical transmission devices, and optical transmission systems

Abstract

To improve DGD tolerance without increasing the power consumption of the FIR filter. [Solution] The signal processing circuit includes a first FIR filter 122 with n taps that performs adaptive control processing of the received signal, a second FIR filter 131 with N taps (where N is a value greater than n) used to update the tap coefficients of the first FIR filter, and a coefficient adaptive control processing unit 133 that sets the tap coefficients of the first FIR filter 122 and the second FIR filter 131. The coefficient adaptive control processing unit 133 sets the tap coefficients of the first FIR filter 122 based on the output data of the second FIR filter 131.

Description

Technical Field

[0001] The present invention relates to a signal processing circuit, an optical transmission device, and an optical transmission system.

Background Art

[0002] In an optical transmission device such as an optical transceiver that transmits and receives optical signals of optical digital coherent, the baud rate handled tends to increase for large capacity. DGD (Differential Group Delay, which is the amount by which two signal components separated by polarization mode dispersion are separated) is adaptively equalized and compensated by an adaptive equalization processing unit (AEQ: Adaptive EQualizer) provided in the optical receiver. The AEQ calculates tap coefficients following fluctuations in the transmission line characteristics by a blind equalization algorithm based on input data and output data, and performs equalization processing by multiplying the input data by the tap coefficients using a FIR filter. Polarization Mode Dispersion is PMD. The FIR filter is a Finite Impulse Response (finite impulse response) filter, which has a large circuit scale and power consumption and depends on the number of taps.

[0003] When the baud rate of optical transmission increases, in order to cope with DGD, it is necessary to increase the number of taps of the FIR filter of the AEQ for compensating DGD, and the power consumption of the FIR filter increases.

[0004] In the prior art, techniques for compensating for DGD include, for example, a CD equalizer, a least-squares mean-squares (LMS) module, and an adaptive PMD equalizer that compensate for chromatic dispersion and polarization mode dispersion. There are also systems that measure the coupling energies of first and second subsets of filter taps and perform adaptive equalization of polarization mode dispersion by providing a centroid adjustment module that shifts the centroid if the coupling energy of the first subset exceeds the coupling energy of the second subset by a threshold amount. Furthermore, there are systems that calculate the centroid position of the filtering determined by the tap coefficients of the adaptive equalization processing unit during initial training before communication starts, and adjust the tap coefficients to bring them closer to the tap center so that the difference from the tap center determined by the number of taps in the adaptive equalization processing unit is minimized. There are also systems that include a first filter that compensates for polarization-independent signal distortion, a second filter that compensates for polarization-dependent signal distortion with an adaptive equalization filter, and update the tap coefficients of the first filter section based on the transfer function corresponding to the polarization-independent signal distortion in the adaptive equalization filter (see, for example, Patent Documents 1 to 4 below). [Prior art documents] [Patent Documents]

[0005] [Patent Document 1] U.S. Patent Application Publication No. 2019 / 0036615 [Patent Document 2] U.S. Patent No. 8705977 [Patent Document 3] Japanese Patent Publication No. 2012-119923 [Patent Document 4] Japanese Patent Publication No. 2014-233039 [Overview of the project] [Problems that the invention aims to solve]

[0006] For example, as the baud rate of optical transmission increases, the number of taps required for the FIR filter to maintain DGD tolerance increases, leading to increased power consumption.

[0007] In one aspect, the present invention aims to improve the DGD tolerance of the FIR filter without increasing its power consumption. [Means for solving the problem]

[0008] According to one aspect of the present invention, the signal processing circuit includes a first FIR (Finite Impulse Response) filter with n taps that performs adaptive control processing of a received signal, a second FIR filter with N taps (where N is a value greater than n) used to update the tap coefficients of the first FIR filter, and a coefficient adaptive control processing unit that sets the tap coefficients of the first FIR filter and the second FIR filter, wherein the coefficient adaptive control processing unit is required to set the tap coefficients of the first FIR filter based on the output data of the second FIR filter. [Effects of the Invention]

[0009] According to one aspect of the present invention, the DGD tolerance can be improved without increasing the power consumption of the FIR filter. [Brief explanation of the drawing]

[0010] [Figure 1] Figure 1 is a diagram showing a signal processing circuit according to Embodiment 1. [Figure 2] Figure 2 shows a signal processing circuit as an example. [Figure 3] Figure 3 shows an example of AEQ. [Figure 4] Figure 4 is a diagram showing an example of the tap coefficients of an FIR filter during blind equalization. [Figure 5] Figure 5 is a diagram showing an example of tap centroid correction processing according to Embodiment 1. [Figure 6] Figure 6 is an explanatory diagram of the operation during initial retraction according to Embodiment 1. [Figure 7] Figure 7 is an explanatory diagram of the operation during normal operation according to Embodiment 1. [Figure 8]FIG. 8 is a flowchart showing a signal processing example according to Embodiment 1. [Figure 9] FIG. 9 is a diagram showing a signal processing circuit according to Embodiment 2. [Figure 10] FIG. 10 is an explanatory diagram of the operation at the time of initial pull-in according to Embodiment 2. [Figure 11] FIG. 11 is an explanatory diagram of the operation during normal operation according to Embodiment 2. [Figure 12] FIG. 12 is a flowchart showing a signal processing example according to Embodiment 2. [Figure 13A] FIG. 13A is an explanatory diagram of the tap selection operation of the FIR filter. (Part 1) [Figure 13B] FIG. 13B is an explanatory diagram of the tap selection operation of the FIR filter. (Part 2) [Figure 14] FIG. 14 is a diagram showing a configuration example of an optical receiver. [Figure 15] FIG. 15 is a diagram showing a configuration example of an optical transmission system.

MODE FOR CARRYING OUT THE INVENTION

[0011] Hereinafter, embodiments of the disclosed signal processing circuit, optical transmission device, and optical transmission system will be described in detail with reference to the drawings. The signal processing circuit of the embodiment is applied to optical digital coherent optical transmission, and compensates for DGD by an AEQ provided in an optical transmission device such as an optical receiver. In the embodiment, at the time of startup (initial pull-in) and operation of an optical transmission device including AEQ in a state where the transmission line characteristics of the transmission line are unknown, the DGD tolerance is improved without increasing the power consumption of the FIR filter.

[0012] In the embodiment, for example, quadrature amplitude modulation (QAM) is used for optical communication transmission.

