JP2008282477A - Data playback device and method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a data playback device and a data playback method by which data is separated accurately and can be taken out by excluding interference from a signal of an adjacent recording track. <P>SOLUTION: N<SB>r</SB>pieces of playback signals are equalized by filters of equalizers 23, 24, and interference from the adjacent track is excluded using an estimated value or an already known value. At the time, interference of gain by the equalizers 23, 24 and gain adjusters 21, 22 and interference of phase by the equalizers 23, 24 and ITRs 25, 26 is suppressed by equalizing frequency response in a fixed state. Then, data is separated accurately and taken out by controlling respectively amplitude of playback signals read out from Nr pieces of playback heads of gain controllers 31, 32 and filter coefficients of adaptive equalization controllers 33, 34 based on output signals in which bit synchronization is performed and its detected result. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a data reproducing apparatus and a data reproducing method for reproducing data recorded in multitrack.

  In recent years, research and development of multi-track recording has been conducted for high-density recording and high transfer rate in magnetic recording and optical recording. Specifically, research has been conducted to realize high-density recording by narrowing the track width of the recording track. Also, research has been conducted to realize a high transfer rate by reproducing a plurality of recording tracks at a time and performing parallel processing.

  However, for example, in a hard disk recording / reproducing apparatus that does not read out signals of adjacent recording tracks, it is necessary to make the width of the reproducing head smaller than the width of the recording head. End up. Therefore, a technique for reproducing a plurality of recording tracks at a time and performing parallel processing has attracted attention.

  Incidentally, in the communication field, there is a technique called MIMO (Muti-Input Multi-Output) as one of methods for transmitting a large amount of information at a time. This MIMO can be applied to parallel processing of multitrack recorded data during reproduction.

  MIMO indicates the following relationship, where s is the vector of the transmission signal, r is the vector of the reception signal, and H is the matrix representing the channel.

The received signal is a state in which several transmission signals are mixed by channels. Here, if the matrix H is obtained on the receiving side and the inverse matrix H ⁻¹ exists, the original transmission signal can be extracted as in the following equation.

  By the way, T.W. Conway et al. Have proposed an equalization method in which the width of the reproducing head is made larger than the width of the recording track to remove interference from the signal of the adjacent recording track (see, for example, Patent Document 1).

Hereinafter, a conventional equalization method will be described. Here, it is assumed that _Nw recording tracks are reproduced by _Nr reproducing heads. Further, the N _w pieces of recording tracks in this case as one group. Further, between groups, flanked by non-recording region called guard band, satisfies the condition N _w ≦ N _r.

  FIG. 7 is a block diagram showing a configuration of a data reproducing apparatus in conventional multitrack recording. The data reproducing apparatus 100 adjusts the gain of an amplified reproduction signal, an input unit 101 to which a plurality of reproduction signals read from a recording medium are input, a reproduction amplifier 102 that amplifies the inputted reproduction signal A variable gain amplifier 103, a low-pass filter (LPF) 104 that removes aliasing distortion when quantizing, an AD converter 105 that quantizes the reproduction signal, and equalization of the reproduction signal and signals from adjacent recording tracks An equalizer 106 that eliminates interference, an ITR (Interpolated Timing Recovery) 107 that performs bit synchronization of the reproduction signal, a detector 108 that detects the reproduction signal, an error is calculated from the reproduction signal that is bit-synchronized and the detection result. An error calculator 109, a gain controller 110 for adjusting the gain of the reproduction signal, an adaptive equalization controller 111 for calculating a filter coefficient of the equalizer 106, Consists of

In this conventional data reproduction apparatus, the reproduction signal input to the equalizer 106 at time n is x _i (n), i = 0, 1,..., N _r −1. Are expressed as follows, where y _k (n), k = 0, 1,..., N _w −1.

Here, L is the length of an FIR (Finite Impulse Response) filter, and h _k (i, j) is the filter coefficient.

