A method for detecting a readback signal sample sequence during optical disk data reading

Abstract

The present invention belongs to the field of data storage, and specifically relates to a method for detecting a readback signal sample sequence in a process of reading optical disc data, comprising: performing equalization filtering processing on the signal-to-signal crosstalk in the readback signal sample sequence; predicting and eliminating the remaining noise in the readback signal sample sequence after the equalization filtering processing to obtain a noise-eliminated readback signal sample sequence; performing sequence detection on the noise-eliminated readback signal sample sequence to obtain a detection result, wherein a state transition grid for sequence detection is constructed by an equivalent complete channel partial response target coefficient vector t including three processes of equalization filtering, noise prediction and elimination, t is represented by the difference between g and gp, g represents a preset partial response target coefficient vector, and p represents a preset prediction tap coefficient vector for prediction. The present invention can avoid the problem of high computational complexity caused by the inclusion of noise prediction and elimination operations in branch metrics.

Description

Technical Field

[0001] The present invention belongs to the field of data storage, and more specifically, relates to a method for detecting a readback signal sample sequence in a process of reading optical disc data. Background Art

[0002] The development of new technologies such as big data, 5G communications, and artificial intelligence has led to an explosive growth in the total amount of global data, and a sharp increase in the demand for data storage. According to the frequency of data access, data can be divided into hot data, warm data, and cold data, of which warm and cold data accounts for the vast majority of the total amount of data. Warm and cold data has become an important digital oil asset in the era of artificial intelligence. Magnetic storage and optical storage are important solutions for long-term data storage due to their advantages in capacity-price ratio and data reliability. Partial response maximum likelihood sequence detection technology has been widely used in various types of warm and cold data storage devices, such as hard disk drives (HDDs), tapes, and optical discs.

[0003] As the recording density of data storage devices increases, the noise interference in the data reading process of the data storage system becomes more and more serious. The readback signal contains various colored noises such as intersymbol interference (ISI), intertrack interference (ITI) and interlayer interference (ILI). In addition, the readback signal is also affected by various additive white Gaussian noises (AWGN) of mechanical and electronic systems. In order to ensure the reliable reading of storage system data, balanced denoising technology can effectively eliminate various signal crosstalk and noise.

[0004] The Partial Response Maximum Likelihood (PRML) detection technology has become a common technical standard for HDD and Tape, and has the best detection performance in the AWGN channel. Common detection algorithms include Maximum Likelihood Sequence Detection (MLSD) and Maximum A Posteriori (MAP) detection. The best algorithm implementation of MLSD is the Viterbi algorithm (A.Viterbi, "Error bounds for convolutional codes and an asymptoticallyoptimum decoding algorithm," in IEEE Transactions on Information Theory, vol.13, no.2, pp.260-269, April 1967, doi:10.1109 / TIT.1967.1054010.), and the best implementation of MAP is the BCJR algorithm (L.Bahl, J.Cocke, F.Jelinek and J.Raviv, "Optimal decoding of linear codes for minimizing symbol error rate (Corresp.)," in IEEE Transactions on Information Theory, vol.20, no.2, pp.284-287, March 1967). 1974, doi:10.1109 / TIT.1974.1055186.). Due to various noise interferences, it is difficult to directly recover the original information from the noisy sample signal, and it is usually necessary to perform equalization and denoising before detection. Existing equalization and denoising technologies mainly use equalizers and noise predictors to process the readback signal samples, among which the equalizer is mainly used to eliminate the crosstalk between signals, while the noise predictor is mainly used to process the fine-grained stray colored noise contained between signals.

[0005] Jonathan D. Coker et al. designed a noise prediction maximum likelihood (NPML) sequence detection technology based on the noise predictor (JD Coker, E. Eleftheriou, RLGalbraith and W. Hirt, "Noise-predictive maximum likelihood (NPML) detection," in IEEE Transactions on Magnetics, vol. 34, no. 1, pp. 110-117, Jan. 1998, doi: 10.1109 / 20.663468.), which embeds the process of noise prediction and whitening into the calculation of the detector branch parameters, and realizes noise prediction and elimination through the feedback random read-only memory lookup table. However, NPML detectors usually require complex numerical calculations, especially when processing high-dimensional data, the computational complexity will increase significantly.

