A file management system and method based on cloud storage

Abstract

This application relates to the field of cloud storage technology and discloses a file management system and method based on cloud storage, including: a data encryption module for homomorphically encrypting user-uploaded files and storing the encrypted files in the cloud; an algebraic geometry mapping module for mapping the encrypted files to an algebraic variety and optimizing the storage path according to the algebraic geometry structure; a query processing module for receiving user query requests, performing geometric measurement calculations on the encrypted files, and returning matching results while in an encrypted state; and an adaptive storage management module for dynamically adjusting the file storage location according to the file access frequency, enabling fast retrieval of frequently accessed files. This invention uses homomorphic encryption technology to encrypt user-uploaded files, achieving the technical effect of effectively protecting data privacy during cloud storage and computation.

Description

Technical Field

[0001] This invention relates to the field of cloud storage technology, specifically to a file management system and method based on cloud storage. Background Technology

[0002] With the continuous development of information technology, especially driven by cloud computing and big data, cloud storage has become an indispensable core component of various application systems. Cloud storage technology not only provides on-demand scalable storage space, but also effectively supports functions such as data sharing, backup, and disaster recovery. However, with the increase in storage capacity and the rise in data security requirements, traditional file management technologies have gradually revealed many shortcomings, especially in terms of data privacy protection and query efficiency.

[0003] Traditional file management systems typically rely on storing and processing files in plaintext. Although encryption technology can protect the data, data encrypted in the cloud usually needs to be decrypted before querying and processing. The decryption process not only increases the computational burden but also leads to potential data leakage risks. When users query encrypted data stored in the cloud, they must first decrypt the data before they can perform query operations or calculations, thus exposing sensitive user data to cloud service providers.

[0004] With the continuous growth of data volume, the query efficiency of file management systems has become a pressing issue. Traditional file management systems rely on static storage and linear search methods for file retrieval. This leads to a sharp increase in query time complexity as the number of files increases, impacting the overall system response speed and user experience. Furthermore, since the relationship between file content and query conditions is often complex, queries based on keywords or other basic retrieval methods often fail to meet the requirement of precise matching, resulting in decreased accuracy and efficiency of query results. Traditional file management systems typically lack optimized storage and query strategies for large-scale data, leading to performance bottlenecks in high-concurrency queries and big data environments. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a file management system and method based on cloud storage, which solves the problem of how to improve the efficiency of large-scale data query and optimize storage management while ensuring the privacy of encrypted data.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a cloud storage-based file management system, comprising:

[0007] The data encryption module is used to perform homomorphic encryption on user-uploaded files and store the encrypted files in the cloud;

[0008] The algebraic geometry mapping module is used to map encrypted files to algebraic clusters and optimize the storage path according to the structure of algebraic geometry; the query processing module is used to receive user query requests, perform geometric measurement calculations on encrypted files, and return matching results in the encrypted state.

[0009] The adaptive storage management module is used to dynamically adjust the storage location of files based on their access frequency, enabling fast retrieval of frequently accessed files.

[0010] The distributed computing and storage module is used to distribute encrypted files across multiple cloud nodes.

[0011] Preferably, the data encryption module uses homomorphic encryption technology for encryption. The homomorphic encryption supports addition and multiplication operations and can perform calculation operations on the encrypted file in the encrypted state to avoid data leakage. The homomorphic encryption technology can be selected as Paillier encryption or other homomorphic encryption schemes that support addition and multiplication operations. Moreover, this encryption method can ensure the security of file data and support incremental updates of encrypted data.

[0012] Furthermore, the data encryption module is responsible for homomorphic encryption of user-uploaded files. By employing homomorphic encryption technology, it ensures that file content remains encrypted throughout the transmission and storage process, preventing data leakage during cloud storage and processing. During encryption, this module can use homomorphic encryption algorithms that support addition and multiplication operations, such as the Paillier encryption scheme or other similar homomorphic encryption schemes. These schemes, while ensuring file data security, support incremental updates of encrypted data. That is, when file content changes, only the changed parts are encrypted and updated, without completely decrypting the file, thus improving processing efficiency. Secondly, the algebraic geometry mapping module maps encrypted files to algebraic varieties and optimizes the storage path based on the algebraic geometry structure. This fully utilizes the characteristics of algebraic geometry, improving file storage efficiency and retrieval performance through optimized storage paths. Based on this, the query processing module receives user query requests and performs geometric measurement calculations on encrypted files. It returns matching results while the files remain encrypted, ensuring user data privacy is not compromised and improving query efficiency and accuracy. To further enhance system performance, the adaptive storage management module dynamically adjusts file storage locations based on access frequency, storing frequently accessed files in areas easily accessible for quick retrieval, thereby improving query speed and system responsiveness. Finally, the distributed computing and storage module distributes encrypted files across multiple cloud nodes, supporting parallel computing, further improving query efficiency and enhancing system scalability and fault tolerance. Through these technical solutions, this invention not only ensures file data privacy but also significantly improves system performance and efficiency through intelligent storage and parallel computing, providing users with an efficient, secure, and intelligent file management solution.

