A method for detecting a signal read in parallel from adjacent tracks in a vector magnetic recording

Abstract

The present application belongs to the technical field of data storage, in particular to the technical field of magnetic recording, and specifically relates to a method for detecting parallel read signals of adjacent tracks in vector magnetic recording, comprising: simultaneously recovering user data of two adjacent tracks by using the orthogonal complementary relationship between a magnetoresistance read head and a vector read head, obtaining a comprehensive read-back signal in a direction perpendicular to a recording plane and a comprehensive read-back signal in a direction horizontal to the recording plane of the two adjacent tracks, and respectively performing adaptive equalization filter processing to eliminate inter-symbol interference; performing joint detection on the two signals after equalization filter processing: M1 mode based on dynamic threshold three-level quantization and logic decision, or enhanced detection mode based on energy competition decision. The present application can realize parallel detection of two-track data, avoid the limitation that the existing multi-in single-out equalization detection cannot simultaneously recover multi-track signals, and does not need to seriously depend on a maximum likelihood sequence detector, and has low computational complexity.

Description

Technical Field

[0001] This invention belongs to the field of data storage, particularly magnetic recording technology, and more specifically, relates to a method for detecting parallel readout signals of adjacent magnetic tracks in vector magnetic recording. Background Technology

[0002] Improving the areal density of hard disk drives has always been a core goal in the field of magnetic recording. As perpendicular magnetic recording approaches the superparamagnetic limit, heat-assisted magnetic recording technology has been proposed to achieve higher bit density and track density. However, higher track density means narrower track width. Traditional single-read heads used for signal detection experience a significant decrease in signal-to-noise ratio under narrow tracks, and their detection performance cannot meet the requirements for high-reliability data reading.

[0003] For data readout, the mainstream technologies currently include Two-Dimensional Magnetic Recording (TDMR) and Multi-Sensor Magnetic Recording (MSMR). However, TDMR uses multiple read heads to read two-dimensional signals from adjacent tracks. Limited by its readout method, it can only distinguish between two writing states along the track direction, and cannot distinguish between the positive and negative polarity states of adjacent tracks across the track direction, resulting in an effective areal density of only about 79% of the theoretical value. MSMR uses multiple read heads to read signals from the same track, improving the signal-to-noise ratio through multi-input single-output fusion, but its goal is signal recovery from a single track and does not involve parallel detection across multiple tracks.

[0004] Recently, M. Fatih Erden et al. (M. Fatih Erden, Walter Eppler, Steve Granz, and Stephanie Hernandez, “Vector Recording: A New Read Process at Ultra HighTPIs,” in the 34th Magnetic Recording Conference (TMRC), Minneapolis, Minnesota, Jul. 2023, pp. 1–2.) proposed Vector Recording technology, the core idea of ​​which is to use a traditional magnetoresistive read head ( ) reads the combined magnetic field perpendicular to the surface of the medium, and the vector read head ( The magnetic field is read parallel to the surface of the medium, and both read heads cover the width of two adjacent tracks. (RH Victora et al., K. Hosen, MF Erden, and RH Victora, “Vector Recording: Advancing Areal Density in HAMR With Innovative Read Head Design,”) IEEE Trans. Magn. ( , vol. 62, no. 3, pp.1–8, Mar. 2026, doi: 10.1109 / TMAG.2025.3612279.) Further, an orthogonal read head structure was designed to reduce crosstalk between read heads. However, existing Vector Recording literature focuses on the physical design of the read heads and signal-to-noise ratio analysis, without providing a perspective from... / The specific signal processing algorithm for jointly decoding two tracks in the readback signal.

[0005] Furthermore, existing signal detection methods in magnetic recording readback channels (such as PRML channels, Viterbi detectors, Turbo equalization, etc.) are all single-input single-output or multi-input single-output architectures. The input is the readback signal of a single read head or a fused single-channel signal, and the output is the detection result of a single track. It is impossible to use the positive intercomplementary information of the two read heads in Vector Recording to simultaneously recover the signals of two tracks. Summary of the Invention

[0006] In view of the above-mentioned defects or improvement needs of the existing technology, the present invention provides a method for detecting the parallel read signals of adjacent magnetic tracks in vector magnetic recording. Its purpose is to provide a method for parallel detection of adjacent track writing data in real time and reliably in the face of crosstalk between high-density magnetic recording tracks.