[0013] (Embodiment 1) Figure 1 shows a signal processing circuit according to Embodiment 1. The signal processing circuit of Embodiment 1 corresponds to the receiving DSP 103 (Digital Signal Processor) shown in Figure 1. In the optical transmission system, the transmitting optical transmission device transmits an optical signal via the optical transmission path L, and the receiving optical transmission device (optical receiver) R receives it. The receiving optical transmission device R includes an O / E conversion unit 101, an AD converter (ADC) 102, and a receiving DSP 103. The O / E conversion unit 101 converts the received optical signal photoelectrically and outputs it to the ADC 102. The ADC 102 converts the electrical signal (received signal) after photoelectric conversion from analog to digital and outputs it to the receiving DSP 103.

[0014] The receiving DSP 103 processes the data of the received signal and includes an FEQ (Fixed Equalizer) 111, an Adaptive Equalization Processing Unit (AEQ) 112, a CPR / FOC 113, and a control unit 114. CPR stands for Carrier Phase Recovery, and FOC stands for Frequency Offset Compensation.

[0015] The FEQ111 performs dispersion compensation, linear compensation, and nonlinear compensation. The Adaptive Equalization Processing Unit (AEQ)112 performs compensation for the Differential Group Delay (DGD) of two orthogonal polarization states, as well as residual dispersion compensation. The CPR / FOC113 compensates for the difference between the carrier frequency of the received optical signal and the frequency of the local emission, restoring the carrier phase. The control unit 114 controls the FEQ111, Adaptive Equalization Processing Unit (AEQ)112, and CPR / FOC113.

[0016] The Adaptive Equalization Processing Unit (AEQ) 112 includes a tap coefficient update unit 121 and a first FIR filter (main signal side FIR filter 122). The tap coefficient update unit 121 and the main signal side FIR filter 122 receive the FEQ 111 output of the main signal (received signal) on the optical transmission path L as input. Based on adaptive equalization control, the tap coefficient update unit 121 sets the updated tap coefficients in the main signal side FIR filter 122. The main signal side FIR filter 122 outputs the filtered received signal to the CPR / FOC 113.

[0017] The tap coefficient update unit 121 includes a second FIR filter (coefficient update side FIR filter) 131, a selector 132, a coefficient adaptive control processing unit 133, and a centroid correction processing unit 134.

[0018] The coefficient update side FIR filter 131 receives the received signal (output of FEQ 111) and the tap coefficients determined by the coefficient adaptation control processing unit 133 as input. The coefficient update side FIR filter 131 of the tap coefficient update unit 121 takes a portion of the received signal as input.

[0019] The selector 132 receives input from the output of the coefficient update side FIR filter 131 and the output of the main signal side FIR filter 122. It selects either the output of the coefficient update side FIR filter 131 or the output of the main signal side FIR filter 122 and outputs it to the coefficient adaptive control processing unit 133.

[0020] The coefficient adaptive control processing unit 133 performs tap coefficient adaptive control using a well-known blind equalization algorithm. The coefficient adaptive control processing unit 133 receives the input and output received signals for the main signal side FIR filter 122, calculates updated tap coefficients that follow the fluctuations in the transmission path characteristics, and sets them in the main signal side FIR filter 122. The updated tap coefficients have orthogonal H and V polarization-specific tap coefficients.

[0021] Furthermore, the coefficient adaptive control processing unit 133 determines the tap coefficients of the FIR filter for the coefficient update side FIR filter 131 or the main signal side FIR filter 122 selected by the selector 132, and outputs them to the corresponding coefficient update side FIR filter 131 or main signal side FIR filter 122. Initial values ​​of the tap coefficients are input to this coefficient adaptive control processing unit 133.

[0022] The centroid correction processing unit 134 receives the updated tap coefficients obtained by the coefficient adaptive control processing unit 133. The centroid correction processing unit 134 corrects the centroid for each orthogonal polarization (H,V) of the updated tap coefficients based on the amount of deviation from the tap center. The correction of the tap center deviation and the bias in the convergence of the tap coefficients are disclosed, for example, in the above-mentioned Patent Document 3. The centroid correction processing unit 134 then outputs the corrected tap coefficients to the coefficient adaptive control processing unit 133.

[0023] The control unit 114 selectively activates each part of the AEQ 112 during the initial engagement of the adaptive equalization process and during normal operation after the initial engagement. The functions that the control unit 114 activates / deactivates during the initial engagement and normal operation will be described later.

[0024] In Embodiment 1, the AEQ 112 is equipped with a dedicated FIR filter, the coefficient update side FIR filter 131, as a processing unit for initial pull-in in the tap coefficient update unit 121. Unlike the main signal side FIR filter 122, the coefficient update side FIR filter 131 extracts input data from the received signal at a fixed period and uses it. For example, it extracts a predetermined number of symbols (1 / 32 to 1 / 64 symbols) per frame by decimating them.

[0025] Thus, during the initial pull-in, the coefficient update side FIR filter 131 is used to obtain a "coarse" update tap coefficient corresponding to the unknown transmission path characteristics, enabling the initial pull-in of the adaptive equalization process. When the coefficient update side FIR filter 131 and the main signal side FIR filter 122 have the same number of taps, the power consumption during the initial pull-in can be reduced to 1 / 32 to 1 / 64 of that of the main signal side FIR filter (depending on the constant period for data extraction).

[0026] During normal operation, the initial tap coefficient update unit 121 is stopped. Here, if the number of taps of the coefficient update side FIR filter 131 is N, the number of taps of the main signal side FIR filter 122 used during normal operation can be set to a number of taps n (for example, n = N / 2) which is less than the number of taps N of the coefficient update side FIR filter 131.

[0027] The "coarse" updated tap coefficients obtained by the coefficient adaptation control processing unit 133 during initial engagement allow the number of taps n in the main signal side FIR filter 122 used during normal operation to be less than the number of taps N in the coefficient update side FIR filter 131. In addition, the centroid correction processing of the tap centers in the centroid correction processing unit 134 during initial engagement allows the number of taps n in the main signal side FIR filter 122 to be less than the number of taps N in the coefficient update side FIR filter 131 (for example, n = N / 2, details will be described later).

[0028] In this embodiment, the method of switching the number of taps from N to n (for example, N / 2) using only the main signal side FIR filter 122 is not implemented because it has disadvantages in terms of circuit size and power consumption. This is because the main signal side FIR filter 122 processes all received data, and increasing the number of taps would result in a significant increase in power consumption and circuit size.

[0029] Although the filtering process is the same for the main signal side FIR filter 122 and the coefficient update side FIR filter 131, the number of FIR filters (parallelism) differs because the amount of data processed in one cycle is different. In the above example, the number of parallel main signal side FIR filters 122 is 32 to 64 times that of the coefficient update side FIR filter 131.

[0030] When the transmission characteristics of the optical transmission path L are unknown, and during initial engagement such as the startup of the signal processing circuit (AEQ112), the coefficient adaptive control processing unit 133 performs initial engagement processing based on the initial value of the tap coefficient and the output of the coefficient update side FIR filter 131. During normal operation after initial engagement, the coefficient adaptive control processing unit 133 sets the updated tap coefficient in the main signal side FIR filter 122 and performs FIR filtering processing on the received signal.

