A global deep coverage intelligent data recovery scheduling system

Abstract

The application provides a global depth coverage intelligent data recovery scheduling system, and relates to the technical field of computer storage. The global physical signal acquisition module is used for performing global physical scanning on the storage medium that has experienced data coverage, collecting an original signal data set, and transmitting the original signal data set to the multi-layer residual extraction module. Through the synergistic effect of the global physical signal acquisition module and the multi-layer residual extraction module, the storage medium that has experienced multiple coverages can be deeply physically scanned and multi-band signal separation can be performed, effective data fragments can be accurately extracted from residual signals, the recovery success rate in the coverage data scene is significantly improved, and the problem that a large amount of random codes or complete recovery cannot be recovered in the traditional software recovery is solved.

Description

Technical Field

[0001] This invention relates to the field of computer storage technology, specifically to a full-domain deep coverage intelligent data recovery scheduling system. Background Technology

[0002] In the field of computer storage technology, technological development has evolved from early simple file deletion recovery to data sculpting at the physical level. Early data recovery software primarily relied on the operating system's file system index for recovery, such as Recuva, DiskGenius (normal mode), and EaseUS Data Recovery Wizard. These software programs quickly located and recovered logically deleted or quickly formatted files by reading index information such as the file allocation table, master file table, or directory entries. In practical applications, these tools have good recovery effects for everyday accidental deletion scenarios, are easy to operate, and have a relatively fast recovery speed, thus being widely used by individual users and small businesses for data recovery. With the continuous upgrading of storage media, some tools have gradually added support for TRIM commands and garbage collection mechanisms.

[0003] Although existing technologies can achieve data recovery in some scenarios, the following problems still exist in practical use: Existing data recovery software mainly relies on file system indexes for recovery. When the storage medium is overwritten once or multiple times (such as system reinstallation, new data writing, etc.), the file system index is completely destroyed or overwritten. Traditional software cannot effectively extract the original data from the physical sector remnants and often can only recover a large number of garbled files or disordered fragments, and cannot achieve the recovery of complete files. Existing technologies struggle to accurately identify the starting point and boundaries of data fragments from the physical remnants after overwriting, and are also unable to effectively reconstruct the original directory hierarchy and filename information. As a result, the recovery results are usually a messy pile of files, making it difficult for users to quickly organize and locate the files they need, and significantly reducing the usability of the recovered data. Existing data recovery software lacks effective multi-layer content alignment and integrity repair mechanisms when recovering overwritten data. The recovered documents, pictures, videos and other files generally have structural damage, missing content or cannot be opened normally. Even if some files can be opened, they often have garbled characters or abnormal functions, resulting in low actual usability.

[0004] Therefore, a comprehensive, deep-coverage intelligent data recovery and scheduling system is needed to solve the above problems. Summary of the Invention

[0005] Technical problems to be solved To address the shortcomings of existing technologies, this invention provides a full-domain deep coverage intelligent data recovery and scheduling system, which solves the problems mentioned in the background section above.

[0006] Technical solution To achieve the above objectives, the present invention provides the following technical solution: a full-domain deep coverage intelligent data recovery and scheduling system, comprising a full-domain physical signal acquisition module, a multi-layer residual extraction module, a sector topology directory reconstruction module, a multi-layer content alignment and repair module, and a coordination and scheduling execution module. The full-domain physical signal acquisition module is used to perform a full-domain physical scan on the storage medium that has been covered by data, acquire the original signal dataset, and transmit the original signal dataset to the multi-layer trace extraction module; The multi-layer trace extraction module is used to receive the original signal dataset, extract the covering traces layer by layer through a multi-band signal separation algorithm, generate a layered data fragment sequence, and transmit the layered data fragment sequence to the sector topology directory reconstruction module. The sector topology directory reconstruction module is used to receive the hierarchical data fragment sequence, reconstruct the directory hierarchy and file name information through physical sector location mapping and residual metadata cross-association algorithm, and transmit the reconstructed structured file object to the multi-layer content alignment and repair module. The multi-layer content alignment and repair module is used to receive the structured file object, repair the file through inter-layer boundary alignment algorithm and content pattern matching algorithm, and output a recoverable file that can be opened directly; The coordination and scheduling execution module is used to receive processing status feedback from each module, generate scheduling instructions based on the layered distribution of remnants, and transmit the scheduling instructions to the global physical signal acquisition module, the multi-layer remnant extraction module, the sector topology directory reconstruction module, and the multi-layer content alignment and repair module for coordinated operation.

[0007] Preferably, when performing a full-domain physical scan, the full-domain physical signal acquisition module first determines the initial scanning frequency range and key physical areas based on the number of times the storage medium is covered and the degree of attenuation of the residual signal. Then, it performs multi-band layered scanning, and then collects and generates a standardized raw signal dataset containing physical address markers, signal amplitude markers and scanning level markers in real time. Finally, it transmits the dataset to the multi-layer residual extraction module.

[0008] Preferably, after receiving the original signal dataset, the multi-layer trace extraction module first separates the overlay traces layer by layer based on the differences in attenuation characteristics of signals at different frequencies, then generates a layered data fragment sequence with inter-layer boundary location markers and inter-layer correlation strength markers, then tracks the traces between multiple overlay layers based on the inter-layer correlation strength markers, and finally transmits the processed layered data fragment sequence to the sector topology directory reconstruction module.

[0009] Preferably, after receiving the hierarchical data fragment sequence, the sector topology directory reconstruction module first establishes a residual topology mapping relationship based on the sector physical adjacency relationship and address continuity, then performs cross-association calculations by combining residual directory records and file system log traces, then generates a complete directory hierarchy structure and file name information, and finally transmits the reconstructed structured file object to the multi-layer content alignment and repair module.

[0010] Preferably, after receiving the structured file object, the multi-layer content alignment and repair module first performs boundary alignment processing on the layered data fragments according to the inter-layer boundary position markers, then fills the missing parts through a content pattern matching algorithm while maintaining file format constraints, then completes the usability verification of the file structure and content, and finally outputs a recoverable file that can be opened directly.

[0011] Preferably, in the process of separating overlay layer remnants, the multi-layer remnant extraction module first tracks the remnant association relationship between multiple overlay layers based on the inter-layer association strength marker, then uses a cross-layer splicing algorithm to optimize the splicing of associated fragments, then generates an optimized layered data fragment sequence, and finally transmits it to the sector topology directory reconstruction module.

[0012] Preferably, when the sector topology directory reconstruction module reconstructs the directory hierarchy and file name information, it first performs weighted enhancement processing on the residual metadata based on the signal amplitude marker, then further enhances the directory hierarchy reconstruction accuracy, then generates an enhanced structured file object, and finally transmits it to the multi-layer content alignment and repair module.

[0013] Preferably, when generating scheduling instructions, the coordination and scheduling execution module first identifies the physical characteristics of the current storage medium, and then generates corresponding media-adaptive scheduling instructions according to the medium type. For different levels of mechanical hard drives, a matching method is adopted. Specifically, consumer-grade mechanical hard drives adopt page-level charge trace extraction mode, and enterprise-grade mechanical hard drives adopt track signal enhancement mode. Then, the scheduling instructions are transmitted to the global physical signal acquisition module and the multi-layer trace extraction module to adjust the processing parameters.

[0014] Preferably, after receiving the processing status feedback from each module, the coordination and scheduling execution module first collects the separation completion rate of the multi-layer trace extraction module and the reconstruction completeness of the sector topology directory reconstruction module. Then, it generates an optimized scheduling instruction based on the layered distribution of traces and transmits the scheduling instruction to the global physical signal acquisition module to adjust the frequency parameters and priority of key areas for subsequent scanning.

[0015] Preferably, when the coordination and scheduling execution module performs convergence control on the entire recovery process, it first monitors the directory hierarchy integrity index and the file repair success rate index in real time, then determines whether each index meets the convergence conditions, and then terminates the recovery process when the convergence conditions are met, and triggers the global physical signal acquisition module to perform a new round of multi-band layered scanning to continue optimizing the recovery results when the convergence conditions are not met.

[0016] Beneficial effects This invention provides a full-domain deep coverage intelligent data recovery scheduling system. It has the following beneficial effects: 1. This invention, through the synergistic effect of the full-domain physical signal acquisition module and the multi-layer residual extraction module, can perform deep physical scanning and multi-band signal separation on storage media that have undergone multiple overlays, accurately extract effective data fragments from residual signals, significantly improve the recovery success rate in overlaid data scenarios, and solve the problem of traditional software recovering large amounts of garbled text or being completely unrecoverable.