[0006] Therefore, existing methods embed the noise prediction process into the calculation process of the detector branch metric, which increases the branch calculation overhead. There is an urgent need for a signal detection method with low computational overhead based on balanced denoising. Summary of the invention

[0007] In view of the defects and improvement requirements of the prior art, the present invention provides a method for detecting a readback signal sample sequence in a process of reading data from an optical disc, the purpose of which is to reduce the computational complexity of branch parameters in a signal detector.

[0008] To achieve the above object, according to one aspect of the present invention, a method for detecting a readback signal sample sequence in a process of reading data from an optical disc is provided, comprising:

[0009] S1, performing equalization filtering on the crosstalk between signals in the readback signal sample sequence;

[0010] S2, predicting and eliminating the remaining noise in the readback signal sample sequence after the equalization filtering process, to obtain a noise-eliminated readback signal sample sequence;

[0011] S3. Perform sequence detection on the noise-eliminated readback signal sample sequence to obtain a detection result, wherein a state transition grid used for sequence detection is constructed by an equivalent complete channel partial response target coefficient vector t including three processes of equalization filtering, noise prediction and elimination, and the equivalent complete channel partial response target coefficient vector t is expressed as the difference between g and gp, g represents a preset partial response target coefficient vector, and p represents a preset prediction tap coefficient vector used for the prediction.

[0012] Furthermore, the equalization filtering process is performed by using a partial response equalization or an adaptive equalization method.

[0013] Further, the prediction tap coefficient vector p used for the prediction is constructed in the following manner:

[0014]

[0015] In the formula, R w (m) represents the autocorrelation matrix of the noise sequence w with a delay of m. The noise sequence w is the difference between the equalized output y and the ideal output d based on the partial response target coefficient vector without noise: w = yd, Represents convolution calculation, and a represents known encoded writing data.

[0016] The present invention also provides a signal detection method based on decision feedback in a process of reading optical disc data, comprising:

[0017] Using the above-mentioned method for detecting a readback signal sample sequence in a process of reading data from an optical disc, a sequence detection is performed on a current readback signal sample sequence to obtain a detection result;

[0018] Based on the detection result, respectively update the equalization tap coefficient vector f used to perform the equalization filtering process and the prediction tap coefficient vector p used to perform the prediction;

[0019] Based on the updated equalization tap coefficient vector f and the prediction tap coefficient vector p, performing the sequence detection on the next readback signal sample sequence;

[0020] The above process is repeated until the detection of all readback signal sample sequences is completed.

[0021] Further, the equalization tap coefficient vector f is determined based on the ideal channel output and the equalization filter output;

[0022] Wherein, when performing the initial sequence detection, the ideal channel output is obtained by convolving the known coded write data a and the partial response target coefficient vector g;

[0023] When performing subsequent sequence detection, the ideal channel output is obtained by convolving the currently obtained sequence detection output and the partial response target coefficient vector g.

[0024] Further, the prediction tap coefficient vector p is obtained by the autocorrelation of a noise sequence constituted by the difference between the equalization filter output and the ideal channel output;

[0025] Wherein, when performing the initial sequence detection, the ideal channel output is obtained by convolving the known coded write data a and the partial response target coefficient vector g;

[0026] When performing subsequent sequence detection, the ideal channel output is obtained by convolving the currently obtained sequence detection output and the partial response target coefficient vector g.

[0027] The present invention also provides a computer-readable storage medium, which includes a stored computer program, wherein when the computer program is executed by a processor, the device where the storage medium is located is controlled to execute the above-mentioned method for detecting a readback signal sample sequence in a CD data reading process and / or the above-mentioned method for detecting a signal based on decision feedback in a CD data reading process.