[0031] (Example assignment) Next, I will explain the assignment using an example.

[0032] Figure 2 shows a signal processing circuit based on an example. Figure 2 shows an example of the overall configuration of an optical transmission system. The transmitting optical transmission device (optical transmitter) T transmits an optical signal via the optical transmission path L, and the receiving optical transmission device (optical receiver) R receives the optical signal via the optical transmission path L.

[0033] The optical transmitter T includes a transmitting DSP 201, a digital-to-analog converter (DAC) 202, and an E / O conversion unit 203. The transmitting DSP 201 includes a PCS unit 211 that converts input data into PCS, a bit / symbol conversion unit 212 that maps the bits of the PCS-converted input data to symbols, and a transmitting frame generation unit 213 that generates a transmission frame from the input data after symbol conversion. The DAC 202 converts the input data from digital to analog and outputs it to the E / O conversion unit 203. The E / O conversion unit 203 converts the input data into an optical signal and sends it to the optical transmission path L.

[0034] The optical receiver R includes an O / E conversion unit 221, an ADC 222, and a receiving DSP 223. The O / E conversion unit 221 converts the received optical signal photoelectrically and outputs it to the ADC 222. The ADC 222 outputs the electrical signal (received signal) after photoelectric conversion to the receiving DSP 223.

[0035] The receiving DSP223 processes the received signal data and includes an FEQ231, an adaptive equalization processing unit232, and a CPR / FOC unit233. In the reference example, the adaptive equalization processing unit232 includes an AEQ241, a frame synchronization and initial tap coefficient generation unit242.

[0036] During the initial pull-in of the AEQ241, the frame synchronization and initial tap coefficient generation unit 242 generates frame synchronization and the initial tap coefficient of the AEQ241.

[0037] Figure 3 shows an example AEQ. The AEQ241 includes a tap coefficient update unit 301 and an FIR filter 302. The tap coefficient update unit 301 includes a coefficient adaptation control circuit 311. The coefficient adaptation control circuit 311 uses the input and output data of the received signal to the FIR filter 302 to determine the updated tap coefficients based on a blind equalization algorithm and sets the updated tap coefficients in the FIR filter 302. This FIR filter 302 corresponds to the main signal side FIR filter 122 in Embodiment 1 (Figure 1).

[0038] As the baud rate of optical transmission increases, the number of taps in the AEQ241's FIR filter 302, which compensates for DGD, needs to be increased to handle the same amount of DGD. For example, to handle twice the baud rate while maintaining DGD tolerance (performance), the number of taps in the FIR filter 302 needs to be doubled, which increases the power consumption of the FIR filter 302. In an environment with double the baud rate, the amount of DGD that can be handled by one tap interval of the FIR filter 302 is halved compared to an environment with a single baud rate, so twice the number of taps is needed in the FIR filter 302 to handle the same amount of DGD. As a result, the circuit size of the FIR filter 302 doubles compared to an environment with a single baud rate, and when combined with the doubling of the baud rate, the power consumption of the FIR filter 302 quadruples.

[0039] As shown in Figure 3, the AEQ has 241 units, and the tap coefficient update unit 301 receives the input data and the output data from the FIR filter 302. It calculates tap coefficients that track the fluctuations in the transmission line characteristics, and the FIR filter 302 performs equalization by multiplying the input data and the tap coefficients. To accommodate, for example, a doubling of the baud rate while maintaining DGD tolerance, it is necessary to double the number of taps in the FIR filter 302, which increases power consumption.

[0040] (Regarding the shift in the center of gravity of taps during blind equalization) Figure 4 is a diagram showing an example of the tap coefficients of an FIR filter during blind equalization. The blind equalization process performed by the coefficient adaptive control circuit 311 of the AEQ241 may cause the center of gravity of the taps to shift relative to the tap center, and the increased interpolarization delay due to DGD increases the likelihood that the weight of the tap coefficients will shift towards the edges of the taps. Designing the FIR filter 302 to operate even when the center of gravity of the taps is shifted to the maximum extent will increase the number of taps.

[0041] In Figure 4, the horizontal axis represents the tap number of the FIR filter 302, and the vertical axis represents the amplitude of each polarization (X-axis polarization tap coefficients: HH, VH; Y-axis polarization tap coefficients: HV, VV). In the example shown in Figure 4, the FIR filter 302 has a total of 15 taps, with the tap center being tap number 8. Due to the interpolarization delay caused by DGD, the peaks of each tap coefficient are shifted relative to the tap center: HH is at tap number 8, VH is at tap number 12, HV is at tap number 9, and VV is at tap number 13. Additionally, the weights of the tap coefficients are shifted towards the ends of the tap (towards tap number 15).

[0042] When addressing the interpolarization delay using DGD as shown in Figure 4, existing technologies require a large number of taps in the FIR filter 302 (15 taps), making low-power operation difficult. Under normal operation, a large number of taps in the FIR filter 302 increases power consumption, making it difficult to reduce the number of taps required in the FIR filter 302, thus hindering low-power operation.

[0043] (Regarding the centroid correction of the tap coefficient in Embodiment 1) In Embodiment 1, in order to address DGD while suppressing power consumption, the centroid correction processing unit 134 corrects the centroid for each orthogonal polarization (H,V) of the updated tap coefficients obtained by the coefficient adaptive control processing unit 133, based on the amount of deviation from the tap center.

[0044] As shown in Figure 4, if the weight of the tap coefficient is concentrated towards the edge of the tap, an equalization residue of AEQ112 occurs. Therefore, it is desirable that the weight of the tap coefficient be concentrated towards the center of the tap (tap number 8 in the case of 15 taps).

[0045] Figure 5 is a diagram showing an example of tap centroid correction processing according to Embodiment 1. The centroid correction processing unit 134 calculates the centroid value of the tap coefficient. For example, as shown in Figure 5(a), the centroid correction processing unit 134 groups HH and VH together as H-side coefficients and HV and VV together as V-side coefficients, and calculates the centroid of the H-side coefficients and the centroid of the V-side coefficients, respectively. Then, the centroid correction processing unit 134 shifts the deviation from the tap center toward the tap center.

[0046] DGD increases the probability that the tap centroid of each polarization will be closer to the edge of the tap. However, the centroid correction processing by the centroid correction processing unit 134 moves the entire tap towards the center of the tap, thereby reducing the number of taps used in the main signal side FIR filter 122. In the example in Figure 5(b), the main signal side FIR filter 122 uses 7 taps, tap numbers 5 to 11 (about half of the total 15 taps). All 8 taps, tap numbers 1 to 4 and tap numbers 12 to 15, are not used. As described above, when the number of taps in the coefficient update side FIR filter 131 is N, the number of taps in the main signal side FIR filter 122 can be less than the number of taps N in the coefficient update side FIR filter 131 (for example, N / 2).

