JPH10229358A - Waveform equivalent coefficient generator and method therefor - Google Patents

Abstract

PROBLEM TO BE SOLVED: To simultaneously reduce the scale of a circuit and power consumption of a waveform equivalent coefficient generator, to generate an equivalent coefficient of each tap of a digital filter to equalize a signal waveform. SOLUTION: In a data-selecting means 22 in a complex operation part 20, either data Cn ,r of a real number part or the data Cn ,i of an imaginary number part of the equivalent coefficient is selected by a first data selector 22a, either data Xr of the real number part or data Xi of the imaginary number part of a signal is selected by a second data selector 22b, and either error data E1 of an I-axis or error data EQ of a Q-axis is selected by a third data selector 22c. Update of the data of both the real number part and the imaginary number part of the equalizing coefficient are calculated by a multiplier 24a and an adder 24f by a product-sum means 24.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

[0001] 1. Field of the Invention [0002] The present invention relates to an apparatus and a method for generating a waveform equalization coefficient used in multilevel digital microwave communication.

[0002]

2. Description of the Related Art In recent years, modulation and demodulation techniques of the digital microwave communication system have tended to be multi-valued in order to make effective use of frequencies. For example, QPSK, 16Q
In addition to AM, 64QAM, 256QAM and the like have begun to be used.

As the modulation and demodulation becomes more multi-valued as described above, the influence of signal distortion or the like generated on the transmission path becomes relatively large, and a technique for guaranteeing a correct signal on the receiving side becomes more and more important. Therefore, an automatic adaptive equalizer that performs transmission path equalization on the receiving side has been proposed.

First, equalization of a transmission path will be described.

FIG. 16 is a diagram showing a model of a transmission path and its equalization. As shown in FIG. 16, the signal transmitted from the transmitter changes according to the characteristics of the transmission path, and is received with noise added. On the receiving side, the received signal X ₀ is equalized to a desired signal Z ₀ for the receiver by an equalizer inserted in series before the receiver. When the noise is very small, an equalizer having an inverse characteristic of the transfer function of the transmission path may be used. However, when the noise is large to a certain extent, it is necessary to design the equalizer in consideration of the noise. The actual equalizer is
An equalizer configured using a digital filter and configured using a digital filter is called a digital equalizer.

FIG. 17 is a block diagram showing an example of the configuration of a digital filter. In FIG. 17, X ₀ is a received signal input via a transmission path, and X _{1 to} X _m are received signals X
₀ is delayed by each delay element, and C _{0 to} C _m are equalization coefficients. The received signal X ₀ is multiplied by an equalization coefficient C ₀ by a multiplier. Similarly, the signals X _{1 to} X _m are respectively multiplied by the equalization coefficients C _{1 to} C _m by the multiplier. The multiplication result of each multiplier is added by the adder, and the equalized signal Z
Output as ₀ .

In a digital filter, a mechanism for multiplying a certain delay signal by an equalization coefficient is called a tap. By adding the multiplication results at each tap, the equalized signal Z
₀ is obtained. At this time, a waveform equalizing coefficient generating device to calculate the optimal equalization coefficients C ₀ -C _m to recover the signal.

Next, an algorithm for generating an equalization coefficient will be described.

As described above, the signal transmitted from the transmitter changes according to the characteristics of the transmission path and is transmitted to the receiver in a form in which noise is added. If the characteristics of the transmission line are constant,
What is necessary is just to calculate the inverse characteristic of the transmission path and use a constant equalization coefficient for realizing the calculated inverse characteristic. However, in a system in which the influence and characteristics of noise change every moment, it is necessary to sequentially update the equalization coefficient according to the state of the received signal. What is used for updating the equalization coefficient is a so-called automatic adaptive algorithm. Actually, the next equalization coefficient is calculated based on the equalization coefficient of the previous state, but in this case, an evaluation index is set and the equalization coefficient is updated so that the value becomes minimum. Go on. A typical example of such an algorithm is the LMS algorithm.

The LMS (Least Mean Square) algorithm uses a mean square error as an evaluation index of an equalization coefficient. Specifically, the equalization coefficient is determined as shown in the following equation. C _{n + 1, m} = C _{n, m} −α × X _m × e ₀ (1) n: Number of updates of the equalization coefficient m: Tap number of the equalization coefficient e ₀ : Z ₀ −χ ₀ (χ ₀ signal before transmission) alpha is: step size, where the signal X _m and the error data _{e 0, X m = X m} (r) -jX m (i) e 0 = e 0 (r) + je 0 (r) ((R) is the real part data,
(i) denotes the imaginary part data, hereinafter the _{same), X m × e 0 =} (X m (r) × e 0 (r) + X m (i) × e 0 (i)) + j
(X _{m (r)} × e _{0 (i)} −X _{m (i)} × e _{0 (r)} ), and the equation (1) becomes as follows. C _{n + 1, m (r)} = C _{n, m (r)} −α × (X _{m (r)} × e _{0 (r)} + X _{m (i)} × e _{0 (i)} ) (2) C _{n +1, m (i)} = Cn _{, m (i)-.} Alpha. * (Xm _(r) * e0 _(i) -Xm _(i) * e0 _(r) ) (3)

However, in the case of an actual transmission system, the signal χ ₀ before transmission is not known on the receiving side, so that the error data e ₀
Cannot be used to calculate Therefore, a signal before transmission is estimated on the receiving side, and waveform estimation is performed using the estimated value as a reference signal. This is called a blind algorithm.

Although it may be difficult to understand in practice, if the update is repeated several thousand times under a certain constraint using the blind algorithm, the equalization coefficient converges, and the equalization of the signal waveform is realized. The conventional waveform equalization coefficient generation device updates the equalization coefficient by executing the operations shown in Expressions (2) and (3).

[0013]

However, the conventional waveform equalization coefficient generator has the following problems.

When updating the equalization coefficient according to the equations (2) and (3), it is necessary to prepare error data for each signal data in advance. For this reason, there is a problem that the capacity of the storage means necessary for storing the error data becomes enormous, and the circuit scale becomes large.

Further, in order to update the equalization coefficient, a large number of arithmetic units are required. For example, in order to execute the operations as shown in Expressions (2) and (3), six multipliers and a total of four adders or subtractors are required. For this reason, the conventional waveform equalization coefficient generation device has a problem that the circuit scale is large and the power consumption is large.

In view of the above problems, it is an object of the present invention to provide a waveform equalization coefficient generation device having a small circuit size and low power consumption.

[0017]

Means for Solving the Problems In order to solve the above-mentioned problems, a solution taken by the invention of claim 1 is to equalize a waveform of a signal transmitted by multilevel QAM modulation and transmitted to a waveform of a signal before transmission. In the digital equalizer that performs equalization, the update of the equalization coefficient is divided into real part data and imaginary part data as a waveform equalization coefficient generator that generates the equalization coefficient of each tap of the digital filter that equalizes the signal waveform. Then, the operation for updating the real part data and the imaginary part data is performed by a common arithmetic unit by appropriately selecting the data required for the operation, thereby greatly simplifying the device configuration and reducing the circuit scale. And power consumption is reduced.

According to a second aspect of the present invention, the waveform equalization coefficient generating device according to the first aspect is configured to perform
An error data generation unit that generates error data necessary for updating the equalization coefficient, and a real part data and an imaginary part based on the input signal and the error data output from the error data generation unit. And a complex operation unit that updates the error data separately from the data. The error data generation unit generates the error data by dividing the error data into I-axis direction error data and Q-axis direction error data. The operation unit performs an operation for updating the real part data and the imaginary part data of the equalization coefficient by appropriately selecting one of the real part data and the imaginary part data of the input signal, and By appropriately selecting one of the axial direction error data and the Q-axis direction error data, the processing is performed by a common arithmetic unit.

Further, in the invention according to claim 3, the complex operation unit in the waveform equalization coefficient generation device according to claim 2 is configured such that one of the real part data and the imaginary part data of the equalization coefficient before update is input to the input unit. A data selection control signal provided by applying one of the real part data and the imaginary part data of the output signal and one of the I-axis direction error data and the Q-axis direction error data output from the error data generator. Multiplying the I-axis direction error data or the Q-axis direction error data by the real part data or the imaginary part data of the input signal output from the data selection means. By adding the data obtained based on the result of the multiplication, the real part data or the imaginary number of the equalization coefficient before update outputted from the data selection means is obtained. Assume that a sum of products means for updating the data.

According to a fourth aspect of the present invention, the data selection means in the waveform equalization coefficient generating apparatus according to the third aspect receives the real part data and the imaginary part data of the equalization coefficient before updating,
A first data selector for selectively outputting one of the real part data and the imaginary part data of the pre-updated equalization coefficient according to the data selection control signal; and a real part data and an imaginary part of the input signal. A second data selector which receives data as input, selects and outputs one of real part data and imaginary part data of the input signal according to the data selection control signal, and outputs the I-axis direction error data and the Q-axis A third data selector which receives direction error data as input, and selects and outputs one of the I-axis direction error data and the Q-axis direction error data in accordance with the data selection control signal.

According to a fifth aspect of the present invention, the product sum means in the waveform equalization coefficient generation device according to the third aspect outputs reciprocal data of a step size indicating a degree of update of the equalization coefficient and the data output from the data selection means. A multiplier that multiplies the real part data or the imaginary part data of the input signal by the I-axis direction error data or the Q-axis direction error data output from the data selection unit, and outputs a multiplication result; A first and second data holder for holding the multiplication result data of the multiplier; and a multiplication result data for the multiplier, the one of the first and second data holders in accordance with the data selection control signal. A data splitter for inputting to one side, and the second
A reciprocal data generator for converting data output from the data retainer according to the data selection control signal as it is or by converting it into reciprocal data, and an equalization coefficient before update output from the data selection means And the data output from the first data holder and the data output from the reciprocal data generator are added, and the addition result is output data of the product-sum means. And an adder for outputting as

According to a sixth aspect of the present invention, the error data generation section in the waveform equalization coefficient generation apparatus according to the second aspect converts the error data necessary for updating the equalization coefficient into a STOP & GO function for the input signal. LMS (Least Me
an Square) algorithm based on the STOP & GO algorithm.

According to a seventh aspect of the present invention, the error data generating section in the waveform equalization coefficient generating apparatus according to the sixth aspect determines coordinate data indicating a position occupied by the input signal in the phase diagram of the multilevel QAM modulation. According to the obtained coordinate data, a stop data signal for instructing the stop of the output of the error data and a coordinate determiner that outputs one of the address data corresponding to the coordinate data, and the error data corresponding to each coordinate data are stored. Storage means for outputting error data stored at an address specified by the address data output from the coordinate determiner, and outputting a predetermined value when a stop data signal is output from the coordinate determiner, When the stop data signal is not output from the coordinate determination unit, the error data output from the storage unit is output. And that a data selector.