[0028] The present invention also provides an optical disc data reading device, characterized in that it comprises: a processor, a transceiver, and the computer-readable storage medium as described above, wherein:

[0029] The transceiver is used to send and receive data under the control of the processor;

[0030] When the processor executes the computer program on the computer-readable storage medium, the processor implements the above-mentioned method for detecting a readback signal sample sequence in a process of reading optical disc data and / or the above-mentioned method for detecting a signal based on decision feedback in a process of reading optical disc data.

[0031] In general, the above technical solutions conceived by the present invention can achieve the following beneficial effects:

[0032] (1) In the present invention, when performing readback signal sample sequence detection in the process of optical disk data reading, the prediction and elimination of noise are divided into two levels. First, the partial response equalization or adaptive equalization method is used to process the signal crosstalk contained in the readback signal sample sequence to effectively reduce the signal crosstalk. Then, the remaining noise in the readback sample sequence after equalization filtering is predicted and eliminated. After that, the sequence detection operation is performed independently of the noise elimination, wherein the target coefficient vector used for state transition grid construction is the target coefficient vector corresponding to the noise prediction and elimination steps deducted, that is, the state transition grid used for sequence detection is constructed by the equivalent complete channel partial response target coefficient vector t including the three processes of equalization filtering, noise prediction and elimination. The equivalent complete channel partial response target coefficient vector t is expressed as the difference between g and gp, expressed as g(1-p), g represents the preset partial response target coefficient vector, and p represents the preset prediction tap coefficient vector for noise prediction. This operation mode avoids the problem of high computational complexity caused by the inclusion of noise prediction and elimination operations in the branch metric. The method of the invention can be used in the data reading process of a high-density optical disk storage system, and can handle the requirements of crosstalk and noise under the condition of high recording density.

[0033] (2) When the present invention detects the readback signal sample sequence during the optical disk data reading process, the detector's decision output feedback is used to calculate the prediction tap coefficient vector for noise prediction, adaptively matching the signals in different storage areas of the memory and predicting the noise, which is conducive to alleviating the signal noise caused by memory medium defects and contamination. The noise predictor coefficient calculated through the feedback process is used to determine the optimal state transition grid, further improving the reliability of maximum likelihood sequence detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 A flowchart of a method for detecting a readback signal sample sequence in a process of reading data from an optical disc provided by an embodiment of the present invention;

[0035] Figure 2 A schematic diagram of equalizer equalization filtering provided by an embodiment of the present invention;

[0036] Figure 3 A schematic diagram of a noise prediction and noise elimination method provided by an embodiment of the present invention;

[0037] Figure 4 A schematic diagram of an initialization equalization tap coefficient vector and a prediction tap coefficient vector provided by an embodiment of the present invention;

[0038] Figure 5 A state transition grid diagram when a partial response target coefficient vector with a length of 5 and a 17PP coding constraint is used according to an embodiment of the present invention;

[0039] Figure 6 A structural diagram of a data reading method of an optical disc read-back system provided by an embodiment of the present invention;

[0040] Figure 7 A flow chart of signal equalization, noise prediction elimination and signal detection based on decision feedback in the optical disc data reading process provided by an embodiment of the present invention;

[0041] Figure 8 A schematic diagram of signal equalization, noise prediction elimination and detection in practical applications provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0042] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. In addition, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0043] Embodiment 1

[0044] A method for detecting a readback signal sample sequence during a CD data reading process, such as Figure 1 As shown, including:

[0045] S1, performing equalization filtering on the crosstalk between signals in the readback signal sample sequence;

[0046] S2, predicting and eliminating the remaining noise in the readback signal sample sequence after the equalization filtering process, to obtain a noise-eliminated readback signal sample sequence;

[0047] S3. Perform sequence detection on the noise-eliminated readback signal sample sequence to obtain a detection result, wherein a state transition grid used for sequence detection is constructed by an equivalent complete channel partial response target coefficient vector t including three processes of equalization filtering, noise prediction and elimination, and the equivalent complete channel partial response target coefficient vector t is expressed as the difference between g and gp, g represents a preset partial response target coefficient vector, and p represents a preset prediction tap coefficient vector for noise prediction.