[0047] (Operation during initial retraction) Figure 6 is an explanatory diagram of the operation during initial pull-in according to Embodiment 1. The control unit 114 controls each part of the AEQ 112, and during the initial pull-in of the adaptive equalization process, as shown in Figure 6, it operates the tap coefficient update unit 121 of the AEQ 112 and controls the main signal side FIR filter 122 to stop.

[0048] The control unit 114 then sets an initial value for the tap coefficient in the coefficient adaptive control processing unit 133 of the tap coefficient update unit 121, and causes it to calculate the updated tap coefficient using adaptive equalization control based on the initial value. At this time, the control unit 114 switches a selector 132 located on the input side of the coefficient adaptive control processing unit 133 to select the output data of the coefficient update side FIR filter 131 and output it to the coefficient adaptive control processing unit 133. The coefficient adaptive control processing unit 133 receives the received signal (input data of the main signal side FIR filter 122) and the output data of the coefficient update side FIR filter 131. The signal path during initial pull-in is shown by a thick line.

[0049] Here, during initial pull-in, the coefficient update side FIR filter 131 receives only the signal selected at a fixed period (for example, a predetermined number of symbols per frame (1 / 32 to 1 / 64 symbols)). The coefficient adaptive control processing unit 133 determines the update tap coefficient at the time of initial pull-in based on the initial value of the tap coefficient, the received signal, and the output of the coefficient update side FIR filter 131 selected by the selector 132. The calculated update tap coefficient is then corrected by the centroid correction processing unit 134 based on the amount of deviation of the update tap coefficient from the tap center, and the correction is returned to the coefficient adaptive control processing unit 133. The updated tap coefficient after center deviation correction is fed back to the coefficient update side FIR filter 131. This allows the number of taps in the update tap coefficient determined during initial pull-in to be multiplied by N / 2.

[0050] Furthermore, the coefficient-updating FIR filter 131 used during initial pull-in has a tap count N (for example, N=15), but since it operates only for inputs of signals selected at a constant interval, power consumption during initial pull-in can be reduced.

[0051] (Operation during normal operation) Figure 7 is an explanatory diagram of the operation during normal operation according to Embodiment 1. The control unit 114 controls each part of the AEQ 112, and during normal operation of the adaptive equalization process, as shown in Figure 7, it stops the coefficient update side FIR filter 131 and the centroid correction processing unit 134 of the tap coefficient update unit 121 of the AEQ 112, and controls the operation of the main signal side FIR filter 122.

[0052] As a result, during normal operation, the signal path (thick line in the diagram) is the same as that of existing technology. During normal operation, the control unit 114 switches the selector 132 on the input side of the coefficient adaptive control processing unit 133 to select the output data of the main signal side FIR filter. As a result, the coefficient adaptive control processing unit 133 determines the update tap coefficient of the main signal side FIR filter 122 to be used during normal operation based on the input data and output data for the main signal side FIR filter 122. The main signal side FIR filter 122 only requires N / 2 taps, which reduces power consumption.

[0053] Figure 8 is a flowchart showing an example of signal processing according to Embodiment 1. An example of signal processing for each part of the AEQ 112 controlled by the control unit 114 will be explained. During initial pull-in, the control unit 114 sets the initial value of the tap coefficient to the coefficient update side FIR filter 131 (step S801).

[0054] Then, the control unit 114 causes the coefficient update side FIR filter 131 to perform processing and the coefficient adaptive control processing unit 133 to determine the updated tap coefficient using the output data of the coefficient update side FIR filter 131 (step S802). At this time, the control unit 114 stops the operation of the main signal side FIR filter 122 and causes the centroid correction processing unit 134 to correct the tap center misalignment of the updated tap coefficient (step S803).

[0055] Subsequently, during normal operation, the control unit 114 instructs the coefficient adaptive control processing unit 133 to perform processing by the main signal side FIR filter 122 and coefficient adaptive control using the output data of the main signal side FIR filter 122 (step S804). During normal operation, the control unit 114 stops the operation of the coefficient update side FIR filter 131 and the centroid correction processing unit 134.

[0056] According to Embodiment 1, the tap coefficient update unit 121 is used to update the tap coefficient only during initial draw-in. At this time, the centroid correction processing unit 134 corrects the centroid of the updated tap coefficient based on the amount of deviation from the tap center. The number of taps in the coefficient update side FIR filter 131 is N, compared to the number of taps n (e.g., N / 2) in the main signal side FIR filter 122. However, the input data is extracted from the received signal at a constant period and the FIR filter is performed. This makes it possible to improve DGD tolerance without increasing power consumption during initial draw-in. In addition, the number of taps in the main signal side FIR filter 122 used during normal operation can be set to a smaller number of taps n (e.g., n=N / 2) than the normal number of taps N. The increase in power consumption of the coefficient update side FIR filter 131 during initial draw-in is short-term and less than that of the main signal side FIR filter 122, and does not affect the overall increase in power consumption. Furthermore, according to Embodiment 1, even if the baud rate doubles, the power consumption can be doubled (compared to four times in the conventional case). The power consumption reduction effect of this embodiment is twice that of existing technologies, resulting in a significant reduction in power consumption.

[0057] (Embodiment 2) Figure 9 shows a signal processing circuit according to Embodiment 2. In the AEQ112 shown in Figure 9, components similar to those in Embodiment 1 (Figures 6 and 7) are denoted by the same reference numerals. Embodiment 2 differs mainly in that the selector 132 described in Embodiment 1 is omitted. The updated tap coefficients output by the coefficient adaptive control processing unit 133 are input to the main signal side FIR filter 122 and the coefficient update side FIR filter 131. The coefficient update side FIR filter 131 operates continuously during initial engagement and normal operation, but the number of taps differs between initial engagement and normal operation.

[0058] (Operation during initial retraction) Figure 10 is an explanatory diagram of the operation during initial engagement according to Embodiment 2. The control unit 114 controls each part of the AEQ 112, and during the initial engagement of the adaptive equalization process, as shown in Figure 10, it operates the tap coefficient update unit 121 of the AEQ 112 and controls the main signal side FIR filter 122 to stop. The control unit 114 controls each part of the AEQ 112, and during the initial engagement of the adaptive equalization process, as shown in Figure 6, it operates the tap coefficient update unit 121 of the AEQ 112 and controls the main signal side FIR filter 122 to stop.