From the relationship between the equations (2) and (3), the inverse matrix H ^{−1 in the} equation (2) is included in the FIR filter coefficients. Then, by obtaining this by adaptive equalization, it is possible to remove interference of signals from adjacent recording tracks.

International Publication No. 06/048849 Pamphlet

  However, if adaptive equalization is performed while maintaining the equation (3), the gain and phase of the filter change, and therefore, interference of gain with the gain adjuster and phase interference with the bit synchronizer occur. There was a problem that the characteristics shifted.

  The present invention has been made in view of the above-described problems, and provides a data reproducing apparatus and a data reproducing method capable of removing interference from signals of adjacent recording tracks and accurately separating and extracting data. With the goal.

To solve the problems described above, the data reproducing apparatus according to the present invention, the N _w pieces of recording tracks on a recording medium, N _r pieces of read by the reproducing head _{(N w ≦ N r),} N w pieces of in the data reproducing apparatus to retrieve the data recording tracks are separated, respectively, and a signal amplifier that amplifies N _r pieces of the amplitude of the N _r number of reproduction signal read from the reproduction head, respectively, N _r number that is the amplified A quantization unit that quantizes each of the reproduced signals, and an equalizing unit that equalizes the quantized N _r reproduced signals with a filter with a fixed frequency response, and outputs N _w output signals; the equalized N _w number of output signals of bit synchronization and a synchronization portion to take each a detector for detecting the bit synchronization has taken the N _w pieces of output signals, respectively, to the detection result of the detecting unit Based on A gain control unit for controlling the respective amplitudes of the reproduction signal read out from the N _r number of the reproducing head, a filter coefficient of the filter based on the detection result of said output signal bit synchronization is taken with the detector And an adaptive equalization control unit for calculating each.

The data reproducing method according to the present invention, the N _w pieces of recording tracks on a recording medium, N read out in the _r reproducing head (N _w ≦ N _r), respectively separating the data of N _w pieces of recording tracks and a data reproducing method retrieved, N _r pieces of a signal amplifying step of amplifying each amplitude of N _r number of reproduction signal read from the reproduction head, the amplified N _r number of reproduced signals, respectively quantization a quantization step of the N _r number of reproduction signal the quantized equalized by the filter in a state of fixing the frequency response, and equalization step of outputting the N _w pieces of output signals, which is the equalized N a synchronization step that takes _w number of output signals of the bit synchronization, respectively, a detection step of detecting the bit synchronization has taken the N _w pieces of output signals, respectively, the N _r number of on the basis of the detection result in the detecting step Adaptation for calculating the filter coefficients of the filter based on the gain control step for controlling the amplitude of the reproduction signal read from the raw head, the output signal with the bit synchronization and the detection result in the detection step, respectively And an equalization control step.

In the present invention, _Nr reproduced signals are equalized by a filter with a fixed frequency response, so that both gain and phase are maintained. By controlling each N _r pieces of the reproduction signal read from the reproduction head the amplitude and filter coefficient of the filter based on the output signal and the detection result of the bit synchronization has been taken, the adjacent recording tracks The interference from the signal can be removed, and the data can be separated and extracted with high accuracy.

Hereinafter, specific embodiments to which the present invention is applied will be described in detail with reference to the drawings. A data reproducing apparatus shown as a specific example of the present invention reads N _w recording tracks on a recording medium with N _r reproducing heads (N _w ≦ N _r ). Then, those taken out by separating the data of the N _w pieces of recording tracks respectively by N _w pieces of the equalizer. Incidentally, since the cumbersome number _{N w} of the number _{N r} and the recording track of the reproducing head is increased, _hereinafter, be described as _{N r = 2, N w =} 2.

