JPH03280708A - Adaptive equalizer - Google Patents

Abstract

PURPOSE:To prevent divergence of an equalizer coefficient by releasing the control of revision of the equalizer coefficient at a prescribed time after the abnormal amplitude of a reception signal is not detected to control the equalizer coefficient revision processing. CONSTITUTION:An equalizer coefficient revision control section is limited from a time of the presence of an abnormal amplitude is detected in a reception signal and an equalizer coefficient revision calculation section 3 makes calculation. The limit of the revision of the equalizer coefficient is released at a prescribed time after the abnormal amplitude is not detected. Then the divergence of the equalizer coefficient is prevented by controlling the equalizer coefficient revision processing.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an adaptive equalizer, and more particularly, to various modulation and demodulation systems, facsimile machines used in ring systems of facsimile machines, communication modulation and demodulation systems, and mobile equalizers. It is a power applied to communications, etc.

J = Jutsu - P S K (yen) ase 5hift Keyj, n
B: Phase modulation method) 7% formula%: The modulation method of the quantile phase modulation method corrects the frequency characteristics of the received signal on the transmission path in order to restore the transmitted data from the received signal. It is necessary to correct the difference in carrier frequency (frequency offset) and phase jitter between the receiving side and the receiving side, extract a synchronization signal from the received signal, and then reproduce the transmitted data in synchronization with this synchronization signal. .

In order to correct the frequency characteristics of the transmission path, a transversal type adaptive equalizer (adapti
v eEqualizer) is used, and its equalizer coefficients are sequentially updated at the symbol rate so that the characteristics are inverse to the transmission path characteristics. The first thing used at this time. The equation for updating the isospecific heat coefficient is valid under the condition that there are no instantaneous interruptions or amplitude hits in the input. Therefore, in a transmission system where instantaneous interruptions and phase hits occur frequently, symbols 1. , /-1-, when the equalizer coefficients are updated according to a predetermined calculation formula, the equalizer coefficients diverge, and there is a problem that the desired transmission characteristics; ψ characteristics cannot be obtained.

FIG. 34 shows a configuration diagram of a conventional adaptive equalizer, which has a 1..Lance Bazaar type filter configuration. In the figure, 81a-=81g is a delay element, to 82a-
8211 is a multiplier, and 83 is an integrator.

In general, an adaptive equalizer is used to cancel the frequency characteristics of a transmission path, and is used to cancel the frequency characteristics of a transmission path. The inverse frequency characteristic of the transmission line is automatically realized by the l-1 noning signal. Furthermore, in the case of an adaptive equalizer, the filter coefficients are sequentially changed while receiving the information using the information of the transmitted signal, and the change in the transmission path is handled in step 7.

The process of changing the output of the adaptive equalizer and the tap coefficients (equalizer coefficients) is performed by the following calculation. However, the change process involves various calculations, and the following is one example.

The filter output q of the adaptive equalizer is expressed as the input time series ai to the adaptive equalizer - the output C (is expressed as .

Further, each tap coefficient is updated for each symbol rate (T) using a fifth-order calculation formula.

C("Ci 1"'e' a I"""(2
) However, e = q − p ; −43)
P: Correct signal value (judgment value) r: Convergence coefficient of equalization amount As shown in equation (2) above, the tap coefficients are updated.

This is done by converging so as to reduce the error e between the reproduction output value q at the point reproduced from the transmission signal and the determination value P at the determination point. In this case, if the communication signal is known in advance, the error is calculated using equation (3) with the value as P. If it is not known, the error is calculated by referring to the sent value q and setting the signal point closest to q (either O or 1 in a transmitting/receiving system where only 1 is sent) as P.

In addition, V, 27ter (International Telegraph and Telephone Advisory Committee: CC
ITT recommendation number) The latter is used for modem facsimiles, etc. Furthermore, the equations (1) to (3) above may be complex number operations.

The important points here are: ■ If the error e is large, convergence is poor. ■ The transmission signal values ai sequentially given to this filter must be random. ■ The convergence coefficient r of the equalizer is the speed of convergence. If the value is small, the convergence speed will be slow; if the value is large, the convergence speed will be fast, but it will oscillate due to a small error.