[0017] 2. This invention uses a sector topology directory reconstruction module, which utilizes physical sector location mapping and residual metadata cross-association algorithms to accurately reconstruct the original directory hierarchy and filenames. This allows the recovered data to be presented in the user's original folder organization format, greatly improving the manageability and ease of use of the recovered data.

[0018] 3. This invention uses a multi-layer content alignment and repair module, employing inter-layer boundary alignment algorithms and content pattern matching algorithms to accurately repair files. Combined with the integrity verification and iterative optimization of the coordination and scheduling execution module, it ensures that the final output recovered file has a complete structure and usable content. Various types of files, such as documents, pictures, and videos, can be opened and used normally, fundamentally improving the actual usability of the recovered data. Attached Figure Description

[0019] Figure 1 This is a system framework diagram of the present invention; Figure 2 This is a flowchart of the process steps of the present invention; Figure 3 This is a schematic diagram of the main interface of the system of the present invention; Figure 4 This is a schematic diagram of the system monitoring interface of the present invention; Figure 5 This is a schematic diagram of the system log of the present invention. Detailed Implementation

[0020] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Specific Implementation Example 1: like Figure 1 As shown, a full-domain deep coverage intelligent data recovery scheduling system is provided, including a full-domain physical signal acquisition module, a multi-layer trace extraction module, a sector topology directory reconstruction module, a multi-layer content alignment and repair module, and a coordination and scheduling execution module. These five core functional modules work together to complete trace acquisition, trace extraction, directory reconstruction, file repair, and global intelligent scheduling operations on multi-layer covered storage media. A closed-loop operating architecture of unidirectional main data flow + bidirectional feedback scheduling data flow is constructed between the five modules, such as... Figure 2 As shown, a flowchart is provided. The main data flow is transmitted step by step in the order of full-domain physical signal acquisition, multi-layer trace extraction, sector topology directory reconstruction, and multi-layer content alignment and repair. The coordination and scheduling execution module independently establishes a two-way communication link with the other four modules, issues control commands in real time, and collects operation status data to ensure the collaborative and synchronous operation of multiple modules.

[0022] The global physical signal acquisition module is a hardware and software integrated unit for front-end signal acquisition in the system. It is used to perform a global physical scan on the storage medium that has been covered by data, acquire the raw signal dataset, and transmit the raw signal dataset to the multi-layer residual extraction module. When performing a global physical scan, the global physical signal acquisition module first determines the initial scanning frequency range and key physical areas based on the number of times the storage medium has been covered and the degree of residual signal attenuation. Then, it performs multi-band layered scanning, and then acquires and generates a standardized raw signal dataset containing physical address markers, signal amplitude markers, and scan level markers in real time. Finally, it transmits the dataset to the multi-layer residual extraction module. The global physical signal acquisition module is suitable for mechanical hard drives, external hard drives, and virtualized storage media. It runs on a dedicated data recovery workstation. The hardware components include a multi-core high-performance CPU processor for overall task coordination and control, a dedicated storage interface controller for connecting different types of storage media such as SATA, SAS, and NVMe, a high-precision signal acquisition card for low-level raw signal reading, a programmable voltage bias circuit for dynamically adjusting the reading voltage at different levels, a high-speed analog-to-digital converter for converting analog electrical signals to digital signals, and a temperature compensation sensor to eliminate the impact of ambient temperature fluctuations on signal acquisition accuracy in real time. All of this hardware is integrated within the dedicated data recovery workstation. Each hardware unit is interconnected via a PCIe 4.0 high-speed bus and operates independently and redundantly, effectively avoiding signal acquisition interruptions caused by single hardware failures. The high-precision signal acquisition card achieves a sampling accuracy of up to 1μV, and the high-speed analog-to-digital converter supports 16 / 24-bit bidirectional switching of sampling levels, adapting to the weak residual signal acquisition needs of different types of storage media.

[0023] In practical applications, when a user brings in a 1TB hard drive that has been overwritten twice due to system reinstallation, the full-domain physical signal acquisition module first establishes a stable physical connection with the hard drive through a dedicated storage interface controller. The connection process defaults to read-only access mode, locking write permissions to the storage medium at the hardware level to prevent secondary overwriting or damage to the original residual data during the scanning process, thus maximizing the protection of the original data integrity. Then, it reads the hard drive's SMART information and preliminary low-precision scan results to estimate the number of overwrites (e.g., estimated as 2 overwrites) and the degree of residual signal attenuation (e.g., attenuation rate reaching 65%). Following the principle that higher overwrite counts result in more severe attenuation, the initial scan frequency range is set to prioritize low-frequency bands. (Starting from 80MHz and gradually increasing to 750MHz), the key physical area is set as the sector range corresponding to the original file system partition (e.g., sectors 0-5000000). Then, multi-band layered scanning is performed. Specifically, low-frequency high-voltage signals are used to scan the outermost layer of overlay data first, and then the frequency is gradually increased and the voltage is decreased to penetrate deeper overlay layers. The voltage adjustment range for a single frequency band scan is 0.8V-3.5V. The high voltage on the surface layer ensures complete acquisition of the surface overwrite data, while the low voltage in the deep layer is combined with high-frequency signals to reduce the risk of voltage breakdown of flash memory chips and disk tracks. At the same time, the original analog signals are acquired in real time and converted into digital signals through a high-speed analog-to-digital converter, and finally a standardized original signal dataset is generated. The physical address marker refers to the precise logical block address (LBA) number corresponding to each sampling point (e.g., LBA 12345678), the signal amplitude marker refers to the specific value of the signal strength acquired at that sampling point (in millivolts, e.g., 125.6mV), and the scan level marker indicates which overlay layer the current sample belongs to (e.g., layer 1 represents the outermost layer of newly covered data, layer 3 represents the deepest layer of original remnants). The generated standardized raw signal dataset is transferred to the multi-layer remnant extraction module via the high-speed PCIe 4.0 data bus inside the workstation using DMA direct memory access. The DMA transfer mode can bypass CPU redundancy scheduling and directly write the dataset to the corresponding hardware memory, reducing transmission latency by more than 40%, making it suitable for high-speed transmission scenarios of gigabyte-level massive raw signal datasets.

[0024] The multi-layer debris extraction module receives the raw signal dataset and runs in the workstation's GPU computing cluster environment. The hardware components include a high-performance GPU accelerator card for large-scale parallel signal processing, an FPGA real-time signal processing board for high-speed filtering and feature extraction, a large-capacity DDR4 cache memory (at least 128GB) for temporarily storing massive amounts of intermediate data, and a high-speed data interface for receiving data transmitted from preceding modules. The FPGA board and GPU accelerator card have clearly defined roles: the FPGA handles basic operations such as fixed-logic filtering and signal transformation, while the GPU handles complex parallel operations such as multi-level debris clustering and similarity comparison. The 128GB DDR4 cache memory uses a partitioned, isolated storage mode, with signal data from different overlay layers stored independently to avoid data cross-contamination.

[0025] In practical applications, when the full-domain physical signal acquisition module transmits the original signal dataset, the multi-layer residual extraction module first receives the dataset completely through a high-speed data interface. The receiving process incorporates a built-in data packet verification and retransmission mechanism. It verifies the integrity of data packets using a CRC32 checksum algorithm and automatically initiates retransmission requests to upstream modules for damaged or missing packets, ensuring the input data is error-free. Then, it extracts the coverage residual layer by layer using a multi-band signal separation algorithm, generating a layered data fragment sequence. The specific calculation logic of the multi-band signal separation algorithm is as follows: first, the received original signal dataset is Fourier transformed to the frequency domain. The signal is processed by setting bandpass filters according to preset frequency ranges (low frequency 80-200MHz, mid frequency 200-450MHz, high frequency 450-750MHz) to separate residual signals from different coverage depth layers. Then, the filtered signals are converted back to the time domain using inverse Fourier transform. Finally, threshold segmentation is performed on each layer to extract valid residual fragments. The threshold segmentation uses an adaptive dynamic threshold mechanism, automatically removing noise and blank fragments below the threshold based on the average amplitude of each layer's signal, retaining only valid residual fragments with data recovery value. The input is the raw signal dataset transmitted by the full-domain physical signal acquisition module, and the output is a sequence of layered data fragments with inter-layer boundary location markers and inter-layer correlation strength markers. The calculation formula for the multi-band signal separation algorithm is as follows: The specific calculation formula for the multi-band signal separation algorithm is as follows: ; in: This represents the frequency domain residual signal after the separation of the l-th layer; This indicates the Fast Fourier Transform operation; This represents the input time-domain raw signal dataset; This represents the frequency response function of the bandpass filter designed for the l-th layer; This indicates that the original phase information is preserved; Represents frequency variables; Indicates the number of the currently separated overlay ( This formula takes the original time-domain signal dataset as input. The frequency domain residual signals after the separation of each covering layer were calculated. After undergoing inverse Fourier transform, a time-domain layered data fragment sequence is obtained, which is used to achieve accurate separation of traces with different coverage depths.