[0048] In the method of this embodiment, the execution body of step S1 can be called an equalizer. Assuming the readback signal sample sequence r and the equalization tap coefficient vector f used for equalization filtering processing, the process of equalization filtering processing is to convolve the readback signal sample sequence with the equalization tap coefficient vector and output y, which is expressed as: like Figure 2 shown.

[0049] The execution subject of step S2 can be called a noise predictor, which predicts the noise n contained in the equalization output y and eliminates the predicted noise. The prediction tap coefficient vector (tap coefficient comes from English taps) p used for noise prediction, the process of noise prediction is to convolve the equalization output y with the prediction tap coefficient vector p and output the predicted noise n, which is expressed as: like Figure 3 shown.

[0050] The above noise elimination method is to eliminate the corresponding prediction noise from the equalization output y, and the resulting signal after eliminating the prediction noise is: In the formula, y n represents the nth element in y, y′ n Represents the nth element in the signal y′ after eliminating the prediction noise.

[0051] The method of this embodiment can be used in the data reading process of a high-density optical disk storage system, and can handle the requirements of crosstalk and noise under high recording density. The prediction and elimination of noise are divided into two levels, that is, the method of this embodiment has two levels of denoising processing. First, the crosstalk between signals contained in the readback signal sample sequence is processed by using a partial response equalization or adaptive equalization method to effectively reduce the crosstalk between signals, and then the residual noise in the readback sample sequence after equalization filtering is predicted and eliminated.

[0052] It should be noted that before actually executing the method of this embodiment, it is necessary to pre-initialize the equalization tap coefficient vector and the prediction tap coefficient vector. The specific method can be: after constrained encoding, the pre-generated pseudo-random data is input into the simulated optical disc readback channel to simulate the acquisition of the readback signal sample; the present invention does not limit the construction method of the simulated optical disc readback channel, and all the methods for calculating the initialization tap coefficients of the present invention shall be regarded as the invention content of the present invention. Further, based on the known coded write data a and the given partial response target coefficient vector g, the known coded write data and the partial response target coefficient vector g are convolved to obtain the ideal channel output, and the equalization tap coefficient vector is obtained by minimizing the mean square error or minimizing the mean square of the equalization output and the ideal channel output, and the prediction tap coefficient vector is calculated according to the difference between the equalization output and the ideal channel output, so as to realize the initialization of the equalization tap coefficient vector and the prediction tap coefficient vector.

[0053] Based on the detection result of the readback signal sample sequence r corresponding to the known coded write data a and the known coded write data a, the bit error rate (BitErrorRate, BER) obtained is compared to optimize and adjust the adopted partial response target coefficient vector g (including adjusting the coefficient value or vector length), select the partial response target coefficient vector g that makes the BER lowest, and under this condition, calculate the equalization tap coefficient vector f, f=[f1,f2,…,f N ] T , and the prediction tap coefficient vector p, p = [p1, p2, ..., p M ] T As the initialization coefficient of the signal equalization detection method for noise prediction and elimination in the actual optical disc read-back process.

[0054] Regarding the determination of the equalization tap coefficient vector f, illustratively, it can be obtained by minimizing the mean square error or the least mean square criterion, which can effectively eliminate the mutual interference between signals in the readback signal sample sequence. The partial response equalization method and the adaptive equalization method are specifically introduced below.

[0055] 1. Partial Response Equalization Method

[0056] When executing the signal equalization detection method of this embodiment, when the partial response equalization method is used for equalization filtering processing, then, if Figure 4 As shown, the calculation process of the equalization tap coefficient vector f is:

[0057] Read the readback signal sample sequence r corresponding to the known original write signal a, using r k Represents the discrete readback signal corresponding to the kth bit in r, using y k represents the kth equalized tap coefficient f in the equalized tap coefficient vector f of the signal sequence composed of the discrete readback signals from the kKth bit to the k+Kth bit k The output obtained after equalization filtering; correspondingly, using d k Represents the known original write signal a k (i.e., known coded written data) passes through the kth partial response target coefficient g in the partial response target coefficient vector g k The ideal output signal obtained after filtering is:

[0058]

[0059] Among them, the length of the equalization tap coefficient vector f is 2K+1, f=[f -K ,…,f0,…,f K ] T ; g i is the i-th partial response target coefficient in the given partial response target coefficient vector, the length of the partial response target coefficient vector g is L, g = [g0, g0, …, g L-1 ] T .