In other words, the convergence operation of the filter is expressed as the filter output q
This means that training data P is given to the filter coefficients so that the filter output q becomes close to the training data p.

Therefore, if there is an abnormal amplitude input that includes a momentary interruption or amplitude fluctuation in the adaptive equalization amount with the above-mentioned characteristics, the error e will increase according to equation (3), the tap coefficient will increase according to equation (2), and therefore Furthermore, the value of the filter output q increases according to equation (1), and this value is fed back to (3), so that the error e becomes larger and the equalization amount diverges.

The present invention was made in view of the above-mentioned circumstances.
When an abnormal amplitude such as momentary interruption or amplitude fluctuation is observed in the received signal, the equalizer coefficient changing process is controlled to prevent the equalizer coefficient from divergence and to obtain the opposite characteristic of the desired transmission characteristic. The purpose of this design was to provide an adaptive equalizer with

In order to achieve the above object, the present invention provides (1) delay means for delay-shifting a received signal at regular time intervals, and multiplication means for multiplying each tap output of the delay means by a predetermined tap gain. and an accumulation means for summing up the results of each multiplication by the multiplication means, and automatically adjusts the tap gain based on a predetermined calculation to remove the frequency characteristics of the transmission path superimposed on the received signal, etc. The equalizer includes an abnormal amplitude determining section that detects a received signal with an abnormal amplitude due to instantaneous interruption or amplitude fluctuation, and an equalizer coefficient changing process from the time when the abnormal amplitude of the received signal is detected by the abnormal amplitude determining section. Equalization that controls the equalizer coefficient changing process by canceling the restriction on changing the equalization amount coefficient and restarting the equalizer coefficient changing process after a predetermined time from the time when abnormal amplitude is no longer detected. (2) a delay means for delay-shifting a received signal at fixed time intervals; a multiplication means for multiplying each tap output of the delay means by a predetermined tap gain; an integrating means for summing up the results of each multiplication by the means, and an adaptive equalizer that removes frequency characteristics of a transmission path superimposed on the received signal by automatically adjusting a tap gain based on a predetermined calculation,
an abnormal amplitude determination unit that detects a received signal with an abnormal amplitude such as instantaneous interruption or amplitude fluctuation, and restricts equalizer coefficient changing processing only when the abnormal amplitude of the received signal is detected by the abnormal amplitude determination unit; (3) comprising an equalizer coefficient change control unit that controls the equalizer coefficient change process by not placing any restrictions on the change of the equalizer coefficients at the time when no abnormal amplitude is detected; or (3)
A delay means for delay-shifting a received signal at regular time intervals, a multiplication means for multiplying each tap output of the delay means by a predetermined tap gain, and an integration means for summing each multiplication result by the multiplication means. In adaptive isospecific heat, which automatically adjusts the tap gain based on a predetermined calculation to remove the frequency characteristics of the transmission path superimposed on the received signal, the received signal with abnormal amplitude such as instantaneous interruption or amplitude fluctuation. of,
From the time when an abnormality in the equal fixture coefficient is detected by the error determination section that determines based on the error between the playback output value and the judgment value, and the difference determination section, the change processing of the equal fixture coefficient is also limited. having an equal specific heat coefficient change control unit that controls the equal specific heat coefficient changing process by canceling the change of the equal fixture coefficient and restarting the equal specific heat coefficient changing process after a predetermined period of time from the time when no abnormality is detected; , (4) a delay means for delay-shifting the received signal at regular time intervals, and a multiplication means for multiplying each tap output of the delay means by a predetermined tap gain;
6. Adaptive equalization that removes the frequency characteristics of the transmission path superimposed on the received signal by automatically adjusting the tap gain based on a predetermined calculation. The device includes an error determining section that determines a received signal with an abnormal amplitude 4 such as clarity or amplitude fluctuation based on the error between the reproduction output value and the judgment value, and an error determining section that determines the received signal with abnormal amplitude 4 such as clarity and amplitude fluctuation, and at the time when an abnormality in the fixture coefficient is detected by the error determining section. Control the process of changing the constant heat coefficient by restricting the process of changing the constant heat coefficient that contributes to the error in the constant heat coefficient calculation at the child time 7, and not restricting the change of the constant constant heat coefficient at other times, etc. (5) An abnormality is detected in the specific heat coefficient, and the process of changing the specific heat coefficient is stopped by setting the convergence coefficient of the specific heat coefficient involved in heating to zero. , furthermore, (6) when an abnormality is detected in the equalizer coefficient, the level of the input signal input to the equal specific heat is clipped until the abnormality disappears thereafter; furthermore, (7) The present invention is characterized in that the abnormality in the equalizer coefficient is determined from an error used when changing the equal heat coefficient, and further, (8) the abnormality in the equalizer coefficient is determined from the overall shape of the constant heat coefficient. Hereinafter, the present invention will be explained based on examples.