[0026] The layered data fragment sequence is then transmitted to the sector topology catalog reconstruction module. After receiving the original signal dataset, the multi-layer remnant extraction module first separates the overburden layer remnants layer by layer based on the differences in attenuation characteristics of signals at different frequencies. Then, it generates a layered data fragment sequence with inter-layer boundary location markers and inter-layer correlation strength markers. Next, it tracks the remnants between multiple overburden layers based on the inter-layer correlation strength markers. Finally, the processed layered data fragment sequence is transmitted to the sector topology catalog reconstruction module. During the separation of overburden layer remnants, the multi-layer remnant extraction module first tracks the remnant correlation relationships between multiple overburden layers based on the inter-layer correlation strength markers, and then employs a cross-layer splicing algorithm. The algorithm optimizes the splicing of correlated fragments, generating an optimized hierarchical data fragment sequence. The specific calculation logic of the cross-layer splicing algorithm is as follows: First, it calculates the feature similarity of each corresponding sampling point between adjacent overlay layers. Then, it sets a correlation strength threshold of 0.75. When the similarity exceeds the threshold, it is marked as a splicable fragment. Next, it performs boundary smoothing and overlapping region fusion on the splicable fragments, finally generating a continuous, unbroken optimized hierarchical data fragment sequence. For fragmented data below the 0.75 threshold, the module automatically classifies them as invalid or isolated fragments and archives them separately, avoiding interference from invalid data in subsequent calculations and facilitating subsequent manual review and screening. The algorithm takes a hierarchical data fragment sequence with inter-layer correlation strength markers as input and outputs an optimized hierarchical data fragment sequence. The specific calculation formula for the cross-layer splicing algorithm is as follows: The specific calculation formula for the cross-layer splicing algorithm is as follows: ; in: Indicates the position after splicing The fusion signal value at the location; Indicates the first Layer in position Weighting coefficients at each location; Indicates the first Layer in position Weighting coefficients at each location; Indicates the first Layer in position Signal characteristic values ​​at the location; Indicates the first Layer in position Signal characteristic values ​​at the location; This represents the sampling point index at the current stitching position. The formula takes as input the signal feature value sequences of the two adjacent stitchable regions and their corresponding weight coefficients, and calculates the fused smooth signal value. It is used to eliminate boundary abrupt changes at cross-layer splicing points and achieve continuous optimization of the trace.

[0027] The sector topology directory reconstruction module receives hierarchical data fragment sequences, reconstructs the directory hierarchy and filename information through physical sector location mapping and residual metadata cross-association algorithms, and transmits the reconstructed structured file objects to the multi-layer content alignment and repair module. Upon receiving the hierarchical data fragment sequences, the sector topology directory reconstruction module first establishes a residual topology mapping relationship based on sector physical adjacency and address continuity. Then, it performs cross-association calculations combining residual directory records and file system log traces, generating a complete directory hierarchy and filename information. Finally, it transmits the reconstructed structured file objects to the multi-layer content alignment and repair module. When reconstructing the directory hierarchy and filename information, the sector topology directory reconstruction module first performs weighted enhancement processing on the residual metadata based on signal amplitude markers, then further enhances the accuracy of directory hierarchy reconstruction, generates enhanced structured file objects, and finally transmits them to the multi-layer content alignment and repair module.

[0028] The sector topology directory reconstruction module runs on the workstation's high-performance computing unit. The hardware includes a high-frequency multi-core CPU processor for complex topology calculations, a large-capacity memory (at least 128GB) for storing topology map data, a GPU accelerator card for parallel association calculations, and a large-capacity hard disk cache array for temporary storage of intermediate results. The hard disk cache array uses a RAID1 redundant array mode to back up the topology map, metadata, and other intermediate computational data in real time, preventing data loss due to unexpected power outages during computation and ensuring the stability of the directory reconstruction operation. The specific calculation logic of the residual metadata cross-association algorithm is as follows: First, it receives the hierarchical data fragment sequence transmitted by the multi-layer residual extraction module; then, it establishes a topology mapping relationship based on the physical sector location; next, it reads residual directory records and file system log traces for cross-matching calculations; finally, it outputs the reconstructed directory hierarchy structure and filenames. The algorithm's input is the hierarchical data fragment sequence and residual metadata, and the output is a structured file object; the specific calculation formula for the residual metadata cross-association algorithm is as follows: ; in: This represents the correlation strength value between the i-th layer remnant and the j-th layer remnant; This represents the signal feature value of the k-th sampling point in the i-th layer; This represents the signal feature value of the k-th sampling point in the j-th layer; This represents the total number of overlapping sampling points between two layers. The input to this formula is a sequence of signal feature values ​​from sampling points in different coverage layers, used to calculate the inter-layer correlation strength value, which is used to determine whether cross-layer splicing and directory association should be performed. Based on the calculation results of this formula, the module presets a threshold of 0.6 for directory association determination. If the value is higher than the threshold, the two layers of fragments are determined to belong to the same file directory and the hierarchical directory relationship is automatically bound. If the value is lower than the threshold, the fragmented data is classified separately.

[0029] The multi-layer content alignment and repair module receives structured file objects and repairs them using inter-layer boundary alignment and content pattern matching algorithms, outputting a directly openable recovered file. After receiving the structured file object, the multi-layer content alignment and repair module first aligns the layered data fragments according to the inter-layer boundary position marks. Then, while maintaining file format constraints, it fills in the missing parts using a content pattern matching algorithm. Next, it verifies the usability of the file structure and content, and finally outputs a directly openable recovered file.

[0030] The multi-layer content alignment and repair module runs in a GPU-accelerated environment on the workstation. The hardware components include a high-performance GPU computing card for parallel pattern matching, a dedicated file format parsing processor for recognizing formats such as JPEG, DOCX, and MP4, a large-capacity memory buffer for temporarily storing intermediate repair data, and a file integrity verification hardware accelerator for final verification. The dedicated file format parsing processor has built-in feature libraries for over twenty mainstream file formats, covering four major categories: images, documents, audio / video, and executable programs. It also supports custom import of feature codes for less common formats, adapting to diverse file repair needs across multiple scenarios. The file integrity verification hardware accelerator simultaneously supports MD5 and SHA-256 dual hash verification mechanisms. The specific calculation logic of the inter-layer boundary alignment algorithm is as follows: First, it receives a structured file object; then, it extracts the key feature sequences of each layer of data; next, it calculates the difference value one by one within the possible offset range, finding the offset position with the smallest difference value; finally, it aligns and splices the layer fragments according to this offset position. The default offset retrieval range is set to ±128 bytes, which can be manually adjusted by the operator based on the storage medium type and file size, balancing alignment accuracy and computational efficiency. The specific computational logic of the content pattern matching algorithm is as follows: First, extract file format features (e.g., SOI tags in JPEG files, central directory structure in ZIP files); then, search for known pattern fragments within the format constraints for matching; finally, intelligently fill in the missing parts of the successfully matched files and verify their usability. The algorithm takes a structured file object as input and outputs a repaired, directly openable file.

[0031] The calculation formula for the inter-layer boundary alignment algorithm is as follows: ; in: This represents the calculated optimal alignment offset; This indicates the offset value being attempted. This represents the feature value of the first layer of data at position k; This represents the feature value of the second layer of data at position k after offset x. The formula takes the feature value sequence of the two layers of data as input and calculates the optimal offset that minimizes the difference between the two layers, thus achieving precise boundary alignment.