According to an eighth aspect of the present invention, in the waveform equalization coefficient generation apparatus of the seventh aspect, when a stop data signal is output from the coordinate determiner, the update of the equalization coefficient in the complex operation unit is performed. Calculation is stopped.

According to a ninth aspect of the present invention, the error data generating section in the waveform equalization coefficient generating apparatus according to the sixth aspect is configured such that the error data generation section determines whether the input signal occupies a positive or negative coordinate data indicating a position occupied in a phase diagram of the multilevel QAM modulation. When the coordinate data is positive, the coordinate data is output as it is as the coordinate data for determination, while when the coordinate data is negative, the coordinate data is bit-inverted and used as the coordinate data for determination. A coordinate data selection unit that outputs an inverted display signal indicating whether or not the coordinate data for determination is bit-inverted from the coordinate data, and a plurality of error evaluation data respectively corresponding to the positive coordinate data. The error evaluation data corresponding to the determination coordinate data output from the coordinate data selection unit is stored in a plurality of stored error evaluations. An error evaluation data storage for selectively outputting data from the data, and an error display data output from the coordinate data selection unit, when the coordinate data for determination indicates that the coordinate data is not bit-inverted of the coordinate data, While the error evaluation data output from the evaluation data storage is output as it is as error data, if the inverted display signal indicates that the determination coordinate data is a bit-inverted version of the coordinate data, the error evaluation data is output. And an error data selection unit that outputs data as error data after bit inversion of the data.

According to the ninth aspect of the present invention, the error data generation section responds to the input signal based on the error evaluation data corresponding to the positive coordinate data stored in the error evaluation data storage in advance. Error data can be generated. That is, the data that needs to be stored in advance to generate the error data is only the error evaluation data corresponding to the positive coordinate data. Therefore, the storage capacity required by the error data generation unit can be significantly reduced, and the waveform equalization coefficient generation device can be realized with a small circuit configuration.

According to a tenth aspect of the present invention, in the waveform equalization coefficient generating apparatus according to the sixth aspect, the error data generating section includes: Sato error reference data indicating coordinates of a Sato error reference point in a phase diagram of the multilevel QAM modulation; A reference data storage for storing coordinate determination reference data indicating a slice level value in a multilevel QAM modulation phase diagram, and error determination reference data indicating a coordinate of a signal reference point in the multilevel QAM modulation phase diagram; The input signal is converted into a phase diagram of the multi-level QAM modulation based on the coordinate data indicating the position occupied by the input signal in the phase diagram of the multi-level QAM modulation and the coordinate determination reference data output from the reference data storage. To determine which of the areas divided by each slice level belongs to, and determine the position of the signal reference point in the determined area. An LMS error generator that generates LMS error data for the input signal based on the error determination reference data and the coordinate data output from the reference data storage, and an LMS error generator output from the reference data storage. A SATO error generator for generating SATO error data for the input signal based on the SATO error reference data and the coordinate data, LMS error data generated by the LMS error generator, and SATO error generation Input the sign bit of the SATO error data generated by the
When the sign bit of the S error data is equal to the sign bit of the SATO error data, the LMS error data is selectively output as error data, and when the SMS error data is different, the predetermined value data is selectively output as error data. And a selection unit.

According to the tenth aspect of the present invention, the error data generating unit is configured to input the signal based on the SATO error reference data, the coordinate determination reference data and the error determination reference data stored in advance in the reference data storage. Can be generated according to the error data. That is, only the reference data is required to be stored in advance in order to generate the error data. Therefore, the storage capacity required by the error data generation unit can be significantly reduced, and the waveform equalization coefficient generation device can be realized with a small circuit configuration.

In a digital equalizer for equalizing the waveform of a signal transmitted by multi-level QAM modulation to the waveform of a signal before transmission, a solution implemented by the invention of claim 11 is to equalize a signal waveform. As a waveform equalization coefficient generation device that generates an equalization coefficient for each tap of a digital filter that performs an LMS (Least Mean S) using a STOP & GO function on an input signal, error data necessary for updating the equalization coefficient is used.
quare) an error data generating unit that generates an error based on a STOP & GO algorithm, and a complex operation unit that updates an equalization coefficient based on the input signal and error data output from the error data generating unit. The error data generating section obtains coordinate data indicating a position occupied by the input signal in the phase diagram of the multi-level QAM modulation, and, in accordance with the obtained coordinate data, a stop data signal for instructing to stop outputting error data. And a coordinate determiner that outputs one of the address data corresponding to the coordinate data, and error data corresponding to each coordinate data are stored, and an address specified by the address data output from the coordinate determiner is Storage means for outputting the stored error data; While when the signal is output for outputting a predetermined value, when from the coordinate determination unit does not output a stop data signal is assumed to have a data selector for outputting error data output from said storage means.

According to a twelfth aspect of the present invention, in the waveform equalization coefficient generating apparatus according to the eleventh aspect, the error data generating unit receives the error data output from the data selector as input, and outputs the input error data to an external device. And a data splitter for outputting one of the I-axis direction error data and the Q-axis direction error data in accordance with the data selection control signal given from.

According to a thirteenth aspect of the present invention, the coordinate determinator in the waveform equalization coefficient generating device according to the eleventh aspect comprises:
The error free area where the error data is a predetermined value is multi-valued Q
It is set in advance on the phase diagram of the AM modulation, and outputs the stop data signal when the position indicated by the obtained coordinate data is within the error-free area, and outputs the stop data signal when the position is not within the error-free area. Assume that corresponding address data is output.

Further, in the invention according to claim 14, the coordinate judging device in the waveform equalization coefficient generating device according to claim 13 comprises:
Multi-level QA according to the instruction of the given communication mode selection signal
It has a function of selecting an M modulation mode and setting an error-free area according to the selected mode.

[0033] In the invention of claim 15, the storage means in the waveform equalization coefficient generating apparatus of claim 11 stores only error data for coordinate data outside the error free area.

According to a sixteenth aspect of the present invention, in the waveform equalization coefficient generation device of the eleventh aspect, when a stop data signal is output from the coordinate determiner, the update of the equalization coefficient in the complex operation unit is performed. Calculation is stopped.

In a digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, the present invention provides a digital equalizer for equalizing a signal waveform. As a waveform equalization coefficient generation device that generates an equalization coefficient for each tap of a digital filter that performs an LMS (Least Mean S) using a STOP & GO function on an input signal, error data necessary for updating the equalization coefficient is used.
quare) an error data generating unit that generates an error based on a STOP & GO algorithm, and a complex operation unit that updates an equalization coefficient based on the input signal and error data output from the error data generating unit. Wherein the error data generation unit determines whether the coordinate data indicating the position occupied by the input signal in the phase diagram of the multi-level QAM modulation is positive or negative. If the coordinate data is positive, the error data generation unit determines the coordinate data. While the coordinate data is output as it is as it is, when the coordinate data is negative, the coordinate data is output as a determination coordinate data after bit inversion of the coordinate data, and the determination coordinate data is a bit inversion of the coordinate data. And a coordinate data selection unit that outputs an inverted display signal indicating whether or not the data is positive. A plurality of error evaluation data corresponding thereto are stored, and error evaluation data corresponding to the coordinate data for determination output from the coordinate data selection unit is selected and output from the plurality of stored error evaluation data. When the inverted display signal output from the evaluation data storage unit and the coordinate data selection unit indicates that the determination coordinate data is not a bit-inverted version of the coordinate data, the inverted evaluation signal is output from the error evaluation data storage unit. While the error evaluation data is output as it is as error data, if the inverted display signal indicates that the determination coordinate data is a bit-inverted version of the coordinate data, the error evaluation data is bit-inverted and an error is output. And an error data selection unit for outputting the data as data.

According to the seventeenth aspect of the present invention, the error data generating section has previously stored in the error evaluation data storage.
Based on the error evaluation data corresponding to the positive coordinate data,
Error data corresponding to the input signal can be generated. That is, the data that needs to be stored in advance to generate the error data is only the error evaluation data corresponding to the positive coordinate data. Therefore, the storage capacity required by the error data generation unit can be significantly reduced, and the waveform equalization coefficient generation device can be realized with a small circuit configuration.

In the eighteenth aspect of the present invention, in the waveform equalization coefficient generating device according to the seventeenth aspect, the error evaluation data storage is configured such that a position indicated by the determination coordinate data in the multi-level QAM phase diagram is an error. It is determined whether or not the data is within an error-free area having a predetermined value, and outputs an error-free signal indicating whether or not the position indicated by the coordinate data for determination is within the error-free area. The data selection unit, when the error-free signal output from the error evaluation data storage indicates that the position indicated by the determination coordinate data is within the error-free area,
It is assumed that data of a predetermined value is output as error data.

According to the eighteenth aspect of the present invention, only the error evaluation data corresponding to the positive coordinate data indicating the position outside the error free area is required to be stored in advance to generate the error data. Therefore, the storage capacity required by the error data generation unit can be further reduced as compared with the invention of claim 17.

In a digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, the present invention provides a digital equalizer for equalizing a signal waveform. As a waveform equalization coefficient generation device that generates an equalization coefficient for each tap of a digital filter that performs an LMS (Least Mean S) using a STOP & GO function on an input signal, error data necessary for updating the equalization coefficient is used.
quare) an error data generating unit that generates an error based on a STOP & GO algorithm, and a complex operation unit that updates an equalization coefficient based on the input signal and error data output from the error data generating unit. Wherein the error data generating unit includes: Sato error reference data indicating coordinates of a Sato error reference point in a multi-level QAM modulation phase diagram; and coordinate determination reference data indicating a slice level value in a multi-level QAM modulation phase diagram. And a reference data storage for storing error determination reference data indicating the coordinates of a signal reference point in the phase diagram of the multilevel QAM modulation, and coordinate data representing the position of the input signal in the phase diagram of the multilevel QAM modulation And the input signal based on the coordinate determination reference data output from the reference data storage. Belongs to an area divided by each slice level in the phase diagram of the multilevel QAM modulation, and an error determination criterion output from the reference data storage indicating coordinates of a signal reference point in the determined area. An LMS error generator that generates LMS error data for the input signal based on the data and the coordinate data, and a Sato error reference data and the coordinate data output from the reference data storage, A Sato error generator for generating Sato error data for the input signal;
The LMS error data generated by the S error generator and the sign bit of the Sato error data generated by the Sato error generator are input to the LMS error data.
When the sign bit of the error data and the sign bit of the SATO error data are equal, the LMS error data is selected and output as error data, and when they are different, a predetermined value data is selected and output as error data. Unit.

