Dynamic storage system, dynamic storage method, electronic equipment and storage medium

The dynamic storage system for SSDs addresses the lack of adaptability in current SSDs by integrating environmental sensing and user behavior prediction to enhance reliability and reduce error rates through adaptive storage policies.

CN120315645APending Publication Date: 2025-07-15INSPUR SUZHOU INTELLIGENT TECH CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202510457484.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-04-11
Publication Date
2025-07-15

AI Technical Summary

Technical Problem

Currently, solid-state drives (SSDs) lack active adaptability in data storage and cannot dynamically adjust storage strategies, resulting in high bit error rate and short service life in extreme environments, making it difficult to meet the real-time needs of scenarios such as edge computing and autonomous driving.

Method used

The environment perception module is used to obtain the operating environment parameters, combine the user behavior prediction module to generate the environment survival weight and user behavior weight, and dynamically adjust the SSD storage strategy through the storage strategy generation module, including independent decision-making and bionic hardware design, to realize the hardware-level self-repair and adaptive functions.

Benefits of technology

It improves the active adaptability of SSD, reduces the bit error rate, extends the service life, reduces power consumption, and meets the real-time needs of scenarios such as edge computing and autonomous driving.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN120315645A_ABST
    Figure CN120315645A_ABST
Patent Text Reader

Abstract

The invention discloses a dynamic storage system, a dynamic storage method, electronic equipment and a storage medium, and relates to the technical field of data storage. The dynamic storage system comprises an environment perception module, a user behavior prediction module and a storage strategy generation module. The environment sensing module is used for acquiring operation environment parameters of the solid state disk and generating an environment survival weight based on the operation environment parameters; the user behavior prediction module is used for acquiring historical access data of the solid state disk and generating a user behavior weight based on the historical access data; the storage strategy generation module is used for determining a dynamic storage strategy of the solid state disk according to the environment survival weight and the user behavior weight, the technical problem that a current solid state disk lacks active adaptive capacity in the aspect of data storage is solved, and the dynamic storage strategy of the solid state disk is obtained according to environment perception and user behavior prediction. The technical effects of dynamically adjusting the storage strategy of the solid state disk, improving the active adaptive capacity of the solid state disk, reducing the bit error rate and improving the reliability of data storage are achieved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This application relates to the technical field of data storage, and in particular, to a dynamic storage system, a dynamic storage method, an electronic device, and a storage medium. Background Art

[0002] As the world enters the digital economy era, data has become one of the core production factors leading economic development. Data centers of various scales, as the infrastructure of the digital economy, make data storage technology a crucial cornerstone technology. Solid State Drives (SSDs) are widely used in consumer electronics, data centers, industrial control, and other fields. However, current SSDs lack the ability to actively adapt in data storage. Summary of the Invention

[0003] This application provides a dynamic storage system, a dynamic storage method, a dynamic storage device, an electronic device, a computer-readable storage medium, and a computer program product to at least solve the problem that SSDs lack the ability to actively adapt in data storage in related technologies.

[0004] This application provides a dynamic storage system, including:

[0005] An environment perception module, configured to obtain the operating environment parameters of the solid state drive and generate an environment survival weight based on the operating environment parameters;

[0006] A user behavior prediction module, configured to obtain the historical access data of the solid state drive and generate a user behavior weight based on the historical access data;

[0007] A storage policy generation module, configured to determine the dynamic storage policy of the solid state drive according to the environment survival weight and the user behavior weight.

[0008] This application also provides a dynamic storage method, including:

[0009] Obtain the operating environment parameters of the solid state drive and generate an environment survival weight based on the operating environment parameters;

[0010] Obtain the historical access data of the solid state drive and generate a user behavior weight based on the historical access data;

[0011] Determine the dynamic storage policy of the solid state drive according to the environment survival weight and the user behavior weight.

[0012] This application also provides a dynamic storage device, including:

[0013] A first acquisition module, configured to obtain the operating environment parameters of the solid state drive and generate an environment survival weight based on the operating environment parameters;

[0014] A second acquisition module, configured to acquire historical access data of the solid-state drive and generate a user behavior weight based on the historical access data;

[0015] A policy determination module, configured to determine a dynamic storage policy of the solid-state drive according to the environmental survival weight and the user behavior weight.

[0016] This application also provides an electronic device, including: a memory, configured to store a computer program; a processor, configured to implement the steps of the above dynamic storage method when executing the computer program.

[0017] This application also provides a computer-readable storage medium, in which a computer program is stored, and when the computer program is executed by a processor, the steps of the above dynamic storage method are implemented.

[0018] This application also provides a computer program product, including a computer program, and when the computer program is executed by a processor, the steps of the above dynamic storage method are implemented.

