JPH04207708A - Adaptive filter control method - Google Patents

Abstract

PURPOSE:To offer the coefficient control method which is scarcely influenced by a non-linear distortion by fetching a decision error from in an output signal and updating a tap coefficient, only in the case a code of an output signal of a transversal filter is different from one of at least a code of immediately before or immediately after. CONSTITUTION:A signal subjected to binary decision by a deciding device 7 from an output signal of a transversal, filter 2 is sent to a subtracter 8 and a code inversion detecting circuit 9. The code inversion detecting circuit 9 detects a code insersion position from the output signal of the deciding device 7, and operates so as to close a switch 12 against an error signal corresponding to a code before and after the inversion. In the code inversion detecting circuit 9, there is a time delay of a unit time T, therefore, a delay element 3 becomes necessary between the subtracter 8 and the switch 12. Also, in connection therewith, a delay quantity of a filter input signal inputted to a correlator 13 also increases by time T. The code inversion detecting circuit 9 sends only an error signal of a code in which a code is inverted immediately before or immediately after to the correlator 13. Therefore, a tap gain coefficient is converged so as to minimize only an error before and after the code inversion.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to characteristic control of an adaptive filter for removing interference components from a reproduced signal of an optical disc in which code amount interference has occurred.

(Prior Art) In an optical disc, in order to remove waveform distortion caused in a signal that has passed through a recording and reproducing system, an equalizer is provided in a reproducing circuit to compensate for code amount interference in a recording sequence.

When the transfer characteristics in the recording/reproduction system change or cannot be specified, an adaptive equalization method is used in which the waveform distortion is estimated from the reproduced signal and the equalizer characteristics are determined. .

A conventionally used adaptive filter is shown in FIG.

The equalizer signal is input from input terminal 1. The input signal has waveform distortion removed by a transversal filter 2 composed of a delay element 3, a multiplier 4, and an adder 5, and is outputted from an output terminal 6. This output signal is fed back to extract the error signal used to update the tap gain.

If the tap gain is not set to an optimal value, a signal with waveform distortion remaining is output from the output terminal.

The determiner 7 performs a binary determination on the signal in which the waveform distortion remains, and outputs a binary signal having a reference amplitude. The tap gain coefficient is changed in a direction that minimizes the power of the error signal obtained by the difference between the output signal of the transversal filter 2 and the output signal of the determiner 7. Therefore, after the tap coefficient converges to an appropriate value, the filter output signal takes on the binary value given as the determiner output.

The error signal is sent to a correlator 13 together with an appropriate time-delayed filter input signal, and the correlator 13 outputs the correlation strength between the two. The output signal of the correlator 13 is integrated by an integrator 14 and given as a tap gain coefficient to the multiplier of each tap.

In the above example, using only the reproduced signal with waveform distortion,
Since data is judged and an error signal is output, if the distortion is large, an error will occur in the output of the judge 7 itself.
Correct output will no longer be obtained. To avoid this, in many cases, a training signal having a predetermined pattern is added to the input signal following the synchronization signal. The tag gain coefficient is modified in such a way that the reproduced signal obtained following the synchronization signal matches the pattern of the training signal. FIG. 5 shows an example of a circuit that updates coefficients using a training signal.

The configurations of the transversal filter 2, correlator 13, and integrator 14 are similar to those in the example shown in FIG.

The synchronization signal detection circuit 15 includes a transversal filter 2
6. Monitor the output of the training signal and prompt the training signal generation circuit 16 to generate the same training pattern as that included in the reproduced signal for a certain period of time from the voice that detects the synchronization signal that appears before the training signal.6. By simultaneously closing the switch 12, the error signal extracted as the difference between the output signal from the transversal filter 2 and the training signal is output to the correlator 13. On the other hand, while a signal other than the training signal is being input, the switch is open and the error signal input of the correlator 13 is set to O, thereby prohibiting updating of the coefficients.

Equalization using a transversal filter is originally aimed at removing linear distortion. This waveform distortion consisting only of linear distortion is effectively removed from the reproduced signal.

However, for example, in the case of an optical disk in which thermal recording is performed using a light beam, if recording and recording are performed successively, pits overlap as shown by diagonal lines in FIG. 6. Therefore, the reproduced signal has a reproduced amplitude as shown by the solid line 20, with a waveform as shown by the broken line 21 obtained by superimposing the reproduced signals 19 by a single pit. The nonlinear distortion components that appear here cannot be eliminated by the transversal filter, so they remain in the output signal and become a factor in worsening the error rate.

In addition to the component caused by overlapping pits shown in Figure 6, the nonlinear component contained in the reproduced signal appears because the sizes of the leading and trailing edges of the recording pit change due to the flow of heat on the recording medium. The ingredients are also known. However, this nonlinear component can be removed by recording power control as shown in Japanese Patent Application No. 1-322340.

(Problem to be solved by the invention) In optical recording and optical chronological recording, information is stored by changing the shape and direction of magnetization of the recording medium, so as the recording density increases, the input/output characteristics become non-linear. Gender appears.

On the other hand, a transversal filter circuit that equalizes the reproduced signal performs various functions such as delaying the reproduced signal, multiplication and addition of coefficients, etc.
Since it is constructed using only linear operations, if the signal has nonlinear distortion, that component cannot be removed. Therefore, when a normal coefficient control method is used here, there is a drawback that the equalization ability is significantly reduced due to the influence of this nonlinearity, leading to an increase in determination errors.