According to the nineteenth aspect of the present invention, the error data generating unit is configured to input the signal based on the SATO error reference data, the coordinate determination reference data and the error determination reference data stored in advance in the reference data storage. Can be generated according to the error data. That is, only the reference data is required to be stored in advance in order to generate the error data. Therefore, the storage capacity required by the error data generation unit can be significantly reduced, and the waveform equalization coefficient generation device can be realized with a small circuit configuration.

According to a twentieth aspect of the present invention, in the waveform equalization coefficient generating device according to the nineteenth aspect, the reference data storage stores absolute values of the SATO error reference data, coordinate determination reference data, and error determination reference data. When the coordinate data is negative, the LMS error generator generates LMS error data based on the coordinate determination reference data and the error determination reference data obtained by inverting the sign. When the coordinate data is negative, the Sato error generator generates Sato error data based on the sign of the Sato error reference data inverted.

According to the twentieth aspect, only the absolute value of each reference data needs to be stored in advance in order for the error data generation unit to generate the error data. The required storage capacity can be further reduced as compared to the nineteenth aspect.

According to the twenty-first aspect of the present invention, the reference data storage in the waveform equalization coefficient generating apparatus according to the nineteenth aspect includes a Sato error reference data and a coordinate determination reference respectively corresponding to a plurality of multilevel modes of QAM. Data and error judgment reference data, and store Sato error reference data, coordinate judgment reference data and error judgment reference data corresponding to the multi-value mode of QAM specified by a mode selection signal given from outside. Output from the selected data.

According to the twenty-first aspect, error data can be generated corresponding to each multi-level mode of multi-level QAM.

In a digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, the present invention provides a digital equalizer for equalizing a signal waveform. As a waveform equalization coefficient generation method for generating an equalization coefficient of each tap of a digital filter to be performed, error data necessary for updating an equalization coefficient is input to an input signal by LMS (Least Mean Meaning) using a STOP & GO function.
Error data generating step for generating based on a STOP & GO algorithm which is a square) algorithm, and a complex operation step for updating an equalization coefficient using the input signal and error data output from the error data generating section. In the error data generating step, an error-free area in which the error data has a predetermined value is set in advance on the phase diagram of the multi-level QAM modulation, and the error data corresponding to each coordinate data outside the error-free area is respectively set. A first step of obtaining coordinate data indicating a position occupied by the input signal in the phase diagram of the multilevel QAM modulation, and a position indicated by the coordinate data obtained in the first step being within an error-free area. , The second step of setting the error data to a predetermined value, and the position indicated by the coordinate data obtained in the first step. And a third step of selecting error data corresponding to the coordinate data from the prepared error data when the position is not within the error-free area.

In a digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, the present invention provides a digital equalizer for equalizing a signal waveform. As a waveform equalization coefficient generation method for generating an equalization coefficient of each tap of a digital filter to be performed, error data necessary for updating an equalization coefficient is input to an input signal by LMS (Least Mean Meaning) using a STOP & GO function.
Error data generating step for generating based on a STOP & GO algorithm which is a square) algorithm, and a complex operation step for updating an equalization coefficient using the input signal and error data output from the error data generating section. The error data generating step determines whether the coordinate data indicating the position occupied by the given signal in the phase diagram of the multi-level QAM modulation is positive or negative, and when the coordinate data is positive, the coordinate data is used as it is for the determination coordinates. On the other hand, when the coordinate data is negative, the coordinate data is bit-inverted and the coordinate data is selected as the coordinate data for determination, and the coordinate data for determination selected in the coordinate data selecting step is selected. An error evaluation data setting step of setting error evaluation data corresponding to When the coordinate data is selected as the determination coordinate data in the coordinate data selecting step, the error evaluation data set in the setting step is selected as it is as the error data, while the coordinate data is selected in the coordinate data selecting step. If the bit is inverted and selected as the judgment coordinate data,
And an error data selecting step of selecting as error data after bit inversion.

According to a twenty-fourth aspect of the present invention, in the waveform equalization coefficient generating method according to the twenty-third aspect, the error evaluation data setting step is such that the position indicated by the determination coordinate data in the multilevel QAM phase diagram is error data Is included in an error-free area having a predetermined value, and the error data selecting step is a step in which the position indicated by the determination coordinate data is in the error-free area in the error evaluation data setting step. , It is assumed that the method includes a step of selecting data of a predetermined value as error data when it is determined that the error data is present.

In a digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, a solution means according to the invention of claim 25 is to equalize a signal waveform. As a waveform equalization coefficient generation method for generating an equalization coefficient of each tap of a digital filter to be performed, error data necessary for updating an equalization coefficient is input to an input signal by LMS (Least Mean Meaning) using a STOP & GO function.
Error data generating step for generating based on a STOP & GO algorithm which is a square) algorithm, and a complex operation step for updating an equalization coefficient using the input signal and error data output from the error data generating section. , The error data generating step includes a multi-value QA
Sato error reference data indicating the coordinates of the Sato error reference point in the M modulation phase diagram, coordinate determination reference data indicating the slice level value in the multi-value QAM modulation phase diagram,
A reference data setting step of setting error determination reference data indicating the coordinates of a signal reference point in the phase diagram of the multilevel QAM modulation; and coordinate data indicating the position of the given signal in the phase diagram of the multilevel QAM modulation. Using the coordinate determination reference data, it is determined which of the regions divided by each slice level the given signal belongs to in the phase diagram of the multilevel QAM modulation, and the coordinates of the signal reference point in the determined region are determined. An LMS error generating step of generating LMS error data for the given signal using the error determination reference data, and a Sato error for the given signal using the coordinate data and the Sato error reference data. A Sato error generating step of generating data, the LMS error data and the Sato error Comparing the sign with the data and selecting the LMS error data as error data when the signs are equal, and selecting a data of a predetermined value as error data when the signs are different. It is assumed that

According to a twenty-sixth aspect of the present invention, in the waveform equalization coefficient generating method according to the twenty-fifth aspect, the reference data setting step includes the steps of calculating the absolute values of the SATO error reference data, the coordinate determination reference data, and the error determination reference data. Setting, when the coordinate data is negative, the LMS error generating step is to generate an LMS error based on the sign determination reference data and the error determination reference data obtained by inverting the sign. In the Sato error generation step, when the coordinate data is negative, the Sato error data is generated based on the polarity of the Sato error reference data inverted.

According to a twenty-seventh aspect of the present invention, the reference data setting step in the waveform equalization coefficient generating method according to the twenty-fifth aspect includes the step of determining the SATO error reference data and the coordinate determination in accordance with the designated QAM multi-value mode. It is assumed that reference data and error determination reference data are set.

[0051]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a waveform equalization coefficient generator according to an embodiment of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram showing the overall configuration of a signal demodulation LSI using a waveform equalizer (digital equalizer). The input analog modulation signal is converted into a digital modulation signal by an A / D converter, and then separated into an In-phase signal (I data) and a Quadrature signal (Q data) through an I / Q detector. From the I data and Q data, a clock synchronized with the original signal before modulation is extracted by the clock regenerator.

The I data and Q data are shaped into a waveform through a roll-off filter, and then input to a waveform equalizer (digital equalizer) to be equalized. AFC for the signal once equalized by the waveform equalizer
(Auto Frequency Control) and APC (Auto Phase C)
ontrol) to remove the frequency offset and the phase offset. The signal subjected to AFC / APC is returned to the waveform equalizer again. This series of waveform equalization and AFC / A
PC is repeated until the equalization converges. on the other hand,
AGC (Auto Gain Control) is performed based on the signal whose waveform has been shaped by the roll-off filter.

The I data and Q data thus converged are converted into ordinary digital data by the inverse mapper.

FIG. 2 is a block diagram schematically showing the configuration of the digital equalizer. As shown in FIG. 2, the digital equalizer includes a filter unit 1 and a waveform equalization coefficient generation device 2. The waveform equalization coefficient generator 2 calculates and outputs an equalization coefficient at each tap of the digital filter 4 constituting the filter unit 1, and the filter unit 1 operates each digital filter 4 using the equalization coefficient. , And converts the received signal X into an equalized signal Z. Digital filter 4
Is composed of a delay element 5, a tap 6, and an adder 7.

Here, usually, the equalized signal Z is (K is a variable representing a tap number), but when the equalized signal Z is considered in a complex expression, X = X (r) + jX (i), C
= C (r) + jC (i), Becomes Therefore, the filter unit 1 requires a total of four digital filters 4. Each digital filter 4
Are output by the adder / subtracter 8 and output as equalized signals Z (r) and Z (i). The equalized signals Z (r) and Z (i) output from the filter unit 1 are input to the AFC / APC unit 3.

In the present invention, STO is used to determine the equalization coefficient.
A method called a P & GO algorithm is used. STOP &
The GO algorithm is a method of determining whether or not to update an equalization coefficient depending on the direction of a vector between an LMS error and a Sato error, and is one of blind algorithms for performing waveform equalization without using a reference signal. .

First, the LMS error is defined as follows. Assuming that the LMS error is LMSER and the reference signal is D, LMSER = Z ₀ -D.

Next, the Sato error is defined as follows. Assuming that the reference value of the SATO error is B, B is defined as follows. _{B = E (| A n |} 2) / E (| A n |) A n: reference signal vector E (): If the average Sato error and Sater, the SATER = Z ₀ -B.

FIG. 3 is a diagram showing the first quadrant of the 64QAM phase diagram, showing the relationship between the LMS error and the Sato error. In FIG. 3, black circles are signal point positions, and black circles surrounded by circles are reference points of Sato error. Equalized signal Z ₀
Is located at the position of the white circle in FIG. 3, the LMS error and the Sato error are the vectors shown in FIG. There are four reference points for the SATO error in each quadrant, that is, a total of four reference points.
In the case of QAM, there are a total of 64 points.