[0019] Through this application, since the storage policy generation module determines the dynamic storage policy of the solid-state drive according to the environmental survival weight and the user behavior weight, it is possible to solve the technical problem that the current solid-state drive lacks the ability of active adaptation in data storage, and achieve the technical effect of dynamically adjusting the storage policy of the solid-state drive according to the perception of the environment and the prediction of the user behavior, improving the active adaptation ability of the solid-state drive, reducing the error rate, and improving the reliability of data storage. Description of the Drawings

[0020] To more clearly illustrate the embodiments of this application, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of this application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0021] Figure 1 A schematic diagram of the modules of a dynamic storage system provided by an embodiment of this application;

[0022] Figure 2 A schematic diagram of the modules of an environment perception module provided by an embodiment of this application;

[0023] Figure 3 A schematic diagram of the working process of a dynamic storage system provided by an embodiment of this application;

[0024] Figure 4 A schematic diagram of the flow of a dynamic storage method provided by an embodiment of this application;

[0025] Figure 5 A schematic diagram of a module of a dynamic storage device provided by an embodiment of the present application;

[0026] Figure 6 A schematic diagram of a module of an electronic device provided by an embodiment of the present application;

[0027] Figure 7 A schematic diagram of a module of a computer-readable storage medium provided by an embodiment of the present application. Detailed implementation manners

[0028] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0029] It should be noted that in the description of the present application, the terms "include", "comprise" or any other variant thereof are intended to cover a non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements but also includes other elements not explicitly listed, or further includes elements inherent to such a process, method, article or device. The terms "first", "second", etc. in the present application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0030] In order to enable those skilled in the art of the present technology to better understand the solution of the present application, the present application will be further described in detail below in conjunction with the accompanying drawings and specific implementation manners.

[0031] As is well known, the world is accelerating into the digital economy era. In this process, data, as the fifth production factor, its strategic position has become increasingly prominent and has gradually become the core driving force leading economic development. At the same time, various large, medium and small data centers, as the cornerstone facilities of the digital economy, are comprehensively supporting the huge demand for digital transformation. Against this background, data storage technology, as a key support for new infrastructure projects, is of great importance.

[0032] At present, the solid state drive (SSD) industry is welcoming a dual opportunity period of technology iteration and market expansion. Driven by policy support and technological breakthroughs, domestic enterprises are accelerating their rise, but the industry still faces the dual challenges of core technology shortcomings and international competition. Looking to the future, the deep integration of artificial intelligence (AI) technology, the popularization of green storage concepts, and the promotion of domestic substitution strategies will become the core driving force for the continued growth of the SSD industry.

[0033] From the perspective of application scenarios, SSDs have been widely used in diversified fields such as consumer electronics, data centers, and industrial control. Among them, the rapid development of AI servers and autonomous driving technologies is driving the explosive growth of enterprise-level SSD market demand. However, current SSDs still have obvious limitations, such as the lack of active adaptability to the external environment and high reliance on host instructions to execute storage strategies, making it difficult to meet the stringent requirements of edge computing and extreme application scenarios.

[0034] Research has found that current SSDs mostly use a global wear leveling system to optimize storage life through multi-dimensional data, but they do not take into account the impact of environmental variables and lack a dynamic response mechanism to environmental variables.

[0035] In view of this, the embodiment of the present application provides a bionic dynamic storage system 100 based on environment perception to solve at least one of the following technical problems existing in current SSDs:

[0036] (1) Insufficient environmental adaptability: Current SSDs lack the ability to dynamically perceive and adapt to changes in external environments such as temperature and vibration, and are unable to dynamically adjust storage strategies. In extreme scenarios, the lack of a real-time voltage compensation mechanism can easily lead to a sudden increase in the original bit error rate or a significantly increased risk of data loss.

[0037] (2) Lack of hardware-level optimization: The relevant technology mainly relies on software algorithm optimization, and does not introduce bionic principles to design physical media, making it difficult to achieve self-repair and adaptive functions at the hardware level. For example, the data retention period of quad-level cell (QLC) storage particles in a high temperature environment of 70°C is sharply shortened from 3 months at normal temperature to 72 hours; the current wear leveling algorithm is only optimized based on the distribution of program / erase (P / E) cycles, and does not incorporate environmental stress factors, that is, it does not consider the impact of environmental stress on storage life.

[0038] (3) Weak edge intelligence decision-making ability: Currently, SSDs rely on host instructions to execute storage management, resulting in excessively high decision-making latency and making it difficult to meet the real-time requirements of scenarios such as autonomous driving and industrial Internet of Things. For example, the autonomous driving system requires a storage access latency of less than 100 μs, while the current cloud decision-making architecture cannot meet this latency requirement; industrial Internet of Things devices are limited by network bandwidth and cannot upload a complete environmental data set in real time to support cloud decision-making.

