JPH03119838A - Adaptive automatic equalizer with frequency offset elimination function - Google Patents

Abstract

PURPOSE:To automatically detect the occurrence of frequency offset in an analog signal in a moment and to perform equalization even when it occurs by estimating the frequency offset in a training mode period by calculating the mean of the phase error of the frequency offset in the training mode period. CONSTITUTION:The received analog signal r(t) of a carrier frequency band area is supplied to a multiplier 12 simultaneously to an intensity detecting part 32, and two signals, the inverse of ak and sk supplied to an error signal detecting part 19 are supplied to a phase difference detecting means 31. The output of the intensity detecting part 32 and that of the phase difference detecting part 31 are multiplied at a multiplier 35, and a mean value computing means 34 calculates the mean value of the output of the multiplier 35 from a time when the number of times tap coefficients of almost double the number of taps of a filter included in an adaptive automatic equalization part 15 is updated from a time t1 when the reception of a training signal 1 is started, and generates a DC voltage proportional to the mean value i.e., the DC voltage proportional to the frequency offset.

Description

[Detailed Description of the Invention] [Objective of the Invention] (Industrial Application Field) The present invention relates to an adaptive automatic equalizer, and more specifically, it relates to an adaptive automatic equalizer, and more specifically, it relates to an adaptive automatic equalizer, and more specifically, it relates to an adaptive automatic equalizer, and more specifically, to an adaptive automatic equalizer that is suitable for applications where transmission path characteristics change quickly over time and frequency offsets occur frequently. The present invention relates to an adaptive automatic equalizer that performs sufficient adaptive automatic equalization of a transmission path and has a frequency offset removal function.

(Prior Art) As is well known, an automatic equalizer uses a training signal to automatically equalize the transmission path characteristics in order to equalize distortion in the transmission path before transmitting an information signal to the transmission path. It is a filter that converts The adaptive automatic equalizer continues to control the filter characteristics even after the transmission path reaches the equalized state in this manner.

The signal received by the adaptive automatic equalizer consists of a training signal 1 and a data signal 2, as shown in FIG.
2) is called the training mode period, and the period (t2 to t3) during which the data signal 2 is received is the judgment feedback mode (Decision Feedback Mode).
e) It is called a period. Training signal 1 is known at the receiver or receiver, whereas data signal 2 is unknown at the receiver or receiver.

On the other hand, if a receiver equipped with an adaptive automatic equalizer is moving, such as a car or mobile phone,
Frequency offsets tend to occur frequently due to the so-called Doppler effect or multiwave propagation, and adaptive automation of the transmission path is not possible, so conventionally, adaptive automatic equalizers have been provided with a frequency offset removal function.

As shown in FIG. 7, a conventional adaptive automatic equalizer having a frequency offset removal function applies a received carrier frequency band analog signal r(t) to a multiplier 12 using a phase-lock loop (hereinafter referred to as "Phase Lock Loop"). ,
(simply referred to as rPLLJ) 13 and converted into an intermediate frequency band or base band signal, sampled by the first switch 14 at an appropriate time interval such as the transmission symbol interval, and transmitted as a digital signal rk. The signal is supplied to an adaptive automatic equalization unit 15 including an adaptive filter such as a versal filter or a lattice filter.

The adaptive automatic equalizer 15 performs waveform equalization on the human input signal r by appropriate signal processing, and then outputs the signal as r. The waveform-equalized signal is subjected to a determination in the determining section 16, and is output as a quantized digital signal a.

The training signal generator 17 generates a training signal I,
generate.

The switching means 18 as a second switch is a training switch.
In the mode period, it is switched to be connected to the output side of the training signal generator 17, and the training signal I8 outputted by the training signal generator 17 is used as the output signal sk, and in the judgment feedback mode period, it is switched to be connected to the determination section 16, and the The signal a outputted by 16 is assumed to be the output signal S&.

The output S of the adaptive automatic equalization section 15 and the output S of the switching means 18 are supplied to one error signal detection section, and the error signal detection section 19 switches to the output of the adaptive automatic equalization section 15. By determining the difference between the outputs S, the means 18 generates an error signal e, 1, and outputs it to the tap coefficient control section 20.