[0013] Preferably, the algebraic geometry mapping module embeds encrypted data into the geometric space through polynomial equations in algebraic clusters to form a hierarchical storage structure, and optimizes the storage through the geometric properties of curves or clusters in algebraic geometry. The algebraic clusters can be divided into multiple levels according to the characteristics of the data, and the encrypted data in each level is dynamically adjusted according to the frequency and properties of the queries.

[0014] Furthermore, by embedding encrypted data into the geometric space of algebraic clusters and using polynomial equations within these clusters to map the data, a hierarchical storage structure is formed. This process not only ensures data security under encryption but also optimizes the storage path through the geometric properties of curves or clusters in algebraic geometry, thereby improving data storage efficiency and retrieval speed. The structural characteristics of algebraic geometry allow file data to be divided into multiple layers based on its inherent characteristics and query requirements. Encrypted data within each layer is dynamically adjusted according to query frequency, nature, and access patterns. Specifically, frequently accessed data is stored in high-efficiency access layers, while infrequently accessed data is moved to lower storage layers. This ensures that frequently queried files can be responded to quickly, improving the overall query performance and storage efficiency of the system. This algebraic geometry-based optimized storage strategy allows the system to flexibly adjust the storage layout according to the access requirements of different data and maximize the use of cloud storage resources. At the same time, it greatly improves the efficiency and response speed of file query and storage operations while ensuring privacy protection.

[0015] Preferably, the query processing module calculates the similarity between the query conditions and the encrypted file in the geometric space of the encrypted data according to the user's query conditions, and returns the encrypted file with high similarity. The similarity calculation adopts a geometric measurement method, which includes but is not limited to Euclidean distance, cosine similarity, and algebraic curve similarity measurement. Moreover, the calculation method supports being performed in the encrypted state. The query results are sorted according to the similarity score, and the encrypted file most relevant to the query conditions is returned.

[0016] Furthermore, upon receiving a user's query request, the query processing module performs similarity calculations in the geometric space of the encrypted data based on the query conditions to determine the degree of matching between the query conditions and the encrypted files. This module employs geometric metrics to calculate similarity, including but not limited to Euclidean distance, cosine similarity, and algebraic curve similarity metrics. These geometric metrics map the feature vectors of the query conditions and encrypted files into the geometric space for calculation, thereby achieving efficient matching and retrieval without decrypting the data. Specifically, Euclidean distance can be used to measure the direct spatial distance between data points, while cosine similarity can assess the angular similarity between the query conditions and the file content. Algebraic curve similarity metrics leverage the structural characteristics of algebraic geometry to calculate more complex data matching. Through these geometric metrics, the query processing module can perform accurate similarity calculations on encrypted files and sort the files according to the calculation results, ensuring that the most relevant encrypted files are returned first. This process is performed entirely in an encrypted state, without decrypting the original file data, thus protecting data privacy and improving query processing efficiency. This allows users to quickly obtain the most matching files, achieving efficient and secure file management and retrieval.

[0017] Preferably, the adaptive storage management module automatically maps frequently accessed files to low-dimensional algebraic clusters based on file access frequency and query patterns, while storing infrequently accessed files in high-dimensional algebraic clusters to save storage space. The adaptive storage management module also includes a dynamic adjustment algorithm, which automatically readjusts the position of files in algebraic clusters based on the real-time load of the system and query frequency, thereby achieving a balance between storage optimization and query performance.

[0018] Furthermore, the adaptive storage management module intelligently maps frequently accessed files to low-dimensional algebraic clusters by analyzing file access frequency and query patterns to improve query efficiency. Low-dimensional algebraic clusters typically have a more compact geometric structure, thus accelerating data retrieval, especially when frequently accessed files require fast response times. Additionally, to optimize storage space and reduce computational complexity, the module stores infrequently accessed files in high-dimensional algebraic clusters. While high-dimensional clusters offer more space, they effectively reduce unnecessary computational burdens caused by frequent access. To ensure the system can dynamically optimize based on actual load and query demands, the adaptive storage management module also includes a dynamic adjustment algorithm. This algorithm monitors system load, file access frequency, and query patterns in real time and automatically adjusts the file's storage location within the algebraic clusters. Through this algorithm, the system can flexibly find the optimal balance between storage optimization and query response based on changes in file access and query performance requirements. This mechanism not only improves storage resource utilization efficiency but also effectively reduces the storage and computational overhead of infrequently accessed files while maintaining query performance, thereby minimizing storage costs while ensuring system efficiency.

[0019] Preferably, the distributed computing and storage module stores encrypted files through distributed nodes, supports parallel computing to accelerate query processing, performs similarity calculations on encrypted data on different computing nodes, and distributes query requests to multiple computing nodes for parallel processing, reducing the computing load on individual nodes. The distributed computing and storage module can automatically and dynamically adjust task allocation according to the load of nodes to ensure stable system operation.