  FIG. 1 is a block diagram showing a configuration of a data reproducing apparatus according to an embodiment of the present invention. The data reproduction apparatus 10 includes input units 11 and 12 to which two reproduction signals read from a recording medium are input, reproduction amplifiers 13 and 14 for amplifying the inputted reproduction signal, and an amplified reproduction signal. Variable gain amplifiers 15 and 16 for controlling the gain, low-pass filters (LPF) 17 and 18 for removing aliasing distortion when quantizing, AD converters 19 and 20 for quantizing a reproduction signal, and quantization Gain adjusters 21 and 22 that adjust the gain of the reproduced signal, equalizers 23 and 24 that eliminate the equalization of the reproduced signal and the interference of the signals of the adjacent recording tracks, and ITR ( Interpolated Timing Recovery) 25, 26, detectors 27, 28 for detecting a reproduction signal, error calculators 29, 30 for calculating an error from the bit-synchronized reproduction signal and its detection result, and a reproduction signal A gain controller 31 for controlling the gain, and and an adaptive equalization controller 33 which calculates the filter coefficients of the equalizer 23 and 24.

  Reproduction signals read from the recording medium using the two reproduction heads are input to the input units 11 and 12, respectively. The reproduction signals output from the input units 11 and 12 are amplified by the reproduction amplifiers 13 and 14, respectively, and are amplified to desired values by the variable gain amplifiers 15 and 16, respectively.

  The reproduced signals amplified by the variable gain amplifiers 15 and 16 are band-limited to prevent aliasing distortion when quantized by the low-pass filters 17 and 18, and are quantized by the AD converters 19 and 20.

  The reproduction signal quantized by the AD converter 19 is sent to the equalizer 23 and the gain adjuster 22. In the gain adjuster 22, the reproduction signal is adjusted to the gain of the variable gain amplifier 16 and sent to the equalizer 24.

  Similarly, the reproduction signal quantized by the AD converter 20 is sent to the equalizer 24 and the gain adjuster 21. In the gain adjuster 21, the reproduction signal is adjusted to the gain of the variable gain amplifier 15 and sent to the equalizer 23.

  The reproduction signals that have been equalized by the equalizers 23 and 24 and from which interference from the signals of adjacent recording tracks has been removed are bit-synchronized by the ITRs 25 and 26. The reproduction signals that are bit-synchronized with the ITRs 25 and 26 are sent to the detectors 27 and 28, and the signals are detected.

  The error calculators 29 and 30 calculate an error from the detection results of the detectors 27 and 28 and the reproduction signal bit-synchronized with the ITRs 25 and 26. The results of the error calculators 29 and 30 are also sent to the adaptive equalization controllers 33 and 34, respectively, and used for calculating the filter coefficients. The detection results of the detectors 27 and 28 are sent to the gain controllers 31 and 32.

  According to such a data reproducing apparatus 10, it is possible to separate and take out a reproduction signal read over a plurality of recording tracks by the reproducing head.

  Here, as shown in FIG. 2, two adjacent recording tracks 41 a and 41 b form one group 41 and are in contact with other groups 42 across a non-recording area called a guard band 43. Here, consider the case where the multi-playback head 51 is off-track. That is, this is a case where there is interference of signals of adjacent recording tracks. Reproduction is performed by the reproduction head 51a and the reproduction head 51b of the multi-reproduction head 51. During reproduction, all the recording tracks in the group are necessarily reproduced. Here, it is assumed that the reproducing head width is equal to the recording track width, but the present invention is not limited to this.

  In the above equation (2), a received signal vector r is assumed to be a reproduction signal equalized by an FIR filter.

  Here, t represents transposition. Further, the output y of the equalizers 23 and 24 is expressed by equation (5).

Since the outputs y of the equalizers 23 and 24 correspond to s in the above equation (2), if H ⁻¹ in the above equation (2) is defined as the equation (6), it is expressed by the following equation (7). .

  That is, in the above equation (3), the following equation (7) is obtained if the following is placed.

  By the way, if each recording track is not written at a different timing within a group of recording tracks, it will not be affected by head wear or the like under different conditions, so that responses between adjacent recording tracks will not be received. There is no significant shift. Therefore, when placed as in equation (9), equation (7) above is expressed by equation (10).