The tap coefficient control unit 20 uses the error signal e detected by the error signal detector 19 to generate a tap coefficient control signal for sequentially updating the tap coefficients of the filter included in the adaptive automatic equalization unit 15. , and based on this, the tap coefficients of the filter included in the adaptive automatic equalizer 15 are successively corrected at the sampling timing of the first switch 14.

As the tap coefficient updating algorithm of the tap coefficient control unit 20, for example, a gradient algorithm, a Kalman algorithm, etc.
Algorithms such as the fast Kalman ◆ algorithm and square number root Kalman algorithm are known.

Furthermore, the filter included in the adaptive automatic equalization unit 15 is of the so-called decision feedback type (Decision feedback type).
ack), the output of the switching means 18 is routed to the path 21.
is fed back to the filter included in the adaptive automatic equalization unit 15.

Further, the output Sk of the switching means 18 and the output of the adaptive automatic equalizer 15 are supplied to the PLL.

The PLL 13 includes a phase difference detection section 31o, a low-pass filter (hereinafter abbreviated as "LPFJ") 32, and a voltage controlled oscillator (V
oltage Control O5clat
or (hereinafter referred to as rVCOJ)) 33, the phase difference detector 31 detects the phase difference between the output signal SK of the switching means 18 and the output signal section of the adaptive automatic equalization section 15, and the low-pass filter 32 The phase difference detector 31 generates a DC voltage proportional to the detected phase difference, and the voltage controlled oscillator 33 adjusts its output oscillation frequency using this DC voltage.

(Problem to be Solved by the Invention) However, the above-mentioned conventional adaptive automatic equalizer equipped with a frequency offset removal function has a slow tracking characteristic of the PLL, so it does not depend on the transmission line characteristics like a wired line. In a line with slow time fluctuations, the PLL can be tracked sufficiently, but in cases where the transmission path characteristics, including frequency offset, change from moment to moment, a sufficiently long training period is required for the PLL to follow the fluctuations in the transmission path characteristics. A signal is required. This makes it difficult to make the training signal long enough, such as in burst-like data communications, or when a receiver equipped with an adaptive automatic equalizer moves, especially in car or mobile telephones. In other words, when frequency offsets occur frequently due to the so-called Doppler effect or multiple wave propagation, there is a problem in that adaptive automatic equalization of the transmission path cannot be performed sufficiently.

Therefore, in view of the above-mentioned actual circumstances of the adaptive automatic equalizer, the present invention automatically calculates the Moreover, it detects this instantly,
The present invention attempts to provide an adaptive automatic equalizer having a frequency offset removal function that can perform equalization.

[Configuration of the Invention] (Means for Solving the Problems) In order to achieve the above-mentioned problems, one of the basic configurations of an adaptive automatic equalizer having a frequency offset removal function according to the present invention is shown in FIG. As shown, an adaptive automatic equalizer 15,
a tap coefficient control unit 20 that sequentially determines the tap coefficients of the adaptive filter of the adaptive automatic equalization unit 15; and an adaptive automatic equalization unit 15;
Means for detecting and quantizing signals output from 1.6 o
and an error signal detection section 19 that performs subtraction processing on the signals on the input side and output side of the quantization means 16o and inputs the difference signal to the tap coefficient section 20, in which: a phase difference detection means 31 for detecting a reproduced carrier phase difference from the two human signals of the error signal detection section 19; 1-step average calculation means that determines a coefficient, weights the phase error signal output from the error signal detection means 31 using the determined weighting coefficient, and calculates the average value of the phase of the weighted error signal; 34, and frequency conversion means 33 for adjusting the oscillation frequency according to the average value of the phase difference signal outputted from the average value calculation means 34.
It is characterized by providing o.

