JP6586840B2 - Receiver and tap coefficient optimization method - Google Patents

Description

  The present invention relates to a receiver including an electrical dispersion compensator and a tap coefficient optimization method suitable for use in the receiver.

  In the next-generation PON (Passive Optical Network) system, the coverage area by the station side device (OLT: Optical Line Terminal) is expanded by multi-branching and lengthening of the transmission distance, that is, by integrating the stations. Low power consumption drive is being studied. In order to realize this station integration, it is necessary to mitigate the influence of chromatic dispersion, which is one of the problems caused by extending the transmission distance.

  Currently, the core network uses a dispersion compensating fiber to mitigate the effects of chromatic dispersion. However, since the PON system has a configuration in which a plurality of subscriber terminals (ONU: Optical Network Unit) having different distances from the OLT share the same fiber, a compensation method using a dispersion compensating fiber is not realistic.

  On the other hand, there is a proposal using an electric dispersion compensation method (EDC: Electric Dispersion Compensating) (see, for example, Patent Document 1). EDC is a technique for correcting distortion by digital signal processing for a signal waveform distorted by chromatic dispersion.

  One of the EDC technologies is DFE (Decision Feedback Equalizer) (see, for example, Patent Document 2). DFE is configured so that the sampled signal passes through multiple taps in order to compensate for the distortion component that appears in the sampled signal, and the distortion component is canceled by the delay between taps and the tap coefficient for weighting. It is an equalizer that operates as follows. Since DFE includes a decision feedback circuit, it is possible to reduce the compensation error.

  As a method for autonomously determining the optimum value of the tap coefficient, a method of monitoring the quality of the eye opening of the received signal and optimizing the tap coefficient so as to reduce the error can be considered. FIG. 3 shows an image when the eye opening is monitored. When the horizontal axis direction is the phase and the vertical axis direction is the intensity (amplitude), the error count at each measurement point is obtained with an arbitrary number of points and point intervals. There are few, and an error becomes large gradually toward an outer ring part. For example, when the error count is obtained in the same measurement range (indicated by II in FIG. 3) before and after the DFE tap coefficient is changed, the one having the smaller error count is determined as the higher quality. . In this case, the optimum value of the DFE tap coefficients can be obtained by comparing the total values of the error counts when the DFE tap coefficients are sequentially changed. By optimizing this in the same manner as taps 2, 3, and 4 in order from tap 1, optimized tap coefficients can be obtained for all taps.

JP 2006-287695 A JP 2002-344822 A

  In order to fully optimize the tap coefficient of DFE, it is necessary to measure brute force with all error count measurement points and tap coefficients as parameters. However, in the assumed next-generation PON system, the time given to the initial setting on the receiving side is limited. For this reason, it is necessary to shorten the time required for optimizing the tap coefficient.

  In view of this, a method is proposed in which points with large errors are omitted and only the necessary points at the center of the eye opening are measured. An example of the measurement range and measurement point is shown in FIG. FIG. 4 shows a received waveform by simulation when 80 km transmission is performed. FIG. 4A shows a waveform without DFE, and FIG. 4B shows a waveform when only DFE tap 1 is optimized. Before optimization of tap 1, as shown in FIG. 4A, the position of the phase where the eye opening is maximum (indicated by I in FIG. 4A) is along the center line. The area along the line is the measurement range. Here, due to the characteristics of DFE, the optimization of the tap 1 operates so as to correct the falling distortion. For this reason, after optimizing the tap 1, as shown in FIG. 4B, the position of the phase where the eye opening is maximum (indicated by II in FIG. 4B) is the center of the measurement range. Deviate. Therefore, if the measurement is performed centering on the phase position where the eye opening before the optimization is maximum, the error count cannot be measured accurately, and the tap coefficients after tap 2 cannot be optimized accurately.

  The present invention has been made in view of the above-described problems. An object of the present invention is to provide a receiver including an electrical dispersion compensator that can shorten the time for optimizing the DFE tap coefficient by reducing the number of measurement points, and a tap coefficient optimization method suitable for use with the receiver. It is to provide.