[0007] To achieve the above objectives, according to one aspect of the present invention, a method for detecting parallel readout signals of adjacent tracks in vector magnetic recording is provided, comprising: Acquire the combined readback signal of two adjacent tracks perpendicular to the recording plane. and the integrated readback signal horizontally to the recording plane. ;right , Both readback signals are filtered separately; timed recovery is performed on both filtered readback signals, and the recovered readback signal sample sequence is synchronized with the recorded bit sample; based on the recovered two readback signal sample sequences, the following two-track joint detection is performed: M1 mode: compares the signal samples of each signal at each time step with a preset positive threshold. Perform a size comparison; when the sample is larger than... When the absolute value of the sample is less than a certain value, the sample is quantized into a symbol 'a'. When the sample is greater than a certain value, the sample is quantized into a symbol b. When the value is negative, the sample is quantized to the sign c; when Quantized to a or c and If the quantization is b, then determine the current two magnetic tracks. Polarity of the estimated write value corresponding to the next read point at any given time , Same polarity; when Quantized to b and If quantized to a or c, then determine , Opposite polarities; when When the average quantization is b, soft arbitration is initiated: if The absolute value is not less than The absolute value of then polarity as , The polarity, otherwise, polarity as The polarity will The opposite polarity as polarity; The corresponding track is obtained by subtracting the readout signals of the current two tracks. The track from which the readout signal is subtracted, and the other track is... The corresponding magnetic track; among which, , Representing respectively , middle The quantization symbol of the signal sample recovered at a given time; preset positive threshold. Satisfies: Based on the M1 mode, the polarity of each point on both tracks can be determined, and the values ​​are determined according to... , The features were determined; Alternatively, the M2 mode: i.e., the soft arbitration, if The absolute value is not less than The absolute value of then polarity as , The polarity, otherwise, polarity as The polarity of, will The antipolarity as The polarity of.

[0008] Furthermore, a magnetoresistive read head is used to read back the signal across two adjacent tracks to obtain a composite readback signal perpendicular to the recording plane; or, two magnetoresistive read heads are used to read back the signal from the two adjacent tracks respectively, and the two readback signals are added together to obtain a composite readback signal from the two adjacent tracks perpendicular to the recording plane. A vector read head is used to read back the signal across two adjacent tracks to obtain a combined readback signal horizontal to the recording plane; or, two magnetoresistive read heads are used to read back the signal from the two adjacent tracks respectively, and the two readback signals are subtracted to obtain a combined readback signal horizontal to the recording plane from the two adjacent tracks.

[0009] Furthermore, an adaptive equalizer is used to perform the filtering process.

[0010] Furthermore, the method for determining the preset positive threshold is as follows: according to The noise variance of the two tracks is determined accordingly. When the noise variances are the same, the positive threshold is... When the noise variances are not the same, the positive threshold In the formula, the lower bound of the ideal observation value in the integrated readback signal is... The ideal upper bound for observation values , and These are the gain coefficients of the two magnetic tracks, respectively. , Indicates according to The noise standard deviation of the upper bound of the integrated readback signal and the noise standard deviation of the lower bound of the integrated readback signal are determined by combining the two factors.

[0011] Furthermore, when using the M2 detection mode, the method also includes: Calculate at each time point and Difference between This serves as a confidence index for the M2 detection mode; the confidence index at each time point is less than a preset threshold. and , are denoted as low-confidence sample points; A first preset number of consecutive low-confidence sample points are merged into a low-confidence segment; for each low-confidence segment, based on the cache, a second preset number of sample points are extended before and after it, and a joint Viterbi sequence detector is run on the extended sample point sequence to obtain a new two-track decision result. Replace the decision results of the corresponding sample points obtained by the M2 detection mode with the new two-track decision results of each low-confidence segment that does not contain extended sample points; For confidence indices higher than a preset threshold and The decision result of the M2 detection mode is retained, and the complete dual-track detection sequence is finally output.

[0012] further, ,in, It is a constant to prevent division by zero.

[0013] According to another aspect of the invention, an electronic device is provided, including a memory and a processor, the memory storing a computer program, the processor executing the computer program to implement the steps of the method described above.