  As can be seen by comparing this equation (10) with the above equation (3), in equation (10), the coefficient α can be set according to the amount of interference of the signal of the adjacent recording track. It is possible to remove the interference from the signal and to separate and extract the data with high accuracy.

  In the above equation (10), the coefficient of the FIR filter (Finite Impulse Response Filter) represented by the following (11) can be obtained by adaptive equalization.

Further, if α _k (k = 0, 1) is known by some method, its value may be given. For example, α may be set according to the off-track amount. Even if it is not understood, if adaptive equalization is performed using an LMS (Least Mean Square) algorithm or the like, interference from signals of adjacent recording tracks can be removed.

  Further, the equalizers 23 and 24 fix (restrict) only certain frequencies (excluding DC and Nyquist frequencies) in the frequency response. By equalizing with the frequency response fixed in this way, both gain and phase are maintained, and gain interference with the gain adjusters 21 and 22 and phase interference with the ITRs 25 and 26 can be prevented. That is, when obtaining the amplitude error in the gain controllers 31 and 32 or obtaining the phase error in the ITRs 25 and 26, the response characteristics of the filter need not be affected.

  For example, taking a hard disk as an example, the basic recording format consists of a preamble used for gain adjustment and pulling in bit synchronization, SYNC data as a synchronization signal, and user data as data to be recorded. In the case of equalization to a partial response PR4 (Partial Response 4) system, the preamble uses a continuous pattern of half-Nyquist frequencies. In this case, it is fixed (constrained) to the response of the preamble frequency as the training signal.

  Then, the structure of the equalizers 23 and 24 mentioned above is demonstrated. Since the equalizer 23 and the equalizer 24 have the same configuration, the equalizer 23 will be described here.

FIG. 3 is a block diagram illustrating a configuration example of the equalizer 23. The equalizer 23 includes an input unit 61, delay elements 62 _{1 to} 62 _L−1 , multipliers 63 _{0 to} 63 _L−1 , an adder 64, and a first FIR filter. Unit 71, delay elements 72 _{1 to} 72 _L−1 , multipliers 73 _{0 to} 73 _L−1 , a second FIR filter including adder 74, and the amount of interference from adjacent recording track signals Is constituted by a multiplier 75 and an adder 76.

The reproduction signal x ₀ (n) input to the input unit 61 is a first FIR composed of delay elements 62 _{1 to} 62 _L−1 , multipliers 63 _{0 to} 63 _L−1, and an adder 64. The signal is equalized by the filter and input to the adder 76.

On the other hand, the reproduction signal x ₁ (n) input to the input unit 71 is a second signal composed of delay elements 72 _{1 to} 72 _L−1 , multipliers 73 _{0 to} 73 _L−1, and an adder 74. Are equalized by the FIR filter and input to the multiplier 75. The multiplier 75 multiplies a coefficient α ₀ corresponding to the amount of interference from the signal of the adjacent recording track.

Then, the adder 76 adds the signal equalized by the first FIR filter and the signal equalized by the second FIR filter, and the operation result is the output y ₀ (n) of the equalizer 23. It becomes. That is, the equalizer 23 outputs y ₀ (n) represented by the above equation (10).

  Next, a method for calculating the coefficients of the FIR filter in the adaptive equalization controllers 33 and 34 will be described. In the following description, bit synchronization is assumed for simplicity. That is, the phase error is zero.

  The frequency response F of the FIR filter expressed by the above equation (11) can be expressed by the following equation (12).

  Here, T is a sampling interval, and hereinafter T = 1. Since the response of the half Nyquist frequency is f = 1 / 4T, the above equation (12) is expressed by the following equation.

  Here, when the filter coefficient is set as equation (14), the above equation (13) is expressed by equation (15).

  Since the phase change is not considered in the FIR filter, Expression (16) is satisfied.

  Further, since the gain g is constant, the equation (17) is satisfied.