  In order to achieve the above-described object, a receiver including the electrical dispersion compensator of the present invention has the following configuration.

  A decision feedback equivalent circuit that branches an input electrical signal to the number of taps L, gives a delay that is a time difference of the bit period T to each other, performs weighting given by a tap coefficient wx, and then subtracts the input electrical signal to generate an output signal And an eye quality monitor unit that measures an error count in comparison with the test pattern before transmission at a predetermined phase and amplitude of the eye opening obtained by inputting the test pattern after transmission to the decision feedback equivalent circuit, and a control unit And is configured.

The control unit sequentially selects taps for determining tap coefficients from tap 1, means for setting tap coefficients for the selected taps, and error counts acquired from the eye quality monitor unit for a predetermined measurement range. Means for obtaining the total value, means for determining the tap coefficient with the smallest error count total value as the optimum value, and determining whether the error count total value improves by shifting the center of the measurement range from the initial value In the case of improvement, there is provided means for updating the center of the measurement range at the time of improvement as a subsequent initial value.

In the tap coefficient optimization method of the present invention, the input electrical signal is branched to the tap number L, given a delay that is a time difference of the bit period T, weighted by the tap coefficient wx, A decision feedback equivalent circuit that generates an output signal by subtraction, and an error compared to the test pattern before transmission at a predetermined phase and amplitude of the eye opening obtained by inputting the test pattern after transmission to the decision feedback equivalent circuit Performed by a receiver having an electrical dispersion compensator, comprising an eye quality monitor unit that measures a count, and a control unit, and a process of selecting taps for determining tap coefficients in order from tap 1 , In the process of setting the tap coefficient and the eye opening obtained by performing data reception using the test pattern, in a preset measurement range The process of acquiring the total value of the error count and the process of determining whether or not the total value of the error count has been acquired for all the tap coefficients of the selected tap, and if not, the process of setting the tap coefficient again If it has been acquired, the process of determining the tap coefficient with the smallest total error count as the optimum value, the process of determining whether or not the optimum value has been set for all taps, and the end And an error count condition extraction process performed when not. In the error count condition extraction process, the center of the measurement range is shifted from the initial value to determine whether or not the total value of the error count is improved. If so, the improved center is updated as a subsequent initial value. .

  According to the receiver and the tap coefficient optimization method of the present invention, the error count condition extraction time for optimizing the center of measurement is added. As a result, in order to fully optimize the DFE tap coefficient, it is necessary to measure brute force in the error count measurement range, DFE tap number, and DFE tap coefficient adjustment range. With a small number of measurement points, highly reliable tap coefficients can be optimized, and the total calculation time can be shortened.

It is a typical block diagram of a receiver. It is a schematic diagram of DFE with which a receiver is provided. It is a figure which shows an image when eye opening is monitored. It is a figure which shows an example of a measurement range and a measurement point. It is a flowchart (1) for demonstrating the tap coefficient optimization method. It is a flowchart (2) for demonstrating the tap coefficient optimization method. It is a figure which shows an example of the tap coefficient optimization method.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the arrangement relationship of each component is merely schematically shown to the extent that the present invention can be understood. In the following, a preferred configuration example of the present invention will be described. However, the numerical conditions of each component are merely preferred examples. Therefore, the present invention is not limited to the following embodiments, and many changes or modifications that can achieve the effects of the present invention can be made without departing from the scope of the configuration of the present invention.

(Receiver)
With reference to FIG.1 and FIG.2, the receiver provided with an electrical dispersion compensator is demonstrated. FIG. 1 is a schematic block diagram of a receiver. FIG. 2 is a schematic diagram of a DFE that is an electrical dispersion compensator provided in the receiver. The receiver 10 includes an optical receiver 20, a PMA (Physical Media Attachment) processing unit 40, a PCS (Physical Coding Sublayer) processing unit 60, and a MAC (Media Access Control) processing unit 80.