[0039] Please refer to Figure 1 , the dynamic storage system 100 of the embodiments of the present application includes an environmental perception module 10, a user behavior prediction module 20, and a storage policy generation module 30. The environmental perception module 10 is used to obtain the operating environment parameters of the solid-state drive and generate an environmental survival weight based on the operating environment parameters. The user behavior prediction module 20 is used to obtain the historical access data of the solid-state drive and generate a user behavior weight based on the historical access data. The storage policy generation module 30 is used to determine the dynamic storage policy of the solid-state drive according to the environmental survival weight and the user behavior weight.

[0040] In the dynamic storage system 100 of the embodiments of the present application, the environmental perception module 10 is used to obtain the operating environment parameters of the solid-state drive and generate an environmental survival weight based on the operating environment parameters; the user behavior prediction module 20 is used to obtain the historical access data of the solid-state drive and generate a user behavior weight based on the historical access data; the storage policy generation module 30 is used to determine the dynamic storage policy of the solid-state drive according to the environmental survival weight and the user behavior weight. In this way, the storage policy of the solid-state drive can be dynamically adjusted according to the perception of the environment and the prediction of user behavior, improving the active adaptation ability of the solid-state drive, reducing the error rate, and improving the reliability of data storage.

[0041] Specifically, please combine Figure 2 and Figure 3 , the environmental perception module 10 is also the environmental perception layer. The environmental perception module 10 can collect the operating environment parameters of the SSD in real time by setting embedded sensors (such as a temperature sensor 11, a vibration sensor 12, an electromagnetic interference detection unit 13, etc.) to generate an environmental survival weight based on the operating environment parameters.

[0042] The user behavior prediction module 20 is also the user behavior prediction layer. The user behavior prediction module 20 can adopt an AI model deployed on the main control chip of the SSD, such as a Tiny Machine Learning (TinyML) model. The user behavior prediction module 20 generates a user behavior weight based on the historical access data by obtaining the historical access data of the solid-state drive.

[0043] The storage policy generation module 30, i.e., the data fusion and policy generation layer. The storage policy generation module 30 combines the environmental survival weight and the user behavior weight, and can autonomously generate and trigger the storage policy of the solid-state drive. In one example, the storage policy generation module 30 can adopt an edge intelligent decision engine. This engine has the ability of online learning, can dynamically update the weight matrix through the reinforcement learning framework, supports local autonomous generation and triggering of storage policies, can reduce decision latency, and reduce the dependence on the host.

[0044] In summary, the dynamic storage system 100 of the embodiment of the present application can dynamically adjust the storage policy of the solid-state drive according to the perception of the environment and the prediction of user behavior, improving the active adaptation ability of the solid-state drive. In addition, the dynamic storage system 100 of the embodiment of the present application can also determine a more accurate and reasonable storage policy, thereby extending the life of the storage unit.

[0045] Next, the hardware composition of the environment perception module 10 will be introduced.

[0046] Please refer to Figure 2 and Figure 3 , in some embodiments, the environment perception module 10 includes a temperature sensor 11, a vibration sensor 12, and an electromagnetic interference detection unit 13. The temperature sensor 11 is used to detect the temperature parameter of the solid-state drive. The vibration sensor 12 is used to detect the vibration frequency parameter of the solid-state drive. The electromagnetic interference detection unit 13 is used to detect the noise fluctuation parameter of the solid-state drive.

[0047] In this embodiment, by comprehensively detecting various operating environment parameters of the solid-state drive, including temperature parameters, vibration frequency parameters, noise fluctuation parameters, etc., multi-dimensional environment perception can be realized, the adaptability to complex environments can be improved, and the storage policy of the solid-state drive can be dynamically adjusted based on various operating environment parameters.

[0048] Specifically, the temperature sensor 11 can adopt a negative temperature coefficient thermistor or an integrated temperature sensor, etc. The temperature sensor 11 can be set on the back side of the printed circuit board (PCB) of the main control chip of the SSD to collect the temperature parameter of the SSD in real time.

[0049] The temperature sensor 11 may have a first operating mode and a second operating mode. In the first operating mode and the second operating mode, the sampling frequencies of the temperature sensor 11 for collecting temperature parameters are different. In an example, the temperature parameter corresponding to the second operating mode is higher than that corresponding to the first operating mode, and the sampling frequency of the temperature sensor 11 in the second operating mode is higher than that in the first operating mode. The first operating mode is, for example, a normal mode, and the second operating mode is, for example, a high-temperature warning mode. Thus, within a normal temperature range, the temperature sensor 11 can operate at a lower sampling frequency to periodically monitor the change in the operating temperature. When the temperature rises to a certain level, the temperature sensor 11 will switch to the high-temperature warning mode, and the sampling frequency will be significantly increased to more frequently monitor the change in the operating temperature, so as to timely adjust the storage strategy of the SSD and ensure the stable operation of the SSD.