Using the STOP & GO algorithm,
The equation for updating the equalization coefficient is as follows. _{C n + 1, m (r} ) = C n, m (r) -α × (X m (r) × e 0 (r) × f r + X m (i) × e 0 (i) × f i) .. (4) C _{n + 1, m (i)} = C _{n, m (i)} −α × (X _{m (r)} × e _{0 (i)} × f _i −X _{m (i)} × e _{0 (r )} × _fr )… (5)

[0062] Here, f _r, f _i is the real axis is a flag that is calculated independently for each of the imaginary axis, is defined by the following conditions. f _r = 1; sgn (I component of LMSER) = sgn (I component of SATER) 0; (I component of LMSER) sgn (I component of _{SATER) ≠ sgn f i = 1} ; sgn (Q component of LMSER) = sgn (Q component of SATER) 0; sgn (Q component of LMSER) ≠ sgn (Q component of SATER)

Here, e _{0 (r)} × f _r is EI, and e _{0 (i)} × f
_Assuming that _i is EQ, equations (4) and (5) are, respectively, C _{n + 1, m (r)} = C _{n, m (r)} -α × (X _{m (r)} × EI + X _{m (i)} × EQ) (6) Cn _{+ 1, m (i)} = Cn _{, m (r)} −α × (Xm _(r) × EQ− _{Xm (i)} × EI) (7) . Here, EI and EQ are called error data.

The apparatus for generating a waveform equalization coefficient according to the present embodiment updates the equalization coefficient according to equations (6) and (7).

(First Embodiment) FIG. 4 is a block diagram showing a configuration of a waveform equalization coefficient generation device according to a first embodiment of the present invention. In FIG. 4, reference numeral 10 denotes an error data generation unit, which outputs error data EI, EI,
Generate EQ. A complex operation unit 20 performs a complex operation using the error data EI and EQ output from the error data generation unit 10 to update the equalization coefficient.

The error data generator 10 is provided with the coordinate determiner 1
1, a ROM 12 as a storage means, a data selector 13, and a data brancher 14.

The complex operation unit 20 includes a data branching unit 21, a data selection unit 22, a reciprocal data generator 23, and a product-sum unit 2.
4, and third to eighth data holders 25a, 25b, 2
6a, 26b, 27a and 27b. The data selection means 22 includes a first data selector 22
a, a second data selector 22b and a third data selector 22c. The product-sum unit 24 includes a multiplier 24a, a data branching unit 24b, and a first data holding unit 2.
4c, second data holder 24d, reciprocal data generator 2
4e and an adder 24f. Further, the waveform equalization coefficient generator shown in FIG.
8 is provided.

Here, it is assumed that the tap number m of the equalization coefficient is fixed. At this time, equations (6) and (7) are as follows. C _{n + 1, r} = C _{n, r} −α × (X _r × EI + X _i × EQ) (8) C _{n + 1, i} = C _{n, i} −α × (X _r × EQ−X _i × EI)… (9)

The operation of the waveform equalization coefficient generator shown in FIG. 4 will be described assuming that the equalization coefficient is updated in accordance with equations (8) and (9).

First, the operation of the error data generator 10 will be described.

First, according to the instruction of the mode selection control signal (not shown) externally applied, 1
Select either 6QAM, 64QAM or 256QAM. The coordinate determiner 11 sets a signal point position and a slice level on the phase diagram of the selected multi-valued QAM, and determines at which coordinates on the phase diagram the input signal is located.

FIG. 5 is a phase diagram showing an example of signal point positions and slice levels set when 64QAM is selected. In FIG. 5, black circles indicate signal point positions, broken lines indicate slice levels, and black circles indicate reference points for Sato error. As shown in FIG. 5, the coordinates of the signal point position are -252, -1
80, -108, -36, 36, 108, 180, and 252, and the coordinates of the slice level are -216, -144, -72, 0, 72, 144, and 21.
6 In addition, the coordinates of the reference point of the Sato error are (247,
247), (-247,247), (-247, -247), (247, -247)
And

Here, considering only the positive I-axis data to simplify the description, the positions of signal points and slice levels are as shown in FIG. Here, the value of the flag f _r is as follows. 0-36 when f _r = 1 180 ~216 when f _r = 0 144 ~180 when f _r = 1 108 ~144 when f _r = 0 72 ~108 when f _r = 1 36 ~72 when f _r = 0 216 ~247 when f _r = 1 247 ~252 when f _r = 0 252 when the ~ f _r = 1

In FIG. 6, the range where f _r = 0 (referred to as an error-free area) is indicated by oblique lines. For example, when the real part data (I-axis data) X _r of the input signal is "110", and f _r = 0 because the range of 108-144. Further, when the I-axis data X _r is "151", and f _r = 1 because the range of 144-180. When _fr = 0, the error data EI becomes zero.

[0075] coordinate determiner 11 determines the value of the flag f _r from the I-axis data X _r of the input signal, while outputting the stop data signal 30 to the data selector 13 when f _r = 0, f _r = and it outputs the specified address data corresponding to the I-axis data X _r when 1 to ROM 12. ROM 12
Error data corresponding to the I-axis data within the range where _fr = 1 is stored in advance, and when the address data is output from the coordinate determiner 11, the error data stored at the address indicated by the address data is output. I do.

When the stop data signal 30 is not output from the coordinate discriminator 11, the data selector 13
While the error data output from M12 is output as it is, when the stop data signal 30 is output from the coordinate determiner 11, "0" is output as error data.

Similarly, the coordinate determiner 11 determines the value of the flag f _i from the imaginary part data (Q-axis data) X _i of the input signal. When f _i = 0, the data selector 13 sends the stop data signal 30 When f _i = 1, specific address data corresponding to the Q-axis data X _i is output to the ROM 12. The ROM 12 previously stores error data corresponding to Q-axis data within a range where f _i = 1, and when address data is output from the coordinate determiner 11, the error stored at the address indicated by the address data is stored. Output data.

When the stop data signal 30 is not output from the coordinate discriminator 11, the data selector 13 outputs
While the error data output from M12 is output as it is, when the stop data signal 30 is output from the coordinate determiner 11, "0" is output as error data.

The data splitter 14 converts the error data output from the data selector 13 into error data EI or EQ.
Output as In this way, the error data EI or EQ is output from the error data generator 10.

Next, the operation of the complex operation unit 20 will be described.

The equalization coefficient output from the complex operation unit 20 is input to the data branching unit 21 as an equalization coefficient before updating. The data branching unit 21 divides the input equalization coefficient into real part data C _{n, r} and imaginary part data C _{n, i} and outputs them. The real part data C _{n, r} of the equalization coefficient is held by the third data holder 25a, while the imaginary part data C _{n, i} of the equalization coefficient is held by the fourth data holder 25b.

After the real part data X _r and the imaginary part data X _i of the input signal are delayed by the data delay 28,
It is input to the complex operation unit 20. Data delay unit 28, the timing of the input of the real part data X _r and imaginary part data X _i of the input signal, in order to adjust the delay time required for the error data is generated by the error data generation unit 10 Is provided. Real part data X of input signal
_r is held by the fifth data holder 26a, while the imaginary part data X _i of the input signal is held by the sixth data holder 26b.

The error data EI output from the error data generator 10 is held by the seventh data holder 27a, while the error data EQ is held by the eighth data holder 27b.

The first data selector 22a outputs the real part data C _{n, r} of the equalization coefficient held by the third data holder 25a and the fourth data in accordance with the data selection control signal 31.
And selectively outputs one of the imaginary part data C _{n, i} of the equalization coefficient held by the data holding unit 25b. Second
The data selector 22b, according to the data selection control signal 31, the real part data X _r and sixth data cage of the fifth input signal held by the data holding unit 26a of 26
b) selects and outputs one of the imaginary part data X _i of the input signal held by b. Third data selector 22c
Selects and outputs one of the error data EI held by the seventh data holder 27a and the error data EQ held by the eighth data holder 27b in accordance with the data selection control signal 31. The reciprocal data generator 13 generates and outputs reciprocal data having a step size α given from the outside.

First to third data selectors 22a to 22c
And the output data of the reciprocal data generator 23 are output to the product-sum means 24.

The multiplier 24a is connected to the second data selector 22
b, the output data of the third data selector 22c, and the output data of the reciprocal data generator 23, and outputs the multiplication result to the data branching unit 24b. The data branching unit 24b outputs the input data to one of the first data holding unit 24c and the second data holding unit 24d according to the data selection control signal 31.

After holding the data output from the data branching unit 24b, the first data holding unit 24c
Output to 4f. The second data holding unit 24d holds the data output from the data branching unit 24b, and then outputs the data to the reciprocal data generator 24e. Reciprocal data generator 24
“e” receives the data output from the second data holding unit 24 d as input and generates and outputs the reciprocal data of the input data according to the data selection control signal 31 or outputs the input data as it is.

The adder 24f is connected to the first data selector 22
a, the output data of the first data holder 24c and the output data of the reciprocal data generator 24e, and
The addition result is output as a new equalization coefficient.

More specifically, the operation of updating the equalization coefficient according to equations (8) and (9) is as follows.

[0090] First, while inputting the real part data X _r of the signals, calculates the error data EI by the error data generation unit 10. Then, “(−
α) × X _r × EI ”is calculated, and the calculation result is held by the first data holding unit 24c.

Next, the imaginary part data X _i of the signal is input, and the error data generator 10 calculates the error data EQ. Then, “(−
α) × X _i × EQ ”is calculated, and the calculation result is held by the second data holding unit 24d.

Finally, “C _{n, r} +
_{(-Α) × X r × EI} + (- α) × X i × EQ "is calculated, and the real part data C _{n + 1, r} of the new equalization coefficients. By these operations, the real part data of the equalization coefficient is updated according to the equation (8).

[0093] In addition, inputs the real part data X _r of the signals, calculates the error data EQ by the error data generation unit 10. Then, “(−
α) × X _r × EQ ”is calculated, and the calculation result is held by the first data holding unit 24c.

Next, the imaginary part data X _i of the signal is input, and the error data generator 10 calculates the error data EI. Then, “(−
α) × X _i × EI ”, the result of the calculation is held by the second data holding unit 24d, and the reciprocal data generating unit 24e
Is given by “α × X _i × EI” (= − (− α) × X _i × E
Find I).

Finally, “C _{n, i} +
_{(-Α) × X r × EQ} + α × X i × EI "is calculated, and the imaginary part data C _{n + 1, i} of the new equalization coefficients. By these operations, the imaginary part data of the equalization coefficient is updated according to the equation (9).

FIG. 7 is a timing chart showing the operation of the equalization coefficient device shown in FIG. As shown in FIG.
The data selector 22a selects one of the real part data and the imaginary part data of the equalization coefficient as appropriate according to the data selection control signal 31, and the second data selector 22b selects the input signal according to the data selection control signal 31. , The third data selector 22c selects one of the error data EI and EQ according to the data selection control signal 31. The reciprocal data generator 24e switches whether to generate reciprocal data of the data held in the second data holder 24d according to the data selection control signal 31. By such an operation, the real part data and the imaginary part data of the equalization coefficient are output alternately.