[0020] Furthermore, by distributing encrypted files across multiple distributed computing nodes, the high scalability and computing resources of the cloud environment can be fully utilized to support parallel computing and accelerate the query processing. In this module, encrypted data is stored across multiple distributed nodes. Each node stores a portion of the file data and can perform independent computations. When a user initiates a query request, the request is intelligently distributed across multiple computing nodes for parallel processing. This significantly reduces the computational load on individual nodes and improves the overall system's query response speed. Through parallel computation, the system can simultaneously perform similarity calculations on multiple encrypted data sets, avoiding the performance bottlenecks of traditional single-threaded computation methods and improving the efficiency of large-scale data processing. Furthermore, the distributed computing and storage module also features dynamic task scheduling, automatically adjusting task allocation based on the real-time load of each node to ensure load balancing and stable system operation. Specifically, when the computational load on some nodes is high, the system can automatically migrate tasks to nodes with lower loads, ensuring a reasonable distribution of the burden on each computing node. This avoids wasting computing resources and potential performance bottlenecks. This dynamic adjustment mechanism not only improves query processing efficiency but also enhances the system's fault tolerance and stability, ensuring smooth operation even under high concurrency and high load environments, providing users with efficient and reliable query services.

[0021] Preferably, the query processing module calculates the similarity of encrypted data using geometric measurement methods, including Euclidean distance or algebraic curve similarity measurement, and retrieves encrypted files based on the similarity results. The geometric measurement methods can be directly calculated on encrypted data, and data privacy is guaranteed during the query process through homomorphic encryption. The similarity calculation supports vectorized calculation and clustering algorithm technology, and can adaptively adjust the calculation strategy according to changes in query conditions.

[0022] Furthermore, the query processing module calculates the similarity of encrypted data using geometric metrics (such as Euclidean distance and algebraic curve similarity metrics) to efficiently retrieve encrypted files based on user query conditions. In this process, geometric metrics can be calculated directly on the encrypted data without prior decryption, effectively ensuring data privacy and security during the query process. This process is achieved through homomorphic encryption technology, ensuring data confidentiality and integrity even when querying and calculating encrypted data in the cloud, preventing data leakage or unauthorized access. The similarity calculation method supports vectorized computation and clustering algorithms, which can efficiently process large-scale data, optimize query efficiency, and reduce computation time. Regarding vectorized computation, by converting encrypted data into vector form, vector computation techniques can be used to perform batch calculations on the encrypted data, thereby improving processing speed. In terms of clustering algorithms, the system can intelligently group data based on its characteristics, further improving query accuracy and efficiency. In addition, the query processing module can adaptively adjust its calculation strategy according to changes in query conditions. When query conditions become more complex or query volume increases, the module can dynamically select the most appropriate calculation method and strategy to ensure the efficiency and flexibility of query processing. Through these technologies, the query processing module not only improves the accuracy of encrypted file retrieval, but also significantly enhances the system's response speed and user experience, optimizing query performance while ensuring data privacy.

[0023] Preferably, the algebraic geometry mapping module supports low-dimensional embedding of high-dimensional data. By using the embedding theorem of algebraic geometry, high-dimensional data is mapped to low-dimensional algebraic clusters for query optimization. The embedding method can effectively reduce the query overhead of high-dimensional data and optimize data processing efficiency while ensuring query accuracy. The embedding of high-dimensional data can be dynamically adjusted based on the properties of the query conditions to optimize the storage and query process.

[0024] Furthermore, the algebraic geometry mapping module supports low-dimensional embedding of high-dimensional data to effectively reduce query computation complexity. It maps high-dimensional data to low-dimensional algebraic varieties using embedding theorems in algebraic geometry, thereby optimizing queries. High-dimensional data often suffers from the curse of dimensionality, and the complexity of query computation increases exponentially with the increase of data dimensionality. Therefore, using low-dimensional embedding methods can significantly reduce the query overhead of high-dimensional data, while improving overall data processing efficiency while ensuring the accuracy of query results. By using embedding theorems in algebraic geometry, the module can map complex high-dimensional data structures to lower-dimensional spaces through the geometric properties of algebraic varieties, thus reducing the complex computation of large-scale data during queries. To meet computational needs, embedded low-dimensional data not only accelerates the query process but also optimizes the data storage structure, making storage and query operations more efficient. Furthermore, this embedding method can dynamically adjust based on the nature of the query conditions to adapt to different query patterns and data access requirements. For example, for requests with relatively simple query conditions, the system can choose to embed the data into a low-dimensional space for fast processing; while for complex queries, the embedding strategy can be flexibly adjusted to balance query accuracy and processing efficiency. This dynamic adjustment mechanism ensures the optimization of the storage and query process and enables the system to flexibly configure resources according to real-time needs, thereby improving the performance and scalability of the entire system in high-dimensional data environments.

[0025] Preferably, the adaptive storage management module further includes a query caching mechanism for caching encrypted files for repeated queries. The query caching mechanism can cache query conditions and their corresponding encrypted file results. When a user submits the same query, it directly returns the encrypted data in the cache without performing repeated calculations. This caching mechanism supports efficient cache update strategies and can dynamically adjust the storage and cleanup of cached content according to system load and query patterns.