[0050] The vibration sensor 12 can be a piezoelectric or capacitive vibration sensor, etc. The vibration sensor 12 communicates with the main control chip of the SSD through the Serial Peripheral Interface (SPI) to detect the vibration frequency parameter of the SSD in real time.

[0051] The electromagnetic interference detection unit 13 can use a voltage / current monitoring chip, and based on the voltage / current monitoring chip, the noise fluctuation parameter in the power supply line of the SSD can be detected in real time.

[0052] Some corresponding controls can be performed on the solid-state drive according to the above temperature parameter, vibration frequency parameter, and noise fluctuation parameter.

[0053] For example, when it is determined according to the temperature parameter that the operating temperature rises, the data storage rate of the whole SSD is limited, thereby reducing the overall power consumption of the SSD and lowering the temperature of the SSD. Further, when the operating temperature rises to a critical value, the read or write of the SSD is stopped, thereby reducing the risk of data errors.

[0054] For another example, when it is determined according to the vibration frequency parameter that the vibration frequency is too high, interface reinforcement technology is used, such as adding shock-absorbing materials at the interface, to reduce the conduction of physical vibration; for the PCB board, anti-seismic materials are used, or an interleaved capacitor array layout is adopted on the PCB board; in terms of algorithm processing, when the vibration frequency increases to a certain extent, the dynamic mapping algorithm is started, or the background garbage collection (GC) operation is paused to ensure the stability of reading and writing, or when the vibration frequency reaches a higher level, some encryption algorithms are used to prevent data leakage caused by physical side-channel attacks and increase the encryption frequency during vibration, etc.

[0055] For another example, when it is determined that the electromagnetic interference is too high based on the noise fluctuation parameter, a shielding layer is set at the PCB board and key interfaces for protection, and the shielding layer is used to absorb or reflect to reduce the interference; or a fully metal casing is used to enable the SSD to conduct electricity continuously; in terms of firmware algorithms, parity check integrity verification and related schemes for signal integrity compensation are used to prevent interference.

[0056] However, the above control of temperature parameters, vibration frequency parameters, and noise fluctuation parameters is mainly processed at the physical level and algorithm level, and all are separate processing strategies. Some processing strategies may also need to combine with the server side or the host side to select the strategy. However, in actual applications, the environmental situation is usually more complex, and there may be various combinations of environmental data. In the case of combinations, weight allocation is required, and the SSD is not dependent on the host side to select the strategy and performs self-machine learning. The following introduces the scheme of the environmental perception module 10 for generating environmental survival weights based on the operating environment parameters.

[0057] Please refer to Figure 1 and Figure 3 , in some embodiments, the environmental perception module 10 is used to determine the temperature weight based on the temperature parameter, the vibration weight based on the vibration frequency parameter, the electromagnetic interference weight based on the noise fluctuation parameter, and determine the environmental survival weight according to the temperature weight, vibration weight, and electromagnetic interference weight.

[0058] In this embodiment, the corresponding weights, namely the temperature weight, vibration weight, and electromagnetic interference weight, are determined based on the temperature parameter, vibration frequency parameter, and noise fluctuation parameter respectively, and then the weight fusion is performed to comprehensively determine the environmental survival weight, rather than adopting a separate processing strategy for each aspect of the parameters as in the related art, which can further improve the adaptability to complex environments.

[0059] Specifically, in an example, the formula for determining the temperature weight based on the temperature parameter is as follows:

[0060] w_temp = 1 / (1 + exp(0.1*(temp - T)))

[0061] Among them, w_temp is the temperature weight, temp is the temperature parameter obtained in real time, and T is the set temperature threshold, such as 70°C. In this embodiment, the temperature weight is a non-linear model, and the high-temperature penalty coefficient is higher.

[0062] In one implementation, the formula for determining the vibration weight based on the vibration frequency parameter is as follows:

[0063] w_vib = 1 - min(1, vibration_freq / V)

[0064] Among them, w_vib is the vibration weight, vibration_freq is the vibration frequency parameter obtained in real time, and V is the set vibration frequency threshold, such as 2000 Hz. In this embodiment, the vibration weight is more sensitive to high-frequency vibrations, and above 2000 Hz, the vibration weight is zeroed.