[0059] The control unit 114 then sets an initial value for the tap coefficient in the coefficient adaptive control processing unit 133 of the tap coefficient update unit 121, and causes it to calculate the updated tap coefficient using adaptive equalization control based on the initial value of the tap coefficient. At this time, the output data of the coefficient update side FIR filter 131 is output to the coefficient adaptive control processing unit 133. The coefficient adaptive control processing unit 133 receives the received signal (input data of the main signal side FIR filter 122) and the output data of the coefficient update side FIR filter 131. The signal path during initial pull-in is shown by a thick line.

[0060] Here, during initial pull-in, the coefficient update side FIR filter 131 receives only the signal selected at a fixed period (for example, a predetermined number of symbols per frame (1 / 32 to 1 / 64 symbols)). The coefficient adaptive control processing unit 133 determines the updated tap coefficient at the time of initial pull-in based on the initial value of the tap coefficient, the received signal, and the output of the coefficient update side FIR filter 131. The calculated updated tap coefficient is then returned to the coefficient adaptive control processing unit 133 after the centroid correction processing unit 134 corrects the centroid based on the amount of deviation of the updated tap coefficient from the tap center. The updated tap coefficient after center deviation correction is fed back to the coefficient update side FIR filter 131. As a result, the number of taps based on the updated tap coefficient determined during initial pull-in can be multiplied by, for example, N / 2.

[0061] Furthermore, the coefficient-updating FIR filter 131 used during initial pull-in has a tap count N (for example, N=15), but since it operates only for inputs of signals selected at a constant interval, power consumption during initial pull-in can be reduced.

[0062] (Operation during normal operation) Figure 11 is an explanatory diagram of the operation during normal operation according to Embodiment 2. The control unit 114 controls each part of the AEQ 112, and during normal operation of the adaptive equalization process, as shown in Figure 11, it stops the centroid correction processing unit 134 of the AEQ 112 and controls the operation of the main signal side FIR filter 122. In addition, only signals selected at a fixed period of time (for example, a predetermined number of symbols per frame (1 / 32 to 1 / 64 symbols)) are input to the coefficient update side FIR filter 131 of the tap coefficient update unit 121 of the AEQ 112. The control unit 114 then operates the coefficient update side FIR filter 131 only for a portion of the number of taps n (n=N / 2) out of N taps. As will be described later, by controlling the tap coefficients of unused tap numbers to be fixed at 0, taps can be operated selectively.

[0063] As a result, during normal operation, the signal path (thick line in the diagram) is the same as that of existing technology. During normal operation, the coefficient adaptive control processing unit 133 determines the update tap coefficient of the main signal side FIR filter 122 used during normal operation based on the input data to the main signal side FIR filter 122 and the output data of the coefficient update side FIR filter 131. The number of taps in the main signal side FIR filter 122 can be reduced, for example to N / 2 taps, thereby reducing power consumption.

[0064] Figure 12 is a flowchart showing an example of signal processing according to Embodiment 2. An example of signal processing for each part of the AEQ 112 under the control of the control unit 114 will be explained. During initial pull-in, the control unit 114 sets the initial value of the tap coefficient to the coefficient update side FIR filter 131 (step S1201).

[0065] Then, the control unit 114 causes the coefficient update side FIR filter 131 to perform processing and the coefficient adaptive control processing unit 133 to determine the updated tap coefficient using the output data of the coefficient update side FIR filter 131 (step S1202). At this time, the control unit 114 stops the operation of the main signal side FIR filter 122 and causes the centroid correction processing unit 134 to correct the tap center misalignment of the updated tap coefficient (step S1203).

[0066] Subsequently, during normal operation, the control unit 114 instructs the coefficient adaptive control processing unit 133 to perform processing by the main signal side FIR filter 122, processing by the coefficient update side FIR filter 131, and coefficient adaptive control using the output data of the coefficient update side FIR filter 131 (step S1204). During normal operation, the control unit 114 operates the coefficient update side FIR filter 131 for only N / 2 taps out of N taps and stops the centroid correction processing unit 134. The main signal side FIR filter 122 operates with N / 2 taps.

[0067] (Regarding the tap selection operation of FIR filters) Figures 13A and 13B are explanatory diagrams of the tap selection operation of the FIR filter. Figure 13A shows an example of the internal configuration of the coefficient update side FIR filter 131 and the main signal side FIR filter 122.

[0068] In the FIR filter shown in Figure 13A, N is the number of taps, and α1 to αn (N=n, where N,n is 15 in the above example) are the tap coefficients of the multiplier at each tap. The tap coefficients α0 to αn are multiplied by a series-connected delay element Z(Z) in relation to the input intensity. -1 These are coefficients that are multiplied by each delay amount. The number of taps N indicates the number of stages in the delay unit Z, and the tap coefficients α0 to αn corresponding to the number of taps N are added and output from the adder A.

[0069] The coefficient update side FIR filter 131 performs FIR filtering using the total number of taps N during initial pull-in in Embodiment 1, as shown in Figure 13A. In addition, during normal operation in Embodiment 2, when using N / 2 taps, the taps near the center (tap coefficients α5 to α11) are used, taking into account the shift in the tap centroid, and the tap coefficients of the taps near both ends (tap coefficients α1 to α4, α12 to α15) are set to 0. The control unit 114 performs the control to set the tap coefficients α1 to α15 to 0. The number of N / 2 taps that the main signal side FIR filter 122 has is the 7 taps used in Figure 13B.

[0070] The above explanation described an example of changing the total number of taps N in an FIR filter to n (for example, n = N / 2). Multiplying the number of taps by N / 2 is just one example, and we will now explain the considerations for designing how many taps to reduce. For example, as seen in Figure 5B(b), the spread of the tap coefficients for one polarization (the distribution in which the tap coefficients have effective values) is approximately 7 taps. This spread of tap coefficients is determined by the DGD amount. Considering the shift of the tap centroid, initial pull-in is possible with a number of taps of approximately DGD ± (DGD / 2) = DGD × 2.

[0071] The center of gravity correction processing unit 134 corrects the center of gravity, and by shifting the tap center towards the middle, only the number of taps equal to DGD is required. Therefore, n can be reduced to half the number of taps of N.

[0072] In Embodiment 2, the coefficient update side FIR filter 131 operates continuously during initial pull-in and normal operation. Therefore, although power consumption is slightly higher compared to Embodiment 1, the selector 132 used in Embodiment 1 is omitted, resulting in less processing delay, thus enabling faster tap coefficient updates compared to Embodiment 1.

[0073] Furthermore, during initial engagement, the coefficient update side FIR filter 131 has N taps, but the input data is extracted from the received signal at a constant period and filtered using the FIR filter. This allows for improved DGD tolerance during initial engagement without increasing power consumption. Also, during normal operation, the coefficient update side FIR filter 131 and the main signal side FIR filter 122 have N / 2 taps, which helps to suppress the overall increase in power consumption.