Another basic configuration of the adaptive automatic equalizer having a frequency offset removal function according to the present invention is as shown in FIG. a tap coefficient control unit 20 that sequentially determines the tap coefficients of the filter; a means 16o that detects and quantizes the signal output from the adaptive automatic equalization unit 15;
An adaptive automatic equalizer comprising an error signal detection section 19 that performs a subtraction process on the input side and output side signals of the converting means 16o and inputs the difference signal to the tap coefficient control section as the tap coefficient correction information. , a tap angular velocity that extracts only information related to a predetermined tap from a series of tap coefficient correction information output from the tap coefficient control unit 20 each time a tap coefficient is corrected, and detects the amount of angular change of the predetermined tap; a detection section 42; and a frequency conversion means 33o that adjusts the frequency of a local oscillator required when converting an analog signal from high frequency to baseband according to the output of the tap angular velocity detection section 42.
It is characterized by having the following.

(Operation) With the above configuration, the frequency offset can be estimated in the training mode period by calculating the average of the phase errors in the first adaptive automatic equalizer mode period. In addition, when the signal strength of the received signal is small, the reliability of the detected phase error is low, and when the strength of the received signal is high, the reliability of the phase error is high. By weighting, it is possible to obtain a highly reliable average value of phase information, that is, information about the frequency offset, which can be used to control the frequency conversion means and compensate for transmission path distortion even when the training signal is short. I can do it.

Also, in the case of the second adaptive automatic equalizer according to the present invention, if the tap coefficient control signal is updated, the average angular velocity of the predetermined tap coefficients that are sequentially updated in the training mode can be calculated. , the frequency offset can be estimated during the training mode, and by controlling the frequency conversion means, it is possible to compensate for transmission path distortion even when the training signal is short.

(Example) Next, a practical example of an adaptive automatic equalizer having a frequency offset removal function according to the present invention will be described based on the drawings.

FIG. 1 is a schematic block diagram of an embodiment of an adaptive automatic equalizer having a frequency offset removal function according to the present invention. However, since this embodiment has the same configuration as the first basic configuration of the present invention, it will be described based on the same drawings. The received signal received by the present adaptive automatic equalizer is composed of a training signal 1 and a data signal 2, as shown in FIG. Then, the period (t
1 to t2) is called the training mode period, while the period during which data signal 2 is being received (t2 to 13) is called the decision feedback mode period. The training signal 1 is known at the receiver or at the receiver;
Data signal 2 is unknown to the receiver or receiver.

In FIG. 1, the received FNA log signal r(t) in the carrier frequency band is multiplied by the output of the V C033o as a frequency converting means in the multiplier 12, and is converted into a signal in the intermediate frequency band or base band. The signal is sampled by the switch 14 at an appropriate time interval such as the transmission symbol interval, and is supplied as a discrete time signal rk to an adaptive automatic equalization unit 15 including an adaptive filter such as a transversal filter or a lattice filter. In the adaptive automatic equalization section 15, the input signal rk is subjected to waveform equalization through appropriate signal processing, and then outputted as a digital signal head. The quantization means 16 performs a determination on the waveform equalized signal redundancy and outputs it as a signal ah. Training signal generator 17 generates training signal IK. During the training mode period, the switching means 18 is connected to the training signal generator 17, and outputs the training signal (1) outputted from the training signal generator 17 as the output signal S, while during the judgment feedback moto period, the switching means 18 outputs the training signal S. 160, and the signal a output from the quantization means 160 is referred to as an output signal Sk. The output of the adaptive automatic equalization section 15 and the output S of the switching means 18 are supplied to an error signal detection section 19, which outputs the output of the adaptive automatic equalization section 15 and the output S of the switching means 1 -8 outputs S, to generate an error signal ek, and the tap coefficient control section 20 generates the error signal e detected by the error signal detection section 19.
, to generate a tap coefficient control signal for sequentially updating the tap coefficients of the filter included in the adaptive automatic equalization unit 15, and based on this, at the sampling timing of the first switches 1 and 4, The tap coefficients of the filter included in the adaptive automatic equalization unit 15 are successively modified. Here, when the tap coefficient control unit 20 starts receiving the training signal 1, the tap coefficient control unit 20 calculates the training signal 1 from an appropriate initial value by a cyclic least squares (RLS) method such as a Kalman algorithm, a fast Kalman algorithm, or a square root Kalman algorithm. , sequentially updates the tap coefficient control signal.