[0060] Equalizer output k The ideal output d of the target channel with partial response k The mean square error between It is expressed as:

[0061]

[0062] In the formula, R r , R a are the autocorrelation matrix of the readback signal sample sequence r and the autocorrelation matrix of the original write sequence a, R ra is the cross-correlation matrix of r and a, and the readback signal sample sequence r is the sequence read by the sensor for the original write sequence a.

[0063] When calculating the equalization tap coefficient vector f and the partial response target coefficient vector g, the equalization error The constraint term 2λ(c Tg-1)=0, λ is the scalar to be determined, c is the known constraint vector, and the process of solving the equalization tap coefficient vector f and the partial response target coefficient vector g is avoided to obtain a trivial solution.

[0064] According to the constrained mean square error, the derivatives of λ, g, and f are obtained respectively:

[0065]

[0066] In particular, the partial response target coefficient vector g may not be updated, and the given partial response target coefficient vector g may be directly substituted into the calculation formula of f to calculate the equalization tap coefficient vector f.

[0067] (II) Adaptive equalization method

[0068] When executing the signal equalization detection method of this embodiment, when the adaptive equalization method is used for equalization filtering processing, then Figure 4 As shown, the adaptive update process of the equalization tap coefficient vector f is:

[0069] Given the initial adaptive equalization tap coefficient vector f, the equalized discrete readback signal r is equalized and filtered:

[0070]

[0071] e k =d k -y′ k

[0072] Among them, y′ k , They are the adaptive equalization output at time k, the adaptive equalization tap coefficient vector with a length of N at time k, and the readback sample sequence with a length of N at time k, where N represents the length of the adaptive equalization window, e k is the adaptive equalization error at time k, d k is the ideal partial response output at time k, specifically the original written signal a k The ideal output signal obtained after filtering by the partial response target coefficient vector g.

[0073] Exemplarily, the k+1 equalization tap coefficient vector update process at the next moment is: Wherein, μ is the learning rate. In particular, the adaptive equalizer update process can also be expressed in other equivalent variants, such as dynamic adjustment of the learning rate, or simplified error.

[0074] Further, when executing the signal equalization detection method of this embodiment, the execution subject of step S3 may be called a detector, which involves the use of a state transition grid. Then, the design and construction of the state transition grid required by the detector is specifically performed as follows:

[0075] The prediction tap coefficient vector p used for noise prediction is used to identify the noise in the equalized output y, and the prediction noise is removed by (1-p). The equivalent complete channel partial response target coefficient vector t including the three processes of equalization, noise prediction and elimination is calculated:

[0076] t=g(1-p)

[0077] According to the equivalent complete channel partial response target coefficient vector t and the binary symbol b (such as "0 and 1", "+1 and -1"), a state transition grid is constructed, and sequence detection is performed based on the state transition grid determined by the equivalent complete channel partial response target t. The construction method is as follows: if the length of the equivalent complete channel partial response target coefficient vector t is L t , then t can be expressed as A state vector s corresponding to each time k in the state transition grid k All by L t -1 symbol data representation, Among them, b k Arrives at s at time k k The input of the state, the output of the grid at this time (that is, the measurement result of the current branch parameter) o k =s k t, each input binary symbol b k+1 , the arrival state of k+1 at the next moment is The final state transfer grid is the state transfer grid of the equivalent complete channel.