[0032] The calculation formula for the content pattern matching algorithm is as follows: The specific calculation formula for the content pattern matching algorithm is as follows: ; in: This indicates the matching score between the current segment and the template; This represents the value of the m-th feature point in the standard template library; This represents the value of the m-th feature point in the segment to be repaired; This represents the total number of feature points involved in the matching; This represents the normalized structural difference distance (ranging from 0 to 1). The formula takes the feature vector of the segment to be repaired and the feature vector of the standard template as input, and calculates the matching score. This is used to determine whether to perform content filling and the confidence level of the filling. The preset matching score threshold for this module is 0.85. A score ≥0.85 indicates a successful segment match, and the missing area is automatically filled; a score ≤0.6 indicates a different match. If the value is less than 0.85, a manual review will be requested. If the value is less than 0.6, the match is considered to have failed, and the segment is discarded and automatically repaired.

[0033] The coordination and scheduling execution module receives processing status feedback from each module, generates scheduling instructions based on the layered distribution of residual traces, and transmits these instructions to the global physical signal acquisition module, multi-layer residual trace extraction module, sector topology directory reconstruction module, and multi-layer content alignment and repair module for coordinated operation. When generating scheduling instructions, the coordination and scheduling execution module first identifies the physical characteristics of the current storage medium, then generates corresponding media-adaptive scheduling instructions based on the medium type. For consumer-grade hard disk drives, it uses a page-level charge residual trace extraction mode; for enterprise-grade hard disk drives, it uses a track signal enhancement mode. The scheduling instructions are then transmitted to the global physical signal acquisition module and the multi-layer residual trace extraction module to adjust processing parameters. After receiving feedback on the processing status of each module, the separation completion rate of the multi-layer trace extraction module and the reconstruction completeness of the sector topology directory reconstruction module are collected first. Then, an optimized scheduling instruction is generated based on the layered distribution of traces. The scheduling instruction is then transmitted to the global physical signal acquisition module to adjust the frequency parameters and priority of key areas for subsequent scans. When the coordination scheduling execution module performs convergence control on the entire recovery process, it first monitors the directory level integrity index and file repair success rate index in real time. Then, it determines whether each index meets the convergence conditions. When the convergence conditions are met, the recovery process is terminated. When the convergence conditions are not met, the global physical signal acquisition module is triggered to execute a new round of multi-band layered scanning to continue optimizing the recovery results.

[0034] The coordination and scheduling execution module runs on the main control unit of the workstation. Its hardware components include a main control CPU processor for overall decision-making, real-time status monitoring sensors for collecting operational data from each module, a large-capacity log storage hard drive for recording the entire process, and multiple high-speed communication buses for command issuance. In actual operation, the coordination and scheduling execution module continuously receives processing status feedback data from the other four modules, and then intelligently generates scheduling commands based on the hierarchical distribution of residual data, ultimately achieving dynamic optimization and closed-loop control of the entire system. The real-time status monitoring sensors can collect more than ten operating parameters in real time, such as module computing power utilization, signal acquisition signal-to-noise ratio, data transmission latency, and hardware temperature, avoiding hardware overheating and computing power overload issues. Convergence conditions can be customized by the staff; the system's default basic convergence thresholds are: directory hierarchy integrity ≥90% and file repair success rate ≥85%. The job terminates when both indicators are simultaneously met. All scheduling commands, operation logs, and job parameters are synchronously stored on the log hard drive for easy traceability and iterative parameter optimization.

[0035] like Figure 3 Provides a schematic diagram of the system's main interface, showing a partial view of the main interface's command options. Users can select the storage medium type and overwrite count on the main interface. The corresponding module will display "Started" when it is working and "Pending Start" when it is not working. like Figure 4Provides a schematic diagram of the system monitoring interface, which can display the system recovery process in real time; like Figure 5 The present invention provides a schematic diagram of the system log, which is used to record the key processes and status changes of a single data recovery operation. Specific Implementation Example 2: This embodiment, based on the full-domain deep coverage intelligent data recovery scheduling system disclosed in Embodiment 1, conducts differentiated data recovery operations for four typical application scenarios: consumer-grade mechanical hard drives, enterprise-grade mechanical hard drives, industrial control-specific storage media, and virtualized storage images, verifying the system's media adaptability and scenario universality. All operation processes, algorithm logic, and hardware invocation methods are consistent with Embodiment 1, with only the coordination and scheduling execution module switching adaptation modes according to media type and scenario requirements.

[0037] Scenario 1: Data recovery from overwrite loss on a consumer-grade hard drive: In the consumer-grade mechanical hard drive (HDD) scenario, designed for everyday use by ordinary users, the cause of failure is the writing of numerous new files to existing partitions, directly overwriting existing documents, photos, and videos, and simultaneously corrupting the directory table. The coordination and scheduling execution module, after identifying the HDD media, automatically activates the page-level charge remnant extraction mode. The global physical signal acquisition module connects to the hard drive via an NVMe interface controller, shielding all write permissions and dynamically adjusting the voltage bias range between 2.8V and 3.6V based on the characteristics of the flash memory chips. It focuses on scanning the user data partition (typically accounting for 70% to 85% of the total hard drive capacity), collecting multi-layer charge remnant signals. The multi-layer remnant extraction module uses a multi-band signal separation algorithm to separate remnants from different overwrite layers, while simultaneously initiating a cross-layer splicing algorithm to match fragmented file remnants. The sector topology directory reconstruction module establishes the remnant topology mapping relationship and reconstructs the original folder hierarchy and filenames through metadata weighted enhancement processing. The multi-layer content alignment and repair module performs inter-layer boundary alignment and content pattern matching and filling on images and office documents, ultimately outputting files that can be opened normally. In this scenario, the system can effectively restore files that have not been completely overwritten on the surface as well as files that have been overwritten in the middle and deep layers, meeting the recovery needs of individual users after accidental data overwriting and deletion.

[0038] Scenario 2: Batch data recovery from enterprise-grade hard drives: In the enterprise-grade hard drive scenario, targeting large-capacity devices used in servers, the failure was caused by accidental formatting followed by batch writing of new data, resulting in severe fragmentation and overlay of historical business files, financial ledgers, and audio / video backups. The coordination and scheduling execution module switched to track signal enhancement mode. The full-domain physical signal acquisition module extended the low-frequency signal scanning time to 1.8 times the standard duration, enhancing the reading intensity of track remnants and completing a full-coverage scan of the entire disk cylinder and read / write head. The multi-layer remnant extraction module increased the cross-layer splicing threshold to 0.78 to prioritize matching large, continuous business file fragments. The sector topology directory reconstruction module utilized a large-capacity memory cache to obtain a complete topology map, combining historical log traces to batch reconstruct multi-level business directory structures. The multi-layer content alignment and repair module expanded the offset retrieval range for audio / video and spreadsheet files, completing batch file unified repair output. The coordination and scheduling execution module monitored hardware temperature and computing power status throughout the process, executing recovery jobs in segments to avoid prolonged high-load operation. This scenario verified the system's adaptability to long-term, high-load recovery operations on enterprise-grade large-capacity hard drives.

[0039] Scenario 3: Recovery from power failure and overwrite of industrial control storage media: The industrial control storage medium scenario targets embedded storage devices used in industrial control equipment. The cause of the failure is that after an abnormal power outage, the system automatically restarts and writes initialization data, overwriting the operation logs and configuration parameter files. The coordination and scheduling execution module narrows the scanning sector range to within 35% of the total capacity of the log and configuration partitions, focusing on critical areas to reduce unnecessary computing power consumption. The global physical signal acquisition module, combined with a temperature compensation sensor, compensates for temperature fluctuations in the confined space in real time, stably acquiring weak parameter remnant signals. The multi-layer remnant extraction module prioritizes extracting remnants from small configuration files. The sector topology directory reconstruction module simplifies the directory reconstruction logic, focusing on reconstructing the log file directory structure. The multi-layer content alignment and repair module reduces the processing priority of large files, quickly completing the extraction of remnants, directory reconstruction, and content filling of small configuration files and log files, ultimately outputting parameter files that can be directly imported into industrial control equipment. This scenario verifies the system's adaptability and rapid recovery performance to small-capacity, special protocol media used in industrial control systems.