Furthermore, if the filter included in the adaptive automatic equalization section 15 is a so-called decision feedback type, the output of the switching means 18 is
The signal is fed back to a filter included in the adaptive automatic equalizer 15 via a path 21.

On the other hand, the received carrier frequency band analog signal r(t)
is supplied to the intensity detector 32 at the same time as the multiplier 1-2. The intensity detection unit 32 detects the received analog signal “(
t) is detected and output.

Further, the two signals sK and sK supplied to the error signal detection section 19 are supplied to the phase difference detection means 31. The phase difference detection means 31- extracts phase difference information between the two signals η and SK and outputs it. The output of the intensity detection section 32 and the output of the phase difference detection section 31 are multiplied by a multiplier 35, and the multiplier 35 supplies the result to the average value calculation means 34. The average value calculation means 34 updates the multiplier 35 from the time t when it started receiving the training signal 1 to the time when the same number of tap coefficients, which is approximately twice the number of taps of the filter included in the adaptive automatic equalization section 15, is updated. calculates the average value of the output of and generates a DC voltage proportional to this, that is, a DC voltage proportional to the frequency offset. Third switch 36
is connected to the switching means 18, and is released during the training mode period when training signal 1 is received, but is short-circuited during the decision feedback mode period when data signal 2 is received, and the frequency conversion Means 33. .

A DC voltage which is the output of the average value calculation means 34 is supplied to the average value calculation means 34. Using this DC voltage, the frequency conversion means 33 finely adjusts the oscillation frequency that is its output.

FIG. 3 shows another embodiment of the adaptive automatic equalizer having a frequency offset removal function according to the present invention.

In FIG. 3, the received analog signal "(t) in the carrier frequency band is multiplied by the output of VC033 as a frequency conversion means in the multiplier 12, converted into a signal in the intermediate frequency band or base band, and The signal is sampled by the switch 14 at an appropriate time interval such as the transmission symbol interval, and is supplied as a discrete time signal rk to an adaptive automatic equalization unit 15 including an adaptive filter such as a transversal filter or a lattice filter. In the adaptive automatic equalizer 15, the input signal rk is subjected to waveform equalization through appropriate signal processing, and then outputted as a digital signal 20. A decision is made in the digital signal am
is output as The training signal generator 17 generates a training signal (2). Switching means] 8 is a training signal generator 4-1 during the training mode period.
A training signal generator connected to the training signal generator 17 and outputted by the training signal generator 17 is defined as an output signal S, and on the other hand,
In the decision feedback mode period, the signal a, which is connected to the quantization means 16° and outputted by the quantization means 16o, is converted into an output signal Sk.
shall be. The output 20 of the adaptive automatic equalization section 15 and the output S of the switching means 18 are supplied to an error signal detection section 19, and the error signal detection section 19 receives the output 20 of the adaptive automatic equalization section 1-5.
The tap coefficient control unit 2o generates an error signal e by determining the difference between the output S of the switching means]8, and the tap coefficient control unit 2o uses the error signal e detected by the error signal detection unit 19 to perform an adaptive A tap coefficient control signal for sequentially updating the touch coefficient of the filter included in the automatic equalization unit 15 is generated, and based on this, the tap coefficient control signal is transmitted to the adaptive automatic equalization unit 15 at the sampling timing of the first switch 14. Modify the tap coefficients of the included filters one by one. Here, when the tap coefficient control unit 20 starts receiving the training signal 1, from an appropriate initial value,
The tap coefficient control signal is sequentially updated by the cyclic least squares method (RLS method) such as the Kalman φ algorithm, fast Kalman skewer algorithm, square root Kalman algorithm, etc.Furthermore, the filter included in the adaptive automatic equalization unit 15 In the case of a so-called decision feedback type, the output of the switching means 18 is fed back to the filter included in the adaptive automatic equalizer 15 via a path 21.

