JPS60210032A - Automatic equalizer - Google Patents

Abstract

PURPOSE:To set the optimum tap weight by providing an automatic gain controller between the output of an automatic equalizer and a deciding device in order to correct the attenuation of the data output. CONSTITUTION:The input signal given from an input terminal 1 of an automatic equalizing device is supplied to an automatic equalizer 3. Then an equalized signal is supplied to a deciding circuit 4 from an output terminal 6 via an automatic gain control circuit 15. The circuit 4 decides the closest transmission data point and delivers it to a decision output terminal 2 and also to a terminal at one side of a subtractor 5. While the output of the equalizer 3 is supplied to the other terminal of the subtractor 5. Then an error signal given from the subtractor 5 is applied to a tap weight controller of the equalizer 3. The attenuation of the data output of the equalizer 3 is corrected by the circuit 15. Thus the characteristics of the automatic equalizing device are improved in the case of large circuit distortions.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical field to which the invention pertains] The present invention relates to a baseband automatic equalization device for data transmission using a least square error method.

[Description of prior art]

A conventional baseband automatic equalization method for data transmission using the least squares error method will be explained with reference to the drawings. Figure 1 is a diagram showing a baseband automatic equalization method using the least squares error method, where 1 is an input terminal, 2 is a judgment output terminal, 3 is an automatic equalizer, 4 is a judgment circuit, and 5 is a subtracter. be. A baseband signal that has been subjected to linear distortion due to the line is input from an input terminal 1 and is equalized by an automatic equalizer 3. Its output is sent to the decision circuit 4, where the nearest transmitted data point is found and output to the decision output terminal 2. on the other hand,
A subtracter 5 takes the difference between the output of the automatic equalizer 3 and the output of the determination circuit 4 to create an error signal, which is supplied to the automatic equalizer 3 and used to modify the tap weights.

FIG. 2 is a detailed diagram of the automatic equalizer 3, in which 1 is a blade terminal;
6 is an output terminal, 7 is an error signal input terminal, 8-1 to 8-3
is a delay circuit, 9-1 to 9-3 are multipliers, 10-1 to 10
-3 is tap weight, 11 is coefficient unit, 12-1 to 12-3
is a multiplier, 13-1 to 13-3 are subtracters, and 14 is an adder. The signal input from input terminal 1 is sent to delay circuit 8-
The signals are sequentially shifted from 1 to 8-3 and input. Each delay circuit 8
-1 to 8-3 outputs each tap weight 10-1 to 10-3
and their respective outputs are added by the adder 14 and output to the output terminal 6. On the other hand, the error signal input from the error signal input terminal 7 is multiplied by a constant by the coefficient multiplier 11. The outputs of multipliers 12-1 to 12-
3 is multiplied by the output of each delay circuit 8-1 to 8-3, and the subtracters 13-1 to 13-3 are multiplied by the output value.
The weights of each tap 10-1 to 10-3 are corrected by subtraction.

Therefore, the conventional automatic equalization method using the least squares error method has the disadvantage that some residual intersymbol interference remains at the output of the automatic equalizer, and the original data is also attenuated, increasing the error rate of the transmitted signal. had.

[Purpose of the invention]

SUMMARY OF THE INVENTION An object of the present invention is to provide an automatic equalization method that eliminates the above drawbacks and sets optimal tap weights.

[Features of the invention]

In the present invention, the root mean square error between the equalization layer output and the transmission signal is used as an evaluation function, and the tap coefficients are successively modified so as to minimize this. In the automatic equalization method using the least squares error method, attenuation is also seen in the original data output due to the mean square error in the case of optimal tap weights. This level fluctuation causes line noise, overload, and noise deterioration, so by adding the automatic gain adjuster of the present invention to the output of the automatic equalizer, it is possible to reduce the level as much as possible. Reduce fluctuations,
The present invention is characterized by correcting the attenuation of the above-mentioned original data signal.

That is, an automatic equalizer, a determination circuit for determining its output, and means for generating an error signal from the output of the automatic equalizer and the output of the determination circuit to control the transmission characteristics of the automatic equalizer. In the baseband automatic equalization device for data transmission, the automatic equalizer includes means for equalizing transmission characteristics using the least squares error method, and the automatic equalizer has a transmission characteristic between the output of the automatic equalizer and the determiner. It is characterized by being equipped with an automatic gain adjuster.