[0026] Furthermore, the adaptive storage management module includes a query caching mechanism. This mechanism caches encrypted files for repeated queries, significantly reducing computational overhead and improving query response speed. The query caching mechanism stores query conditions and their corresponding encrypted file results. When a user submits the same query request, the system can directly return the cached encrypted data without repeating the same calculation process. This mechanism effectively improves system response efficiency, especially when facing high-frequency, repeated queries, reducing unnecessary computational burden and query latency. In addition, this query caching mechanism supports efficient cache update strategies, ensuring that cached content is always up-to-date and preventing outdated data from affecting query results. The system can dynamically adjust the storage and cleanup strategies of cached content based on real-time system load, query patterns, and access frequency to achieve optimal utilization of cache resources. When the system load is high or the query pattern changes, the caching mechanism can automatically clean up cached data that is no longer frequently accessed and reallocate cache space according to new query requirements. Through this intelligent cache management, the system not only improves query processing efficiency but also optimizes storage resource utilization, ensuring that overall system performance and stability are maximized while maintaining query accuracy.

[0027] A file management method based on cloud storage includes:

[0028] The user-uploaded files are homomorphically encrypted and stored in the cloud.

[0029] The encrypted file is mapped to an algebraic cluster, and the storage path is optimized based on the file's geometric characteristics;

[0030] Upon receiving a query request, calculate the geometric measure of the encrypted data based on the query criteria and return an encrypted file similar to the query criteria.

[0031] The file storage location is dynamically adjusted based on the file access frequency, so that frequently accessed files can be retrieved quickly.

[0032] Encrypted files are distributed and stored across multiple cloud nodes, and query response speed is improved through parallel computing;

[0033] Low-dimensional embedding of high-dimensional data.

[0034] Furthermore, firstly, user-uploaded files undergo homomorphic encryption to ensure privacy and security during cloud storage, and the encrypted files are then stored in the cloud. Subsequently, algebraic geometric mapping technology is used to map the encrypted files to algebraic varieties, and the storage path is optimized based on the file's geometric characteristics to improve storage efficiency and retrieval performance. Next, upon receiving a user query request, the system calculates the geometric measure of the encrypted data based on the query conditions and uses a geometric similarity calculation method to return encrypted files similar to the query conditions. This entire process is performed under encryption, avoiding the risk of data leakage. Simultaneously, the system dynamically adjusts the file storage location based on the file's access frequency, moving frequently accessed files to storage areas that are easier to retrieve quickly, thereby improving query response speed and optimizing storage space utilization, further enhancing the system's capabilities. To improve system performance, the encrypted files are distributed across multiple cloud nodes, and parallel computing technology accelerates the query process, avoiding excessive load on a single node and improving system processing efficiency and stability. Furthermore, to address the complexity of high-dimensional data, this method employs low-dimensional embedding technology, mapping high-dimensional data to a low-dimensional space using algebraic geometry embedding theorems. This reduces the complexity of query computation, optimizes query performance, and ensures high-efficiency query and processing speed even with large-scale datasets. Through these comprehensive measures, this method achieves a balance between data security and efficiency, providing users with more efficient, secure, and intelligent cloud storage and file management services.

[0035] This invention provides a file management system and method based on cloud storage. It has the following beneficial effects:

[0036] 1. This invention employs homomorphic encryption technology to encrypt user-uploaded files, effectively protecting data privacy during cloud storage and computation. Compared to traditional encryption methods in existing technologies, which often require decryption before file processing or querying, this invention uses homomorphic encryption to enable computation and querying while the file is encrypted. This avoids potential leakage risks, ensures the security and privacy of user data, and significantly reduces the possibility of data leakage.

[0037] 2. This invention employs an algebraic geometric mapping method to map encrypted files into algebraic clusters and optimizes the storage path based on the file's geometric characteristics, achieving both optimized storage structure and improved query efficiency. Compared to common file storage methods in the prior art, which typically use static storage layouts and struggle to dynamically optimize storage paths, this invention utilizes embedding theorems and geometric metric calculations based on algebraic geometry to make file storage more intelligent and efficient. This effectively reduces the computational complexity during queries and optimizes the file layout during storage, allowing for dynamic adjustments based on the file's query characteristics, thereby significantly improving query response speed.

[0038] 3. This invention combines adaptive storage management and distributed computing technologies, supporting dynamic adjustment of file storage locations based on file access frequency. Encrypted files are distributed and computed in parallel across multiple cloud nodes, achieving improved query response speed and system processing capabilities. Compared to existing technologies that typically employ single-node storage and computation, this invention effectively avoids the problems of excessive single-node load and slow query response through distributed storage and parallel computing. Furthermore, by dynamically adjusting file locations, it ensures rapid retrieval of frequently accessed files, thereby significantly improving system performance and stability, enhancing overall computational efficiency, and reducing storage and computational costs. Attached Figure Description

[0039] Figure 1 This is a system framework diagram of the present invention;

[0040] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0041] The technical solutions in 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.

[0042] Please see the appendix Figure 1 This invention provides a cloud storage-based file management system, comprising:

[0043] The data encryption module is used to perform homomorphic encryption on user-uploaded files and store the encrypted files in the cloud;

[0044] The algebraic geometry mapping module is used to map encrypted files to algebraic clusters and optimize the storage path according to the structure of algebraic geometry; the query processing module is used to receive user query requests, perform geometric measurement calculations on encrypted files, and return matching results in the encrypted state.

[0045] The adaptive storage management module is used to dynamically adjust the storage location of files based on their access frequency, enabling fast retrieval of frequently accessed files.

[0046] The distributed computing and storage module is used to distribute encrypted files across multiple cloud nodes.