On the other hand, information regarding a predetermined tap coefficient, such as a center tap coefficient, in the output of the tap coefficient control section 20 is supplied to the tap angular velocity detection section 42 . The tap angular velocity detection unit 42 detects the angular velocity at which a predetermined tap coefficient, which is sequentially updated, rotates on a complex plane, and outputs a DC voltage proportional to the detected angular velocity. The third switch 36 is associated with the switching means 18 and is open during the training mode period when the training signal 1 is being received, but during the decision feedback mode period when the data signal 2 is being received. shorted, VC
033 is supplied with a DC voltage which is the output of the tap angular velocity detection section 42. Using this DC voltage, VC033 finely adjusts the oscillation frequency that is its output.

FIG. 4 shows a third embodiment of the invention.

In FIG. 4, the received analog signal r(t) in the carrier frequency band is multiplied by the output of The signal is sampled at an appropriate time interval such as the transmission symbol interval, and is supplied as a discrete time signal rk to an adaptive automatic equalization unit comprising a transversal filter. The transversal filter of the adaptive automatic equalization unit is a decision feedback type filter with five taps, and includes four delay elements T1, T2 .
T3. T4 and the human power signal X1 from the first switch 14
and delay elements Tl, T2. Output signal x of T3 and T4
2. x 3. x 4. 5 multipliers 52 that multiply and combine X 5 and 5 tap coefficients CII c21 c31 c41 c5, and the outputs C of the 5 multipliers 52,
X1. C2X 2. C,X 3゜c4x4. c, x5
and an adder 53 that adds together and outputs the sum.

As a result, after the waveform equalization of the input signal ``,'' is performed,
It is output as a digital signal 20. The waveform-equalized signal 20 is subjected to a judgment in the judgment section 16 and outputted as a digital signal ak. A training signal generator 17 generates a training signal I.

The switching means 18 is connected to the training signal generator 17 during the training mode period, and uses the training signal IK outputted by the training signal generator 17 as the output signal sk, while during the judgment feedback mode period, the switching means 18 outputs the training signal IK output from the training signal generator 17 as the output signal sk.
The signal a connected to the determination unit 16 and output by the determination unit 16 is referred to as an output signal S. The output 20 of the adaptive automatic equalization section 15 and the output Sk of the switching means 18 are supplied to an error signal detection section 19.
, and the output sk of the switching means 18 to generate an error signal e, and the tap coefficient control section 20 performs adaptive automatic equalization using the error signal e detected by the error signal detection section 19. A tap coefficient control signal for sequentially updating the tap coefficients of the filter included in the adaptive automatic equalization unit 15 is generated, and based on this signal, the tap coefficient control signal of the filter included in the adaptive automatic equalization unit 15 is generated at the sampling timing of the first switch 14. Modify the tap coefficients one by one. Here, when the tap coefficient control unit 20 starts receiving the training signal 1, the tap coefficient control unit 20 performs tap coefficient control from an appropriate initial value by an RLS method such as a Kalman algorithm, a fast Kalman algorithm, or a square root Kalman algorithm. Update the signal sequentially. Furthermore, since the transversal filter is of the so-called decision feedback type, the output Sk of the switching means 18 is fed back to the filter included in the adaptive automatic equalizer 15 via the path 21.

On the other hand, information regarding the center tap coefficient in the output of the tap coefficient control section 20 is supplied to the fifth delay element T5 and the phase difference detection section 62. The phase difference detection section 62 includes a fifth
is stored in delay element T5.
Extract each phase information from the information about the tap coefficient and the information about the center tap coefficient at the current time,
The difference is determined and supplied to the average value calculation means 34. The average value calculating means 34 calculates the training signal 1. The average value of the output of the phase difference detection unit 62 is calculated from the time when the tap coefficients are updated approximately twice as many times as the number of taps of the transversal filter, that is, approximately 10 times, from the time t1 when reception of the signal is started. , generates a DC voltage proportional to this, that is, a DC voltage proportional to the frequency offset. The third switch 36 is connected to the switching means 18 and is open in the training mode when the training signal 1 is being received, but is short-circuited during the decision feedback mode when the data signal is being received. , supplies the DC voltage that is the output of the average value calculation unit 34 to VC033. Using this DC voltage, VC033 finely adjusts the oscillation frequency that is its output.

FIG. 5 shows a fourth embodiment of the invention.