[0078] For example, when the partial response target coefficient vector is g = [1, 2, 2, 2, 1], its length is 5, and the corresponding state transition grid is as follows: Figure 5 As shown, the number of possible states at each moment is 2 5-1=16. In particular, before data is written, a specific constraint code is usually required to prevent the generation of states that do not meet the constraint conditions. Therefore, when designing the state transition grid, some states and branches should be eliminated according to the constraint code used. When the 17PP code (Miyauchi, Toshiyuki, et al. "Soft-output decoding of 17parity preserve / prohibit repeated minimum transitionrunlength(PP)code." Japanese journal of applied physics 43.7S(2004):4882.) constraint is used, since it restricts the bit pattern of some records, 10 state nodes remain after removing the restricted state nodes and branches.

[0079] Based on the state transition grid of the equivalent complete channel, the maximum likelihood sequence detection or maximum a posteriori probability detection algorithm is used to perform sequence detection of the original written data. The method of this embodiment improves the state transition grid required by the detector, which is designed by the equivalent complete channel partial response target coefficient vector of the three processes of equalization, noise prediction and elimination, avoiding the problem of increasing the calculation complexity of the detector branch metric due to the need to eliminate the prediction noise.

[0080] In general, a state transition grid is used to implement sequence detection. The state transition grid is determined by the product vector of the prediction tap coefficient vector p and the given partial response target coefficient vector g. The product vector is expressed as: g(1-p). Since the method of this embodiment separates the noise prediction and elimination operation from the branch metric operation and operates in steps, the prediction and elimination operation is in step S2, and the branch metric operation is in step S3. Therefore, the target coefficient vector used for state transition grid construction is the target coefficient vector corresponding to the noise prediction and elimination step minus, that is, the state transition grid used for sequence detection is constructed by the equivalent complete channel partial response target coefficient vector t including the three processes of equalization filtering, noise prediction and elimination. The equivalent complete channel partial response target coefficient vector t is expressed as the difference between g and gp, expressed as g(1-p), g represents the preset partial response target coefficient vector, p represents the preset prediction tap coefficient vector for noise prediction, and this operation mode avoids the problem of high computational complexity caused by the inclusion of noise prediction and elimination operations in the branch metric. The method of this embodiment can be used in the data reading process of a high-density optical disk storage system, and can handle the requirements of crosstalk and noise under high recording density.

[0081] As a preferred implementation, the prediction tap coefficient vector p for residual noise prediction is constructed in the following manner:

[0082]

[0083] In the formula, R w (m) represents the autocorrelation matrix of the noise sequence w with a delay of m. The noise sequence w is the difference between the equalized output y and the ideal output d based on the partial response target coefficient vector without noise: w = yd, Represents convolution calculation, a represents known coded write data, and in the decision feedback scenario, a can also represent the most recent sequence prediction output result.

[0084] The noise sequence w is defined as the difference between the equalizer output y and the ideal output d based on the partial response target coefficient vector without noise: w = yd;

[0085] The noise sequence corresponding to the delay of m bits is expressed by w k-m Represents that the autocorrelation R of the noise sequence w delayed by m w (m) is: R w (m) = E{w k w k-m}, where E{·} represents expectation.

[0086] Given that the length of the prediction tap coefficient vector p is M, the prediction tap coefficient vector p is calculated as:

[0087]

[0088] This prediction tap coefficient vector p only needs to be calculated and adjusted according to the input of the detection process, and the prediction noise is eliminated in time before detection.

[0089] Figure 6 A relatively complete system for reading optical disc data is shown. An optical pick-up detector obtains a readback electrical signal from an optical disc, and a weak readback signal is amplified by a preamplifier for subsequent processing. An analog filter is used to filter high-frequency and low-frequency signals except for the effective readback signal. The analog electrical signal sampling is converted into a quantized digital signal by an analog-to-digital (A / D) converter, and then the distorted signal envelope is compensated and adjusted by the automatic gain control (AGC). Then, an approximately unbiased readback sample of the corresponding position of the recording bit is obtained by the timing recovery (Timing recovery) technology. The processed readback signal sample is passed through an equalizer to eliminate crosstalk between signals, and then the noise predictor is used to predict and eliminate the remaining noise in the signal sample. Then, the detector performs sequence detection on the equalized denoised output signal, and the detection result is used for subsequent decoder processing to restore the original user data stored in the storage medium.