[0014] According to another aspect of the invention, a computer-readable storage medium is provided, the computer-readable storage medium including a stored computer program, wherein, when the computer program is run by a processor, it controls the device where the storage medium is located to perform the steps of the method described above.

[0015] According to another aspect of the invention, a computer program product is provided, comprising a computer program or instructions that, when executed by a processor, implement the steps of the method described above.

[0016] In summary, compared with the prior art, the technical solutions conceived by this invention have the following main advantages: 1. This invention proposes a method for detecting parallel readout signals of adjacent tracks in vector magnetic recording. Utilizing... Positive intercomplementary structure, acquiring the combined readback signal perpendicular to the recording plane from two adjacent tracks. and the integrated readback signal horizontally to the recording plane. By inputting a signal data sample at a given moment through either M1 or M2 mode, the system can output the detection results of the corresponding points in the two tracks, enabling real-time, parallel detection of the two tracks. This is a 2-input, 2-output multi-input multi-output detection method that breaks through the limitations of traditional multi-input, single-output architectures. Furthermore, the M1 and M2 detection modes employ symbol-level hard decision-making, eliminating the need for sequence detectors such as Viterbi, resulting in low computational resource overhead and enabling real-time detection. It is suitable for high-density magnetic records with inter-track crosstalk. In particular, the M2 detection mode has no threshold parameter, and the decision result is dynamically determined by the signal energy (i.e., the absolute value of the recovered signal sample), exhibiting better robustness.

[0017] 2. This invention further proposes a hybrid detection mode. Using the M2 detection results, a confidence index is calculated to identify regions where M2 decision-making may be flawed. The Viterbi sequence detector is activated only in these regions, while the fast decision result of M2 is directly applied in regions where the M2 confidence is sufficiently high. Combining fast M2 detection with accurate Viterbi detection significantly reduces computational complexity while maintaining near-Viterbi detection accuracy. Attached Figure Description

[0018] Figure 1 A flowchart illustrating a method for detecting parallel readout signals of adjacent magnetic tracks in vector magnetic recording, provided in an embodiment of the present invention. Figure 2 The present invention provides a block diagram of a vector magnetic recording read channel maximum likelihood partial response detection system, wherein (a) is a schematic diagram of the construction principle of the positive interactive complement integrated write data, and (b) is a mapping table of the original two adjacent track write data and the positive interactive complement integrated write data; Figure 3 This is a flowchart of the M1 detection mode provided in an embodiment of the present invention; Figure 4 This is a schematic diagram of vector magnetic recording readback provided in an embodiment of the present invention; Figure 5 This is a schematic diagram of vector magnetic recording readback of a conventional magnetoresistive read head array provided in an embodiment of the present invention; Figure 6 This is a flowchart of the M2 detection mode provided in an embodiment of the present invention; Figure 7 A flowchart of the hybrid detection mode provided in an embodiment of the present invention; Figure 8 The partial response target provided for the embodiments of the present invention is Mesh diagram of a two-track Viterbi detector; Figure 9 The experimental results of the detection bit error rate of the vector magnetic recording multi-track parallel readout signal detection method provided in the embodiments of the present invention under different signal-to-noise ratio conditions when the inter-symbol interference is low; Figure 10 The experimental results of the detection bit error rate of the vector magnetic recording multi-track parallel readout signal detection method provided in the embodiments of the present invention under different signal-to-noise ratio conditions when inter-symbol interference increases. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0020] Example 1 A method for detecting parallel readout signals of adjacent tracks in vector magnetic recording, such as... Figure 1 As shown, it includes: Acquire the combined readback signal of two adjacent tracks perpendicular to the recording plane. and the integrated readback signal horizontally to the recording plane. ;right , Both readback signals are filtered separately; timed recovery is performed on both filtered readback signals, and the recovered readback signal sample sequence is synchronized with the recorded bit sample; based on the recovered two readback signal sample sequences, the following two-track joint detection is performed: M1 mode: compares the signal samples of each signal at each time step with a preset positive threshold. Perform a size comparison; when the sample is larger than... When the absolute value of the sample is less than a certain value, the sample is quantized into a symbol 'a'. When the sample is greater than a certain value, the sample is quantized into a symbol b. When the value is negative, the sample is quantized to the sign c; when Quantized to a or c and If the quantization is b, then determine the current two magnetic tracks. Polarity of the estimated write value corresponding to the next read point at any given time , Same polarity; when Quantized to b and If quantized to a or c, then determine , Opposite polarities; when When the average quantization is b, soft arbitration is initiated: if The absolute value is not less than The absolute value of then polarity as , The polarity, otherwise, polarity as The polarity of, will The antipolarity as polarity; The corresponding track is obtained by subtracting the readout signals of the current two tracks. The track from which the readout signal is subtracted, and the other track is... The corresponding magnetic track; among which, , Representing respectively , middle Signal samples recovered at a set time; preset positive threshold Satisfies: Based on the M1 mode, the polarity of each point on both tracks can be determined, and the values ​​are determined according to... , The features were determined; Alternatively, the M2 mode: i.e., the aforementioned soft arbitration, if The absolute value is not less than The absolute value of then polarity as , The polarity, otherwise, polarity as The polarity of, will The antipolarity as The polarity of.