In FIG. 5, the received carrier frequency band analog signal "(t) is multiplied by the output of VC033 in the multiplier 12, converted into an intermediate frequency band or base band signal, and then sent to the first switch]- 4, the automatic equalizer 15 includes an adaptive filter such as a transversal filter or a lattice filter.
supplied to The adaptive automatic equalization unit 15 performs waveform equalization on the human input signal rk through appropriate signal processing, and then outputs it as a signal 20. The waveform-equalized signal is judged by the judgment unit 16, and the digital signal a
It is output as k. Training signal generator 1-7 generates training signal IK. During the training/moto period, the switching means 18 is connected to the training signal generator 17, and outputs the training signal 3 output from the training signal generator 17 as the output signal S.
In the decision feedback mode period, it is connected to the quantization means 16,
The signal a output by the determination unit 16 is defined as an output signal Sk.

The output signal of the adaptive automatic equalization section 15 and the output s of the switching means 18 are supplied to one error signal detection section, and the error signal detection section 1-9 is supplied to the output 20 of the adaptive automatic equalization section 15. Output S of switching means 18.

The tap coefficient control unit 20 uses the error signal e detected by the error signal detection unit 19 to generate an error signal e by calculating the difference between the filters included in the adaptive automatic equalization unit 1-5 generates a tap coefficient control signal for sequentially updating the tap coefficients of the adaptive automatic equalizer 15, and based on this signal, sequentially corrects the tap coefficients of the filter included in the adaptive automatic equalizer 15 at the sampling timing of the first switch 14. . Here, when the tap coefficient control unit 20 starts receiving the training signal 1, it selects a Kalman algorithm, fast Kalman algorithm, square root Kalman algorithm, etc. from an appropriate initial value.
The tap coefficient control signal is sequentially updated by an RLS method such as an algorithm. Furthermore, if the filter included in the adaptive automatic equalization section 15 is a so-called decision feedback type, the output of the switching means 18 is transmitted via the path 21 to the adaptive automatic equalization section].
It is fed back to the filter included in 5.

On the other hand, the received carrier frequency band analog signal "(t
) is supplied to the multiplier 12 as well as the intensity detector 32 . The intensity detection unit 32 detects the received analog signal “(
t) is detected and output.

Furthermore, information regarding a predetermined tap coefficient, such as a center tap coefficient, in the output of the tap coefficient control section 20 is supplied to the fifth delay element T5 and the phase difference detection means 62. The phase difference detection means 62 includes a fifth
extract each phase information from the information about the predetermined tap coefficient one time ago and the information about the predetermined tap coefficient at the current time stored in the delay element T5, and find the difference between them. and output it. The output of the intensity detection section 32 and the output of the phase difference detection means 62 are multiplied by a multiplier 82, and the multiplier 82 supplies the result to the average value calculation means 34. The average value calculation means 34 updates the tap coefficients approximately twice the number of taps of the filter included in the adaptive automatic equalization section 15 from the time t1 when it starts receiving the training signal 1. The average value of the 62 outputs is calculated, and a DC voltage proportional to the average value, that is, a DC voltage proportional to the frequency offset is generated. The third switch 36 is connected to the switching means 18 and is connected to a training switch receiving the training signal 1.
During the mode period, it is released, but during the decision feedback mode period when receiving the data signal, it is short-circuited and VC03
3 is supplied with a DC voltage which is the output of the average value calculating means 34. With this DC voltage, vC033 finely adjusts the oscillation frequency that is its output.

By doing this, when the signal strength of the received signal is low, the reliability of the error signal is low, and when the signal strength of the received signal is high, the reliability of the error signal is high. By weighting the phase information of the signal, a highly reliable average value of the phase information, that is, information regarding the frequency offset can be obtained.

In addition, the above-mentioned embodiments (Fig. 1, Fig. 3, Fig. 4, Fig. 5)
VC033 as the frequency conversion means 330 in FIG.
Although the frequency conversion means of the present invention is not limited to one using such CO, for example, an analog/digital conversion section 51 as shown in FIG. , a sine wave/cosine wave ROM 52, an accumulative adder 53, an LPF 54, and the like, as long as it performs the frequency conversion function.