[0090] Therefore, the method of this embodiment can perform signal equalization, noise prediction and elimination on the optical disc data reading process in the optical disc storage system, and realize more universal sequence detection according to the equivalent complete channel partial response target coefficient vector, thereby realizing efficient reading of the optical disc data.

[0091] Embodiment 2

[0092] A signal detection method based on decision feedback in the process of reading optical disc data, such as Figure 7 As shown, including:

[0093] Using the method for detecting a readback signal sample sequence in an optical disc data reading process as described in the first embodiment above, a sequence detection is performed on the current readback signal sample sequence to obtain a detection result;

[0094] Based on the detection result, respectively update the equalization tap coefficient vector f used to perform the equalization filtering process and the prediction tap coefficient vector p used to perform the prediction;

[0095] Based on the updated equalization tap coefficient vector f and the prediction tap coefficient vector p, the above sequence detection is performed on the next readback signal sample sequence;

[0096] The above process is repeated until the detection of all readback signal sample sequences is completed.

[0097] This embodiment proposes an equalization denoising and noise prediction method based on detector decision feedback. The direct benefit of decision feedback is that it can adaptively calculate the equalizer and noise predictor coefficients according to real-time detection signal characteristics, thereby achieving more accurate noise equalization and prediction, which is beneficial to solving signal distortion caused by medium defects and impurity contamination during the operation of the optical disc medium.

[0098] As a preferred implementation, the equalization tap coefficient vector f is determined based on the ideal channel output and the equalization filter output;

[0099] Wherein, when performing the initial sequence detection, the ideal channel output is obtained by convolving the known coded write data a and the partial response target coefficient vector g;

[0100] When performing subsequent sequence detection, such as Figure 8 As shown, the ideal channel output is obtained by convolving the current sequence detection output and the partial response target coefficient vector g.

[0101] That is, the equalizer includes a partial response equalizer and an adaptive equalizer; wherein, the equalization tap coefficient vector of the partial response equalizer is determined based on the minimum mean square error of the ideal channel output and the equalizer output, and the equalization tap coefficient vector of the adaptive equalizer is determined based on the minimum mean square of the ideal output signal and the equalizer output; given a partial response target coefficient vector g, the ideal output signal is obtained by convolving the detector decision output a′ (known encoded write data) and the partial response target coefficient vector g.

[0102] As a preferred implementation, the prediction tap coefficient vector p is obtained by the autocorrelation of a noise sequence formed by the difference between the equalization filter output and the ideal channel output;

[0103] Wherein, when performing the initial sequence detection, the ideal channel output is obtained by convolving the known coded write data a and the partial response target coefficient vector g;

[0104] When performing subsequent sequence detection, such as Figure 8 As shown, the ideal channel output is obtained by convolving the current sequence detection output and the partial response target coefficient vector g.

[0105] That is, the prediction tap coefficient vector of the noise predictor is obtained by the autocorrelation of the noise sequence constituted by the difference between the equalized output and the ideal output; wherein, given the partial response target coefficient vector g, the ideal output signal is obtained by convolving the detector decision output a′ (or known encoded write data) and the partial response target coefficient vector g.

[0106] Embodiment 3

[0107] A computer-readable storage medium includes a stored computer program, wherein when the computer program is executed by a processor, the device where the storage medium is located is controlled to execute the above-mentioned method for detecting a readback signal sample sequence in a process of reading data from an optical disc and / or the above-mentioned method for detecting a signal based on decision feedback in a process of reading data from an optical disc.

[0108] The relevant technical solution is the same as that in Embodiment 1 and will not be described in detail here.

[0109] Embodiment 4

[0110] An optical disc data reading device, comprising: a processor, a transceiver, and the computer-readable storage medium as described above, wherein:

[0111] The transceiver is used to send and receive data under the control of the processor;

[0112] When the processor executes the computer program on the computer-readable storage medium, the processor implements the above-mentioned method for detecting a readback signal sample sequence in a process of reading optical disc data and / or the above-mentioned method for detecting a signal based on decision feedback in a process of reading optical disc data.