[0065] In one embodiment, the formula for determining the electromagnetic interference weight based on the noise fluctuation parameter is as follows:

[0066] w_emi = 1, if emi < E, else 0.5

[0067] Among them, w_emi is the electromagnetic interference weight, emi is the noise fluctuation parameter obtained in real time, and E is the noise fluctuation threshold, such as 0.1 V. In this embodiment, when the noise fluctuation parameter emi obtained in real time is less than 0.1 V, the electromagnetic interference weight w_emi is set to 1; when the noise fluctuation parameter emi obtained in real time is greater than or equal to 0.1 V, the electromagnetic interference weight w_emi is set to 0.5 to suppress voltage fluctuations.

[0068] The environmental survival weight can be expressed as def env_weight(temp, vibration, emi). In one embodiment, the formula for determining the environmental survival weight based on the temperature weight, vibration weight, and electromagnetic interference weight is as follows:

[0069] def env_weight = A * w_temp + B * w_vib + C * w_emi.

[0070] Among them, A is the coefficient of the temperature weight, B is the coefficient of the vibration weight, and C is the coefficient of the electromagnetic interference weight. In an example, A > C, B > C. For example, A = 0.4, B = 0.4, C = 0.2.

[0071] It has been found through research that the bit error rate of SSDs is usually more strongly correlated with temperature and vibration than with electromagnetic interference. For example, at high temperatures (>70 °C), the lifespan decay rate of SSDs can be more than twice that of electromagnetic interference (>50 dB). Therefore, relatively larger coefficients can be assigned to the temperature weight and vibration weight, while a relatively smaller coefficient can be assigned to the electromagnetic interference weight. In this way, by reasonably allocating the weight coefficients, the adaptability and reliability of the storage strategy can be improved.

[0072] In one embodiment, the environmental perception module 10 can also adopt a multi-modal data fusion technology for the temperature weight, vibration weight, and electromagnetic interference weight, and combine a time series analysis algorithm to predict the trend of the operating environment parameters, thereby forming an environmental risk index, that is, the environmental survival weight, to quantify the instantaneous reliability of the storage medium.

[0073] The process by which the user behavior prediction module 20 obtains the historical access data of the solid-state drive and generates user behavior weights based on the historical access data will be introduced below.

[0074] Please refer to Figure 1 and Figure 3 In some embodiments, the user behavior prediction module 20 includes a Long Short-Term Memory (LSTM) model. That is to say, in this embodiment, the TinyML model can adopt a lightweight LSTM model.

[0075] The training data of the LSTM model includes the number of Input / Output Operations Per Second (IOPS), file size distribution, and access time interval of the solid-state drive within a predetermined time period before the current moment. In one example, the IOPS, file size distribution, and access time interval can be obtained through the historical access records of the SSD main control chip.

[0076] The predetermined time period is, for example, 24 hours. That is to say, the IOPS, file size distribution, and access time interval (for example, 90% of the accesses are concentrated in small files <4KB) within the past 24 hours are collected as the training data of the LSTM model.

[0077] The label definition of the training data can include hot data and cold data, etc. Hot data is, for example, data with a probability of being accessed >80% within the next 1 hour, and cold data is, for example, data with a probability of being accessed <5% within the next 1 hour.

[0078] The LSTM model can be trained using the training data and the corresponding label definition. After the LSTM model is trained, the historical access data of the solid-state drive is input into the LSTM model. By analyzing the historical access pattern or user access habit, the cold / hot data distribution, burst load, etc. can be predicted to output the corresponding user behavior weights. It can be understood that the data categories in the historical access data can be the same as the data categories in the training data.

[0079] In this embodiment, the user behavior prediction module 20 adopts an LSTM model to generate user behavior weights based on the historical access data of the solid-state drive. The LSTM model can process multi-dimensional time series data (such as access time, frequency, data volume, etc.), accurately capture the time series dependence relationship, and identify complex behavior patterns, so as to achieve accurate prediction of user behavior and generate the corresponding user behavior weights.

[0080] Combining the environmental survival weight and the user behavior weight can jointly determine the dynamic weight matrix to dynamically adjust the storage strategy of the solid-state drive.

[0081] Please refer toFigure 1 and Figure 3 In some embodiments, the storage policy generation module 30 is configured to determine a dynamic weight matrix based on the environmental survival weight and the user behavior weight, and dynamically adjust the mapping relationship between the logical address and the physical address of the solid state drive according to the dynamic weight matrix.

[0082] The dynamic weight matrix may be composed of the environmental survival weight env_weight and the user behavior weight data_heat. Based on the environmental survival weight env_weight, it is possible to distinguish whether it is a high-risk environment. Based on the user behavior weight, logical storage partitions can be divided, such as a hot data area, a cold data area, an emergency cache area, etc.