[0040] Scenario 4: Virtualized storage image data overwrite recovery: The virtualized storage image scenario targets virtualized storage files on cloud computing platforms. The causes of failure are partition overwriting within the virtual disk and accidental deletion of multiple generations of snapshots. The coordination and scheduling execution module initiates a virtualized media-specific mode, skipping the physical interface connection step and directly processing the image file using logical addressing. The global physical signal acquisition module reads raw binary signals from the underlying image file of the virtual disk, simulates the physical layered scanning process, and generates a standardized signal dataset. The remaining four modules follow the algorithm logic of Example 1, adapting to the virtual LBA addressing rules, to complete multi-layer snapshot remnant extraction, virtual directory reconstruction, and image file content repair, ultimately outputting a complete virtual disk image file that can be directly mounted and used. This scenario verifies the system's processing capabilities for non-physical virtualized storage media and its cross-platform applicability. Specific Implementation Example 3: Based on the system architecture of Embodiment 1, this embodiment further optimizes the scheduling logic of the coordination and scheduling execution module, adds a manual and automatic dual-mode switching mechanism, and supports two operating platforms: high-configuration dedicated workstations and low-configuration lightweight chemical control hosts.

[0042] The five functional modules, hardware architecture, and algorithm logic of this embodiment are completely consistent with those of Embodiment 1. The core improvement lies in the addition of job mode adjustment and hardware adaptive computing power allocation functions to the coordination and scheduling execution module.

[0043] The coordination and scheduling execution module offers two working modes: automatic scheduling and manual fine-grained scheduling. Operators can switch between these modes with a single click on the system's main interface. Automatic scheduling mode is suitable for ordinary users and common data loss scenarios. The system autonomously completes media type identification, scan frequency range setting, remnant extraction parameter configuration, directory reconstruction weight selection, and convergence condition determination. No manual intervention is required for parameter adjustment throughout the process, making it simple and suitable for rapid recovery needs. Manual fine-grained scheduling mode is suitable for professional technicians and high-difficulty recovery scenarios. Operators can manually adjust the scan frequency range (80MHz to 850MHz), voltage bias range (2.5V to 4.0V), fragment association strength threshold (0.65 to 0.92), directory reconstruction weighting coefficient, file repair matching score threshold (above 0.82), and job convergence conditions. It supports manually defining key scanning areas and masking invalid blank partitions, enabling targeted deep recovery of deep, multi-layered overlay data.

[0044] The coordination and scheduling execution module also features hardware adaptive computing power allocation. In a high-configuration dedicated data recovery workstation environment, the system automatically enables a fully parallel computing mode using GPU and FPGA, allocating all computing resources to meet the high-speed recovery needs of large-capacity media and batch files. In a low-configuration lightweight chemical control host environment, the system automatically reduces the number of parallel threads to 40% of the original setting, breaking down complex computational tasks into multiple sequential execution stages. It completes the signal acquisition, trace extraction, and file repair processes step by step, significantly reducing CPU, memory, and GPU utilization while ensuring recovery quality, ensuring that even low-configuration equipment can stably run data recovery jobs on small to medium-sized media.

[0045] In both modes, the five functional modules complete signal acquisition, trace separation, fragment splicing, directory reconstruction, and file repair operations according to the algorithm formula disclosed in Example 1. The dual-mode switching mechanism and computing power allocation logic only operate at the global scheduling level, without changing any underlying algorithms, data marking rules, or hardware / software interaction logic. This ensures system stability while meeting the dual needs of easy operation and professional, sophisticated tasks. Specific Implementation Example 4: A data storage company installed this system and used this solution to assist in data recovery for various storage media used by multiple companies and individuals. The entire process relies on the five core modules, algorithm logic, and scheduling mechanism of this invention, strictly corresponding to actual fault conditions and recovery results. Specific examples are as follows: Case 1: Case Background A technology service company, while undertaking computer maintenance work for a medical institution, had its operators fail to back up the equipment data beforehand. They directly formatted and reinstalled the Windows 7 system C drive on a 3.5-inch 500GB Seagate hard drive, resulting in the medical institution's desktop files, accumulated over more than ten years, being overwritten by the new system data. After the failure, the technology service company commissioned several local and out-of-town professional data recovery agencies and tried various mainstream commercial recovery software programs, but all failed to retrieve any usable original data, and the recovery attempts all failed. The medical institution held them accountable for the data loss, and their cash compensation offer was rejected, leading to a dispute that lasted for over a month. Subsequently, the technology service company commissioned this company to perform data recovery. This company first used commercially available recovery software to scan and detect the data, but failed to obtain any usable original data. Ultimately, they relied on this invention's comprehensive deep-coverage intelligent data recovery scheduling system to perform a deep recovery operation.

[0047] Implementation process: 1. Media access and intelligent scheduling adaptation: The company's staff connected the Seagate hard drive under test to a dedicated data recovery workstation via a dedicated storage interface controller, maintaining a read-only hardware access mode throughout the process to lock write permissions at the lowest level and prevent secondary data overwriting. The system coordination and scheduling execution module automatically identified the characteristics of the hard drive's media, determining that the fault was a multi-layer overwrite fault caused by system formatting and reinstallation. It adaptively switched to the hard drive's dedicated track signal enhancement mode, and combined this with the degree of signal attenuation from media traces, designated the original system C drive and the corresponding sectors of the desktop as key scanning areas, generating dedicated scheduling instructions.

[0048] 2. Layered acquisition of physical signals across the entire domain: The global physical signal acquisition module sets the appropriate scanning frequency range and dynamic voltage control interval according to the scheduling instructions, performs multi-band layered scanning on key physical areas, and collects new data signals from the surface system and raw file remnant signals from the deep layer respectively. It generates a standardized raw signal dataset with physical address markers, signal amplitude markers, and scan level markers in real time, and pushes it to the multi-layer remnant extraction module via DMA transmission through the PCIe 4.0 high-speed bus.

[0049] 3. Multi-layer residue separation and fragment optimization: The multi-layer remnant extraction module receives the complete dataset via a high-speed interface and runs a multi-band signal separation algorithm based on the FPGA board and GPU computing cluster. It filters and transforms the frequency domain signals, peeling away surface system overwrite signals layer by layer to accurately extract deep-layer original file remnants. Noise and blank segments are removed using adaptive dynamic thresholding, generating a layered data fragment sequence with inter-layer boundary location markers and inter-layer correlation strength markers. Simultaneously, a cross-layer splicing algorithm is invoked to smooth the boundaries and merge highly correlated fragments, optimizing the discrete fragment structure. The optimized sequence is then transmitted to the sector topology directory reconstruction module.

[0050] 4. Topology mapping and directory structure reconstruction: The sector topology directory reconstruction module constructs a global residual topology mapping map based on the physical adjacency relationship and address continuity of sectors. It calculates the inter-layer fragment association strength through a residual metadata cross-association algorithm and completes cross-matching by combining residual log traces from the file system. At the same time, it performs weighted enhancement processing on residual metadata based on signal amplitude markers to improve the accuracy of directory reconstruction, fully restore the directory hierarchy, folder structure and file name information of the original files, generate standardized structured file objects and transmit them to the multi-layer content alignment and repair module.

[0051] 5. Content alignment repair and closed-loop convergence control: The multi-layer content alignment and repair module calculates the optimal offset using an inter-layer boundary alignment algorithm to achieve precise alignment of multi-layer fragments. Then, relying on a content pattern matching algorithm, it retrieves corresponding file format feature templates to intelligently fill and repair incomplete file areas, while simultaneously performing MD5 and SHA-256 dual integrity checks. Throughout the process, the coordination and scheduling execution module collects real-time operational parameters such as the separation completion rate of each module and the directory reconstruction integrity rate, dynamically optimizing scan priorities and collection parameters until the indicators meet the system convergence conditions, at which point the iterative operation terminates.

[0052] Case Summary: Using the system invented in this invention, the company successfully recovered 39GB of original, important data from a hard drive partition that had undergone multiple overwrites. Remote verification by medical personnel confirmed that core business files and historical data were intact and usable, effectively resolving the cooperation dispute caused by this data failure and fully validating the system's deep data recovery capabilities in scenarios involving system reinstallation and overwrite, and the failure of traditional technologies.

[0053] Case 2: Case Background: A user accidentally copied a large number of Windows system installation files to a 64GB USB drive during the operating system installation process. More than half of the USB drive's storage space was overwritten by the new data, leaving only a small amount of space remaining. The USB drive contained warehouse management system software, only a few tens of megabytes in size, which embedded all of the company's accounting data, delivery details, and linked to millions of outstanding payments. There were no paper, cloud, or other electronic backups available, and the data loss would directly lead to the company's bankruptcy. After the incident, the user contacted numerous data recovery agencies nationwide and consulted experts from a well-known domestic data recovery software development company. All of them determined that the underlying data was essentially wiped clean, with only invalid fragments remaining, making recovery impossible. The user ultimately entrusted the faulty USB drive to this company, relying on the system described in this invention to conduct a challenging data recovery operation.