[0113] The relevant technical solutions are the same as those in Embodiment 1 and Embodiment 2 and will not be described in detail here.

[0114] It will be easily understood by those skilled in the art that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting a readback signal sample sequence during an optical disc data reading process, characterized in that: include: S1, performing equalization filtering on the crosstalk between signals in the readback signal sample sequence; S2, predicting and eliminating the remaining noise in the readback signal sample sequence after the equalization filtering process, to obtain a noise-eliminated readback signal sample sequence; S3. Perform sequence detection on the noise-eliminated readback signal sample sequence to obtain a detection result, wherein a state transition grid used for sequence detection is constructed by an equivalent complete channel partial response target coefficient vector t including three processes of equalization filtering, noise prediction and elimination, so as to avoid mixing noise prediction and elimination operations in branch metrics, and the equivalent complete channel partial response target coefficient vector t is expressed as the difference between g and gp, g represents a preset partial response target coefficient vector, and p represents a preset prediction tap coefficient vector for noise prediction.

2. The readback signal sample sequence detection method according to claim 1, characterized in that: The equalization filtering process is performed by using a partial response equalization or an adaptive equalization method.

3. The readback signal sample sequence detection method according to claim 1, characterized in that: The prediction tap coefficient vector p for noise prediction is constructed as follows: In the formula, R w (m) represents the autocorrelation matrix of the noise sequence w with a delay of m. The noise sequence w is the difference between the equalized output y and the ideal output d based on the partial response target coefficient vector without noise: w = yd, Represents convolution calculation, and a represents known encoded writing data.

4. A signal detection method based on decision feedback in the process of reading optical disc data, characterized in that: include: Using the method for detecting a readback signal sample sequence in an optical disc data reading process as claimed in any one of claims 1 to 3, performing sequence detection on a current readback signal sample sequence to obtain a detection result; Based on the detection result, an equalization tap coefficient vector f for performing an equalization filtering process and a prediction tap coefficient vector p for performing noise prediction are respectively updated; Based on the updated equalization tap coefficient vector f and the prediction tap coefficient vector p, performing the sequence detection on the next readback signal sample sequence; The above process is repeated until the detection of all readback signal sample sequences is completed.

5. The signal detection method according to claim 4, characterized in that: The equalization tap coefficient vector f is determined based on the ideal channel output and the equalization filter output; Wherein, when performing the initial sequence detection, the ideal channel output is obtained by convolving the known coded write data a and the partial response target coefficient vector g; When performing subsequent sequence detection, the ideal channel output is obtained by convolving the currently obtained sequence detection output and the partial response target coefficient vector g.

6. The signal detection method according to claim 4, characterized in that: The prediction tap coefficient vector p is obtained by the autocorrelation of a noise sequence formed by the difference between the equalization filter output and the ideal channel output; Wherein, when performing the initial sequence detection, the ideal channel output is obtained by convolving the known coded write data a and the partial response target coefficient vector g; When performing subsequent sequence detection, the ideal channel output is obtained by convolving the currently obtained sequence detection output and the partial response target coefficient vector g.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored computer program, wherein, when the computer program is executed by a processor, the device where the storage medium is located is controlled to execute a readback signal sample sequence detection method in a CD data reading process as described in any one of claims 1 to 3 and / or a decision feedback-based signal detection method in a CD data reading process as described in any one of claims 4 to 6.

8. An optical disc data reading device, characterized in that: include: A processor, a transceiver, and a computer-readable storage medium as claimed in claim 7, wherein: The transceiver is used to send and receive data under the control of the processor; When the processor executes the computer program on the computer-readable storage medium, it implements a readback signal sample sequence detection method in a CD data reading process as described in any one of claims 1 to 3 and / or a decision feedback-based signal detection method in a CD data reading process as described in any one of claims 4 to 6.

Images