As described above, according to the waveform equalization coefficient generating apparatus according to the present embodiment, the real part data or the imaginary part data of the equalization coefficient is selected and updated, and the input signal and the error when updating are selected. By selecting the data appropriately,
The number of multipliers and adders in the device can be greatly reduced.

Further, since it is not necessary to store error data for a signal whose error data has a predetermined value, the storage capacity of the ROM in the error data generator can be greatly reduced.

(Second Embodiment) Hereinafter, a second embodiment of the present invention will be described with reference to the drawings.

FIG. 8 is a block diagram showing a configuration of a waveform equalization coefficient generator according to the second embodiment of the present invention. FIG.
4 is different from the waveform equalization coefficient generation device shown in FIG. 4 in that the stop data signal 30 output from the coordinate determination unit 11 in the error data generation unit 10 is input to the multiplier 24A in the complex operation unit 20. It is a point. Other configurations are the same as those of the waveform equalization coefficient generation device shown in FIG. 4, and common components are denoted by the same reference numerals.

The operation of the waveform equalization coefficient generator shown in FIG. 8 will be described.

The error data generator 10 has the first
This embodiment is the same as the first embodiment except that the stop data signal 30 output from the coordinate determiner 11 is also input to the multiplying means 24A.

Next, the operation of the complex operation unit 20 will be described. The operations of the data splitter 21, the data selection means 22, and the reciprocal data generator 23 are the same as in the first embodiment. The output data of the second data selector 22b, the output data of the third data selector 22c, and the reciprocal data generator 2
The output data of No. 3 is input to the multiplier 24A.

The multiplier 24A includes an error data generator 10
When the stop data signal 30 is output from the coordinate determiner 11 in the above, it is determined that the error data output from the third data selector 22c is “0”, and “0” is output as data without performing multiplication. Is output. On the other hand, the coordinate determiner 1
When the stop data signal 30 is not output from 1, the input data is multiplied and the multiplication result is output. The operation of the other components of the sum of products means 24 is the same as that of the first embodiment.

As described above, according to the waveform equalization coefficient generator according to the present embodiment, the multiplier 24A performs the multiplication when the error data output from the third data selector 22c is “0”. "0" is output without performing. Therefore, the power consumption of the entire device can be significantly reduced. Flag f in STOP & GO algorithm
Since the probability _r or the f _i is 0 is 1/2, the probability that the stop data signal is output is also at 1/2, the power consumption becomes about 1/2.

The waveform equalization coefficient generator according to the first and second embodiments may be provided for each tap of the digital filter, or may be provided so as to be commonly used for each tap.

The error data generation unit 10 according to the first and second embodiments is used not only with the complex operation unit 20 according to the first and second embodiments but also with a complex operation unit having another configuration. For example, it may be used together with a complex operation unit in which the equalization coefficient is updated with the real part data and the imaginary part data by another arithmetic unit.

The same effect can be obtained by using, instead of the input signals X _r and X _i , a signal that has been equalized and returned to the digital equalizer after AFC / APC has been performed.

(First Modification) FIG. 9 is a block diagram showing a configuration of a first modification of the error data generator 10 in the waveform equalization coefficient generator according to the first and second embodiments of the present invention. It is. As shown in FIG. 9, the error data generation unit 10A according to the present modification includes a coordinate data selection unit 51 including a first bit inverter 51a and a first data selector 51b, an error evaluation data storage 52, and The error data selector 53 includes a second bit inverter 53a and a second data selector 53b. Error data generation unit 10A further includes (real axis component of the coordinates data) the real part data of the input signal X _r, (imaginary axis component of the coordinate data) imaginary part data of the input signal buffer that outputs the X _i once to sequentially hold 54, and a buffer 55 for temporarily storing the obtained error data EI and EQ. FIG. 10 is a block diagram showing the internal configuration of the error evaluation data storage 52.

As in the first embodiment, assuming that the equalization coefficient is updated in accordance with equations (8) and (9),
The operation of the error data generation unit 10A shown in FIG. 9 will be described with reference to the flowchart shown in FIG. Here, a case of obtaining the error data EI from the real axis component X _r coordinate data as an example.

On the IQ coordinate plane of the multilevel QAM modulation, the coordinate data of each coordinate point and the STO
When both error data obtained by the P & GO algorithm are expressed in binary numbers, their bit-inverted values also correspond to each other. For example, when the error data corresponding to the coordinate data “00” is “100”, the error data corresponding to the coordinate data “11” is “011”. In this embodiment, error data is processed using such a relationship. Also, by utilizing such a relationship, the error evaluation data storage 52
Means that, for each multi-valued mode, only the error evaluation data for the coordinate data of which the sign of the coordinate data is positive and out of the error free area among all the coordinate points need be stored.

First, in step S31, the multi-value mode of QAM is selected. Now, the error data generation unit 10A shown in FIG.
It is assumed that the configuration is such that it can support any of 56QAM. Mode selection signal SL given from outside
Is 16QAM, 32QAM, 64Q among multi-value QAM
A signal for instructing to select one of AM and 256QAM, and is input to the error evaluation data storage 52. The error evaluation data storage 52 is, as shown in FIG.
Are provided for storing error evaluation data corresponding to each multi-level mode of the QAM. The error data storage bank 52b to be referred to when reading the error evaluation data is provided by the mode selection signal SL. Selected according to instructions. Each error data storage bank 52b
Stores, for each corresponding multi-value mode, only the error evaluation data of the coordinate data of which the sign of the coordinate data is positive and out of the error free area among all the coordinate points.

The coordinate data _Xr is input to the coordinate data selection section 51 (step S32), and the first bit inverter 51
The bit is inverted by a. First data selector 51
b is the most significant bit (MSB) selecting control signal which is the sign bit of the coordinate data X _r, one of the inverted coordinate data / X _r outputted from the coordinate data X _r or the first bit inverter 51a Selectively output (Step S3
(S3, S35), and outputs an inverted display signal S1 indicating whether or not the coordinate data _Xr has been inverted (step S34,
S36). For example, when the selection control signal is "0" (when the coordinate data _Xr is positive), the coordinate data _Xr is selectively output and "0" is output as the inverted display signal S1, while the selection control signal is "1". (When the coordinate data _Xr is negative), the inverted coordinate data / _Xr is selectively output and "1" is output as the inverted display signal S1. The operation up to this point corresponds to determining whether the coordinate data is positive or negative when considering processing in the I-axis direction on the IQ coordinate plane shown in FIG. In FIG. 12, a black circle surrounded by a circle corresponds to a reference point of the SATO error, a black circle corresponds to a signal point whose position is indicated by the error judgment reference data, and a broken line corresponds to a slice level indicated by the coordinate judgment reference data. By such an operation, steps S32 to S3
6 is performed.

[0114] Error evaluation data storage unit 52 receives as input the coordinate data X _r or inverted coordinate data / X _r outputted from the coordinate data selecting section 51, referring to the error data storage banks 52b selected by the mode selection signal SL _Then , the error evaluation data DE corresponding to the coordinate data _Xr or the inverted coordinate data / _Xr assuming the IQ coordinate plane corresponding to the selected multi-value mode is output. However, when the coordinate data X _r or inverted coordinate data / X _r belongs to an error-free region, such as a hatched portion shown in FIG. 13 (S37)
Since the error evaluation data is “0”, the coordinate recognizer 52
a outputs "1" as the error-free signal S2 when the coordinate data _Xr or the inverted coordinate data / _Xr belongs to the error-free area (S38). At this time, the data selector 52c holds the error evaluation data DE output for the immediately preceding coordinate data _Xr or the inverted coordinate data / _Xr . On the other hand, coordinate data _Xr or inverted coordinate data / _Xr
Is not in the error-free area, the coordinate recognizer 52a
Outputs “0” as the error-free signal S2 (S4
0), the data selector 52c outputs the error evaluation data DE that correspond to the coordinate data X _r or inverted coordinate data / X _r (S39). By such an operation, step S37 is performed.
To S40 are performed.

The error evaluation data DE is input to the error data selection section 53, and the bit is inverted by the second bit inverter 53a. The second data selector 53b uses the inverted display signal S1 output from the coordinate data selection unit 51 and the error-free signal S2 output from the error evaluation data storage 52 as selection control signals, and uses the error evaluation data DE and the second And outputs one of the inversion error evaluation data / DE output from the bit inverter 53a and the data "0" (steps S41 to S45). The data selectively output from the second data selector 53b is the error data E
It is output from the error data selection unit 53 as I.

When the error free signal S2 is "1" (when the coordinate data _Xr or the inverted coordinate data / _Xr is in the error free area), data "0" is output as error data EI (step S42). ). When the error free signal S2 is "0" (when the coordinate data _Xr or the inverted coordinate data / _Xr is outside the error free area),
When the inverted display signal S1 is "0", the error evaluation data DE
Is output as it is as error data EI (step S45), while when the inverted display signal S2 is "1", the inverted error evaluation data / DE is output as error data EQ (step S44).

The operation of the error data selection unit 53 as shown in steps S41 to S45 is performed when the inverted display signal S1 is "0" in consideration of the processing in the I-axis direction on the IQ coordinate plane shown in FIG. Since the sign of the coordinate data on the coordinate plane is positive, the error evaluation data DE for the positive coordinate data previously stored in the error evaluation data storage 52 is output as it is as the error data EI, while the inverted display signal S1 is When the value is “1”, the sign of the coordinate data on the coordinate plane is negative, so that the error evaluation data D for the positive coordinate data stored in advance in the error evaluation data storage 52 is stored.
This corresponds to outputting a value / DE obtained by bit-inverting E as error data EI.

[0118] Also the case of obtaining the error data EQ from the coordinate data X _i, can be carried out by the same operation.

As described above, according to the present modification, only the absolute value of the error evaluation data corresponding to the coordinate data existing outside the error-free area needs to be stored, so that the storage capacity required by the error data generation unit is reduced. The waveform equalization coefficient generation device can be realized by a small-scale circuit configuration.

(Second Modification) FIG. 14 is a block diagram showing the configuration of a second modification of the error data generator 10B in the waveform equalization coefficient generator according to the first and second embodiments of the present invention. It is. As shown in FIG. 14, the error data generation unit 10B according to the present modification includes the LMS error generator 4
1, a SATO error generator 42, a reference data storage 43, and an error data selector 44. Error data generation unit 10B further (real axis component of the coordinates data) the real part data of the input signal X _r, (imaginary axis component of the coordinate data) imaginary part data of the input signal buffer that outputs the X _i once to sequentially hold 44, and the obtained error data EI
And a buffer 45 for temporarily holding the EQ.