[0054] Implementation process: 1. Fault analysis and refined scheduling configuration: After the company connected the USB flash drive, the coordination and scheduling execution module identified the attributes of the portable mobile storage medium. In response to the extreme fault characteristics of "overwriting micro target files with ultra-large volume data and extremely low residual trace ratio", the company abandoned the conventional full-disk coarse scanning mode and issued refined weak signal acquisition and scheduling instructions to enhance the ability to capture trace traces at the underlying level.

[0055] 2. Multi-band weak signal acquisition: The global physical signal acquisition module dynamically adjusts the scanning frequency and voltage parameters to weaken the strong signal interference from large-capacity installation files on the surface, focuses on the entire sector of the USB flash drive, collects weak signals from deeply buried software remnants, and generates a standardized raw signal dataset.

[0056] 3. Micro-residual fragment screening and cross-layer reconstruction: The multi-layer residual extraction module runs a multi-band signal separation algorithm to accurately screen out the effective residual fragments of the warehouse software from massive invalid overwrite signals; it tracks the correlation relationship of discrete fragments by marking the inter-layer correlation strength, and completes the reconstruction optimization of micro-fragments by relying on the cross-layer splicing algorithm, retaining all effective data fragments with recovery value.

[0057] 4. Reconstructing the niche software directory topology: The sector topology directory reconstruction module combines USB flash drive sector addressing rules and residual metadata traces to complete topology mapping and cross-association calculations for fragmented and scattered software remnants, accurately reconstructing the exclusive directory structure and program identifier of the warehouse management system software.

[0058] 5. Specialized document repair and iteration efforts: The multi-layer content alignment and repair module adapts to the specific format characteristics of the database and program files, completing fragment boundary alignment, incomplete content filling, and structural integrity restoration. The coordination and scheduling execution module iteratively optimizes scanning parameters and filtering thresholds multiple times based on real-time job feedback, continuously deepening the mining of underlying remnants. After several days of continuous intensive computation, the data repair operation was successfully completed.

[0059] Case Summary: Relying on the extremely strong ability of the system to extract weak traces and reconstruct discrete fragments, the company was able to completely restore the warehouse management system software and all the core data of the embedded accounts and delivery details under the extreme condition of a recovery probability of one in ten million. The software can run normally and is used normally, successfully saving the company from bankruptcy and highlighting the unique technical advantages of this system in ultra-difficult deep overwrite scenarios.

[0060] Case 3: Case Background: While organizing data on a 1TB 2.5-inch external hard drive, a user accidentally performed a full format. After discovering data loss, the user actively wrote new data back to the drive, resulting in a complex multi-layered overwrite failure caused by both formatting and secondary writing. The hard drive contained all of the user's important personal data, resulting in significant data loss. The user consulted multiple online and offline data recovery service providers, all of whom determined that the data was completely overwritten and unrecoverable. Subsequently, the user entrusted the faulty external hard drive to this company for deep recovery using the system described in this invention.

[0061] Implementation process: 1. Media adaptation and fault mode scheduling: After the company connected the portable hard drive, the coordination and scheduling execution module identified the media attributes of the large-capacity portable hard drive. For the composite fault characteristics of two overwrites, it activated the track signal enhancement mode and formulated a layered through-scan operation strategy.

[0062] 2. Full-area layered penetration scanning: The full-domain physical signal acquisition module refines the scanning granularity and expands the frequency scanning range, distinguishing between formatting residual signals, secondary write data signals and original data trace signals. It performs a multi-band layered scan of the entire hard drive sector without blind spots, completely acquiring trace signals at different depths and generating a standardized dataset.

[0063] 3. Multi-layer signal stripping and fragmentation optimization: The multi-layer trace extraction module uses a multi-band signal separation algorithm to peel off the surface overwrite data layer by layer, filter out invalid noise, extract the original data fragments at the bottom layer, and combine it with a cross-layer splicing algorithm to complete large-scale fragment integration and smoothing optimization, eliminating the fragment boundary fault problem.

[0064] 4. Complete directory reconstruction: The sector topology directory reconstruction module relies on the complete sector topology map and residual file system logs to batch reconstruct the original multi-level directory hierarchy, folder classification and original file names of the hard drive, so as to achieve complete restoration of the data structure.

[0065] 5. Full file repair and iterative optimization: The multi-layer content alignment and repair module performs boundary alignment, content filling, and integrity verification for various file formats, including documents, images, and videos. The coordination and scheduling execution module monitors job convergence metrics throughout the process, triggering multiple iterative scans to continuously improve the completeness of data recovery.

[0066] Case Summary: Through long-term, dedicated research and development using this invention's system, the company successfully recovered nearly 1TB of almost all original data from a portable hard drive. All files were structurally intact and could be opened and used normally. Users who remotely verified the data highly praised the recovery results, verifying that this system can efficiently solve the problem of data loss due to complex deep overwrite caused by formatting and secondary writing.

[0067] Case 4: Case Background: A senior executive at a listed company had his 1TB 2.5-inch Western Digital external hard drive accidentally formatted by a subordinate. To evade responsibility, the subordinate privately hired someone to recover the data, and then copied over 400GB of inaccessible and corrupted files back onto the hard drive, causing further damage to the original important data. After the incident, the user consulted dozens of data recovery services locally and online, all of which determined that the data was severely overwritten and unrecoverable. The user learned about deep overwrite data recovery technology through reviews on an e-commerce platform and sent the faulty hard drive to this company, where it was repaired using the proprietary system of this invention.

[0068] Implementation process: 1. Intelligent analysis and scheduling of complex faults: After the company connects the hard drive, the coordination and scheduling execution module identifies the media type and adaptively adjusts the scheduling parameters to address complex faults such as "formatting damage + secondary overwriting of garbled files + multi-layer signal interference," enabling an anti-interference deep recovery mode to avoid signal interference from surface-damaged garbled files.

[0069] 2. Differentiated signal acquisition by region: The global physical signal acquisition module performs layered scanning by region, frequency band, and voltage based on the signal attenuation differences in different areas, and collects surface scrambled code overwrite signals, middle formatted residual signals, and bottom original data remnant signals respectively, accurately distinguishing the signal data of the three layers with different attributes.

[0070] 3. Interference filtering and extraction of underlying traces: The multi-layer trace extraction module uses a multi-band signal separation algorithm to accurately filter abnormal interference signals generated by garbled files, separately peels off and purifies the underlying original data traces, and completes the reorganization and optimization of fragmented traces through a cross-layer splicing algorithm, eliminating invalid and isolated fragments.

[0071] 4. Accurate restoration of the original directory structure: The sector topology directory reconstruction module relies on the physical sector adjacency relationship and historical residual metadata, combined with weighted enhancement processing technology, to completely restore the multi-level directory structure, file storage path and original naming of the hard drive before it was lost, ensuring that the data classification is completely consistent with that before the loss.

[0072] 5. Algorithm Iteration and Full File Repair: The multi-layer content alignment and repair module uses boundary alignment and content pattern matching algorithms to complete the structural repair and content completion of the entire file, and double-verifies file availability. During the operation, the company's technicians relied on the system's optimizable features to continuously iterate and optimize the core algorithm parameters, coordinate the scheduling and execution modules to dynamically adjust the scanning strategy, and deepen the recovery effect through multiple rounds.

[0073] Case Summary: By leveraging the anti-interference signal processing and intelligent algorithm iteration optimization capabilities of the system invented in this invention, the company successfully recovered core original important corporate data from the hard drive. All files retained their original directory structure and could be opened and used normally, perfectly solving the difficult recovery problem caused by multi-layer complex overwriting and abnormal signal interference.

[0074] Case 5: Case Background: During routine bookkeeping, the accounting staff encountered a sudden computer malfunction, displaying a pop-up message indicating a file saving error. In their haste to confirm the operation, the staff mistakenly overwrote the current period's accounting file with the same name from previous years, resulting in the loss of nearly a year's worth of accounting data. This portion of the accounting data had no paper records or other electronic backups, and the data loss created a gap in the financial records. The user visited local data recovery centers and several online paid recovery services, but after more than half a month of unsuccessful attempts, they finally entrusted a 500GB Western Digital 3.5-inch hard drive to this company, where the system described in this invention assisted in accurate data recovery.