SUMMARY OF THE INVENTION An object of the present invention is to solve the above-mentioned problems by providing a coefficient control method that is less susceptible to the effects of nonlinear distortion, and to suppress the determination error of a filter output signal to a low level.

(Means for Solving the Problems) An adaptive filter control method of the present invention is a method for controlling tap coefficients in an adaptive transversal filter for equalizing a binary reproduction signal of an optical disc. The present invention is characterized in that the determination error is extracted from the output signal and the tap coefficient is updated only when the sign of the output signal is different from at least one of the immediately preceding or following sign.

(Function) Nonlinear distortion in the digital recording system of an optical disk is mainly caused by the duplication of recording pits as shown in FIG. 6, so it characteristically appears when the same code continues. On the other hand, when determining the reproduced signal, it is more difficult to identify the position when the same code is inverted than when the same code is consecutive, and it is less susceptible to bit shifts caused by code amount interference. easy.

In the present invention, it is relatively easy to be affected by code amount interference,
In addition, tap coefficients are controlled using only error signals before and after the sign inversion position, which are less affected by nonlinear distortion. This makes it possible to obtain an output signal that has good equalization characteristics before and after sign inversion, which is important in data determination, and is less affected by nonlinear distortion.

(Example) Next, the present invention will be described in detail with reference to FIGS. 1 to 3.

FIG. 1 is a block diagram showing one embodiment. The transversal filter 2 is configured similarly to the conventional example. The signal subjected to binary judgment by the judge 7 from the output signal is sent to a reducer 8 to extract an error signal, and is also sent to a sign inversion detection circuit 9 for detecting the sign inversion position. The sign inversion detection circuit 9 detects the sign inversion position from the output signal of the determiner 7, and operates to close the switch 12 for error signals corresponding to the signs before and after the inversion.

Since the sign reversal detection circuit 9 has a time delay of unit time T, a delay element 3 is required between the subtracter 8 and the switch 12. Additionally, along with this, the amount of delay of the filter input signal input to the correlator 13 also increases by the time T.

The sign reversal detection circuit 9 sends to the correlator 13 only the error signal of the sign whose sign has been reversed immediately before or after. Therefore, the tap gain coefficient converges to minimize only the error before and after sign inversion, regardless of the reproduction amplitude when the same sign continues.

FIG. 2 shows equalized waveforms when using the conventional method and when using the present method. The figure shows a bit length of 0.5μm'''C-
This is an equalization result obtained by simulating a recording waveform recorded in NRZ. Nonlinear distortion is included in the D/U ratio of approximately 10 dB. According to the conventional equalization method, coefficient control is performed so that all decision points match the reference output level determined as the output of the judger, and while the output signal level from the filter approaches the reference level, The eye aperture ratio, which has a large effect on data judgment errors, has become smaller. On the other hand, when coefficient control is performed using this method, the output signal amplitude increases as successive filter output signals with the same sign pass further outside the reference level;
The eye is wider than when using the conventional method, and the judgment margin is increased by about 20%.

Also, the output waveform shown here is when a code with no limit on the minimum run length of the same code is used, but when using an RLL code in which at least two or more of the same code are consecutive, the output waveform is sign-inverted. Because they do not appear consecutively, a higher effect can be obtained.

FIG. 3 shows a block diagram when the equalization method of this system is applied to a recorded signal to which a training signal is added. The synchronization signal detection circuit 15 monitors the output signal from the transversal filter 2 as in the conventional example, and during the training period that appears following the synchronization signal, the output of the training signal generation circuit 16 is subtracted to prompt the generation of a training signal. At the same time, it is sent to a sign inversion detection circuit 9 to detect the sign inversion position.

Switch 1 provided in the error signal feedback path
2 is closed only when the outputs of both the sign inversion detection circuit 9 and the synchronization signal detection circuit 15 are active, and the coefficients can be controlled only by the error signals before and after the sign inversion position during the training signal period. It will be done. This is effective when the waveform distortion of the input signal is large and the determination error obtained from the reproduced signal is large.

In this circuit, the coefficients are controlled only by the training signal. Therefore, when using a recording code such as an RLL code, the training signal also has a similar RL code.
It is more preferable that the L-law is satisfied. Furthermore, it is preferable to record the training signal at the same time as the actual data, since the same recording conditions can be applied to the training signal and the actual data.

(Effects of the Invention) By using the secondary leg method of the present invention, it is possible to extract a good-quality equalized output with less 9-point judgment even when the optical disk reproduction waveform input to the adaptive filter has a lot of nonlinear distortion. can.

[Brief explanation of the drawing]

Fig. 1 is a system diagram showing one embodiment of the present invention, Fig. 2 is a diagram for explaining the output results of this control method, Fig. 3 is a system diagram showing another embodiment of the invention, and Fig. 4 is a system diagram showing an embodiment of the present invention. 5 and 5 are system diagrams for explaining the prior art, and FIG. 6 is a diagram for explaining the causes of nonlinear distortion. In the figure, ■...Input terminal, 2...Transversal filter, 6...Output terminal, 7...Judgment device, 8
... Subtractor, 9... Sign inversion detector, 13... Correlator, 14... Integrator, 15... Synchronization signal detection circuit, 16... Training signal generation circuit, 18...・Record pit 19...Single pit response waveform, 20...
This is the reproduced waveform.

Claims (1)

[Claims] In a method for controlling tap coefficients in an adaptive transversal filter for equalizing a binary playback signal of an optical disc, only when the sign of the output signal of the transversal filter is different from at least one of the immediately preceding or following sign. , an adaptive filter control method characterized in that the determination error is extracted from the output signal and the tap coefficients are updated.