[0047] Step S1: Homomorphic encryption processing and encrypted file storage

[0048] In this invention, the main task of step S1 is to perform homomorphic encryption on the user-uploaded file and store the encrypted file in the cloud. Through this step, this invention ensures the security and privacy protection of user files during storage and processing. The file encryption operation is completed through homomorphic encryption technology, which allows computation on encrypted data without decryption, thereby avoiding potential data leakage problems caused by decryption in traditional methods. The specific implementation method, working principle and technical effects of this step will be described in detail below.

[0049] In this embodiment, the implementation steps of homomorphic encryption processing and storage include the following parts:

[0050] First, user-uploaded files are sent to the cloud via a file transfer system. During this process, the file content is converted into a data format suitable for encryption. Specifically, file content (such as text documents, images, or multimedia files) is converted into a set of numerical data. This data can be represented using hash algorithms, ASCII code conversion, or other standard conversion methods for subsequent encryption processing. It should be noted that the type and format of the file determine how it is effectively converted into a numerical representation. Typically, text files are first converted into a string of numerical representations, while image files can be converted into a matrix representation of their pixel values.

[0051] Subsequently, the file will be encrypted using a homomorphic encryption algorithm. In this embodiment, the Paillier encryption algorithm, which supports addition and multiplication operations, is selected. As a common homomorphic encryption method, the Paillier encryption algorithm has the property of additive homomorphism. Using this algorithm, addition operations can be performed on the encrypted data without decryption. The specific operation is as follows:

[0052] Input file data encryption: The numerical representation of the file (such as the ASCII code after text conversion, the pixel matrix of an image, etc.) will be used as input data and encrypted using the Paillier encryption algorithm. Specifically, the data m will be encrypted using the public key pk to generate ciphertext c. That is, the encryption process can be represented as c = E(m, pk).

[0053] Homomorphic addition support: Through the Paillier encryption algorithm, the system can perform addition operations on encrypted data. For example, assuming there are two encrypted data c1 = E(m1, pk) and c2 = E(m2, pk), the system can directly perform addition operations on the encrypted results to obtain c1 ⊕ c2 = E(m1 + m2, pk) without decrypting the data. In this way, even in the cloud, the content of the file always remains encrypted, and data privacy is effectively protected.

[0054] c1 and c2: These are encrypted data. They are the results of encrypting the encrypted messages m1 and m2 using the public key pk.

[0055] E(m1,pk) and E(m2,pk): These represent the results of encrypting messages m1 and m2 using the public key pk, respectively, i.e., the data after encryption.

[0056] m1 and m2: These are the raw message data, i.e., plaintext data that needs to be encrypted.

[0057] pk: Public key, usually used to encrypt data, ensuring that only the person with the corresponding private key can decrypt it.

[0058] c1⊕c2: This represents an addition operation performed on encrypted data c1 and c2. This operation is used to combine the encrypted results without needing to directly decrypt the data.

[0059] Incremental encryption: When the content of a file changes, the system can update part of the file's content through incremental encryption without re-encrypting the entire file. This incremental encryption method not only reduces the computational burden but also improves the efficiency of file updates. Especially when the file is large or updated frequently, incremental encryption can significantly reduce the system's computational overhead.

[0060] File storage procedure:

[0061] After homomorphic encryption is completed, the file data is stored in a distributed storage system in the cloud. Each encrypted file is split into several file blocks and stored separately on multiple distributed storage nodes. This distributed storage method enables data redundancy, ensuring high availability even if some storage nodes fail. During storage, each encrypted file is assigned a unique identifier (ID) and its metadata (such as file type, creation time, encryption method, etc.) is saved. This metadata is also encrypted to prevent the leakage of any sensitive information.

[0062] It should be noted that the file storage path and storage strategy will be optimized based on factors such as file type, size, and access frequency.

[0063] Small file storage: For smaller files, encrypted files may be stored directly on fewer nodes;

[0064] Large file storage: For large files, the system splits the file into multiple encrypted blocks and distributes them across multiple nodes, thereby optimizing storage space utilization and improving query efficiency.

[0065] Dynamic storage adjustment: Further optimizations include dynamic adjustment based on file access frequency. Frequently accessed files are preferentially stored on storage nodes with faster access speeds, while infrequently accessed files are stored on low-cost storage nodes to save resources and reduce costs.

[0066] Description of effects:

[0067] By employing homomorphic encryption technology, this invention enables the querying and calculation of encrypted files while ensuring data privacy. Compared with existing technologies that require file decryption to perform query operations, this invention, through homomorphic encryption technology, can perform queries and calculations without decrypting the data, thereby avoiding the risk of data leakage. In addition, the homomorphic encryption algorithm supports addition operations on encrypted files, allowing the system to process files efficiently without exposing the file content.

[0068] In terms of storage, this invention provides an efficient storage and data management solution through distributed storage and dynamic optimization storage strategies. Distributed storage of files ensures data redundancy and high availability, while the dynamic storage adjustment mechanism can adjust the storage path in real time according to file access frequency and query patterns, optimize the use of storage resources, and improve query response speed. Encrypted file storage not only ensures data security but also effectively improves the scalability and fault tolerance of the system.