[Effects of the Invention] As is clear from the above description, according to the present invention, when the transmission path characteristics including frequency offset change from moment to moment, or when the training signal is In particular, if the receiver equipped with an adaptive automatic equalizer, such as in a car or mobile telephone, is moving, i.e. the frequency offset may be affected by the Doppler effect or by multiwave propagation. It is possible to realize an adaptive automatic equalizer that can automatically and instantaneously detect transmission path distortion including frequency offset and compensate for it even when it occurs frequently.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the first basic configuration and the configuration of a first embodiment of an adaptive automatic equalizer having a frequency offset removal function according to the present invention, and FIG. 2 is a diagram showing the frequency offset removal function according to the present invention. A second basic configuration diagram of an adaptive automatic equalizer having a frequency offset removal function having the basic configuration shown in FIG. A schematic configuration diagram of a second embodiment of the converter,
FIG. 4 is a schematic configuration diagram of a third embodiment of an adaptive automatic equalizer having a frequency offset removal function with the basic configuration shown in FIG. 2, and FIG. 5 shows a frequency offset removal function with the basic configuration shown in FIG. 2. FIG. 6 is a diagram showing the format of the received signal of the adaptive automatic equalizer, and FIG. 7 is a diagram showing the format of the received signal of the adaptive automatic equalizer. Schematic configuration diagram of an adaptive automatic equalizer with
The figure is a schematic diagram of another embodiment of the frequency conversion means of the adaptive automatic equalizer according to the present invention. 1... Training signal 2... Data signal 12
...Multiplier 13...PLL14...
First switch 15... adaptive automatic equalization unit 16...
- Judgment unit 16o...quantization means 17...
- Training signal generator 18...Switching means 19...Error signal detection section 20...Tap coefficient control section 21...Feedback path 31...Phase difference detection section 32...Intensity detection section 3
3...Voltage controlled oscillator 33o...Frequency conversion means 34...Average value calculation detection means 35...Multiplier 36...Third switch 42...Tap angular velocity detection section representative patent attorney Miyoshi Hidekazu

Claims (3)

[Claims] (1) an adaptive automatic equalization unit, a tap coefficient control unit that sequentially determines tap coefficients of an adaptive filter of the adaptive automatic equalization unit, and means for detecting and quantizing a signal output from the adaptive automatic equalization unit; an adaptive automatic equalizer comprising an error signal detection section that performs subtraction processing on the input side and output side signals of the quantization means and inputs the difference signal to the tap coefficient control section as correction information for the tap coefficient control section; , a reception strength detection means for the received analog signal; a phase error detection means for detecting a reproduced carrier phase error from the output of the error signal detection section; and a weighting coefficient determined according to the reception strength of the detected analog signal. and an average value calculation means for weighting the phase error signal output from the error signal detection section using the determined weighting coefficient and calculating an average value of the phase of the weighted error signal; and from the average value calculation means. What is claimed is: 1. An adaptive automatic equalizer having a frequency offset removal function, characterized in that the adaptive automatic equalizer is provided with a frequency conversion means for adjusting the oscillation frequency according to the average value of the output error signal. (2) an adaptive automatic equalization unit, a tap coefficient control unit that sequentially determines the tap coefficients of the adaptive filter of the adaptive automatic equalization unit, and means for detecting and quantizing the signal output from the adaptive automatic equalization unit; An adaptive automatic equalizer comprising an error signal detection section that performs subtraction processing on the input side and output side signals of the quantization means and inputs the difference signal to the tap coefficient section, a tap angular velocity detection unit that extracts only information regarding a predetermined tap from a series of tap coefficient correction information output from the tap coefficient control unit and detects an angular change amount of the predetermined tap; and a frequency conversion means for adjusting the frequency of a local oscillator necessary for converting an analog signal from high frequency to baseband according to the output of the unit. Maker. (3) In the tap coefficient control section, according to the circular least squares method,
3. The adaptive automatic equalizer having a frequency offset removal function according to claim 1, wherein the tap coefficient control signal is updated.