[0083] The following is an example of dynamically adjusting the mapping relationship between the logical address and the physical address of the solid state drive according to the dynamic weight matrix:

[0084] In one example, the storage policy includes: if env_weight < 0.6 and data_heat > 0.8, strategy = "SLC_CACHE_MIGRATION". That is to say, for a high-risk environment + hot data, it is allocated to the single-level cell (SLC) cache area.

[0085] In one example, the storage policy includes: if elif vibration > 1500Hz and data_heat < 0.3, strategy = "LOW_FREQ_RELOCATION". That is to say, for high-frequency vibration + cold data, it is migrated to the durable block of the 3D triple-level cell (TLC).

[0086] In this embodiment, according to the weighted decision tree (dynamic weight matrix = environmental survival weight & user behavior weight), the mapping rule of the flash translation layer (FTL) can be dynamically adjusted, and the mapping relationship between the logical address and the physical address of the solid state drive can be adjusted in real time, so as to select the optimal storage path. For example, in terms of write optimization, data is preferentially written to the high-temperature resistant block in a high-temperature environment, and the voltage is reduced to reduce error codes; in terms of anti-vibration cache, when continuous vibration is detected, the volatile data is migrated to the SLC cache area.

[0087] In one embodiment, when the storage policy operation is successful, the corresponding environmental survival weight and user behavior weight can be returned to the main control chip for model training to optimize the model, which is beneficial for coping with subsequent similar situations; when the storage policy operation fails, the self-repair or data migration of the solid-state drive can be triggered to achieve data transfer or backup. In this way, the feedback closed-loop control of the dynamic storage system 100 can be realized.

[0088] Please refer to Figure 1 and Figure 3 , in some embodiments, the dynamic storage system 100 further includes a bionic storage module 40. The bionic storage module 40 is used to dynamically adjust the write voltage of the solid-state drive according to the dynamic weight matrix and perform sleep repair.

[0089] Specifically, the bionic storage module 40 is also the bionic storage medium control layer. The bionic storage module 40 can adjust the write voltage of the solid-state drive and trigger sleep repair according to the temperature factor and electromagnetic interference factor in the dynamic weight matrix. In this way, by dynamically adjusting the write voltage, the bit error rate in a high-temperature environment can be reduced, and the power consumption can be reduced; by triggering sleep repair and suspending the write operation, the lifespan of the bionic storage module 40 can be extended.

[0090] In some embodiments, the bionic storage module 40 includes variable resistor flash memory cells. Specifically, the variable resistor flash memory cells can be variable resistor NAND (NOT-AND) flash memory cells. The resistance value of the variable resistor flash memory cells can be dynamically adjusted with the change of the electric field or temperature to reduce the write voltage and bit error rate in a high-temperature environment.

[0091] In this embodiment, a new type of variable resistor flash memory cell is designed by referring to the self-repair mechanism of biological cells. The variable resistor flash memory cells support the dynamic adjustment of the resistance value and the sleep repair function. Its core principle is to imitate the voltage-gated characteristics of the ion channels in the cell membrane and regulate the electron migration rate of the floating gate layer through the electric field, so as to achieve the elastic adjustment of the write voltage.

[0092] In summary, the dynamic storage system 100 of the embodiments of the present application has the following characteristics:

[0093] (1) Environment-behavior joint optimization technology: deeply integrates external environmental variables and user behavior patterns to realize the dynamic generation of cross-layer storage policies.

[0094] (2) Breakthrough in bionic hardware design: Based on the unique principle of biological cell membrane ion channels, a NAND flash memory cell with variable resistor characteristics is developed, breaking through the limitations of current pure algorithm optimization.

[0095] (3) Edge intelligent autonomous system: By embedding a lightweight AI model into the main control chip, an intelligent system that supports local decision-making is constructed, effectively reducing the dependence on the cloud.

[0096] The dynamic storage system 100 of the embodiments of the present application has at least the following beneficial effects:

[0097] (1) Significantly improved environmental adaptability: In complex environments with high temperature (such as above 70 °C), high-frequency vibration (such as above 2000 Hz), and electromagnetic interference, the bit error rate can be effectively reduced, and the reliability of data storage can be greatly improved.

[0098] (2) Effectively extended service life: Through the bionic self-repair mechanism, the service life of NAND flash memory cells is extended.

[0099] (3) Optimized edge computing ability: The latency of local AI decision-making can be reduced to the microsecond level, significantly reducing the dependence on the host, and is especially suitable for the real-time data processing requirements in the autonomous driving scenario. In addition, the dynamic storage system 100 of the embodiments of the present application can also be applied to industrial Internet of Things and aerospace scenarios, and realizes dynamic adjustment of data storage priorities through edge intelligent decision-making.

[0100] (4) Significantly improved energy efficiency ratio: By adopting dynamic voltage adjustment technology, the power consumption is effectively reduced, and it can fully meet the requirements of various scenarios with strict power consumption requirements.