[Explanation based on examples]

Embodiments of the present invention will be described based on the drawings. FIG. 3 is a block diagram of an apparatus in which the embodiment method is used. The input signal input from the input terminal 1 is input to the automatic equalizer 3, and the signal output from the output terminal is further sent to the judgment circuit 4 through the automatic gain adjuster 15, and then output to the judgment output terminal 2. On the other hand, the output of the automatic equalizer 3 and the judgment circuit 4
Both outputs are input to a subtracter 5, and the output thereof is further connected to a tap weight control device of the automatic equalizer 3.

In general, if the transmission symbol is a, 1, the impulse response of the transmission path is 6, the tap weight of the automatic equalizer is Ct, the output of the automatic equalizer is yll, and the error signal is 07, then the input signal of the automatic equalizer is x7 is multiplied by xll-Σakh n-k---(1). Also, y7=Σct X□.

=Σakρn −k−−−−−−−(2) However, ρaΣ
c,h,.

And if there is no judgment error by the judgment circuit, then 6 n = 3' n a n −−−−−−−−(3)
becomes. The least squares error method is an algorithm that minimizes the mean square error of the error signal e7. The objective is to minimize , where E() means the ensemble average.

Through a series of calculations, the root mean square error ε for the optimal tap weights. 1 is ε. □−1−ρ. It can be expressed as: (5).

Here, ρ7 is the total impulse response of the transmission line and automatic equalizer, and ρ. indicates its peak value.

Therefore, in that case, from equation (2), yrl=Σak　ρn−k (5) Substituting Eq. --------・-(6) becomes.

The first term in equation (6) is the original data at the automatic equalizer output, and the second and second terms are the residual intersymbol interference.

The feature of the present invention is that an automatic gain adjuster 15 is inserted between the automatic equalizer 3 and the judgment circuit 4 in the area surrounded by the dashed line in FIG. Attenuated bones in the original data output (
1-ε. pt) and suppress level fluctuations as much as possible to prevent an increase in the error rate.

The input signal input to the input terminal 1 is automatically equalized by the automatic equalizer 3, and its output is gain limited by the automatic gain adjuster 15 so that the average power of the data points is constant. The output is sent to the decision circuit 4, the closest transmission data point is found, and its decision output value is output to the decision output terminal 2.

On the other hand, the output of the automatic equalizer 3 and the output of the determination circuit 4 are subtracted by a subtracter 5 to become an error signal, which is used to correct the tap weight of the automatic equalizer 3.

〔Effect of the invention〕

By providing an automatic gain adjuster at the output of the automatic equalizer, the present invention can correct the attenuation in the original data output and improve the characteristics of the automatic equalizer when line distortion is large. .

[Brief explanation of drawings]

FIG. 1 is a block diagram showing a conventional automatic equalization method. FIG. 2 is a block diagram showing details of the automatic equalizer. FIG. 3 is a block diagram showing an embodiment of the present invention. DESCRIPTION OF SYMBOLS 1... Input terminal, 2... Judgment output terminal, 3... Automatic equalizer, 4... Judgment circuit, 5... Subtractor, 6...
・Output terminal, 7...Error signal input terminal, 8-1 to 8-
3... Delay circuit, 9-1 to 9-3... Multiplier, 10
-1 to 10-3...Tap weight, 11...Coefficient unit,
12-1 to 12-3...multiplier, 13-1 to 13-3
...Subtractor, 14...Adder, 15...Automatic gain adjuster circuit. Patent applicant NEC Co., Ltd. Patent attorney Takashi Ide 1 (21st class Q)

Claims (1)

[Claims] 11) an automatic equalizer; a determination circuit for determining its output; and an error signal generated from the output of the automatic equalizer and the output of the determination circuit, and transmitted by the automatic equalizer. A baseband automatic equalizer for data transmission including means for equalizing transmission characteristics using a least squares error method, wherein the automatic equalizer has an output of the automatic equalizer and An automatic equalization device characterized by comprising an automatic gain adjuster between the determiner and the determiner.