[0075] Implementation process: 1. Targeted and refined scheduling and deployment: After the company connected the hard drive, the coordination and scheduling execution module determined that the fault was a single point of precise overwriting of a file with the same name. Without the need for a full disk scan, it accurately located the partition where the accounting file was located and the corresponding logical sector, and issued a local high-precision scan scheduling command.

[0076] 2. Single-point sector depth signal acquisition: The global physical signal acquisition module focuses on the target file storage sector, improving the accuracy of single-point signal acquisition. It collects residual signals from both old and new accounting files at the same physical address, generating a high-precision local raw signal dataset.

[0077] 3. Separate signals from files with the same name into different layers: The multi-layer trace extraction module relies on the frequency attenuation differences of data signals from different periods to accurately separate the traces of the current accounting documents that are covered, remove signal interference from historical old documents, extract valid accounting data fragments, and complete splicing optimization.

[0078] 4. Metadata Enhancement and Catalog Reconstruction: The sector topology directory reconstruction module combines file system residual operation logs to lock the exclusive directory node of the file with the same name, and strengthens the residual metadata features through a signal amplitude weighting enhancement algorithm to accurately restore the target file identifier and storage location.

[0079] 5. Specific verification and repair of accounting documents: The multi-layer content alignment and repair module adapts to the format characteristics of financial documents, completing fragment boundary alignment and intelligent filling of incomplete content. Double hash verification ensures the integrity and accuracy of accounting data. The coordination and scheduling execution module fine-tunes local acquisition parameters, further improving the accuracy of single-point sector remnant restoration.

[0080] Case Summary: The company successfully recovered core current accounting data that had been overwritten by files with the same name by using the system's pinpoint precision scanning and layered signal separation capabilities. The data was complete and the accounts were continuous, fully meeting the needs of financial reconciliation and efficiently solving the problem of recovering lost critical files with single-point precision overwriting.

[0081] Case Six: Case Background: A company employee made a mistake during operating system installation, incorrectly installing the system, which should have been installed on the solid-state drive (SSD), onto the first partition of the hard drive (HDD). This resulted in the system installation files extensively overwriting all company office files and core business data on that partition. After the incident, the user attempted to recover the data independently using conventional recovery software, but found no valid data. They then consulted several online technicians, all of whom also failed to resolve the issue. The user learned about this deep overwrite data recovery technology through industry channels and sent the hard drive to this company for specialized recovery.

[0082] Implementation process: 1. Heavy overwrite mode adaptation scheduling: After the company connected the mechanical hard drive, the coordination and scheduling execution module identified the media properties, switched the track signal enhancement mode, determined that the first partition of the hard drive was a heavily overwritten area, and issued a penetrating deep scan scheduling command to focus on enhancing the ability to collect deep traces.

[0083] 2. Deep penetration signal acquisition: The full-domain physical signal acquisition module lowers the scanning frequency and optimizes the dynamic reading voltage, penetrates the thick system installation file overwrite layer, deeply collects the weak residual signals of the underlying original office files, and generates a complete standardized signal dataset.

[0084] 3. Business data remnant reconstruction: The multi-layer trace extraction module effectively strips surface system file signals through a multi-band signal separation algorithm, extracts effective fragments of underlying office documents and business data, and integrates and optimizes fragmented data by relying on a cross-layer splicing algorithm.

[0085] 4. Rebuild the partitioned business catalog: The sector topology directory reconstruction module reconstructs the original multi-level business directory structure and file names of the first partition of the hard drive based on the sector topology mapping relationship, restoring the original file classification system of the enterprise.

[0086] 5. Batch office file repair and iteration: The multi-layer content alignment and repair module adapts to various office file formats, performing batch fragment alignment, content repair, and integrity verification. The coordination and scheduling execution module continuously improves the scanning priority of key partitions, and through multiple rounds of iteration and in-depth work, completes the repair of all recoverable data.

[0087] Case Summary: Through the deep penetration scanning and batch data repair capabilities of the system invented in this invention, the company has successfully recovered more than 50G of core office data. All files can be opened and used normally, saving the company from work losses and business risks, and verifying the system's adaptability to scenarios such as system misinstallation partitions and large-scale heavy overwrite.

[0088] Case Seven: Case Background: A user entrusted a third-party maintenance technician to reinstall their computer's operating system. The technician failed to back up the system desktop data beforehand and directly formatted and reinstalled the system, resulting in the complete overwriting of most of the user's important files stored on the desktop. After numerous unsuccessful attempts at conventional recovery methods, the user ultimately handed the faulty hard drive over to this company for data recovery using the system described in this invention.

[0089] Implementation process: 1. Regular overwrite scenario scheduling configuration: After the company connected the 3.5-inch hard drive, the coordination and scheduling execution module started the hard drive's standard operating mode, accurately marked the dedicated logical sector corresponding to the system desktop, and set it as the core scanning area.

[0090] 2. Desktop area directional layered scanning: The global physical signal acquisition module focuses on the desktop file storage sector, performs multi-band layered scanning, and acquires surface system new file signals and deep original desktop file remnant signals layer by layer to generate a standardized raw signal dataset.

[0091] 3. Surface signal stripping and fragmentation optimization: The multi-layer trace extraction module uses a multi-band signal separation algorithm to strip away surface system overwrite traces, extracts valid fragments of the original desktop file, and completes smooth fusion and structural optimization of fragments through a cross-layer splicing algorithm.

[0092] 4. Restore desktop-specific directories: The sector topology directory reconstruction module relies on residual metadata and directory logs to accurately restore the system desktop's unique directory structure, file names, and storage order.

[0093] 5. Unified repair of multiple file types: The multi-layer content alignment and repair module performs boundary alignment, content filling, and availability verification for various file formats such as documents, images, and data stored on the desktop, ensuring that all files are complete and usable.

[0094] Case Summary: Using the system developed in this invention, the company successfully recovered most of the important files from the user's original desktop, with no core data missing and the files fully functional. This case demonstrates that the system possesses stable and efficient recovery performance for high-frequency system reinstallation and desktop file overwriting scenarios for individuals and small businesses.

[0095] Case 8: Case Background: A university faculty member connected a 2TB 3.5-inch Seagate hard drive containing years of academic research materials and graduate lecture notes to a computer. Due to a computer malfunction, the operating system was mistakenly installed on the hard drive, causing the entire drive to be overwritten with system files, resulting in the loss of all research and teaching materials. After the incident, the user consulted numerous high-end data recovery agencies nationwide, all of which determined that the hard drive had undergone a deep overwrite and was beyond recovery value. Desperate to recover years of research results and teaching materials, the user entrusted the hard drive to this company for a intensive recovery operation.

[0096] Implementation process: 1. Large-capacity full-domain overwrite special scheduling: After the company connected to a large-capacity mechanical hard drive, the coordination and scheduling execution module started a high-load operation mode specifically for large-capacity media. In response to the fault characteristics of full disk deep overwrite and extremely weak residual signals, a scheduling strategy of long-term iterative scanning and high-precision signal acquisition was configured.

[0097] 2. Comprehensive, no-blind-spot deep scan: The full-domain physical signal acquisition module performs full-disk, full-domain multi-band layered scanning, dynamically and adaptively adjusting the reading voltage and scanning frequency to maximize the penetration of surface system data overwriting and deeply mine the weak residual signals of teaching and research materials inside the disk tracks.

[0098] 3. Parallel processing of massive amounts of residual data: The multi-layer trace extraction module relies on the massive parallel computing power of GPUs to perform layered separation, noise filtering, and feature extraction on massive full-domain signal data. It accurately separates effective fragments of academic materials and lecture notes from the full-coverage cluttered signal and completes large-scale fragment recombination and optimization.

[0099] 4. Reconstruction of multi-level scientific research catalogs: The sector topology directory reconstruction module constructs a complete topology map, and combines the long-accumulated characteristics of scientific research file storage to restore the multi-level scientific research data directory, lecture handout classification structure and file naming system formed over many years.

[0100] 5. Long-term iterative repair and tackling key challenges: The multi-layer content alignment and repair module adapts to the format characteristics of teaching and research documents and text materials, completing fragment alignment and accurate content repair. The coordination and scheduling execution module monitors hardware load, signal-to-noise ratio, and job convergence indicators throughout the process, continuously launching multiple rounds of iterative deep scanning to continuously improve the completeness of residual extraction.