[0101] Please refer to Figure 4 , the embodiments of the present application also provide a dynamic storage method. The dynamic storage method includes:

[0102] S10: Obtain the operating environment parameters of the solid-state drive, and generate an environmental survival weight based on the operating environment parameters;

[0103] S20: Obtain the historical access data of the solid-state drive, and generate a user behavior weight based on the historical access data;

[0104] S30: Determine the dynamic storage policy of the solid-state drive according to the environmental survival weight and the user behavior weight.

[0105] In some embodiments, obtaining the operating environment parameters of the solid-state drive includes:

[0106] Detect the temperature parameter of the solid-state drive;

[0107] Detect the vibration frequency parameter of the solid-state drive;

[0108] Detect the noise fluctuation parameter of the solid-state drive.

[0109] In some embodiments, generating an environmental survival weight based on the operating environment parameters includes:

[0110] Determine the temperature weight based on the temperature parameter;

[0111] Determine the vibration weight based on the vibration frequency parameter;

[0112] Determine the electromagnetic interference weight based on the noise fluctuation parameter;

[0113] Determine the environmental survival weight according to the temperature weight, vibration weight, and electromagnetic interference weight.

[0114] In some embodiments, generating a user behavior weight based on historical access data includes:

[0115] Generate the user behavior weight based on the historical access data through a long short-term memory network model.

[0116] In some embodiments, the dynamic storage method further includes:

[0117] Obtain the training data of the long short-term memory network model, where the training data includes the number of input / output operations per second, file size distribution, and access time interval of the solid-state drive within a predetermined time period before the current moment;

[0118] Perform model training on the long short-term memory network model through the training data.

[0119] In some embodiments, determining the dynamic storage policy of the solid-state drive according to the environmental survival weight and the user behavior weight includes:

[0120] Determine the dynamic weight matrix according to the environmental survival weight and the user behavior weight;

[0121] Dynamically adjust the mapping relationship between the logical address and the physical address of the solid-state drive according to the dynamic weight matrix.

[0122] In some embodiments, the dynamic storage method further includes:

[0123] Dynamically adjust the write voltage of the solid-state drive and perform sleep repair according to the dynamic weight matrix.

[0124] For the description of the features in the embodiments corresponding to the dynamic storage method, reference can be made to the relevant description of the embodiments corresponding to the dynamic storage system 100, which will not be elaborated here one by one.

[0125] Through the description of the implementation manners of the foregoing dynamic storage system 100, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases the former is a better implementation manner.

[0126] Please refer to Figure 5, embodiments of the present application also provide a dynamic storage device 200. The dynamic storage device 200 includes a first acquisition module 210, a second acquisition module 220, and a policy determination module 230. The first acquisition module 210 is configured to acquire the operating environment parameters of the solid-state drive and generate an environmental survival weight based on the operating environment parameters. The second acquisition module 220 is configured to acquire the historical access data of the solid-state drive and generate a user behavior weight based on the historical access data. The policy determination module 230 is configured to determine the dynamic storage policy of the solid-state drive according to the environmental survival weight and the user behavior weight.

[0127] In some embodiments, the first acquisition module 210 is specifically configured to: detect the temperature parameter of the solid-state drive; detect the vibration frequency parameter of the solid-state drive; detect the noise fluctuation parameter of the solid-state drive.

[0128] In some embodiments, the first acquisition module 210 is specifically configured to: determine the temperature weight based on the temperature parameter; determine the vibration weight based on the vibration frequency parameter; determine the electromagnetic interference weight based on the noise fluctuation parameter; determine the environmental survival weight according to the temperature weight, the vibration weight, and the electromagnetic interference weight.

[0129] In some embodiments, the second acquisition module 220 is specifically configured to generate a user behavior weight based on the historical access data through a long short-term memory network model.

[0130] In some embodiments, the second acquisition module 220 is further configured to: acquire the training data of the long short-term memory network model, where the training data includes the number of input / output operations per second, the file size distribution, and the access time interval of the solid-state drive within a predetermined time period before the current moment; perform model training on the long short-term memory network model through the training data.

[0131] In some embodiments, the policy determination module 230 is specifically configured to: determine a dynamic weight matrix according to the environmental survival weight and the user behavior weight; dynamically adjust the mapping relationship between the logical address and the physical address of the solid-state drive according to the dynamic weight matrix.

[0132] In some embodiments, the policy determination module 230 is further configured to dynamically adjust the write voltage of the solid-state drive and perform sleep repair according to the dynamic weight matrix.

[0133] For the description of the features in the embodiments corresponding to the dynamic storage device 200, reference may be made to the relevant descriptions of the embodiments corresponding to the dynamic storage system 100 and the dynamic storage method, which will not be elaborated here one by one.