[0074] (Example configuration of optical transmission equipment: optical receiver) Figure 14 shows an example of the configuration of an optical receiver. The signal processing circuit described above can be applied to the optical receiver R located on the receiving side of the optical transmission device shown in Figure 14. In Figure 14, the same reference numerals are used for functions that are the same as those in Figure 1. The signal processing circuit shown in Figure 1 corresponds to the function of the adaptive equalization processing unit (AEQ) 112 shown in Figure 14.

[0075] As shown in Figure 14, the optical receiver R receives the coherent detection result (analog electrical signal) of the received signal as input to the ADC 102, which then converts it to digital and outputs it. The dispersion compensation unit 1401 compensates for waveform distortion due to dispersion such as polarization mode dispersion (PMD). The sampling phase detection and adjustment unit 1402 adjusts the phase position when sampling digital data and outputs it to the adaptive equalization processing unit 112. The adaptive equalization processing unit 112 performs coefficient adaptive control processing using the blind equalization algorithm described above.

[0076] The synchronization detection / frequency offset monitor / compensation unit 1403 detects and compensates for the difference (frequency offset) between the carrier frequency of the received signal and the frequency of the station light emission. The carrier phase restoration unit 1404 includes the CPR function described above and restores the phase of the carrier wave. For example, the amount of frequency offset can be detected by a well-known method, and the frequency offset is compensated by reversing the rotation of the constellation at a phase rotation speed corresponding to the detected frequency error.

[0077] The IQ distortion compensation unit 1405 compensates for IQ distortion (such as IQ imbalance or IQ incompleteness) that occurs within the optical receiver R. The received frame synchronization unit 1406 synchronizes the received signal frame. The error correction decoding unit 1407 corrects bit errors using an error correction code from an FEC (Forward Error Correction code) decoder, decodes the received signal, and outputs it.

[0078] (Example of an optical transmission system configuration) Figure 15 shows an example of the configuration of an optical transmission system. The signal processing circuits described in Embodiments 1 and 2 above were explained using an optical receiver located on the receiving side of the optical transmission device as an example. As shown in Figure 15, optical transceivers 1 and 2 (1501 and 1502) are located at both ends of the optical transmission path L, respectively, as optical transmission devices.

[0079] Optical transceiver 1 (1501) transmits an optical signal via the downstream optical transmission path L1 from its transmitting side T, and optical transceiver 2 (1502) receives the optical signal from its receiving side R. On the other hand, optical transceiver 2 (1502) transmits an optical signal via the upstream optical transmission path L2 from its transmitting side T, and optical transceiver 1 (1501) receives the optical signal from its receiving side R.

[0080] To describe the configuration of the transmitting side T of the optical transceiver 1 (1501), the framer 1511 frames the input signal from the client on the optical transceiver 1 (1501) side, and the transmitting DSP of the digital signal processing unit 1512, which consists of a DSP, performs data processing on the transmitted signal. In the optical transceiver unit 1513, the DAC 1521 converts the transmitted signal from digital to analog, the E / O conversion unit 1522 converts the electrical signal to an optical signal, and the optical signal is sent to the optical transmission path L1. The light source 1523 is a local light emitter for generating the optical signal to be transmitted, and the optical signal is transmitted after being subjected to a predetermined optical modulation.

[0081] The configuration of the receiving side R of the optical transceiver 2 (1502) is as follows: The O / E conversion unit 1541 of the optical transceiver unit 1531 converts the optical signal into an electrical signal, and the ADC 1542 converts the received signal from analog to digital. The light source 1543 is a local light emitter for demodulating the received optical signal. The receiving DSP of the digital signal processing unit 1532, which consists of a DSP, performs reception processing. This receiving DSP corresponds to the receiving DSP 103 in Figure 1 described above and includes the functions of the adaptive equalization processing unit 112. The output of the receiving DSP is framed via the framer 1533 and output as an output signal to the client on the optical transceiver 2 (1502) side.

[0082] The configurations of the receiving side R of optical transceiver 1 (1501) are the same as those of the receiving side R of optical transceiver 2 (1502). Similarly, the configurations of the transmitting side T of optical transceiver 2 (1502) are the same as those of the transmitting side T of optical transceiver 1 (1501). In Figure 15, identical components are denoted by the same reference numerals.

[0083] As shown in Figure 15, each optical transmission device (optical transceiver) located at both ends of the optical transmission path L has the functions of an optical transmitter T and an optical receiver R. The adaptive equalization processing unit (AEQ) 112 described in the above embodiment can be applied to the optical receiver R.

[0084] Incidentally, the adaptive equalization processing unit 112 mentioned above requires high-speed signal processing, and currently uses a dedicated DSP. However, it is not limited to this; the adaptive equalization processing unit 112 can also be configured using an ASIC or FPGA that supports high-speed processing. Furthermore, a high-speed CPU could be used as the control unit 114 of the adaptive equalization processing unit 112 in the future. ASIC stands for Application Specific Integrated Circuit, and FPGA stands for Field Programmable Gate Array.

[0085] The signal processing circuit of the embodiment described above includes a first FIR filter with n taps that performs adaptive control processing of the received signal, a second FIR filter with N taps (where N is a value greater than n) used to update the tap coefficients of the first FIR filter, and a coefficient adaptive control processing unit that sets the tap coefficients of the first and second FIR filters. The coefficient adaptive control processing unit sets the tap coefficients of the first FIR filter based on the output data of the second FIR filter. For example, the first FIR filter is used during the initial startup of the equalization control process when the transmission path characteristics are unknown, and the second FIR filter is used during normal operation after the initial startup. This makes it possible to improve the DGD tolerance without increasing the power consumption of the FIR filters.

[0086] Furthermore, the signal processing circuit of this embodiment includes a centroid correction processing unit that corrects the deviation of the second FIR filter from the tap center relative to the tap coefficients obtained by the coefficient adaptive control processing unit. The centroid correction processing corrects the deviation of the second FIR filter from the tap center due to the interpolarization delay of the DGD, and allows the tap numbers of the updated tap coefficients to be brought closer to the overall tap center, so that the number of taps in the FIR filter can be reduced to a tap number n that is less than the total number of taps N. By correcting the deviation of the second FIR filter from the tap center due to the interpolarization delay of the DGD, the number of taps in the second FIR filter can be reduced, thereby reducing power consumption.