[0021] This embodiment is oriented towards vector magnetic recording, where the written data is a bipolar binary value. In the above method, symbols a, b, and c are different symbols, which can be numbers or letters, and are only used to represent ternary discrimination. This embodiment proposes two optional joint detection modes.

[0022] Among them, the M1 basic detection mode is a dynamic threshold three-level quantization joint logic decision.

[0023] First, such as Figure 2 As shown, adaptive equalization filtering is performed on the two integrated readback signals respectively.

[0024] Let the written data of two adjacent tracks be respectively and , This represents the index into which data is written, and its value is a bipolar binary number. After synchronization via magnetic media writing, the combined writing sequence for adjacent tracks corresponding to the magnetoresistive read head and the vector read head is as follows: and , , ,in and These are the gain coefficients for the two tracks. The expected channel outputs for the two read heads are... and From the Partial Response (PR) objective The results are obtained by convolving the combined write sequences of the two adjacent tracks respectively:

[0025]

[0026] The LMS adaptive equalizer (preferred) is used to process the readback signal. and Perform equalization processing to partially respond to the target, such as As an equilibrium objective, the partial response objective can be determined by publicly available algorithms (such as the minimum mean square error algorithm for generalized partial responses), and this invention does not limit it. Indicates the index of the convolution variable, i.e., the equilibrium coefficient. , All represent the equalization coefficients of adaptive learning. The equalizer outputs are calculated as follows:

[0027]

[0028] The error signal is calculated as follows:

[0029] Error signal and Used to update tap coefficients:

[0030] Adaptive step size , Determined by the largest eigenvalue of the autocorrelation matrix of the readback signal:

[0031] The above adaptive equalization filtering uses the LMS adaptive algorithm, which takes the convolution result of the partial response target and the training sequence as the expected response, and the error signal is used to update the equalizer tap coefficients after being amplitude-limited.

[0032] Then, the two filtered readback signals were processed. , All are recovered at regular intervals. The recovered readback signal sample sequence is synchronized with the recorded bit sample to obtain... , .

[0033] Then, the ternary joint logic decision is executed: first, the positive threshold is determined. It should be noted that a preset positive threshold is used. The value is based on , The characteristics are determined, and the value only needs to satisfy "accurately determine the polarity of each point on the two tracks based on the M1 mode". The M1 joint detection logic of this invention is original and feasible, and can perform real-time and reliable parallel detection of adjacent track writing data when facing crosstalk between high-density magnetic recording tracks. As for The specific value of is not explicitly limited in this invention.

[0034] For example, such as Figure 3 As shown, in a group , Below, corresponding settings for:

[0035] Let a be 2, b be 0, and c be -2, then the recovered readback signals at each time point... and Each state is quantized into a ternary state, as follows:

[0036]

[0037] Then based on the quantized state The combination of these factors makes a logical decision (essentially, it's about which probe's actual readback signal is more likely to be trusted). r ):when and Time judgment as The estimated write values ​​corresponding to the readback points under the index are of the same polarity. );when and It is judged as having opposite polarity ( ); otherwise, when and If the result is still ambiguous, then soft arbitration is initiated. The larger the signal amplitude contained in the absolute value, the higher the signal quality. Therefore, comparison... and The absolute value of the magnetic field determines the polarity of the magnetic field, meaning that the equilibrium result with the larger absolute value is more trusted. Specifically, if... The absolute value is not less than The absolute value of then polarity as , The polarity, otherwise, polarity as The polarity of, will The antipolarity as The polarity of. Among them, The corresponding track is obtained by subtracting the readout signals of the current two tracks. The track from which the readout signal is subtracted, and the other track is... The corresponding magnetic track.