FIG. 1 is a block diagram for explaining an embodiment of an adaptive equalizer according to the present invention, and in the figure, ]- is a foreign amplitude determination section;
2 is a coefficient change control section, 3 is an isothermal coefficient update calculation section, 4 is an error calculation section, 5 is a determination section, 68 to 6g are delay elements, 7
a to 7 h let, a multiplier, and 8 is an integrator. The configuration of the adaptive equalizer of the present invention shown in FIG. 1 is the same as that of the conventional adaptive equalizer shown in FIG.
It has a configuration in which a coefficient change control section 2 is added (encircled by a thick line in the figure). In this adaptive equalizer, when an abnormal amplitude such as a momentary interruption or an amplitude fluctuation is detected by the abnormal amplitude determination section 1, the coefficient change control section 2 performs control related to changing the isothermal coefficient.

FIG. 2 is a flowchart for explaining the operation of the coefficient change control section using the adaptive equalizer shown in FIG. Below, each step will be explained in order.

! Operation: When a received signal is input, it is determined whether there is an abnormality in the amplitude value (s) of the received signal. As the permissible amplitude range, a maximum value wax and a minimum value m'n are provided.

If it is determined that the amplitude value (s) is abnormal, that is, mj, n>s, max<s, then the convergence coefficient r is set to O and the isospecific heat coefficient change calculation is performed.

This corresponds to stopping the update of the uniform fixture coefficient C by setting r=o in the equation (2).

stem; When the amplitude value (,) is no longer abnormal in step 1, that is, when it is determined that m]n≦5≦max, the process in Japan for determining whether to resume changing the isospecific heat coefficient is, for example, when the amplitude is abnormal. The number of received signals after the signal disappears is calculated and it is determined whether this exceeds a predetermined threshold. If the predetermined threshold value is not exceeded, the isospecific heat coefficient change calculation is performed through step 2 described above while keeping the convergence coefficient r20.

4; In step 3 above, when the number of received signals exceeds a predetermined threshold, the convergence coefficient r is set to a number other than O, and isospecific heat coefficient change calculation is performed. This is because in a transversal type filter, input signals are sequentially sent to the delay elements, so it takes time for the abnormal amplitude signal value to disappear from the delay elements.

FIG. 3 is a diagram showing another embodiment of the present invention, in which 21
22 is an abnormal amplitude determination unit, 22 is a coefficient change control unit, and 23a to 2
3h is an isospecific heat coefficient update calculation unit (hatched in the figure); 24 is an error calculation unit; 25 is a determination unit; 26a to 26;
g is the first delay element.

27a-27h is a multiplier, 28 is an integrator, 29a~
29f is a second delay element. In this example, first delay elements 26a to 26g for delaying input (received) signals
and a second delay element 29a for delaying the convergence coefficient r.
29f, and when an abnormal amplitude occurs, that is, S>
When wax or smin, coefficient change control unit 22
The convergence coefficient r is set to O, and when the input signal does not have an abnormal amplitude, r≠O is set and the input signal is sent to the second delay element with the convergence coefficient r. The isospecific heat coefficient update calculation units 23a to 23h perform calculations using the convergence coefficient F given to each second delay element.

FIG. 4 is a flowchart for explaining the operation of the coefficient change control section of the embodiment shown in FIG. Below, each step will be explained in order.