In FIG. 14, an LMS error generator 41
Coordinate data X _r of the equalized signal, which generates an LMS error data SLMS from X _i, the first inverse number generator 41a, the first data selector 41b, the coordinate data judging unit 41c and the first It is constituted by an adder 41d. The SATO error generator 42 calculates the coordinate data X of the equalized signal.
_r, to generate the Sato error data from X _i, the most significant bit is the sign bit of the generated Sato error data to output the result as a Sato error code signals SATO, a second inverse number generator 42a, the third Reciprocal generator 42
b, second data selector 42c and second adder 42d
It is constituted by. The reference data storage 43 stores Sato error reference data SS serving as a reference for generating Sato error data, error determination reference data SE serving as a reference for generating LMS error data SLMS, and coordinate data X.
_r, the X _i of I-Q coordinate coordinate determination reference data SC which position the determining criterion in the plane, the mode selection signal SL
Is output in a form corresponding to each multi-level mode of QAM specified by. The error data selection unit 44 receives the LMS error data SLMS output from the LMS error generator 41.
And an error determination using the SATO error code signal SATO output from the SATO error generator 42 to output error data EI and EQ, and is constituted by a third data selector 44a and an exclusive OR circuit 44b. Have been.

As in the first embodiment, it is assumed that the equalization coefficient is updated in accordance with the equations (8) and (9).
The operation of the error data generator 10 shown in FIG. 14 will be described with reference to the flowchart shown in FIG. Here, a case of obtaining the error data EI from the real axis component X _r coordinate data as an example.

First, in step S11, the QAM multi-value mode is selected. The error data generator 10 shown in FIG. 14 now has 16 QAM, 32 QAM, 64 QAM,
It is assumed that the configuration is such that it can support any of 56QAM. Mode selection signal SL given from outside
Is 16QAM, 32QAM, 64Q among multi-value QAM
A signal for instructing to select one of AM and 256QAM, and is input to the reference data storage 43.

Next, in step S12, the Sato error reference data SS and the coordinate determination reference data SC are set according to the multi-value mode selected in step S11.
The reference data storage 43 stores data corresponding to each multi-level mode as shown in Table 1, and a mode selection signal SL
Select and output the SATO error reference data SS and the coordinate determination reference data SC in accordance with the multi-level mode of QAM specified by the above. For example, when 64QAM is specified, three data as shown in Table 1 are output from the reference data storage 43 as the coordinate determination reference data SC.

[0125]

[Table 1]

Next, in step S13, the coordinate data X
Enter _r . LMS error and Sato error is generated respectively for the coordinate data X _r in the following steps.

(LMS Error Generating Step) The LM consisting of steps S14 to S20 is generated by the LMS error generator 41.
An S error generation step is performed. First, the first reciprocal generator 4
1a converts the coordinate determination reference data SC output from the reference data storage 43 into reciprocal coordinate determination reference data -SC, which is data in which the absolute value is equal to each of these data and the sign is inverted, and is output. I do. Next, the first data selector 41b uses the most significant bit (MSB), which is the sign bit of the coordinate data Xr, as a selection control signal and outputs the coordinate determination reference data SC output from the reference data storage 43 or the first counter data. One of the reciprocal coordinate determination reference data-SC output from the number generator 41a is selected and output. For example, when the selection control signal is “0” (when the coordinate data _Xr is positive), the coordinate determination reference data SC is selectively output,
When the selection control signal is "1" (when the coordinate data _Xr is negative), the reciprocal coordinate determination reference data -SC is selectively output.
Steps S14 and S15 are performed by such an operation.

The coordinate data judging unit 41c outputs the coordinate judging reference data S selected and output from the first data selector 41b.
C or reciprocal coordinate determination reference data-SC and coordinate data X
_{Using r} as an input, it was determined which of the coordinate range created by the coordinate determination reference data SC and the reciprocal coordinate determination reference data -SC on the IQ coordinate plane as shown in FIG. The result is output as a coordinate data determination signal SR. In FIG. 12, a black circle surrounded by a circle corresponds to a reference point of the SATO error, a black circle corresponds to a signal point whose position is indicated by the error judgment reference data, and a broken line corresponds to a slice level indicated by the coordinate judgment reference data. The determination of the range of the coordinates of the equalized signal included in the broken line (step S16) is realized by comparing the coordinate determination reference data corresponding to each slice level with the coordinate data of the equalized signal. You.

The coordinate data judgment signal SR output from the coordinate data judgment unit 41c is input to the reference data storage 43, and the reference data storage 43 outputs the coordinate data judgment signal SR.
In accordance with the mode selection signal SL, the error determination reference data SE corresponding to the coordinate range indicated by the coordinate data determination signal SR is selectively output from the stored data. Step S17 is performed by such an operation.

The first reciprocal generator 41a converts the error judging reference data SE output from the reference data storage 43 into reciprocal error judging reference data -SE, which is data having the same absolute value and inverted signs. Convert and output. The first data selector 41b uses the most significant bit (MSB), which is the sign bit of the coordinate data Xr, as a selection control signal and outputs the error determination reference data SE output from the reference data storage 43.
Alternatively, one of the reciprocal error determination reference data -SE output from the first reciprocal generator 41a is selected and output. For example, when the selection control signal is "0" (when the coordinate data _Xr is positive), the error determination reference data SE is selectively output, while when the selection control signal is "1" (the coordinate data _Xr).
Is negative), the reciprocal error determination reference data -SE is selected and output. By such an operation, steps S18 and S1 are performed.
9 is performed.

The first adder 41d is connected to the error judgment reference data SE or the reciprocal error judgment reference data -SE output from the first data selector 41b and the Sato error generator 42.
Number counter output from the coordinate data / X _r (coordinate data X _r
And data having the same absolute value and the positive and negative signs inverted) and outputs the addition result as LMS error data SLMS. Step S20 is performed by such an operation.

(Sato error generation step) The Sato error generator 42 performs a Sato error generation step consisting of steps S21 to S23. First, the second reciprocal generator 4
2a converts the Sato error reference data SS output from the reference data storage 43 into reciprocal Sato error reference data -SS, which is data having the same absolute value and inverted signs. Next, the second data selector 42c
Most significant bit is the sign bit of the coordinate data X _r (MS
Using B) as a selection control signal, one of the Sato error reference data SS output from the reference data storage 43 and the reciprocal Sato error reference data -SS output from the second reciprocal generator 42a is selectively output. I do. For example, when the selection control signal is “0” (when the coordinate data _Xr is positive)
When the SATO error reference data SS is selectively output and the selection control signal is "1" (when the coordinate data _Xr is negative)
Selectively output the reciprocal Sato error reference data -SS. Steps S21 and S22 are performed by such an operation.

The third reciprocal generator 42b generates and outputs reciprocal coordinate data -X _r (data having the same absolute value as the coordinate data X _r but with the positive and negative signs inverted). Second adder 42
d is Sato and error reference data SS or counter number Sato error criterion data -SS and inverse number coordinate data -X _r output from the third inverse number generator 42b selectively output from the second data selector 42c Are added to obtain Sato error data (step S23), and the MSB indicating the sign of the obtained Sato error data is output as a Sato error code signal SATO.

(Error Data Selection Step) The error data selection section 44 performs an error data selection step including steps S24 to S26.

The exclusive OR circuit 44b is configured to output the LMS error data SLMS output from the LMS error generator 41.
MSB and SATO error signal SATO
And the exclusive-OR signal is output to the third
As a selection control signal to the data selector 44a.
The third data selector 44a selects and outputs one of the LMS error data SLMS and the constant data "0" as the error data EI according to the selection control signal output from the exclusive OR circuit 44b. When the selection control signal is “0” (when the sign of the LMS error data and the sign of the SATO error data are the same), the LMS error data S
While the LMS is output as error data EI, when the selection control signal is "1" (when the sign of the LMS error data is different from the sign of the Sato error data), the constant data is "0".
Is output as error data EI. Steps S24 to S26 are performed by such an operation.

[0136] Also the case of obtaining the error data EQ from the coordinate data X _i, can be carried out by the same operation.

As described above, according to this modification, only the absolute values of the coordinate determination reference data, the error determination reference data, and the Sato error reference data corresponding to each multi-value mode are stored, and all the coordinate values are stored. Since error data can be calculated from the data, the storage capacity required by the error data generator can be significantly reduced, and a waveform equalization coefficient generator can be realized with a small-scale circuit configuration.

Note that the error data generators 10A and 10B according to the first and second modifications are used not only in place of the error data generator 10 shown in the first and second embodiments, but also, for example, The update of the equalization coefficient may be used together with a complex operation unit in which the real part data and the imaginary part data are updated by different operation units.

[0139]

As described above, according to the present invention, the storage capacity of the storage means for storing error data can be significantly reduced. Also, the real part data and the imaginary part data of the equalization coefficient can be updated by a common computing unit. Further, when the error data has a predetermined value, the calculation for updating the equalization coefficient can be stopped. Therefore, it is possible to realize a waveform equalization coefficient generation device having a small circuit size and low power consumption.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating a configuration of a signal demodulation LSI including a waveform equalizer.

FIG. 2 is a diagram illustrating a position of a waveform equalization coefficient generation device according to an embodiment of the present invention, and is a block diagram illustrating a configuration of a digital equalizer.

FIG. 3 is a diagram for explaining an LMS error and a Sato error in QAM.

FIG. 4 is a block diagram illustrating a configuration of a waveform equalization coefficient generation device according to the first embodiment of the present invention.

FIG. 5 is a phase diagram showing signal point positions, slice levels, and reference points of Sato error in 64QAM.

6 is a diagram illustrating a positive direction portion of the I-axis data of the phase diagram illustrated in FIG. 4, and is a diagram illustrating a range where a flag _fr is 0.

7 is a flowchart showing an operation of the waveform equalization coefficient generation device shown in FIG.

FIG. 8 is a block diagram illustrating a configuration of a waveform equalization coefficient generation device according to a second embodiment of the present invention.

FIG. 9 is a block diagram illustrating a configuration of an error data generation unit according to a first modification of the present invention.

FIG. 10 is a block diagram showing an internal configuration of an error evaluation data storage in the error data generator shown in FIG.

11 is a flowchart showing the operation of the error data generation unit shown in FIG.