[0038] about The corresponding track will now be further explained: In fact, for the integrated readback signal perpendicular to the recording plane... Optional, such as Figure 4 As shown, a magnetoresistive read head can be used to read back signals across two adjacent tracks to obtain... Or, such as Figure 5 As shown, two magnetoresistive read heads are used to read back signals from two adjacent tracks respectively. The two read-back signals are then added together to obtain the result. For the integrated readback signal horizontally to the recording plane direction ,like Figure 4 As shown, a vector read head can be used to read back the signal across two adjacent tracks to obtain... Or, such as Figure 5 As shown, two magnetoresistive read heads are used to read back signals from two adjacent tracks respectively. The two read-back signals are subtracted to obtain... .

[0039] When using two independent magnetoresistive read heads to read back two tracks, due to It is clearly obtained by subtracting the two readback signals. Therefore, the track corresponding to the subtracted readout signal is taken as the first track. Figure 4 and Figure 5 T1 track in the middle, that is The corresponding track. When the vector read head traverses two adjacent tracks to read back the signal, it obtains... Since a vector read head is essentially manufactured by integrating independent read head components, the signal output depends on which track the subtrahend of the subtraction logic is located on; that track is the first track. Therefore, it is essentially based on subtracting the read signals from the two tracks. The track from which the readout signal is subtracted is designated as the first track, which is also the... Figure 4 and Figure 5 T1 track in the middle, that is The corresponding magnetic track.

[0040] In addition, in the Vector Recording system, this embodiment utilizes a conventional magnetoresistive read head. and vector read heads The positive interactive complementary characteristics are used to obtain the comprehensive readback signal. , It is used to achieve parallel detection of data from two adjacent magnetic tracks.

[0041] The concept of a preset positive threshold can be further elaborated upon. Under the aforementioned principle, namely the preset positive threshold... The value is based on , The characteristics are determined, and the value only needs to satisfy "the polarity of each point on the two tracks can be accurately determined based on the M1 mode". Further, this can be used as a preferred determination method. The preset positive threshold value can be determined as follows: according to The noise variance of the two tracks is determined accordingly. When the noise variances are the same, the positive threshold is... When the noise variances are not the same, the positive threshold In the formula, the lower bound of the ideal observation value in the integrated readback signal is... The ideal upper bound for observation values , and These are the gain coefficients of the two magnetic tracks, respectively. , Indicates according to The noise standard deviation of the upper bound of the integrated readback signal and the noise standard deviation of the lower bound of the integrated readback signal are determined by combining the two factors.

[0042] The theoretical explanation for this determination method is as follows: Ideally, , There are only two observed values, meaning the ideal level has only two amplitudes. (Reverse polarity) (Same polarity), ternary quantization is a two-level hypothesis test. The maximum likelihood detection boundary when the noise variance (standard case) of the two tracks is the same is the midpoint between the two levels. The error rates on both sides are equal, and the judgment margin is also equal. Validity conditions ; Time degenerates into When the noise variance is not uniform (noise varies with level, such as due to media jitter), the standard deviation of the two-level noise is set. , The boundary is the weighted midpoint. That is, shift towards the level with greater variance; return to the level with equal variance. .

[0043] The above describes the process of the M1 basic detection mode.

[0044] Below, the M2 mode – Winner-Takes-All energy competition decision.

[0045] M2 mode directly uses the "Winner-Takes-All" energy competition decision, without the need for preset threshold parameters: such as Figure 6 As shown, when When, they are judged to be of the same polarity ( );when When, it is judged as reverse polarity ( In the formula, The sign function is represented here. The decision result of this mode is dynamically determined entirely by the relative magnitude of the signal energy, thus exhibiting better robustness. This can be expressed as a formula:

[0046]

[0047] The above describes the process of M2 mode.

[0048] Both M1 and M2 modes can simultaneously recover signals from two tracks using low-complexity symbol-level hard decision without requiring a sequence detector.