From Eqs. (16) and (17), constraints on V _r and V _i are given. Therefore, V _r and V _i are set as in equation (18), and the matrix P orthogonal to the matrix V is multiplied by the error signal vector e _k (p) and then updated, thereby satisfying these constraint conditions. The error signal vector e _k (p) is calculated by the error calculators 29 and 30.

  The matrix P is represented by the following equation. Here, I is a unit matrix.

At this time, the update of the filter coefficient is given by the following equation. In addition, u _k (p) is a filter input signal vector, and μ is an update gain.

  The adaptive equalization controllers 33 and 34 input the filter coefficients calculated in this way to the equalizers 23 and 24.

Next, the reproduction signal gain control in the gain controllers 31 and 32 will be described. The gain controller 31 calculates a gain and an equalization error with respect to the reproduction signal of the recording track 41a reproduced by the reproduction head 51a. Further, the gain controller 32 calculates gain and equalization error for the reproduction signal of the recording track 41b reproduced by the reproduction head 51b. If the root mean square of the constraint condition expressed by the above equation (17) is taken, the variance of the signal component is σ _s ² , and the variance of the noise component is σ _w ² , equation (21) is obtained.

Therefore, even if the gain controllers 31 and 32 try to bring the amplitude value closer to a desired value, the signal component becomes (g ² −σ with increasing noise of the input signal in order to satisfy the constraint of the equation (17). _w ² ) The signal is reduced by a factor of ^½, and the signal to noise and noise ratio (SDNR), which is the signal-to-noise power ratio, deteriorates.

  Therefore, an evaluation function for calculating the amplitude error e (p) of the gain controllers 31 and 32 is represented by equation (22).

Here, d (p) is a desired response, and z (p) is an output of the ITRs 25 and 26. Also, (g ² + σ _w ² ) / g ² is the power when there is no noise when the gain g on the right side is increased by the noise power σ _w ² in the constraint condition of the equation (17). This is a standardized value. By multiplying this by the outputs of the ITRs 25 and 26, the gain g in the equation (17) is increased by the amount of noise. By not changing the gain of the signal component in this way, it is possible to prevent the degradation of SDNR.

Further, the gain controllers 31 and 32 calculate the coefficient σ _w ² . Σ _w ² that is the power of the noise component can be expressed by equation (23) using λ as a constant and a square error average value MSE (Mean Square Error). Note that the constant λ is λ = 2 when equalized to PR4, for example. In addition, the noise component can be exemplified by medium noise, head thermal noise, circuit thermal noise, current noise, code interference noise, and the like because interference in adjacent tracks is removed in an ideal state. .

  Furthermore, the followability can be provided by updating the MSE at a certain sample interval.

  Next, the configuration of the gain controllers 31 and 32 described above will be described. Since the gain controller 31 and the gain controller 32 have the same configuration, the gain controller 31 will be described here.

  FIG. 4 is a block diagram illustrating a configuration example of the gain controller 31. The gain controller 31 includes a slicer 81 that calculates a desired response d (n) of the reproduction signal y (n), and an adder 82 that calculates the difference between the reproduction signal y (n) and the slicer output d (n). An absolute value circuit 83 for calculating the absolute value of the reproduction signal y (n), an absolute value circuit 84 for calculating the absolute value of the slicer output d (n), an MSE calculation unit 85 for calculating the coefficient of y (n), y ( a multiplier 86 for multiplying a coefficient of n) by y (n), an adder 87 for calculating a difference between d (n) and the output of the multiplier 86, a gain storage unit 88 for storing a loop gain, and an output of the adder 87 89, the delay element 90 that delays the input by one clock, the adder 91 that calculates the difference between the output from the delay element 90 and the output of the multiplier 89, and the output of the delay element 90 are analog. And a DA converter 92 for converting the signal.