  The light receiver 20 converts the received optical signal into an electrical signal and sends it to the PMA processing unit 40. The PMA processing unit 40 performs predetermined processing on the electrical signal and sends it to the PCS processing unit 60. Details of the PMA processing unit 40 will be described later. The PCS processing unit 60 adds a header to the data generated by the PMA processing unit 40 and sends the data to the MAC processing unit 80. The MAC processing unit 80 sends the data received from the PCS processing unit 60 to other devices (not shown).

  The light receiver 20 can be configured using any suitable light conversion element such as a photodiode (PD). The PMA processing unit 40, the PCS processing unit 60, and the MAC processing unit 80 can be configured using any suitable programmable logic device such as an FPGA manufactured by Altera.

  The PMA processing unit 40 includes a DFE 50, an eye quality (EYE_Q) monitoring unit 42, a control unit 44, and a memory 46. The control unit 44 operates as a CPU and realizes each function by executing a predetermined program. Each function realized by the control unit 44 will be described later.

  The DFE 50 is a decision feedback equivalent circuit. The DFE 50 branches the received electrical signal, adds a delay and weights it, and then superimposes the received signal. The DFE 50 includes a difference circuit 56, L weighting circuits 54-1 to L, and L delay circuits 52-1 to L, where L is the number of taps.

  In the DFE 50, the input signal u (t) is branched into two branches, one being sent to the first delay circuit 52-1, and the other being the output signal y (t) as the PCS processing unit 60 and the EYE_Q monitor unit 42. Sent to. The first delay signal delayed by the first delay circuit 52-1 is branched into two, one is sent to the second delay circuit 52-2, and the other is fed through the first weighting circuit 54-1. Then, it is sent to the difference circuit 56 and subtracted from the input signal. Similarly, the m-th delay signal delayed by the m-th (m is an integer not less than 2 and not more than L-1) delay circuit 52-m is branched into two, and one is the (m + 1) -th delay circuit 52. -(M + 1), and the other is sent to the difference circuit 56 via the m-th weighting circuit 54-m and subtracted from the input signal. The Lth delay signal delayed by the Lth delay circuit 52-L is sent to the difference circuit 56 via the Lth weighting circuit 54-L and subtracted from the input signal.

  The first to L delay circuits 52-1 to L give a delay corresponding to a 1-bit period T to the received signal. The first to L weighting circuits 54-1 to 54-L give the received signals weights with tap coefficients Wx (x is an integer of 1 to L). This tap coefficient Wx is set by an instruction from the control unit 44.

  The EYE_Q monitor unit 42 counts an error count at a predetermined measurement point given from the control unit 44 with respect to the eye opening obtained from the output of the DFE 50. Measurement points are given in pairs of phase and amplitude corresponding to the horizontal and vertical coordinates of the eye opening. The error count is measured by comparing the eye opening obtained by inputting the test pattern after transmission to the DFE 50 with the test pattern before transmission. An error count is added when the values before and after transmission do not match. The measurement result of the error count is sent to the control unit 44.

  In the memory 46, initial setting values necessary for executing the optimization method are stored in a readable manner. This initial set value may be stored in advance, or may be input from the outside by any suitable input means and stored.

(Tap coefficient optimization method)
The tap coefficient optimization method will be described with reference to FIGS. 5 and 6. FIGS. 5 and 6 are flowcharts for explaining the tap coefficient optimization method.

  First, in S10, the control unit 44 acquires an initial setting value by reading from the memory 46 or by inputting from the outside. The initial set values are, for example, the size of the measurement range, the initial value of the measurement center, and the number D of measurement points.

  Next, in S20, the control unit 44 selects a tap for determining a tap coefficient. When the number of taps L is 5, taps are sequentially selected from tap 1 to tap 5, for example.

  Next, in S30, the control unit 44 sets a tap coefficient for the tap selected in S20. The tap coefficient is set in 2 × C + 1 steps from −C to + C. C is 15, for example. In this case, the tap coefficient is set in 31 steps from -15 to +15. The tap coefficient set by the control unit 44 is notified to the DFE 50. The DFE 50 sets the tap coefficient according to the notification from the control unit 44.