When a received signal is input, it is determined whether there is an abnormality in the amplitude value (S) of the received signal. As the permissible amplitude range, a maximum value max and a minimum value win are provided.

If there is an abnormality in the amplitude value (S), that is, m1
n) When it is determined that s, wax<s, the convergence coefficient r=o and other ship coefficient change calculations are performed.

When the amplitude value (S) is no longer abnormal, that is, when it is determined that min≦S≦maX, the convergence coefficient r≠O(
Set r to a number other than O, and perform other ship coefficient change calculations.

FIG. 5 is a diagram showing still another embodiment of the present invention, in which:
31 is an abnormal amplitude determination unit and an amplitude clipping unit, 32 is an isospecific heat coefficient update calculation unit, 33 is an error calculation unit, 34 is a determination unit,
35a to 35g are delay elements, 36a to 36h are multipliers,
37 is an integrator.

In this example, when an abnormal amplitude occurs, the amplitude clipping section 31 first determines whether or not the input signal has an abnormal amplitude,
Add a limit to the amplitude and input it to the equalizer.

FIG. 6 is a flowchart for explaining the operation of the amplitude clipping section of the embodiment shown in FIG. Below, each step will be explained in order.

When a received signal is input, it is determined whether there is an abnormality in the amplitude value (S) of the received signal. As the permissible amplitude range, a maximum value max and a minimum value ll1in are provided. If the amplitude value (s) of the received signal is not abnormal, other ship coefficient change calculations are performed.

Totsuji master: If there is an abnormality in the amplitude value (S), that is, tni
When it is determined that n>s and max<s, first wi
Determine whether n>s.

If min>s, then calculate the other ship coefficient change by setting s=min.

Tori rA-; If win > s, then wax
Determine whether <s.

5; max (If s, calculate the other ship coefficient change by setting s = max. In this case, max = m1
If n=0, it is possible to perform processing equivalent to that shown in FIG.

FIG. 7 is a diagram showing another embodiment of the adaptive equalizer according to the present invention. In the figure, 41 is a coefficient change control section, 42 is an isospecific heat coefficient storage section (buffer), and 43 is an isospecific heat coefficient change section. 44 is an error determining section, 45 is an error calculating section, 46 is a determining section, 47a to 47g are delay elements, 48a to 48h are multipliers, and 49 is an integrator. The configuration of the adaptive equalizer of the present invention shown in FIG. 7 is based on the conventional adaptive equalizer shown in FIG.
er) 42 and a coefficient change control section 41 are added. When a received signal with abnormal amplitude such as instantaneous interruption or amplitude fluctuation is input to this adaptive equalizer, the error e increases according to equation (3) above.

Therefore, this error e is determined by an error determining section.

If this value is greater than or equal to the threshold, the coefficient change control unit performs control related to changing the isospecific heat coefficient. FIG. F3 explains the operation of the coefficient change control unit of the embodiment shown in FIG. This is the final flowchart 1-. Below, each step will be explained in order.

i 轡1 ;Is there an abnormality in the equalization seed coefficient?
Make a judgment based on e). The allowable range of error (e) is
Now, a maximum value of 1 lax is set.

If the error (e) is abnormal, that is, e>
When it is determined that Inax, the equalization seed coefficient is replaced with the equalization seed coefficient stored in the other ship coefficient storage section before the error C increases, and the convergence coefficient r is set to O, and the equalization seed coefficient is used. Perform a calculation to update the coefficients. This is because updating of the equalization seed coefficient is stopped by setting r=o in the above equation (2) [2. By using the same equalization seed coefficient in the abnormal amplitude section, a divergent shell of the equalizer is prevented.

White 4: Next, the human input signals are sequentially input to the equalizer, and the error is determined each time by the error determining section, and when the equalization seed coefficient is no longer abnormal, a determination is made to restart other ship coefficients, etc. The process of restarting the equalizer coefficient change is, for example, by calculating the number of input signals after the equalization seed coefficient is no longer abnormal, and determining whether this exceeds a predetermined threshold. to decide.