FIG. 12 is a phase diagram showing signal point positions, slice levels, and reference points for Sato error in 64QAM.

FIG. 13 is a diagram illustrating an error-free area.

FIG. 14 is a block diagram illustrating a configuration of an error data generation unit according to a second modification of the present invention.

FIG. 15 is a flowchart showing an operation of the error data generation unit shown in FIG.

FIG. 16 is a diagram illustrating a model of a transmission path and its equalization.

FIG. 17 is a diagram illustrating an example of a configuration of a digital filter.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 10 Error data generation part 11 Coordinate determiner 12 ROM 13 Data selector 14 Data brancher 20 Complex operation part 21 Data brancher 22 Data selection means 22a First data selector 22b Second data selector 22c Third data Selector 23 reciprocal data generator 24 sum of products means 24a multiplier 24A multiplier 24b data branching unit 24c first data holder 24d second data holder 24e reciprocal data generator 24f adder 30 stop data signal 31 Data selection control signal 41 LMS error generator 42 Sato error generator 43 Reference data storage 44 Error data selector SS Sato error reference data SC Coordinate determination reference data SE Error determination reference data Xr, Xi coordinate data SLMS LMS error data SATO Sato Error data Sign bit SL mode selection signal EI, EQ error data 51 coordinate data selection unit 52 the error evaluation data storage unit 53 the error data selection portion S1 highlight signal S2 error-free signal DE error evaluation data

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] April 7, 1997

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name]

[Correction method] Change

[Correction contents]

(Second Modification) FIG. 14 is a block diagram showing the configuration of a second modification of the error data generator 10B in the waveform equalization coefficient generator according to the first and second embodiments of the present invention. It is. As shown in FIG. 14, the error data generation unit 10B according to the present modification includes the LMS error generator 4
1, a SATO error generator 42, a reference data storage 43, and an error data selector 44. Error data generation unit 10B further (real axis component of the coordinates data) the real part data of the input signal X _r, (imaginary axis component of the coordinate data) imaginary part data of the input signal buffer that outputs the X _i once to sequentially hold 45 and the obtained error data EI
And a buffer 46 for temporarily holding the EQ.

[Procedure amendment 2]

[Document name to be amended] Drawing

[Correction target item name] FIG.

[Correction method] Change

[Correction contents]

FIG. 14

[Procedure amendment 3]

[Document name to be amended] Drawing

[Correction target item name] FIG.

[Correction method] Change

[Correction contents]

FIG.

Continued on the front page (51) Int.Cl. ⁶ Identification code FI H04L 27/38 H04L 27/00 G

Claims (27)