[0049] As a preferred implementation, a hybrid detection mode—selective Viterbi detection based on the confidence level of the M2 detection results—can be further adopted.

[0050] This embodiment preferably proposes a hybrid detection mode, such as... Figure 7 As shown, it includes: Calculate at each time point and Difference between This serves as a confidence index for the M2 detection mode; the confidence index at each time point is less than a preset threshold. and , are denoted as low-confidence sample points; A first preset number of consecutive low-confidence sample points are merged into a low-confidence segment; for each low-confidence segment, based on the cache, a second preset number of sample points are extended before and after it, and a joint Viterbi sequence detector is run on the extended sample point sequence to obtain a new two-track decision result. Replace the decision results of the corresponding sample points obtained by the M2 detection mode with the new two-track decision results of each low-confidence segment that does not contain extended sample points; For confidence indices higher than a preset threshold and The decision result of the M2 detection mode is retained, and the complete dual-track detection sequence is finally output.

[0051] The first preset quantity needs to be determined by a trade-off between computational efficiency (i.e., real-time performance) and reliability. The second preset quantity is determined based on the detector's minimum reliable detection length and is related to signal quality.

[0052] The core idea of ​​this hybrid detection mode is to use the M2 detection results to calculate a confidence index to identify regions where M2 decisions may be incorrect. The Viterbi sequence detector is activated only in these regions, while the fast decision result of M2 is directly used in regions where the M2 confidence is sufficiently high. Combining fast M2 detection with accurate Viterbi detection significantly reduces computational complexity while maintaining near-Viterbi detection accuracy.

[0053] Confidence index that can be preferred Defined as two magnetic tracks and The degree of difference in absolute value: ,in To prevent division by zero for small constants. The closer the value is to 1, the more dominant the energy of a certain channel is, and the higher the reliability of the M2 decision result; The closer the value is to 0, the more similar the energies of the two magnetic tracks are, and the lower the reliability of the M2 decision result.

[0054] In specific implementation, the following steps can be referenced as an example: Step 1: For and Run the M2 enhanced detection mode to obtain preliminary judgment results. and the recovered signal value of each sample (i.e., the two tracks under a certain index). and ) confidence level ; Step 2: Based on the preset confidence threshold (Typical values ​​range from 0.5 to 0.8), mark all Low-confidence sample points; Step 3: Merge consecutive low-confidence sample points (corresponding to consecutive indices) into low-confidence segments, each segment having a length of at least, for example, 5 samples (empirical value). Step 4: For each low-confidence segment, extend 10 samples (with a linear relationship to some response targets) before and after as boundaries, and run a 16-state (preset number of states, related to the length of some response targets, 3) joint Viterbi sequence detector on it to correct possible erroneous decisions of M2 in this segment by utilizing the path constraint characteristics of the Viterbi algorithm. Step 5: Replace the corresponding positions in the original M2 decision sequence with the Viterbi-corrected decision results, retain the fast decision results of M2 in the high-confidence region, and finally output the complete dual-track detection sequence.

[0055] The branch output calculation method for the 16-state joint Viterbi sequence detector described above is as follows: Let the PR goal be The status is The input is The expected output of the branch is calculated as follows:

[0056]

[0057] The branch metric for Viterbi detection can be calculated based on the sum of squared errors between the expected output of a branch and the timed recovery output, as shown below:

[0058] like Figure 8 As shown, according to Moment and The possible values ​​of pair[ It can be divided into 4 groups ( "–" and "+" represent two possible polarities, respectively. Each group corresponds to... There are also 4 possible values ​​for time, from Time and The combination of values ​​at time points determines the state with a memory length of 2, resulting in a total of 16 possible states S0~S15. time and State combinations as input in( , According to equations (15)~(16), the output out can be calculated. , ).

[0059] Experimental results show that the three detection methods of the present invention offer different trade-offs between detection accuracy and computational complexity, and are suitable for different application scenarios.

[0060] The M1 basic detection mode has a simple structure and low computational load, requiring only threshold comparison and logical decision-making, making it suitable for applications with limited power consumption and hardware resources; as shown in Figure 9, when the gains of adjacent tracks are close ( Furthermore, some response targets are approximately free of inter-symbol interference (e.g.) Performance is good when the quantization threshold is set to 0. When selecting it, it is robust to gain mismatch: Actual measurements show that... , SNR=28dB, target is Under these conditions, the bit error rate is... time Down to .