Park Town-4-: When the number of input signals exceeds a predetermined threshold, the convergence coefficient r is set to a number greater than or equal to 0:'.

If not, change other ship coefficients such as leaving r=o. This is because in the Lancebar-Sal type filter, since the single-handed signal is sequentially sent to the delay elements, it takes time for the abnormal amplitude signal value to disappear from the delay elements. Accumulation in the other ship coefficient accumulation section is performed at predetermined timing until an abnormality occurs. However, this is performed at a predetermined times the rate for each timing. 1. Figure 9 does not show other embodiments of the present invention, and in the figure, 5]
- is a coefficient change control unit, and 52a to 5211 are other ship coefficient storage units and equalization type coefficient update JI calculation units.

53a to 53f are second delay elements; 54 is an error determination unit;
55 is an error calculation section, 56 is a determination section, 57a to 57g are delay elements, 58a to 58h are multipliers, and 59 is an integrator. In this example, τ has a delay element D 7 a to 57g for delaying the human power <A, and a second delay element 53, □1 to 53f for delaying the convergence coefficient, and the other configurations are This is almost the same as FIG. 7, but only when an abnormality occurs in the equalization seed coefficient, that is, when e>l1lax, the coefficient change control unit 5 converges the convergence coefficient r that is particularly related to the error calculation. Let the coefficient r be O, and if not, (4r≠0) and send it to the second delay M = f of the convergence coefficient r.Here, as the convergence coefficient r involved in the 31 calculation of the error C, For example, center tap C.

It is involved in calculations related to subsequent taps C-i. In the other ship coefficient update calculation units 52a to 521i, calculations are performed using the convergence coefficients r given to the second delay elements 53a to 53f.

FIG. 10 is a flowchart 1- for explaining the operation of the coefficient change control section of the embodiment shown in FIG. 9. The steps will be explained below in order.

It is determined whether there is an abnormality in the qieJ ri-2 equalizer coefficient based on the error (e). The maximum value max is set as the allowable range of error (+3).

Village eJ−λ; If the error <e> is a foreign country, that is, (
,! > When it is determined that the convergence coefficient is r-0, other ship coefficient change calculations are performed.

When the error (e) is no longer abnormal, that is, e
When it is determined that ≦may, the convergence coefficient r≠O is set (r is a number other than O), and other ship coefficient change calculations are performed.

This processing can prevent divergence of each filter coefficient by 11 degrees. In addition, the processing of the other ship coefficient accumulation section is as follows:
Same as the figure.

[No.] Coe diagram is a diagram showing still another embodiment of the present invention, in which 6 is an amplitude flip section, 62 is an equal ship coefficient update calculation section, 63 is an error judgment section, and 64 is an error A calculating section, 65 is a determining section, 66a to 66g are delay elements, 67a to 67h are multipliers, and 68 is an integrator. In this example, when an abnormality occurs in the equalization seed coefficients, the amplitude clipping unit 6]-
, and input the amplitude to the equalizer.

FIG. 12 is a flowchart for explaining the operation of the amplitude rip section of the embodiment shown in FIG. 1-1.

Below, each step will be explained in order.

Comparable: When it is determined that the equalizer coefficient is abnormal, that is, the error (e) is abnormal, it is assumed that an abnormality has occurred in the amplitude value of the received signal, and the amplitude value (S ) with a predetermined threshold. As the permissible range of the amplitude value (S), a maximum value max and a minimum value win are provided. Amplitude value (S
) is not abnormal, that is, if win≦S≦wax, equalizer coefficient change calculation is performed.

If there is an abnormality in the amplitude value (S), that is,
When it is determined that ll1in>s and IIIax<s, it is first determined whether win>s.

If win > s, s=mi
Equalizer coefficient change calculation is performed as n.

If it is not 扛弘地;m1n)s, then it is determined whether it is waxes.

If max<s, equalizer coefficient change calculation is performed with s=max.

This can prevent divergence of the equalizer coefficients.

In addition, in FIG. 12, max = min = 0
If so, it is possible to perform processing equivalent to that shown in FIG.

FIG. 13 is a diagram showing still another embodiment of the present invention.