[0069] In summary, by employing homomorphic encryption technology and a distributed storage strategy, this invention successfully achieves privacy protection, efficient storage, and query processing of encrypted files, solving the problems of low data privacy protection and query efficiency in traditional cloud storage. It can provide more secure, reliable, and efficient file management services in cloud storage environments.

[0070] Algebraic geometry mapping and storage path optimization

[0071] In this invention, the main task of step S2 is to map the encrypted file into an algebraic cluster and optimize the storage path based on the file's geometric characteristics. This process effectively improves file storage efficiency and optimizes query processing speed. This step is key to achieving efficient storage and fast querying in this invention. Specifically, file data is mapped into an algebraic cluster using algebraic geometric mapping technology, and query optimization is performed using the spatial characteristics of algebraic geometry. The specific implementation method, working principle, and technical effects of this step will be described in detail below.

[0072] In this embodiment, the algebraic geometric mapping and storage path optimization are implemented as follows:

[0073] First, the content of the encrypted file is geometrically embedded, that is, the encrypted data is transformed into points or curves in an algebraic variety through algebraic geometry mapping methods. An algebraic variety is a geometric object defined by algebraic equations, which has specific geometric properties and can be used to effectively organize and store data. Specifically, file data (such as text, images or other types of file content) is mapped into the geometric space of an algebraic variety through an appropriate algebraic mapping function.

[0074] For example, suppose each data block in a file can be represented as a vector v. i Through the algebraic geometric mapping function f, the file data will be mapped into the algebraic variety C, i.e.;

[0075] f(v i )∈C;

[0076] Among them, v i is the numerical representation of file data blocks, f is the mapping function, and C is the algebraic variety. In this mapping process, data blocks are embedded in a low-dimensional geometric space to reduce the complexity of query calculations and improve data processing and storage efficiency.

[0077] It should be noted that algebraic geometric mapping can not only map data to geometric space, but also optimize data storage paths by utilizing the geometric structure of algebraic varieties. The geometric structure of algebraic varieties has special properties that can optimize storage paths based on the data characteristics of files, making data storage more compact and enabling faster location of relevant data during queries. By analyzing the geometric relationships between data points in algebraic varieties, the system can allocate more suitable storage locations for data, thereby reducing computational and search overhead during queries.

[0078] The key to storage path optimization lies in distributing data across different storage levels based on the characteristics of algebraic geometry to achieve efficient data access. Specifically, data points in an algebraic cluster can be dynamically adjusted to different levels based on their access frequency and query requirements. Frequently accessed data will be stored in more efficient and faster storage levels to shorten query response time; while infrequently accessed data can be stored in lower-level storage areas to save storage costs. For example, file blocks that are close to each other in an algebraic cluster may be stored on the same storage node, while file blocks that are far apart can be distributed across different nodes.

[0079] The specific implementation process includes the following aspects:

[0080] Data mapping: Each data block of a file (such as a text segment or image data block in a file) is mapped algebraically and geometrically according to its numerical representation, into the geometric space of an algebraic cluster. The mapping method can take into account the properties of curves and clusters in algebraic geometry to maximize the use of the geometric structure to optimize storage and query performance.

[0081] Storage path optimization: Once data is mapped into an algebraic cluster, the system optimizes the storage path based on the geometric properties of the data (such as the distance and density between data points). By optimizing the storage path, the layout of the file during storage is improved, which not only increases the utilization of storage space but also reduces the storage areas that need to be traversed during queries, thereby improving query efficiency.

[0082] Dynamic storage adjustment: Based on the access frequency and query pattern of files, the system can dynamically adjust the position of files in the algebraic cluster. Frequently accessed files will be moved to storage areas closer to the query node, while infrequently accessed files will be moved to storage areas further away, thereby optimizing the utilization of storage resources and query response speed.

[0083] Description of effects:

[0084] By using algebraic geometric mapping and storage path optimization, this invention can significantly improve the efficiency of file storage and the response speed of queries. Compared with existing storage systems that typically use static storage and non-optimized query paths, this invention utilizes the spatial characteristics of algebraic geometry to dynamically optimize the storage path based on the geometric features of the file content, reducing unnecessary data access and thus improving the efficiency and accuracy of query processing.

[0085] Specifically, by mapping encrypted files to algebraic clusters and optimizing storage paths, the system can achieve the following effects: Storage optimization: File data is stored in more suitable locations, improving the utilization efficiency of storage space; frequently accessed data is stored in storage areas with faster query speeds, while infrequently accessed data is stored in low-cost storage areas, reducing storage costs.

[0086] Improved query efficiency: The file query process is optimized. Instead of traversing the entire file system, the relevant data is quickly located using the geometric properties of algebraic varieties. This allows the system to maintain high query efficiency even in large-scale data storage environments.

[0087] Data privacy protection: Because files are stored and retrieved in an encrypted state, the entire process does not expose users' sensitive data, ensuring data privacy and security.

[0088] Geometric measurement calculation and encrypted file query

[0089] In this invention, the main task of step S3 is to calculate the similarity between the query conditions and the encrypted file in the geometric space of the encrypted data based on the user's query conditions, and return the encrypted file that is similar to the query conditions. This step calculates the similarity of the encrypted data using geometric measurement methods (such as Euclidean distance and cosine similarity), and at the same time uses homomorphic encryption technology to ensure data privacy during the query process. Specifically, this step ensures that even if the file is in an encrypted state, the system can still perform efficient and secure query operations. The specific implementation of this step will be described in detail below, including its working principle, technical content and beneficial effects.