  Next, in S40, the total value of error counts is acquired.

  S40 includes the processes of S42 to S48. In S <b> 42, the control unit 44 instructs the EYE_Q monitor unit 42 at an arbitrary measurement point within the measurement range. Thereafter, by performing data reception using the test pattern, the EYE_Q monitor unit 42 measures an error count for the measurement point. The error count is measured by comparing the eye opening obtained by inputting the test pattern after transmission to the DFE 50 and the test pattern before transmission. An error count is added when the values before and after transmission do not match. The measurement result of the error count is sent to the control unit 44.

  Next, in S44, it is determined whether measurement at all measurement points in the measurement range is completed. As a result of the determination, when the measurement at all measurement points is not completed (No), the control unit 44 instructs the EYE_Q monitor unit 42 to change the measurement point in S46, and the measurement points S42 and S44 are changed at the changed measurement points. Do the process. On the other hand, as a result of the determination, if the measurement at all the measurement points is completed (Yes), the control unit 44 calculates the total value of the error counts at all the measurement points in S48.

  After acquiring the total value of error counts in S40, it is determined in S50 whether the optimum value of the tap coefficient has been determined. This determination is made based on whether or not the control unit 44 has acquired a total value of error counts for all tap coefficients of the tap selected in S20. When the total value of error counts for all tap coefficients has not been acquired, that is, when the optimum value of the tap coefficient has not been determined (No), the tap coefficient is set in S30, and S40 and S50 are performed again. On the other hand, if the total value of error counts for all tap coefficients has been acquired (Yes), the tap coefficient that minimizes the total value of error counts is determined as the optimum value, and the process of S60 is performed.

  In S60, it is determined whether or not all taps have been set. If all taps have been set (Yes), the optimization method process is terminated. On the other hand, if all taps have not been set (No), error count condition extraction in S100 is performed.

  This S100 includes the processes of S110 to S130.

  In S110, the control unit 44 shifts one or both of the phase and the amplitude of the center of the measurement range by one point with respect to the initial value of the measurement center.

  Next, in S120, the total value of error counts is acquired. This process is performed in the same manner as S40 except that the measurement center is deviated from the initial value.

  Next, in S130, it is determined whether or not there is an improvement effect and whether or not the measurement under all conditions is completed.

  For example, if the total value acquired in S120 is smaller than the total error count measured using the initial value at the measurement center, it is determined that there is an improvement effect. If there is an improvement effect, the measurement center where the improvement effect is obtained is newly changed to the initial value of the measurement center.

  If the improvement effect cannot be obtained, the measurement center is changed and the processes of S110 to S130 are performed, and measurement is performed under all conditions. If the initial value of the measurement center is (0, 0), the measurement under all conditions is (-1, -1), (-1, 0), (-1, 1), (0, -1), It means measurement under eight conditions (0, 1), (1, -1), (1, 0) and (1, 1). If an improvement effect cannot be obtained in any of all the conditions, S100 is terminated without changing the initial value of the measurement center.

  After the error count condition extraction is completed, the tap selection of S20 is performed again.

  FIG. 7 shows a received waveform when only tap 1 is optimized. FIG. 7A shows the measurement range before extracting the error count condition, and FIG. 7B shows an image of the measurement range after extracting the error count condition. Before extracting the error count condition, the center line (indicated by I in FIG. 7A) indicating the phase at which the eye opening is maximum is deviated from the center of the measurement range. After the tap 1 is optimized, the error count condition is extracted to move the center of the measurement range. As a result, a center line (indicated by II in FIG. 7B) indicating the phase at which the eye opening is maximum is aligned with the center of the measurement range.