[Claims] 1. A digital equalizer for equalizing a waveform of a signal transmitted after being subjected to multi-level QAM modulation to a waveform of a signal before transmission, wherein an equalization coefficient of each tap of a digital filter for equalizing the signal waveform is generated. A waveform equalization coefficient generation device that performs updating of equalization coefficients into real part data and imaginary part data, and appropriately selects data required for the calculation for updating the real part data and the imaginary part data. A waveform equalization coefficient generation apparatus characterized in that the waveform equalization coefficient generation apparatus is performed by a common arithmetic unit. 2. The waveform equalization coefficient generation device according to claim 1, wherein: an error data generation unit that generates error data necessary for updating an equalization coefficient for the input signal; A complex operation unit that updates the equalization coefficient into real part data and imaginary part data based on the error data output from the error data generation unit, and wherein the error data generation unit includes the error data Is divided into I-axis direction error data and Q-axis direction error data, and the complex operation unit performs an operation for updating the real part data and the imaginary part data of the equalization coefficient by the input operation. By appropriately selecting either one of the real part data and the imaginary part data of the signal, and appropriately selecting one of the I-axis direction error data and the Q-axis direction error data. By-option, waveform equalizing coefficient generating device, characterized in that is performed by a common calculator. 3. The waveform equalization coefficient generation device according to claim 2, wherein the complex operation unit includes one of a real part data and an imaginary part data of the equalization coefficient before update, and a real number of the input signal. One of the partial data and the imaginary part data, and one of the I-axis direction error data and the Q-axis direction error data output from the error data generator in accordance with an instruction of a given data selection control signal. Multiplying the real part data or the imaginary part data of the input signal output from the data selecting means by the I-axis direction error data or the Q-axis direction error data; A product for updating the real part data or the imaginary part data of the equalization coefficient before update output from the data selection means by adding the data obtained based on the data. A waveform equalization coefficient generation device, comprising: a summing means. 4. The waveform equalization coefficient generation device according to claim 3, wherein the data selection unit receives real part data and imaginary part data of an equalization coefficient before updating and receives the data in accordance with the data selection control signal. A first data selector for selecting and outputting one of the real part data and the imaginary part data of the equalization coefficient before update; and inputting the real part data and the imaginary part data of the input signal and selecting the data A second data selector for selecting and outputting one of the real part data and the imaginary part data of the input signal according to a control signal; and inputting the I-axis direction error data and the Q-axis direction error data, A third data selector for selecting and outputting one of the I-axis direction error data and the Q-axis direction error data in accordance with a data selection control signal. Waveform equalizing coefficient generating device to. 5. The waveform equalization coefficient generation device according to claim 3, wherein the product-sum unit includes step-number reciprocal data indicating a degree of update of the equalization coefficient, and the input output from the data selection unit. Part data or imaginary part data of the output signal and the I-axis direction error data or Q output from the data selection means.
A multiplier that multiplies the axial error data and outputs a multiplication result; first and second data holders that hold the multiplication result data of the multiplier; According to a selection control signal, a data branching unit input to one of the first and second data holding units, and data output from the second data holding unit, according to the data selection control signal, A reciprocal data generator that converts the data into reciprocal data and outputs the data; real part data or imaginary part data of the equalization coefficient before update output from the data selection unit; and output from the first data holder. Waveform adder for adding the data output from the reciprocal data generator and the data output from the reciprocal data generator, and outputting an addition result as output data of the product-sum means. Number generator. 6. The waveform equalization coefficient generation device according to claim 2, wherein the error data generation unit generates error data necessary for updating the equalization coefficient for the input signal using STOP & G.
A waveform equalization coefficient generation device, which is generated based on a STOP & GO algorithm which is an LMS (Least Mean Square) algorithm using an O function. 7. The waveform equalization coefficient generating device according to claim 6, wherein the error data generating unit obtains coordinate data indicating a position occupied by the input signal in a phase diagram of the multilevel QAM modulation, and obtains the obtained coordinate. According to the data, a coordinate determination unit that outputs one of address data corresponding to the coordinate data and a stop data signal instructing the stop of output of error data, and stores error data corresponding to each coordinate data,
A storage unit that outputs error data stored at an address specified by the address data output from the coordinate determiner, and outputs a predetermined value when a stop data signal is output from the coordinate determiner, while outputting the predetermined value. A waveform selector for outputting error data output from the storage means when the stop data signal is not output from the determiner. 8. The waveform equalization coefficient generation device according to claim 6, wherein when a stop data signal is output from the coordinate determiner, the operation for updating the equalization coefficient is stopped in the complex operation unit. An apparatus for generating a waveform equalization coefficient, characterized in that: 9. The waveform equalization coefficient generation device according to claim 6, wherein the error data generation unit determines whether the coordinate data indicating the position occupied by the input signal in the phase diagram of the multilevel QAM modulation is positive or negative. When the coordinate data is positive, the coordinate data is output as it is as the determination coordinate data. On the other hand, when the coordinate data is negative, the coordinate data is bit-inverted and output as the determination coordinate data. A coordinate data selection unit that outputs an inverted display signal indicating whether the coordinate data for determination is bit-inverted from the coordinate data, and stores a plurality of error evaluation data respectively corresponding to the positive coordinate data. The error evaluation data corresponding to the determination coordinate data output from the coordinate data selection unit is stored in a plurality of stored error evaluation data. An error evaluation data storage unit selectively outputting the error evaluation data storage unit when the inverted display signal output from the coordinate data selection unit indicates that the determination coordinate data is not bit-inverted of the coordinate data. While the error evaluation data output from the device is output as it is as error data, when the inverted display signal indicates that the determination coordinate data is a bit-inverted version of the coordinate data, the error evaluation data is output as a bit. A waveform equalization coefficient generation device, comprising: an error data selection unit that outputs the inverted data as error data. 10. The waveform equalization coefficient generating device according to claim 6, wherein the error data generating unit includes: a Sato error reference data indicating a coordinate of a Sato error reference point in a multi-level QAM modulation phase diagram; A reference data storage for storing coordinate determination reference data indicating a slice level value in the phase diagram of the above, and error determination reference data indicating coordinates of a signal reference point in the phase diagram of the multi-level QAM modulation; Based on the coordinate data representing the position occupied in the phase diagram of the multilevel QAM modulation and the coordinate determination reference data output from the reference data storage, the input signal is converted into each slice level in the phase diagram of the multilevel QAM modulation. And the reference indicating the coordinates of the signal reference point in the determined area. An LM for generating LMS error data for the input signal based on the error determination reference data and the coordinate data output from the data storage;
An S error generator; a Sato error generator that generates Sato error data for the input signal based on the Sato error reference data and the coordinate data output from the reference data storage; Input the LMS error data generated by the transmitter and the sign bit of the Sato error data generated by the Sato error generator, and when the sign bit of the LMS error data is equal to the sign bit of the Sato error data, A waveform equalization coefficient generation device, comprising: an error data selection unit that selectively outputs the LMS error data as error data and, when different, outputs data of a predetermined value as error data. 11. A digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, wherein an equalization coefficient of each tap of a digital filter for equalizing the signal waveform is generated. An LMS (Leas (Leas) function using a STOP & GO function for an input signal to generate error data necessary for updating an equalization coefficient.
t Mean Square) An error data generation unit that generates based on a STOP & GO algorithm, and a complex operation unit that updates an equalization coefficient based on the input signal and error data output from the error data generation unit. The error data generation unit obtains coordinate data indicating a position occupied by the input signal in the phase diagram of the multilevel QAM modulation, and according to the obtained coordinate data, a stop data signal for instructing to stop outputting the error data; It stores a coordinate determiner that outputs one of address data corresponding to coordinate data, and error data corresponding to each coordinate data.
A storage unit that outputs error data stored at an address specified by the address data output from the coordinate determiner, and outputs a predetermined value when a stop data signal is output from the coordinate determiner, while outputting the predetermined value. A waveform selector for outputting error data output from the storage means when the stop data signal is not output from the determiner. 12. The waveform equalization coefficient generation device according to claim 11, wherein the error data generation unit receives the error data output from the data selector as input, and sets the input error data as I-axis direction error data. Q
A waveform equalization coefficient generation device, comprising: a data splitter that outputs data separately from axial error data. 13. The waveform equalization coefficient generating device according to claim 11, wherein the coordinate determination unit determines an error-free area in which error data is a predetermined value as a multi-valued Q.
It is set in advance on the phase diagram of the AM modulation, and outputs the stop data signal when the position indicated by the obtained coordinate data is within the error-free area, and outputs the stop data signal when the position is not within the error-free area. A waveform equalization coefficient generation device for outputting corresponding address data. 14. The waveform equalization coefficient generating device according to claim 13, wherein the coordinate determination unit performs multi-level QA according to an instruction of a given communication mode selection signal.
A waveform equalization coefficient generation device having a function of selecting an M modulation mode and setting an error-free area according to the selected mode. 15. The waveform equalization coefficient generating apparatus according to claim 11, wherein said storage means stores only error data for coordinate data outside an error-free area. apparatus. 16. The waveform equalization coefficient generation device according to claim 11, wherein when a stop data signal is output from the coordinate determiner, the operation for updating the equalization coefficient is stopped in the complex operation unit. An apparatus for generating a waveform equalization coefficient, characterized in that: 17. A digital equalizer for equalizing a waveform of a multi-level QAM-modulated transmitted signal to a waveform of a signal before transmission, wherein an equalization coefficient of each tap of a digital filter for equalizing the signal waveform is generated. An LMS (Leas (Leas) function using a STOP & GO function for an input signal to generate error data necessary for updating an equalization coefficient.
t Mean Square) An error data generation unit that generates based on a STOP & GO algorithm, and a complex operation unit that updates an equalization coefficient based on the input signal and error data output from the error data generation unit. The error data generator determines whether the coordinate data indicating the position occupied by the input signal in the phase diagram of the multilevel QAM modulation is positive or negative, and when the coordinate data is positive, the coordinate data is determined as it is On the other hand, when the coordinate data is negative, the coordinate data is output as determination coordinate data after bit inversion of the coordinate data, and the determination coordinate data is obtained by bit inversion of the coordinate data. A coordinate data selector for outputting an inverted display signal indicating whether or not there is an Error evaluation data corresponding to the determination coordinate data output from the coordinate data selecting unit, and selectively outputting the error evaluation data from the stored plurality of error evaluation data. A data storage; and an error output from the error evaluation data storage when the inverted display signal output from the coordinate data selection unit indicates that the coordinate data for determination is not bit-inverted of the coordinate data. While the evaluation data is output as it is as error data, when the inverted display signal indicates that the determination coordinate data is a bit-inverted version of the coordinate data, the error evaluation data is bit-inverted and the error data is output. And an error data selection unit that outputs the waveform equalization coefficient. 18. The waveform equalization coefficient generation device according to claim 17, wherein the error evaluation data storage is configured such that an error in which a position indicated by the determination coordinate data in the phase diagram of the multi-valued QAM indicates that the error data is a predetermined value. It is determined whether or not the position is within the free area, and outputs an error-free signal indicating whether or not the position indicated by the determination coordinate data is within the error-free area. When the error-free signal output from the evaluation data storage indicates that the position indicated by the coordinate data for determination is within the error-free area, data of a predetermined value is output as error data. A waveform equalization coefficient generator. 19. A digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, wherein an equalization coefficient of each tap of a digital filter for equalizing the signal waveform is generated. An LMS (Leas (Leas) function using a STOP & GO function for an input signal to generate error data necessary for updating an equalization coefficient.
t Mean Square) An error data generation unit that generates based on a STOP & GO algorithm, and a complex operation unit that updates an equalization coefficient based on the input signal and error data output from the error data generation unit. The error data generation unit includes: Sato error reference data indicating coordinates of a Sato error reference point in a multi-level QAM modulation phase diagram; coordinate determination reference data indicating a slice level value in a multi-level QAM modulation phase diagram; A reference data storage for storing error determination reference data indicating coordinates of a signal reference point in a phase diagram of the multi-level QAM modulation; and coordinate data indicating a position of the input signal in the phase diagram of the multi-level QAM modulation; Based on the coordinate determination reference data output from the reference data storage, the input signal is In the phase diagram of the value QAM modulation, it is determined which of the areas divided by each slice level belongs, and the error determination reference data output from the reference data storage indicating the coordinates of the signal reference point in the determined area; LM for generating LMS error data for the input signal based on the coordinate data
An S error generator; a Sato error generator that generates Sato error data for the input signal based on the Sato error reference data and the coordinate data output from the reference data storage; Input the LMS error data generated by the transmitter and the sign bit of the Sato error data generated by the Sato error generator, and when the sign bit of the LMS error data is equal to the sign bit of the Sato error data, A waveform equalization coefficient generation device, comprising: an error data selection unit that selectively outputs the LMS error data as error data and, when different, outputs data of a predetermined value as error data. 20. The waveform equalization coefficient generator according to claim 19, wherein the reference data storage stores absolute values of the SATO error reference data, coordinate determination reference data, and error determination reference data, When the coordinate data is negative, the LMS error generator generates LMS error data based on the sign determination reference data and the error determination reference data obtained by inverting the sign. A waveform generator for generating the Sato error data based on the Sato error reference data obtained by inverting the sign when the coordinate data is negative. 21. The waveform equalization coefficient generation device according to claim 19, wherein the reference data storage includes: a Sato error reference data, a coordinate determination reference data, and an error determination reference respectively corresponding to a plurality of multilevel modes of QAM. Data is stored, and Sato error reference data, coordinate determination reference data, and error determination reference data corresponding to the multi-value mode of QAM specified by a mode selection signal given from outside are stored in the stored data. A waveform equalization coefficient generation device for selectively outputting a waveform equalization coefficient. 22. A digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, wherein an equalization coefficient of each tap of a digital filter for equalizing the signal waveform is generated. A method for generating a waveform equalization coefficient that performs error data necessary for updating an equalization coefficient of an input signal using an LMS (LeS
AST Mean Square) STOP & GO
An error data generation step of generating based on an algorithm; and a complex operation step of updating an equalization coefficient using the input signal and error data output from the error data generation unit. Is to set an error-free area where the error data is a predetermined value on the phase diagram of the multilevel QAM modulation, and prepare error data corresponding to each coordinate data outside the error-free area. A first step of obtaining coordinate data indicating a position occupied by the obtained signal in the phase diagram of the multi-level QAM modulation; and when the position indicated by the coordinate data obtained in the first step is within the error-free area, And a second step of setting the predetermined value as a predetermined value, and when the position indicated by the coordinate data obtained in the first step is not within the error-free area, The third step and the waveform equalizing coefficient generation method characterized by comprising a selecting error data corresponding to the coordinate data from the error data being Bei. 23. A digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, wherein an equalization coefficient of each tap of a digital filter for equalizing the signal waveform is generated. A method for generating a waveform equalization coefficient that performs error data necessary for updating an equalization coefficient of an input signal using an LMS (LeS
AST Mean Square) STOP & GO
An error data generation step of generating based on an algorithm; anda complex operation step of updating an equalization coefficient using the input signal and error data output from the error data generation unit. Determining whether the given signal occupies the coordinate data indicating the position occupied in the phase diagram of the multi-level QAM modulation, and when the coordinate data is positive, selects the coordinate data as it is as the determination coordinate data; When the coordinate data is negative, the coordinate data is bit-inverted, and the coordinate data selecting step of selecting the coordinate data as the determination coordinate data; andthe error evaluation data corresponding to the determination coordinate data selected in the coordinate data selecting step is executed. An error evaluation data setting step to be set; and the error evaluation data set in the error evaluation data setting step. When the coordinate data is directly selected as determination coordinate data in the coordinate data selecting step, the coordinate data is directly selected as error data, while the coordinate data is bit-inverted in the coordinate data selecting step, and then the determination coordinate data is used. An error data selecting step of inverting the bit and selecting the error data after the bit is inverted. 24. The method for generating a waveform equalization coefficient according to claim 23, wherein in the error evaluation data setting step, the position indicated by the determination coordinate data in the phase diagram of the multilevel QAM is a predetermined value of the error data. It is provided with a step of determining whether or not it is in an error-free area, the error data selecting step has determined in the error evaluation data setting step that the position indicated by the determination coordinate data is in the error-free area And a step of selecting data of a predetermined value as error data. 25. A digital equalizer for equalizing a waveform of a signal transmitted by multi-level QAM modulation to a waveform of a signal before transmission, wherein an equalization coefficient of each tap of a digital filter for equalizing the signal waveform is generated. A method for generating a waveform equalization coefficient that performs error data necessary for updating an equalization coefficient of an input signal using an LMS (LeS
AST Mean Square) STOP & GO
An error data generation step of generating based on an algorithm; and a complex operation step of updating an equalization coefficient using the input signal and error data output from the error data generation unit. Are the Sato error reference data indicating the coordinates of the Sato error reference point in the multi-level QAM modulation phase diagram, the coordinate determination reference data indicating the slice level value in the multi-level QAM modulation phase diagram, and the multi-level QAM modulation phase diagram A reference data setting step of setting error determination reference data indicating the coordinates of the signal reference point in, using the coordinate determination data and the coordinate determination reference data indicating the position that the given signal occupies in the phase diagram of the multilevel QAM modulation. Where the given signal is divided by each slice level in the phase diagram of the multilevel QAM modulation. And an LMS error generating step of generating LMS error data for the given signal using the error determination reference data indicating the coordinates of the signal reference point in the determined area; and Generating a Sato error data for the given signal using the data and the Sato error reference data; and comparing the sign of the LMS error data with the sign of the Sato error data. An error data selecting step of selecting the LMS error data as error data and selecting data of a predetermined value as error data when the sign is different. 26. The waveform equalization coefficient generating method according to claim 25, wherein the reference data setting step sets absolute values of the SATO error reference data, coordinate determination reference data, and error determination reference data, The LMS error generating step includes, when the coordinate data is negative, generating an LMS error based on inversion of the positive and negative of the coordinate determination reference data and the error determination reference data. In the method, when the coordinate data is negative, the Sato error data is generated based on the polarity of the Sato error reference data inverted. 27. The waveform equalization coefficient generation method according to claim 26, wherein the reference data setting step is performed in accordance with a designated multilevel mode of QAM, the SATO error reference data, the coordinate determination reference data, and the error determination. A method for generating waveform equalization coefficients, wherein reference data is set.