[0090] In this embodiment, the geometric measurement calculation and encrypted file query are implemented as follows:

[0091] First, when a user submits a query request, the query conditions are converted into numerical representations in geometric space. Specifically, the user's query conditions (such as file keywords, file type, and creation time) are converted into vector or numerical forms to facilitate matching and similarity calculations in the geometric space of encrypted files. This conversion process can be performed through text encoding, image feature extraction, or other feature extraction methods suitable for different types of data. For example, query conditions for text files can be converted into vector form using natural language processing techniques, while query conditions for images can be converted into feature vectors using image feature extraction algorithms (such as SIFT and SURF).

[0092] The geometric metrics used in the calculations include Euclidean distance, cosine similarity, and similarity measures of algebraic curves.

[0093] Euclidean distance: Euclidean distance, as a commonly used geometric metric, can be used to measure the similarity between a query condition and an encrypted file. Assuming the query condition is represented by vector q and the feature vector of the encrypted file is represented by f, the Euclidean distance can be calculated using the following formula:

[0094]

[0095] Where, q i and f i These are the i-th dimensions of the query conditions and the encrypted file vector, respectively, where n represents the number of dimensions of the query conditions and the encrypted file vector. By calculating the Euclidean distance, the system can obtain a similarity measure between the query conditions and the encrypted file. The smaller the distance, the higher the similarity.

[0096] Cosine similarity: Another commonly used metric is cosine similarity, which measures similarity by calculating the angle between the query condition and the feature vector of the encrypted file. The formula for calculating cosine similarity is as follows:

[0097]

[0098] Where q·f represents the dot product of vectors, and ∥q∥ and ∥f∥ are the magnitudes of the query condition vector and the encrypted file vector, respectively. The closer the cosine similarity value is to 1, the more similar the two are; the closer it is to 0, the less similar the two are.

[0099] Algebraic curve similarity measure: In some cases, the relationship between query conditions and file data may be represented as data points on an algebraic curve or surface, especially when dealing with images or complex data. In such cases, using an algebraic curve similarity measure can more effectively calculate the similarity between query conditions and encrypted files. For example, by using distance measurement methods in algebraic geometry, the geometric relationship between query conditions and file data points can be calculated.

[0100] It should be noted that all these geometric measurement methods can be calculated directly on encrypted data. The application of homomorphic encryption ensures that even if the file is encrypted, the system can still perform the above geometric measurement calculations on the encrypted data without decrypting the data. During the calculation process, all data remains encrypted, avoiding the risk of data leakage. Specifically, the system can use the addition and multiplication operations supported by homomorphic encryption to perform the above measurement calculations on encrypted data. For example, the system can calculate the dot product between encrypted feature vectors without decrypting them.

[0101] Query processing procedure:

[0102] During the query process, the query request is sent to the cloud system. The system calculates the similarity between the file and the query conditions based on the geometric metric results between the query conditions and the encrypted file, and sorts the encrypted files according to the similarity score. For each query request, the system first loads the encrypted data of the encrypted file related to the query conditions from the storage, then calculates the similarity between each file and the query conditions using the geometric metric method, and finally returns the encrypted file most relevant to the query conditions.

[0103] The similarity calculation results are sorted according to preset sorting rules, usually in descending order of similarity score, so that the most relevant files are returned to the user first. It should be noted that encrypted files in the query results remain encrypted. The system only returns encrypted data and does not expose the original content of the file.

[0104] Description of effects:

[0105] By employing geometric metrics for similarity calculation and combining them with homomorphic encryption, this invention enables efficient and secure encrypted file queries. Unlike traditional file query methods, this invention allows for query operations without decrypting the file, greatly enhancing query security. During the query process, by utilizing geometric metrics such as Euclidean distance and cosine similarity, the system can accurately calculate the similarity between the file and the query conditions, thereby ensuring the accuracy and efficiency of the query results.

[0106] Furthermore, by performing calculations in an encrypted state, the system effectively avoids the risk of data leakage and ensures the privacy of file data. Even when processed in the cloud, file data remains encrypted to prevent unauthorized access. At the same time, the application of geometric measurement methods enables queries to be executed efficiently on large-scale datasets, and query response time can be effectively controlled even with huge amounts of data.

[0107] The file management method based on cloud storage described below can be referred to in correspondence with the file management system based on cloud storage described above.

[0108] Please see the appendix Figure 2 This invention provides a file management method based on cloud storage, comprising:

[0109] The user-uploaded files are homomorphically encrypted and stored in the cloud.

[0110] The encrypted file is mapped to an algebraic cluster, and the storage path is optimized based on the file's geometric characteristics;

[0111] Upon receiving a query request, calculate the geometric measure of the encrypted data based on the query criteria and return an encrypted file similar to the query criteria.

[0112] The file storage location is dynamically adjusted based on the file access frequency, so that frequently accessed files can be retrieved quickly.