  In this way, by extracting the error count condition each time one tap coefficient of DFE is optimized, it is possible to eliminate the deviation between the measurement center and the measurement point caused by tap optimization. Become. Conventionally, in order to fully optimize the DFE tap coefficient, it has been necessary to perform brute force measurement in all of the error count measurement range, DFE tap number, and DFE tap coefficient adjustment range. By adding error count condition extraction time to optimize the center of measurement, it is possible to obtain highly reliable tap coefficients with a small number of measurement points and shorten the total calculation time. There are certain effects.

  To fully optimize the DFE tap coefficient, assuming that the error count measurement point is A (an integer greater than or equal to 1), and the adjustment range of the DFE tap coefficient is ± C (C is an integer greater than or equal to 1), In all cases, the number of points measured in a round robin is A (2C + 1). For example, if the measurement point A is 81 and C is 15, the number of measurement points required for optimization for one tap is 2511 (= A (2C + 1) = 81 × (2 × 15 + 1)).

  On the other hand, when it is considered that the error count can be reduced to the measurement point D (an integer of 1 or more and less than A) according to the present invention, the number of measurement points is D (2C + 1).

  Further, the number of measurement points in the error count condition extraction step is 8D. As a result, the number of measurement points required for optimization for one tap is 624 points.

  Thus, according to the present invention, A (2C + 1)-{D (2C + 1) + 8D} = (AD) (2C + 1) -8D, in this example, the number of measurement points is reduced by 1887 (= 2511-624). can do. Even if the error count measurement point D is 25, the measurement points can be reduced by 60% or more.

DESCRIPTION OF SYMBOLS 10 Receiver 20 Light receiver 40 PMA process part 42 EYE_Q monitor part 44 Control part 46 Memory 50 DFE (judgment feedback equivalent circuit)
52 delay circuit 54 weighting circuit 56 difference circuit 60 PCS processing unit 80 MAC processing unit

Claims (2)

A receiver comprising an electrical dispersion compensator,
A decision feedback equivalent circuit that branches an input electrical signal to the number of taps L, gives a delay that is a time difference of the bit period T to each other, performs weighting given by a tap coefficient wx, and then subtracts the input electrical signal to generate an output signal When,
An eye quality monitor unit that measures an error count in comparison with a test pattern before transmission at a predetermined phase and amplitude of an eye opening obtained by inputting the test pattern after transmission to the decision feedback equivalent circuit;
A control unit,
The controller is
Means for sequentially selecting taps for determining the tap coefficient from tap 1 ;
Means for setting a tap coefficient for the selected tap;
Means for obtaining a total value of error counts obtained from the eye quality monitor unit for a predetermined measurement range;
Means for determining the tap coefficient with the smallest error count as the optimum value;
It is determined whether or not the total value of the error count is improved by shifting the center of the measurement range from the initial value, and when improved, the center of the measurement range at the time of improvement is updated as a subsequent initial value. A receiver characterized by that. A decision feedback equivalent circuit that branches an input electrical signal to the number of taps L, gives a delay that is a time difference of the bit period T to each other, performs weighting given by a tap coefficient wx, and then subtracts the input electrical signal to generate an output signal When,
An eye quality monitor unit that measures an error count in comparison with a test pattern before transmission at a predetermined phase and amplitude of an eye opening obtained by inputting the test pattern after transmission to the decision feedback equivalent circuit;
With control
A tap coefficient optimization method performed by a receiver having an electrical dispersion compensator, comprising:
A process of selecting taps for determining tap coefficients in order from tap 1 ;
A process of setting a tap coefficient for the selected tap;
Using the eye opening obtained by performing data reception using the test pattern, obtaining a total value of error counts in a preset measurement range;
It is determined whether or not the total value of the error count has been acquired for all the tap coefficients of the selected tap. If not, the process of setting the tap coefficient is performed again. The process of determining the tap coefficient with the smallest total value as the optimal value,
A process of determining whether or not the setting of the optimum value for all taps has been completed,
An error count condition extraction process, which is performed when not completed,
The error count condition extraction process, by shifting the center of the measurement range from the initial value to determine whether to improve the total value of the error count, if improved, the initial value after the center of the measuring range with improved The tap coefficient optimization method characterized by updating as follows.

Images