[0061] The M2 enhanced detection mode requires no preset threshold; the decision is determined solely by the relative magnitudes of the equalized signal energies from the two channels. It is naturally robust to gain mismatch and has the same computational complexity as M1 (pure vectorized computation). In low inter-symbol interference scenarios, its performance is close to optimal: the measured target is... hour, This achieves zero bit error rate. At this point, the hybrid mode has the same performance as M2 and does not require activation of Viterbi. Therefore, M2 is suitable as the default fast decision maker in normal scenarios.

[0062] Hybrid detection mode shows the most significant advantages against targets with strong inter-code interference. By adjusting the confidence threshold, a continuous trade-off can be made between computational complexity and detection accuracy, making it suitable for scenarios where bit error rate is a priority and computing power is sufficient.

[0063] Selecting adjacent track gain mismatch ( , And some response targets are The experimental conditions were simulated, as shown in Figure 10. Under the same configuration, when SNR=20dB: the bit error rate of the M1 basic detection mode is... The bit error rate of the M2 enhanced detection mode is [missing information]. Hybrid detection mode at confidence threshold When the bit error rate is 0.5, only about 20% of the samples are activated for Viterbi detection, further reducing the bit error rate to 0.5. It decreased by about 39% relative to M2; when When the bit error rate is 0.6, approximately 38% of the samples are activated, and the bit error rate drops to [a certain value]. This represents a relative reduction of approximately 59%. All three detection modes achieved their optimal performance under the same configuration. The hybrid detection mode achieved a significant performance improvement with a moderate computational increment, demonstrating its flexible trade-off between detection accuracy and computational complexity. It should be noted that the benefit of the hybrid detection mode stems from the decision uncertainty caused by inter-symbol interference introduced by some response targets and track gain mismatch: when the gains of the two tracks are equal ( ) or the target has no inter-symbol interference (e.g.) When M2 decision is close to optimal, the hybrid mode performs similarly to M2. Figure 9 and Figure 10 “Sel-Viterbi” indicates selective Viterbi detection based on a confidence index.

[0064] Therefore, in the simulation, the M1 detection mode, M2 detection mode, and hybrid detection mode all meet the reliability testing requirements when the signal-to-noise ratio is higher than 15dB. M1 is suitable for resource-constrained scenarios, M2 is suitable for conventional fast decision-making scenarios, and M1 is more refined and accurate than M2. The hybrid mode is suitable for scenarios where bit error rate is prioritized and computing power is sufficient. If soft arbitration is enabled in the M1 detection process, Viterbi detection can also be called. That is, if the conditions for enabling soft arbitration are met in the M1 detection process, the hybrid mode is called for simultaneous detection of two tracks. In actual working conditions, the three modes can be flexibly selected according to the target characteristics, signal-to-noise ratio, and hardware constraints.

[0065] Example 2 This application also relates to an electronic device, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described above.

[0066] The electronic device can be a desktop computer, laptop, handheld computer, or cloud server, etc. The processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The memory can be used to store computer programs and / or modules. The processor performs various functions of the electronic device by running or executing the computer programs and / or modules stored in the memory, and by accessing data stored in the memory.

[0067] The relevant technical solutions are the same as above, and will not be repeated here.

[0068] Example 3 This application also relates to a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0069] Specifically, the memory may include high-speed random access memory, as well as non-volatile memory, such as hard disks, RAM, plug-in hard disks, smart media cards (SMC), secure digital (SD) cards, flash cards, at least one disk storage device, flash memory device, or other volatile solid-state storage devices.

[0070] The relevant technical solutions are the same as above, and will not be repeated here.

[0071] Example 4 This application provides a computer program product or computer program that includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the steps of the method described in the above embodiments of this application.

[0072] The relevant technical solutions are the same as above, and will not be repeated here.

[0073] Those skilled in the art will readily understand that the above description is merely 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 within the scope of protection of the present invention.