The slicer 81 makes a tentative determination as to which symbol {-1, 0, 1} the reproduction signal y (n) corresponds to. The slicer 81 has two threshold values, a positive threshold value Th ₊ and a negative threshold value Th ₋ , and outputs “1” when y (n) exceeds Th ₊ , and y (n) becomes Th _−. When it is lower, “−1” is output, and when y (n) is between Th ₊ and Th ₋ , “0” is output. The output of the slicer 81 is a desired value d (n) of y (n).

  The adder 82 calculates an error signal from the reproduction signal y (n) and the output d (n) of the slicer 81. The adder 82 outputs the calculated error signal to the MSE calculator 85.

The MSE calculator 85 calculates the MSE based on the error signal and multiplies the MSE by a constant λ. λMSE corresponds to the variance σ _w ² of the noise component as shown in the above equation (23). Further, the MSE calculator 85 calculates the coefficient (g ² + λMSE) / g ² of the reproduction signal y (n) in the above equation (22) based on the constant g and λMSE, and outputs it to the multiplier 86.

The multiplier 86 multiplies the absolute value | y (n) | of the reproduction signal y (n) by the output (g ² + λMSE) / g ^{2 of the} MSE calculator 85 and outputs the calculation result to the adder 87. The adder 87 takes the difference between the output | d (n) | from the absolute value circuit 84 and the output (g ² + λMSE) / g ² · | y (n) | from the multiplier 86 to obtain an amplitude error e (n). Is calculated.

  The output e (n) from the adder 87 is multiplied by the loop gain by the multiplier 89. The calculation result is sent to a one-tap IIR (Infinite Impulse Response) loop filter constituted by an adder 91 and a delay element 90. The adder 91 calculates the difference between the amplitude error e (n) one clock before output from the delay element 90 and the output of the multiplier 89 and outputs the difference to the delay element 90. The delay element 90 outputs the amplitude error e (n) to the DA converter 92. The DA converter 92 sends a gain control signal obtained by converting and amplifying the amplitude error e (n) to an analog signal to the variable gain amplifier 15. Thus, the noise reproduction characteristic can be improved by increasing the gain g by the amount of noise.

  FIGS. 5 and 6 show the influence of interference from adjacent recording track signals on the bit error rate in a data reproducing apparatus to which the present invention is applied and a conventional data reproducing apparatus that does not have a function of suppressing interference between adjacent recording tracks. FIG. Here, each data reproducing apparatus performs PR4ML detection (Partial Response Class 4 Maximum Likelihood Sequence Detection) generally used in magnetic recording, and uses 16/17 as a recording code. The user linear density was 2.0.

  As can be seen by comparing the results shown in FIGS. 5 and 6, the data reproducing apparatus to which the present invention is applied exhibits a low BER (Bit Error Rate) with respect to the off-track ratio.

  From this, it can be seen that interference from adjacent recording tracks is sufficiently suppressed.

As described above, the data reproducing apparatus to which the present invention is applied equalizes with the frequency response fixed and calculates the amplitude error e (n) by using the reproduction signal y (n) as (g ² + λMSE) / g ^2-fold. That is, the noise reproduction characteristic can be improved by fixing the frequency response, equalizing the gain and the phase, and increasing the gain g by the amount of added noise. Further, it is possible to improve the characteristics of the ITRs 25 and 26, which are the dominant factors of the error characteristics, and to have followability.

  It should be noted that the present invention is not limited to the above-described embodiments, and various modifications can be made without departing from the scope of the present invention. For example, in the above embodiment, the width of the reproducing head is the same as the width of the recording track, but the present invention is not limited to this.

  In the above embodiment, the two recording tracks are read and equalized by the two reproducing heads. However, if the reproducing signals of all the reproducing heads reading one recording track are equalized. Good.

  In the above embodiment, the frequency response of the preamble is fixed. However, stable characteristics can be obtained without using the data reproducing apparatus and preamble to which the present invention is applied. For example, a preamble is not recorded on an optical disc recording medium such as a CD (Compact Disc) or a DVD (Digital Versatile Disc). However, if an appropriate frequency response is fixed, the recording medium is also applicable to reproduction of these recording media. be able to.