In the figure, 71 is a coefficient change control unit, 72 is an equal ship coefficient storage unit (buffer), 73 is an equal other ship coefficient shape determination unit, and 75 is a coefficient change control unit.
76 is an error calculating section, 76 is a determining section, and 79 is an integrator. The configuration shown in FIG. 13 does not judge the abnormality of the equalizer coefficient based on the error (e) as in the configuration shown in FIG. It is something.

As an example of this process, when a coefficient before change and a coefficient after change are compared and there is a tap that has changed significantly, it can be treated as an abnormality occurring in the equalizer coefficient.

Effects---Button 3 As is clear from the above description, the present invention has the following effects.

(1) Delay means for delay-shifting a received signal at regular time intervals, multiplication means for multiplying each tap output of the delay means by a predetermined tap cane, and integration means for summing the results of each multiplication by the multiplication means. In an adaptive equalizer that removes the frequency characteristics of the transmitter superimposed on the received signal by automatically adjusting the tap gain based on a predetermined calculation, the received signal has abnormalities such as instantaneous interruptions and amplitude fluctuations. The process of changing the equalizer coefficient is restricted from the time when the amplitude is detected, and after a predetermined time from the time when the abnormal amplitude is no longer detected, the restriction on changing the equalizer coefficient is canceled and the process of changing the coefficient of other ships is performed. By restarting, etc., it is possible to prevent divergence of the equalizer coefficients by controlling the other ship coefficient change processing.

(2) a delay means for delay-shifting a received signal at regular time intervals; a multiplication means for multiplying each tap output of the delay means by a predetermined tap gain; and an integration means for summing the results of each multiplication by the multiplication means; In an adaptive equalizer that removes frequency characteristics of a transmitter superimposed on the received signal by automatically adjusting tap gain based on a predetermined calculation.

Restricts the process of changing the equalizer coefficients only when abnormal amplitudes such as instantaneous interruptions or amplitude fluctuations are observed in the received signal, and does not impose restrictions on changing the equalizer coefficients when no abnormal amplitudes are detected. In this way, it is possible to prevent divergence of the equalizer coefficients by controlling the process of changing the coefficients of other ships, and to realize an adaptive equalizer that is effective against impulsive noise components.

(3) delay means for delay-shifting the received signal at regular time intervals; multiplication means for multiplying each tap output of the delay means by a predetermined tap gain; and integration means for summing the results of each multiplication by the multiplication means. An adaptive equalizer that removes the frequency characteristics of the transmitter superimposed on the received signal by automatically adjusting the tap gain based on a predetermined calculation, which includes abnormal amplitudes such as instantaneous interruptions and amplitude fluctuations. However, the process of changing the equalizer coefficient is restricted from the time when the received signal is detected by the equalizer coefficient, and the restriction on changing the equalizer coefficient is canceled after a predetermined period of time from the time when abnormal amplitude is no longer detected. By restarting the other ship coefficient changing process, etc., it is possible to prevent the equalizer coefficient from divergence by controlling the other ship coefficient changing process.

(4) Rather than finding abnormalities in the equalizer coefficients from errors, it is possible to respond more quickly by paying attention to fluctuations in the equalizer coefficients of each tap.

[Brief explanation of drawings]

FIG. 1 is a block diagram for explaining an embodiment of the adaptive equalizer according to the present invention, and FIG. 2 is a flowchart for explaining the operation of the coefficient change control section of the adaptive equalizer of FIG.・-Char 1・, FIG. 3 is a diagram showing another embodiment of the present invention, and FIG. 4 is a diagram showing another embodiment of the present invention.
5 is a flowchart for explaining the operation of the coefficient change control section of the embodiment shown in FIG. 3, FIG. 5 is a diagram showing still another embodiment of the present invention, and FIG. FIG. 7 is a flowchart for explaining the operation of the amplitude clipping section of the embodiment shown in FIG. Flowchart 1-1 for explaining the operation of the coefficient change control unit of the illustrated embodiment
FIG. 9 is a diagram showing another embodiment of the present invention, and FIG. 10 is a diagram showing another embodiment of the present invention.
Operation of the coefficient change control section of the embodiment shown in FIG. FIG. 1-1 shows a flowchart for explaining the operation of the amplitude clipping section of the embodiment [-1 FIG. 13 is a diagram showing still another embodiment of the present invention, and FIG. , is a configuration diagram of a conventional adaptive equalizer. ],... Abnormal amplitude determination section, 2... Coefficient change control section, :3
Equalizer coefficient update calculation unit, 4... error calculation unit, 5... determination unit, 68-6g... delay element, 7 a □ 7 h
...multiplier, 8...integrator. -F! Diagram diagram