Claims

1. A method for detecting parallel readout signals of adjacent tracks in vector magnetic recording, characterized in that, include: Acquire the combined readback signal of two adjacent tracks perpendicular to the recording plane. and the integrated readback signal horizontally to the recording plane. ;right , Both readback signals are filtered separately; timed recovery is performed on both filtered readback signals, and the recovered readback signal sample sequence is synchronized with the recorded bit sample; based on the recovered two readback signal sample sequences, the following two-track joint detection is performed: M1 mode: compares the signal samples of each signal at each time step with a preset positive threshold. Perform a size comparison; when the sample is larger than... When the absolute value of the sample is less than a certain value, the sample is quantized into a symbol 'a'. When the sample is greater than a certain value, the sample is quantized into a symbol b. When the value is negative, the sample is quantized to the sign c; when Quantized to a or c and If the quantization is b, then determine the current two magnetic tracks. Polarity of the estimated write value corresponding to the next read point at any given time , Same polarity; when Quantized to b and If quantified as a or c, then determine , Opposite polarities; when When the average quantization is b, soft arbitration is initiated: if The absolute value is not less than The absolute value of then polarity as , The polarity, otherwise, polarity as The polarity of, will The antipolarity as polarity; The corresponding track is obtained by subtracting the readout signals of the current two tracks. The track from which the readout signal is subtracted, and the other track is... The corresponding magnetic track; among which, , Representing respectively , middle Signal samples recovered at a set time; preset positive threshold Satisfies: Based on the M1 mode, the polarity of each point on both tracks can be determined, and the values ​​are determined according to... , The features were determined; Alternatively, the M2 mode: i.e., the soft arbitration, if The absolute value is not less than The absolute value of then polarity as , The polarity, otherwise, polarity as The polarity of, will The antipolarity as The polarity of.

2. The detection method as described in claim 1, characterized in that, A magnetoresistive read head is used to read back signals across two adjacent tracks to obtain a composite readback signal perpendicular to the recording plane; or, two magnetoresistive read heads are used to read back signals from the two adjacent tracks respectively, and the two readback signals are added together to obtain a composite readback signal perpendicular to the recording plane from the two adjacent tracks. A vector read head is used to read back the signal across two adjacent tracks to obtain a combined readback signal horizontal to the recording plane; or, two magnetoresistive read heads are used to read back the signal from the two adjacent tracks respectively, and the two readback signals are subtracted to obtain a combined readback signal horizontal to the recording plane from the two adjacent tracks.

3. The detection method as described in claim 1, characterized in that, The filtering process is performed using an adaptive equalizer.

4. The detection method as described in claim 1, characterized in that, The method for determining the preset positive threshold is as follows: according to The noise variance of the two tracks is determined accordingly. When the noise variances are the same, the positive threshold is... When the noise variances are not the same, the positive threshold In the formula, the lower bound of the ideal observation value in the integrated readback signal is... The ideal upper bound for observation values , and These are the gain coefficients of the two magnetic tracks, respectively. , Indicates according to The noise standard deviation of the upper bound of the integrated readback signal and the noise standard deviation of the lower bound of the integrated readback signal are determined by combining the two factors.

5. The detection method as described in claim 1, characterized in that, When using the M2 detection mode, the method also includes: Calculate at each time step and Difference between This serves as a confidence index for the M2 detection mode; the confidence index at each time point is less than a preset threshold. and , are denoted as low-confidence sample points; A first preset number of consecutive low-confidence sample points are merged into a low-confidence segment; for each low-confidence segment, based on the cache, a second preset number of sample points are extended before and after it, and a joint Viterbi sequence detector is run on the extended sample point sequence to obtain a new two-track decision result. Replace the decision results of the corresponding sample points obtained by the M2 detection mode with the new two-track decision results of each low-confidence segment that does not contain extended sample points; For confidence indices higher than a preset threshold and The decision result of the M2 detection mode is retained, and the complete dual-track detection sequence is finally output.

6. The detection method as described in claim 5, characterized in that, The degree of difference is determined in the following way: ,in, It is a constant to prevent division by zero.

7. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method as described in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored computer program, wherein the computer program, when executed by a processor, controls the device on which the storage medium is located to perform the steps of the method as described in any one of claims 1 to 6.

9. A computer program product, comprising a computer program or instructions, characterized in that, When the computer program or instructions are executed by a processor, they implement the steps of the method as described in any one of claims 1 to 6.

Images