It is a block diagram which shows the structure of the data reproduction apparatus which concerns on one embodiment of this invention. It is a figure which shows the relationship between the recording track on a recording medium, and a group. It is a block diagram which shows the structural example of an equalizer. 3 is a block diagram illustrating a configuration example of a gain controller 31. FIG. It is a figure which shows the influence which the interference from the signal of an adjacent recording track has on a bit error rate in the data reproducing device to which this invention is applied. It is a figure which shows the influence which the interference from the signal of an adjacent recording track has on a bit error rate in the conventional data reproducing apparatus which does not have the function to suppress the interference of an adjacent recording track. It is a block diagram which shows the structure of the conventional data reproduction apparatus.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 10 Data reproduction | regeneration apparatus, 11,12 input part, 13,14 reproduction | regeneration amplifier, 15,16 variable gain amplifier, 17,18 low-pass filter, 19,20 converter, 21,22 gain regulator, 23,24 equalizer , 25, 26 ITR, 27, 28 detector, 29, 30 error calculator, 31, 32 gain controller, 33, 34 adaptive equalization controller, 41 group, 41 recording track, 42 group, 43 guard band, 51 Multi-play head, 61 input unit, 64 adder, 71 input unit, 74 adder, 75 multiplier, 76 adder,

Claims (6)

In a data reproducing apparatus, N _w recording tracks on a recording medium are read out by N _r reproducing heads (N _w ≦ N _r ), and the data of N _w recording tracks are separately extracted,
A signal amplifier for amplifying N _r pieces of the amplitude of the N _r number of reproduction signal read from the reproduction head, respectively,
A quantizing unit for quantizing the amplified N _r reproduction signals,
An equalization unit which equalizes a filter, and outputs the N _w number of output signals in a state in which the N _r number of reproduction signal the quantized fixed frequency response,
A synchronization unit for respectively performing bit synchronization of the equalized N _w output signals;
A detection unit for detecting each of the N _w output signals having the bit synchronization;
A gain control unit for controlling the respective amplitudes of the reproduction signal read out from the N _r number of the reproducing head based on the detection result of the detection unit,
A data reproduction apparatus comprising: an adaptive equalization control unit that calculates filter coefficients of the filter based on the bit-synchronized output signal and the detection result of the detection unit.   2. The data reproduction apparatus according to claim 1, wherein the gain control unit performs control so as to correct a decrease in signal power due to noise included in the reproduction signal.   3. The data according to claim 2, wherein the gain control unit uses a square error average value (MSE) for calculating an amplitude error between the bit-synchronized output signal and the detection result of the detection unit. Playback device. 4. The data according to claim 3, wherein the gain control unit multiplies the output signal by a coefficient (g ² + λMSE) / g ^{2 including} a gain g in a state where the frequency response is fixed and a constant λ. Playback device.   4. The data reproducing apparatus according to claim 3, wherein the gain adjusting unit updates the MSE at a predetermined sample interval. The N _w pieces of recording tracks on a recording medium, N _r pieces of playback head reads _{(N w ≦ N r),} N w pieces of recording tracks of data in the data reproducing method for extracting and separating respectively,
A signal amplifying step of amplifying N _r pieces of the amplitude of the N _r number of reproduction signal read from the reproduction head, respectively,
A quantization step of quantizing each of the amplified N _r reproduction signals;
An equalization step of equalizing by a filter, and outputs the N _w number of output signals in a state in which the N _r number of reproduction signal the quantized fixed frequency response,
A synchronization step of respectively performing bit synchronization of the equalized N _w output signals;
A detection step of detecting each of the N _w output signals in which the bit synchronization is achieved;
A gain control step of controlling the respective amplitudes of the reproduction signal read out from the N _r number of the reproducing head based on the detection result in the detecting step,
A data reproduction method comprising: an adaptive equalization control step of calculating filter coefficients of the filter based on the output signal in which the bit synchronization is achieved and a detection result in the detection step.

Images