[0087] Furthermore, in the signal processing circuit of Embodiment 1, during initial pull-in, the coefficient adaptive control processing unit performs coefficient adaptive control of the second FIR filter based on the input data to the second FIR filter and the output data of the second FIR filter, and during steady-state operation after the completion of the initial pull-in process, the coefficient adaptive control processing unit performs coefficient adaptive control based on the input data and output data of the first FIR filter. In the signal processing circuit of Embodiment 2, during initial pull-in, the coefficient adaptive control processing unit performs coefficient adaptive control of the second FIR filter based on the input data to the second FIR filter and the output data of the second FIR filter, and during steady-state operation after the completion of the initial pull-in process, the coefficient adaptive control processing unit performs coefficient adaptive control based on the input data and output data of the second FIR filter. As a result, since the first digital filter operates with a small number of taps n, the increase in power consumption during steady-state operation can be suppressed, and the DGD withstand capability can be improved without increasing the power consumption of the FIR filter.

[0088] Furthermore, the signal processing circuit of Embodiment 1 includes a selector that selectively switches between the output of the second FIR filter and the output of the first FIR filter. The selector selects the output of the second FIR filter during the initial pull-in process and the output of the first FIR filter during steady-state operation. In this way, initial pull-in and subsequent steady-state operation can be easily performed by switching the signal of the selector.

[0089] Furthermore, in the signal processing circuit of Embodiment 1, during the initial pull-in process, the second FIR filter and the centroid correction processing unit operate, and the first FIR filter stops. During steady-state operation, only the first FIR filter and n of the N taps of the second FIR filter operate, and the centroid correction processing unit stops. Similarly, in the signal processing circuit of Embodiment 2, during the initial pull-in process, the second FIR filter and the centroid correction processing unit operate, and the first FIR filter stops. During steady-state operation, only the first FIR filter and n of the N taps of the second FIR filter operate, and the centroid correction processing unit stops. This allows the processing during the initial pull-in to be completed, and enables a smooth transition to the processing during steady-state operation thereafter.

[0090] Furthermore, in the second embodiment, the signal processing circuit sets the tap coefficients of the tap numbers that are not used in the second FIR filter to 0. By not using the taps in the second FIR filter, power consumption can be reduced.

[0091] Furthermore, in the embodiment, the signal processing circuit sets the number of taps n to N / 2 of the total number of taps N, and the coefficient adaptive control processing unit uses the tap coefficients of the second FIR filter for n taps, including the center tap number among the N taps, to determine the tap coefficients of the first FIR filter. This reduces power consumption by the amount of the reduced number of taps n.

[0092] Furthermore, in the signal processing circuit of this embodiment, the second FIR filter extracts input data from the received signal at a fixed interval and uses it. This makes it possible to reduce the power consumption of the second FIR filter.

[0093] Furthermore, the optical transmission device of this embodiment receives and processes the received signal via an optical transmission path. The optical transmission device includes an O / E conversion unit that converts the received optical signal into an electrical signal, an ADC that converts the received signal after conversion by the O / E conversion unit from analog to digital, and the above-mentioned signal processing circuit that receives and processes the received signal output by the ADC. Thus, the signal processing circuit can be applied to an optical transmission device such as an optical receiver that receives a received optical signal.

[0094] Furthermore, the optical transmission system of this embodiment includes a first optical transmission device that transmits an optical transmission signal to an optical transmission path, and a second optical transmission device that receives and processes an optical reception signal received via the optical transmission path. The first optical transmission device transmits the transmission signal. The second optical transmission device includes the signal processing circuit described above. Thus, this can be applied to an optical transmission system that transmits and receives optical signals between the first optical transmission device and the second optical transmission device.

[0095] With regard to the embodiments described above, the following additional information is disclosed.

[0096] (Note 1) The number of taps is n, and the first FIR (Finite Impulse Response) filter performs adaptive control processing of the received signal, The number of taps is N (where N is a value greater than n), and a second FIR filter is used to update the tap coefficients of the first FIR filter, The system includes a coefficient adaptive control processing unit that sets the tap coefficients of the first FIR filter and the second FIR filter, The coefficient adaptive control processing unit sets the tap coefficients of the first FIR filter based on the output data of the second FIR filter. A signal processing circuit characterized by the following features.

[0097] (Note 2) The signal processing circuit according to Note 1, characterized in that it includes a centroid correction processing unit that corrects the deviation of the tap coefficient of the second FIR filter relative to the tap center with respect to the tap coefficient obtained by the coefficient adaptive control processing unit.

[0098] (Note 3) During the initial connection, The coefficient adaptive control processing unit performs coefficient adaptive control of the second FIR filter based on the input data for the second FIR filter and the output data of the second FIR filter. During normal operation after the completion of the initial pull-in process, The coefficient adaptive control processing unit performs coefficient adaptive control based on the input data and output data of the first FIR filter. The signal processing circuit described in Appendix 2, characterized by the above.

[0099] (Note 4) During the initial pull-in, the coefficient adaptive control processing unit, Based on the input data for the second FIR filter and the output data for the second FIR filter, coefficient adaptive control of the second FIR filter is performed. During normal operation after the completion of the initial pull-in process, The coefficient adaptive control processing unit performs coefficient adaptive control based on the input data and output data of the second FIR filter. The signal processing circuit described in Appendix 2, characterized by the above.

[0100] (Note 5) The signal processing circuit is Includes a selector for selectively switching between the output of the second FIR filter and the output of the first FIR filter, The selector is During the initial pull-in process, the output of the second FIR filter is selected. During the aforementioned steady-state operation, the output of the first FIR filter is selected. The signal processing circuit described in Appendix 3, characterized by the features described herein.

[0101] (Note 6) During the initial pull-in process, the second FIR filter and the centroid correction processing unit operate, and the first FIR filter stops. During the normal operation described above, the first FIR filter operates, and the second FIR filter and the centroid correction processing unit stop. The signal processing circuit described in Appendix 3, characterized by the features described herein.

[0102] (Note 7) During the initial pull-in process, the second FIR filter and the centroid correction processing unit operate, and the first FIR filter stops. During normal operation, only the first FIR filter and n of the n taps of the second FIR filter operate, and the centroid correction processing unit stops. The signal processing circuit described in Appendix 4, characterized by the features described herein.

[0103] (Note 8) The signal processing circuit described in Note 7, wherein the second FIR filter is characterized by setting the tap coefficient of the tap number to be unused to 0.

[0104] (Note 9) Set the number of taps n to N / 2 of the number of taps N, The signal processing circuit according to Appendix 1, characterized in that the coefficient adaptive control processing unit uses the tap coefficients of the second FIR filter for n taps, including the tap number of the center tap among the N taps, to determine the tap coefficients of the first FIR filter.

[0105] (Note 10) The signal processing circuit described in Note 1, characterized in that the second FIR filter extracts and uses input data from the received signal at a fixed interval.