[0101] Case Summary: After a long period of intensive, round-the-clock work, the company successfully recovered most of the core data, including academic research materials and graduate lecture notes, from the hard drive using the system developed in this invention. The key research and teaching materials required by the user were essentially preserved intact. This case fully demonstrates the system's core capabilities in handling large-capacity hard drives with comprehensive, in-depth coverage and recovering extremely complex data.

[0102] Case Nine: Case Background: A company's 500GB 3.5-inch business hard drive had undergone a system reinstallation and subsequently continued to read and write new data for an extended period. The medical insurance system log files for the specific time period requiring recovery were buried under multiple layers of new data, located at the bottom of the storage layer, with extremely severe signal attenuation. After inspection by several nationally renowned data recovery agencies, all determined that the deep log data was untraceable and unrecoverable. Through an industry referral, the user entrusted the hard drive to this company to attempt to recover critical business log evidence data.

[0103] Implementation process: 1. Special scheduling of in-depth niche documents: After the company connected the hard drive, the coordination and scheduling execution module issued special refined scheduling instructions based on the characteristics of log files, such as "deeply buried at the bottom layer, weak signal, small file size, and strong time sequence", to weaken the strong signal interference of data written to the surface layer later.

[0104] 2. Acquisition of weak log signals at the lower level: The full-domain physical signal acquisition module locks onto the dedicated storage partition of the medical insurance system logs, improving signal acquisition accuracy, optimizing the scanning frequency band for weak signals, and focusing on capturing the residual signals of historical logs buried deep in the underlying layer.

[0105] 3. Precise extraction of discrete log fragments: The multi-layer remnant extraction module optimizes the signal separation and fragment splicing logic based on the unique signal characteristics of log text data, and accurately filters and extracts discretely distributed valid log fragments from multi-layer overwrite data.

[0106] 4. Locating and rebuilding the time-series log directory: The sector topology directory reconstruction module combines the storage rules, temporal characteristics, and residual metadata of the medical insurance system logs to accurately locate log files within the target time period and reconstruct the temporal directory and identification information of the log files.

[0107] 5. Log content repair and integrity verification: The multi-layer content alignment and repair module completes log fragment boundary alignment, time-series content completion and format regularization, verifies the continuity and integrity of log content, and ensures that log records can be used for business verification and data traceability.

[0108] Case Summary: The company successfully recovered the medical insurance system log files required for a specified time period by relying on the deep and weak trace extraction and precise location of niche time-series files of the system of this invention. The log content is continuous and complete, which meets the core needs of the unit for business verification and data evidence collection, and verifies the unique advantages of this system in the recovery of deep, hidden and niche key files.

[0109] Overall application summary: These real-world commercial scenarios comprehensively cover all the mainstream and challenging data overwrite faults in the industry, including system reinstallation and overwriting, accidental overwriting of micro core data by large-capacity files, formatting and superimposed secondary writing, secondary interference overwriting of garbled files, accurate overwriting of files with the same name, system misinstallation partition overwriting, full-domain deep overwriting of large-capacity hard drives, and overwriting of deeply buried log files. It is compatible with all types of mainstream storage media such as mechanical hard drives, external hard drives, and USB flash drives.

[0110] It should be noted that, in this document, relational terms such as "first" and "second" are used merely to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising a reference structure" does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes the element.

[0111] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A full-domain deep coverage intelligent data recovery scheduling system, characterized in that, It includes a global physical signal acquisition module, a multi-layer residual extraction module, a sector topology directory reconstruction module, a multi-layer content alignment and repair module, and a coordination and scheduling execution module. The full-domain physical signal acquisition module is used to perform a full-domain physical scan on the storage medium that has been covered by data, acquire the original signal dataset, and transmit the original signal dataset to the multi-layer trace extraction module; The multi-layer trace extraction module is used to receive the original signal dataset, extract the covering traces layer by layer through a multi-band signal separation algorithm, generate a layered data fragment sequence, and transmit the layered data fragment sequence to the sector topology directory reconstruction module. The sector topology directory reconstruction module is used to receive the hierarchical data fragment sequence, reconstruct the directory hierarchy and file name information through physical sector location mapping and residual metadata cross-association algorithm, and transmit the reconstructed structured file object to the multi-layer content alignment and repair module. The multi-layer content alignment and repair module is used to receive the structured file object, repair the file through inter-layer boundary alignment algorithm and content pattern matching algorithm, and output a recoverable file that can be opened directly; The coordination and scheduling execution module is used to receive processing status feedback from each module, generate scheduling instructions based on the layered distribution of remnants, and transmit the scheduling instructions to the global physical signal acquisition module, the multi-layer remnant extraction module, the sector topology directory reconstruction module, and the multi-layer content alignment and repair module for coordinated operation.

2. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 1, characterized in that, When performing a full-domain physical scan, the full-domain physical signal acquisition module first determines the initial scanning frequency range and key physical areas based on the number of times the storage medium is covered and the degree of attenuation of the residual signal. Then, it performs multi-band layered scanning, and then collects and generates a standardized raw signal dataset containing physical address markers, signal amplitude markers and scan level markers in real time. Finally, it transmits the dataset to the multi-layer residual extraction module.

3. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 1, characterized in that, After receiving the original signal dataset, the multi-layer remnant extraction module first separates the overlay remnants layer by layer based on the differences in attenuation characteristics of signals at different frequencies. Then, it generates a layered data fragment sequence with inter-layer boundary location markers and inter-layer correlation strength markers. Next, it tracks the remnants between multiple overlay layers based on the inter-layer correlation strength markers. Finally, it transmits the processed layered data fragment sequence to the sector topology directory reconstruction module.

4. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 3, characterized in that, After receiving the hierarchical data fragment sequence, the sector topology directory reconstruction module first establishes a residual topology mapping relationship based on the sector physical adjacency relationship and address continuity. Then, it performs cross-association calculations by combining residual directory records and file system log traces. Next, it generates a complete directory hierarchy structure and file name information. Finally, it transmits the reconstructed structured file object to the multi-layer content alignment and repair module.

5. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 4, characterized in that, After receiving the structured file object, the multi-layer content alignment and repair module first performs boundary alignment processing on the layered data fragments according to the inter-layer boundary position marks. Then, under the condition of maintaining the file format constraints, it fills the missing parts through the content pattern matching algorithm. Next, it completes the usability verification of the file structure and content, and finally outputs a recoverable file that can be opened directly.

6. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 3, characterized in that, In the process of separating overlay layer remnants, the multi-layer remnant extraction module first tracks the remnant association relationship between multiple overlay layers based on the inter-layer association strength marker, then uses a cross-layer splicing algorithm to optimize the splicing of associated fragments, then generates an optimized layered data fragment sequence, and finally transmits it to the sector topology directory reconstruction module.

7. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 4, characterized in that, When reconstructing the directory hierarchy and file name information, the sector topology directory reconstruction module first performs weighted enhancement processing on the residual metadata based on the signal amplitude marker, then further enhances the accuracy of directory hierarchy reconstruction, then generates an enhanced structured file object, and finally transmits it to the multi-layer content alignment and repair module.

8. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 1, characterized in that, When generating scheduling instructions, the coordination and scheduling execution module first identifies the physical characteristics of the current storage medium, and then generates corresponding media-adaptive scheduling instructions according to the medium type. For different levels of mechanical hard drives, a matching method is adopted. Specifically, consumer-grade mechanical hard drives use page-level charge trace extraction mode, while enterprise-grade mechanical hard drives use track signal enhancement mode. Then, the scheduling instructions are transmitted to the global physical signal acquisition module and the multi-layer trace extraction module to adjust the processing parameters.

9. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 1, characterized in that, After receiving the processing status feedback from each module, the coordination and scheduling execution module first collects the separation completion rate of the multi-layer trace extraction module and the reconstruction completeness of the sector topology directory reconstruction module. Then, it generates an optimized scheduling instruction based on the layered distribution of traces and transmits the scheduling instruction to the global physical signal acquisition module to adjust the frequency parameters and priority of key areas for subsequent scanning.

10. The intelligent data recovery and scheduling system with full-domain deep coverage according to claim 1, characterized in that, When the coordination and scheduling execution module performs convergence control on the entire recovery process, it first monitors the directory hierarchy integrity index and the file repair success rate index in real time. Then, it determines whether each index meets the convergence conditions. If the convergence conditions are met, the recovery process is terminated. If the convergence conditions are not met, the full-domain physical signal acquisition module is triggered to perform a new round of multi-band layered scanning to continue optimizing the recovery results.

Images