[0134] Please refer to Figure 6, an embodiment of the present application further provides an electronic device 300, including a memory 310 and a processor 320. A computer program is stored in the memory 310, and the processor 320 is configured to run the computer program to execute the steps in any of the above-described dynamic storage method embodiments.

[0135] Please refer to Figure 7 , an embodiment of the present application further provides a computer-readable storage medium 400. A computer program 410 is stored in the computer-readable storage medium 400, wherein the computer program 410 is configured to execute the steps in any of the above-described dynamic storage method embodiments when running.

[0136] In an exemplary embodiment, the above computer-readable storage medium 400 may include, but is not limited to: various media such as USB flash drives, read-only memories (ROM), random access memories (RAM), external hard drives, magnetic disks, or optical discs that can store the computer program 410.

[0137] An embodiment of the present application further provides a computer program product. The above computer program product includes a computer program, and when the computer program is executed by a processor, it implements the steps in any of the above-described dynamic storage method embodiments.

[0138] An embodiment of the present application further provides another computer program product, including a non-volatile computer-readable storage medium. The non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, it implements the steps in any of the above-described dynamic storage method embodiments.

[0139] Those skilled in the art can further realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described according to their functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods for each specific application to implement the described functions, but such implementation should not be considered to exceed the scope of the present application.

[0140] The above has introduced in detail a dynamic storage system 100, a dynamic storage method, a dynamic storage device 200, an electronic device 300, a computer-readable storage medium 400, and a computer program product provided by the present application. Specific examples are used herein to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application. It should be noted that for those of ordinary skill in the art of the present technology, without departing from the principle of the present application, several improvements and modifications can be made to the present application, and these improvements and modifications also fall within the protection scope of the claims of the present application.

Claims

1. A dynamic storage system, characterized in that, Comprising: An environment perception module, configured to obtain the operating environment parameters of the solid-state drive, and generate an environment survival weight based on the operating environment parameters; A user behavior prediction module, configured to obtain the historical access data of the solid-state drive, and generate a user behavior weight based on the historical access data; A storage policy generation module, configured to determine the dynamic storage policy of the solid-state drive according to the environment survival weight and the user behavior weight.

2. The dynamic storage system according to claim 1, wherein The environment perception module includes: A temperature sensor, configured to detect the temperature parameter of the solid-state drive; A vibration sensor, configured to detect the vibration frequency parameter of the solid-state drive; An electromagnetic interference detection unit, configured to detect the noise fluctuation parameter of the solid-state drive.

3. The dynamic storage system according to claim 2, wherein The environment perception module is configured to determine a temperature weight based on the temperature parameter, determine a vibration weight based on the vibration frequency parameter, determine an electromagnetic interference weight based on the noise fluctuation parameter, and determine the environment survival weight according to the temperature weight, the vibration weight, and the electromagnetic interference weight.

4. The dynamic storage system according to claim 1, wherein The user behavior prediction module includes a long short-term memory network model, and the training data of the long short-term memory network model includes the number of input / output operations per second, the file size distribution, and the access time interval of the solid-state drive within a predetermined time period before the current moment.

5. The dynamic storage system according to claim 1, wherein The storage policy generation module is configured to determine a dynamic weight matrix according to the environment survival weight and the user behavior weight, and dynamically adjust the mapping relationship between the logical address and the physical address of the solid-state drive according to the dynamic weight matrix.

6. The dynamic storage system according to claim 5, wherein, The dynamic storage system further includes a bionic storage module, and the bionic storage module is configured to dynamically adjust the write voltage of the solid-state drive and perform sleep repair according to the dynamic weight matrix.

7. The dynamic storage system according to claim 6, characterized in that, The bionic storage module includes variable resistance flash memory cells.

8. A dynamic storage method, characterized in that, Comprising: Obtaining the operating environment parameters of the solid-state drive, and generating an environment survival weight based on the operating environment parameters; Obtaining the historical access data of the solid-state drive, and generating a user behavior weight based on the historical access data; Determining the dynamic storage policy of the solid-state drive according to the environment survival weight and the user behavior weight.

9. An electronic device, characterized in that, Comprising: A memory, configured to store a computer program; A processor, configured to implement the steps of the dynamic storage method as claimed in claim 8 when executing the computer program.

10. A computer-readable storage medium, characterized in that, A computer program is stored in the computer-readable storage medium, wherein the computer program implements the steps of the dynamic storage method as claimed in claim 8 when being executed by a processor.

Citation Information

Cited By

  • Intelligent hierarchical storage method and system for cold and hot data of solid state disk

    CN120994139A

  • Solid state disk cold and hot data intelligent layered storage method and system

    CN120994139B