Claims (1)

[Claims] 1. Delay means for delay-shifting a received signal at fixed time intervals, multiplication means for multiplying each tap output of the delay means by a predetermined tap gain, and each multiplication result by the multiplication means In an adaptive equalizer that removes the frequency characteristics of the transmission path superimposed on the received signal by automatically adjusting the tap gain based on a predetermined calculation, the adaptive equalizer is comprised of an integrating means for summing the signal, and eliminates the frequency characteristics of the transmission path superimposed on the received signal. an abnormal amplitude determination unit that detects a received signal with an abnormal amplitude; and an abnormal amplitude determination unit that restricts equalizer coefficient changing processing from the time when the abnormal amplitude of the received signal is detected by the abnormal amplitude determination unit, so that the abnormal amplitude is no longer detected. and an equalizer coefficient change control unit that controls the equalizer coefficient change process by canceling the restriction on the change of the equalizer coefficient and restarting the equalizer coefficient change process after a predetermined time from the point in time. Features an adaptive equalizer. 2. Delay means for delay-shifting the received signal at fixed time intervals, multiplication means for multiplying each tap output of the delay means by a predetermined tap gain, and integration means for summing the results of each multiplication by the multiplication means. In an adaptive equalizer that removes the frequency characteristics of the transmission path superimposed on the received signal by automatically adjusting the tap gain based on a predetermined calculation, the adaptive equalizer removes the frequency characteristics of the transmission path superimposed on the received signal. An abnormal amplitude determination unit detects the signal, and the abnormal amplitude determination unit limits the equalizer coefficient changing process only when an abnormal amplitude of the received signal is detected, and changes the equalizer coefficient at the time when no abnormal amplitude is detected. An adaptive equalizer comprising an equalizer coefficient change control section that controls equalizer coefficient change processing without imposing any restrictions on changes. 3. Delay means for delay-shifting the received signal at regular time intervals, multiplication means for multiplying each tap output of the delay means by a predetermined tap gain, and integration means for summing the results of each multiplication by the multiplication means. In an adaptive equalizer that removes the frequency characteristics of the transmission path superimposed on the received signal by automatically adjusting the tap gain based on a predetermined calculation, abnormal amplitudes such as momentary interruptions and rear width fluctuations occur. an error determining section that determines a received signal based on an error between a reproduced output value and a determination value; and an error determining section that limits equalizer coefficient changing processing from the time when an abnormality in the equalizer coefficient is detected by the error determining section; Equalizer coefficient change that controls the equalizer coefficient change process by canceling the change of the equalizer coefficient and restarting the equalizer coefficient change process after a predetermined time from the time when no abnormality is detected in the equalizer coefficient. An adaptive equalizer comprising a control section. 4. Delay means for delay-shifting the received signal at fixed time intervals, multiplication means for multiplying each tap output of the delay means by a predetermined tap gain, and integration means for summing the results of each multiplication by the multiplication means. In an adaptive equalizer that removes the frequency characteristics of the transmission path superimposed on the received signal by automatically adjusting the tap gain based on a predetermined calculation, the adaptive equalizer removes the frequency characteristics of the transmission path superimposed on the received signal. an error determining unit that determines a signal based on the error between a reproduction output value and a determination value; and an error determining unit that contributes to an error in equalizer coefficient calculation at that time when an abnormality in the equalizer coefficient is detected by the error determining unit. It is characterized by having an equalizer coefficient change control unit that controls the equalizer coefficient change process by restricting the change process of the equalizer coefficient and not imposing restrictions on the change of the equalizer coefficient at other times. An adaptive equalizer.