[0113] Encrypted files are distributed and stored across multiple cloud nodes, and query response speed is improved through parallel computing;

[0114] Low-dimensional embedding is used to embed high-dimensional data into low-dimensional data to reduce the complexity of query computation.

[0115] The method in this embodiment can be used to execute the above method embodiments, and its principle and technical effect are similar, so it will not be described again here.

[0116] 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 cloud storage based file management system, characterized by, The application relates to a cloud storage system based on homomorphic encryption and algebraic geometry, which comprises the following modules: a data encryption module for homomorphic encryption of user-uploaded files and storage of the encrypted files in the cloud; an algebraic geometry mapping module for mapping the encrypted files into algebraic varieties and optimizing the storage path according to the structure of algebraic geometry; a query processing module for receiving user query requests and performing geometric metric calculation on the encrypted files to return matching results in an encrypted state; an adaptive storage management module for dynamically adjusting the storage location of files according to the access frequency of the files to enable quick retrieval of high-frequency access files; a distributed computing and storage module for distributed storage of the encrypted files in multiple cloud nodes. The algebraic geometry mapping module embeds encrypted data into a geometric space through a polynomial equation in an algebraic variety to form a hierarchical storage structure, and optimizes the storage through the geometric properties of curves or varieties in algebraic geometry. The algebraic variety can be divided into multiple levels according to the characteristics of data, and the encrypted data in each level is dynamically adjusted according to the frequency and nature of the query. The adaptive storage management module automatically maps high-frequency access files to low-dimensional algebraic varieties according to the access frequency and query mode of the files, and saves infrequently accessed files in high-dimensional algebraic varieties to save storage space. The adaptive storage management module further comprises a dynamic adjustment algorithm which automatically adjusts the position of the files in the algebraic variety according to the real-time load condition and query frequency of the system, thereby balancing storage optimization and query performance.

2. The cloud storage based file management system of claim 1, wherein, The data encryption module uses homomorphic encryption technology for encryption. The homomorphic encryption supports addition and multiplication operations and can perform calculation operations on encrypted files in an encrypted state to avoid data leakage. The homomorphic encryption technology can be Paillier encryption or other homomorphic encryption schemes supporting addition and multiplication operations, and the homomorphic encryption technology can guarantee the security of file data and support incremental update of encrypted data.

3. The cloud storage based file management system of claim 1, wherein, The query processing module calculates the similarity between the query condition and the encrypted files in the geometric space of encrypted data according to the user query condition, and returns encrypted files with high similarity. The similarity calculation adopts a geometric metric method, which includes Euclidean distance, cosine similarity and algebraic curve similarity measurement. The geometric metric method supports operation in an encrypted state. The query results are sorted according to the similarity score, and the encrypted files most relevant to the query condition are returned.

4. The cloud storage based file management system of claim 1, wherein, The distributed computing and storage module stores encrypted files through distributed nodes, supports parallel computing to speed up query processing, performs similarity calculation of encrypted data on different computing nodes, distributes query requests to multiple computing nodes for parallel processing to reduce the computing load of a single node. The distributed computing and storage module can automatically dynamically adjust task allocation according to the load condition of the nodes to ensure stable operation of the system.

5. The cloud storage based file management system of claim 1, wherein, The query processing module calculates the similarity of encrypted data by a geometric metric method including Euclidean distance or algebraic curve similarity metric, and performs retrieval of encrypted files based on the similarity result, the geometric metric method can directly calculate on encrypted data, and ensures data privacy in the query process through homomorphic encryption, the similarity metric calculation supports vector calculation, clustering algorithm technology, and can adaptively adjust the calculation strategy according to the change of the query condition.

6. The cloud storage based file management system of claim 1, wherein, The algebraic geometric mapping module supports low-dimensional embedding of high-dimensional data, maps high-dimensional data into low-dimensional algebraic varieties for query optimization through the embedding theorem of algebraic geometry, the embedding method can effectively reduce the query overhead of high-dimensional data, optimize data processing efficiency under the premise of ensuring query accuracy, and dynamically adjust the embedding of high-dimensional data based on the properties of the query condition to optimize the storage and query process.

7. The cloud storage based file management system of claim 1, wherein, The adaptive storage management module further includes a query cache mechanism for caching encrypted files of repeated queries, the query cache mechanism can cache query conditions and corresponding encrypted file results, directly return encrypted data in the cache when the user submits the same query, and does not perform repeated calculation, the query cache mechanism supports efficient cache update strategy, and can dynamically adjust the storage and cleaning of cache content according to system load and query mode.

8. A cloud storage-based file management method applied to the cloud storage-based file management system of claims 1-7, characterized in that, Including: Homomorphic encryption processing is performed on the file uploaded by the user, and the encrypted file is stored to the cloud; Map the encrypted file to the algebraic variety, and optimize the storage path according to the geometric characteristics of the file; When receiving a query request, calculate the geometric metric of encrypted data according to the query condition, and return the encrypted file similar to the query condition; Dynamically adjust the storage location of the file according to the access frequency of the file, so that the frequently accessed file can be quickly retrieved; Distributed storage of encrypted files in multiple cloud nodes, and improve the query response speed through parallel calculation; Low-dimensional embedding of high-dimensional data.

Images