[0106] (Note 11) In an optical transmission device that receives and processes a received signal via an optical transmission path, An O / E conversion unit that converts the received optical signal into an electrical signal, The O / E conversion unit includes an ADC that converts the received signal after conversion from analog to digital, The circuit includes a signal processing circuit that receives and processes the received signal output by the ADC, The aforementioned signal processing circuit is A first FIR (Finite Impulse Response) filter with n taps performs adaptive control processing of the received signal, The number of taps is N (where N is a value greater than n), and a second FIR filter is used to update the tap coefficients of the first FIR filter, The system includes a coefficient adaptive control processing unit that sets the tap coefficients of the first FIR filter and the second FIR filter, The coefficient adaptive control processing unit sets the tap coefficients of the first FIR filter based on the output data of the second FIR filter. An optical transmission device characterized by the following features.

[0107] (Note 12) In an optical transmission system including a first optical transmission device that transmits an optical transmission signal to an optical transmission path, and a second optical transmission device that receives and processes an optical reception signal received via the optical transmission path, The first optical transmission device transmits the transmission signal, The second optical transmission device includes a signal processing circuit, The aforementioned signal processing circuit is A first FIR (Finite Impulse Response) filter with n taps performs adaptive control processing of the received signal, The number of taps is N (where N is a value greater than n), and a second FIR filter is used to update the tap coefficients of the first FIR filter, The system includes a coefficient adaptive control processing unit that sets the tap coefficients of the first FIR filter and the second FIR filter, The coefficient adaptive control processing unit sets the tap coefficients of the first FIR filter based on the output data of the second FIR filter. An optical transmission system characterized by the following: [Explanation of Symbols]

[0108] 101 O / E conversion unit 102 ADC 103 Receiving DSP (Signal Processing Circuit) 111 FEQ 112 Adaptive Equalization Processing Unit (AEQ) 113 CPR / FOC 114 Control Unit 122 Main signal side FIR filter (first digital filter) 131 Coefficient-updating FIR filter (second digital filter) 132 Selector 133 Coefficient Adaptive Control Processing Unit 134 Center of gravity correction processing unit 1401 Dispersion compensation section 1402 Sampling Phase Detection and Adjustment Unit 1403 Synchronization Detection / Frequency Offset Monitor / Compensation Unit 1404 Carrier phase restoration section 1405 IQ distortion compensation section 1406 Received frame synchronization section 1407 Error Correction and Decoding Unit 1511, 1533 Frame 1512,1532 Digital signal processing unit 1513,1531 Optical Transceiver Unit 1523,1543 light source A Adder L(L1,L2) Optical transmission line R Optical Receiver T optical transmitter Z delay unit

Claims

1. A first FIR (Finite Impulse Response) filter with n taps performs adaptive control processing of the received signal, The number of taps is N (where N is a value greater than n), and a second FIR filter is used to update the tap coefficients of the first FIR filter, It includes a coefficient adaptive control processing unit that sets the tap coefficients of the first FIR filter and the second FIR filter, The coefficient adaptive control processing unit sets the tap coefficients of the first FIR filter based on the output data of the second FIR filter. A signal processing circuit characterized by the following features.

2. The signal processing circuit according to claim 1, characterized in that it includes a centroid correction processing unit that corrects the deviation of the tap center of the second FIR filter with respect to the tap coefficient obtained by the coefficient adaptive control processing unit.

3. During the initial connection, The coefficient adaptive control processing unit performs coefficient adaptive control of the second FIR filter based on the input data for the second FIR filter and the output data of the second FIR filter. During normal operation after the completion of the initial pull-in process, The coefficient adaptive control processing unit performs coefficient adaptive control based on the input data and output data of the first FIR filter. The signal processing circuit according to feature 2.

4. During the initial pull-in, the coefficient adaptive control processing unit, Based on the input data to the second FIR filter and the output data of the second FIR filter, coefficient adaptive control of the second FIR filter is performed. During normal operation after the completion of the initial pull-in process, The coefficient adaptive control processing unit performs coefficient adaptive control based on the input data and output data of the second FIR filter. The signal processing circuit according to feature 2.

5. The aforementioned signal processing circuit is Includes a selector for selectively switching between the output of the second FIR filter and the output of the first FIR filter, The selector is During the initial pull-in process, the output of the second FIR filter is selected. During the normal operation described above, the output of the first FIR filter is selected. The signal processing circuit according to feature 3.

6. During the initial pull-in process, the second FIR filter and the centroid correction processing unit operate, and the first FIR filter stops. During normal operation, the first FIR filter operates, and the second FIR filter and the centroid correction processing unit stop. The signal processing circuit according to feature 3.

7. During the initial pull-in process, the second FIR filter and the centroid correction processing unit operate, and the first FIR filter stops. During normal operation, only the first FIR filter and n of the n taps of the second FIR filter operate, and the centroid correction processing unit stops. The signal processing circuit according to feature 4.

8. The signal processing circuit according to claim 7, characterized in that the second FIR filter sets the tap coefficient of the tap number to be unused to 0.

9. Set the number of taps n to N / 2 of the number of taps N. The signal processing circuit according to claim 1, characterized in that the coefficient adaptive control processing unit uses the tap coefficients of the second FIR filter for n taps, including the tap number of the center of the N taps, to determine the tap coefficients of the first FIR filter.

10. In an optical transmission device that receives and processes a received signal via an optical transmission path, An O / E conversion unit that converts the received optical signal into an electrical signal, The O / E conversion unit performs analog-to-digital conversion of the received signal after conversion using an ADC, The circuit includes a signal processing circuit that receives and processes the received signal output by the ADC, The aforementioned signal processing circuit is A first FIR (Finite Impulse Response) filter with n taps performs adaptive control processing of the received signal, The number of taps is N (where N is a value greater than n), and a second FIR filter is used to update the tap coefficients of the first FIR filter, It includes a coefficient adaptive control processing unit that sets the tap coefficients of the first FIR filter and the second FIR filter, The coefficient adaptive control processing unit sets the tap coefficients of the first FIR filter based on the output data of the second FIR filter. An optical transmission device characterized by the following features.

11. In an optical transmission system including a first optical transmission device that transmits an optical transmission signal to an optical transmission path, and a second optical transmission device that receives and processes an optical reception signal received via the optical transmission path, The first optical transmission device transmits the transmission signal, The second optical transmission device includes a signal processing circuit, The aforementioned signal processing circuit is A first FIR (Finite Impulse Response) filter with n taps performs adaptive control processing of the received signal, The number of taps is N (where N is a value greater than n), and a second FIR filter is used to update the tap coefficients of the first FIR filter, It includes a coefficient adaptive control processing unit that sets the tap coefficients of the first FIR filter and the second FIR filter, The coefficient adaptive control processing unit sets the tap coefficients of the first FIR filter based on the output data of the second FIR filter. An optical transmission system characterized by the following:

Images