A data encryption sharing method and an efficient data anonymous indexing and retrieval method based on inner product function encryption

Abstract

The application discloses a data encryption sharing method and an efficient data anonymous tracking retrieval method based on inner product function encryption, and comprises the following steps: a data owner extracts the features of personal data and generates corresponding personal data feature vectors; the data owner sends the personal data feature vectors to a key generation center; the key generation center executes a key generation algorithm to generate corresponding private keys for the personal data feature vectors of the data owner; a retrieval execution party collects personal data of a data retrieval party; the retrieval execution party obtains public parameters from the key generation center, encrypts the personal data of the data retrieval party based on the public parameters, and forms vector ciphertext; and the retrieval execution party executes a decryption algorithm on the vector ciphertext based on the private key and the public parameters, and performs retrieval from the private key of the data owner.

Description

Technical Field

[0001] This invention belongs to the field of data security and privacy computing technology, specifically a data encryption and sharing method and an efficient data anonymity retrieval method based on inner product function encryption. Background Technology

[0002] Cloud-based anonymous data retrieval, as a privacy-preserving data retrieval technology for multi-user scenarios, aims to ensure privacy and security during the retrieval process. Specifically, this technology requires that the retrieval executor (such as a cloud service provider) and other participants cannot access any sensitive information related to the retrieval content or process. Data encryption, as a key technology for ensuring data confidentiality, has been widely used in data storage and sharing. Against this backdrop, researchers have proposed numerous cryptography-based cloud-based anonymous data retrieval schemes.

[0003] Based on the different cryptographic techniques employed, existing schemes can be divided into three main categories: schemes based on data feature digests, schemes based on secure multi-party computation, and schemes based on homomorphic encryption. Schemes based on data feature digests achieve efficient retrieval by extracting key data features, but when processing similar data, the repetitive feature digests can lead to decreased retrieval accuracy. Schemes based on secure multi-party computation ensure the security and accuracy of the retrieval process, but require multiple rounds of interactive communication between the participants. In multi-user scenarios, this significantly increases communication overhead and computational complexity, severely impacting retrieval efficiency. In contrast, schemes based on homomorphic encryption have become a hot research focus due to their advantages, such as supporting lossless ciphertext computation, executing complex retrieval logic, and not requiring frequent interactions in multi-user environments. However, these schemes still face the inherent ciphertext expansion and computational efficiency bottlenecks of homomorphic encryption algorithms, placing significant communication, storage, and computational burdens on all participants in the system. Currently available fully homomorphic encryption algorithms (CHEON JH, KIM A, KIM M, et al. Homomorphic encryption for arithmetic of approximate numbers[C] / / International Conference on the Theory and Applications of Cryptology and Information Security. Hong Kong, China, 2017:409-437) suffer from high computational overhead due to the inherent complexity of their underlying mathematical operations. Furthermore, the security of fully homomorphic encryption algorithms relies on difficult mathematical problems in high-dimensional spaces (such as the LWE problem), and the noise mechanisms introduced to support ciphertext operations significantly increase the size of the generated homomorphic ciphertext.

[0004] Therefore, how to overcome the efficiency bottlenecks of existing technical solutions and build a new data retrieval solution that combines high security and high retrieval performance while ensuring the privacy and security of retrieval data and processes has become a key scientific problem that urgently needs to be solved in the field of cloud-based anonymous data retrieval. Summary of the Invention

[0005] Purpose of the invention: To address the problems of high computational overhead and large ciphertext size faced by existing fully homomorphic encryption algorithms, this invention proposes a data encryption and sharing method and an efficient data anonymity retrieval method based on inner product function encryption. Aimed at the field of privacy-preserving data retrieval in multi-user scenarios, this invention can meet the functional and privacy protection requirements in complex data retrieval processes.

[0006] Technical solution: A data encryption and sharing method, comprising the following steps:

[0007] Data owners extract features from their personal data and generate corresponding personal data feature vectors;

[0008] Data owners send their personal data feature vectors to the key generation center;

[0009] The key generation center executes a key generation algorithm to generate a corresponding private key for the data owner's personal data feature vector;

[0010] The key generation center returns the generated private key to the data owner;

[0011] The data owner uploads the private key to the cloud service provider.

[0012] Furthermore, the key generation algorithm includes the following steps:

[0013] The key generation center obtains the security parameter λ and the maximum vector length.

[0014] Using a security parameter λ, the group parameters corresponding to the asymmetric group are generated using a group generation algorithm: in It is a multiplicative cyclic group, and q is the order of the group;

[0015] The key generation center randomly selects generators from the group.

[0016] For any The key generation center selects random numbers. And calculate

[0017] Key generation center generates public parameters and master key

[0018] The key generation center will broadcast the public parameter pp publicly and store the master key msk in a secure area;

[0019] Assuming the data owner's personal data feature vector is x, the key generation center generates the vector according to the following formula. Corresponding private key:

[0020]

[0021] This invention also discloses an efficient data anonymity retrieval method based on inner product function encryption, comprising the following steps:

[0022] Step 1: The key generation center uses a data encryption and sharing method to generate public parameters, master keys, and private keys for the personal data of different data owners; and sends the generated private keys to the corresponding data owners.

[0023] Step 2: Retrieve the personal data of the data collection party;

[0024] Step 3: Obtain public parameters from the execution direction key generation center, and encrypt the personal data of the data retrieval party based on these public parameters to form vector ciphertext;

[0025] Step 4: The retrieval executor performs a decryption algorithm on the vector ciphertext based on the private key and public parameters, and retrieves the data from the data owner's private key;

[0026] The data encryption and sharing method is one of the data encryption and sharing methods disclosed above.

[0027] Furthermore, in step 3, the retrieval execution direction key generation center obtains public parameters and encrypts the personal data of the data retrieval party based on these public parameters. Specific operations include:

[0028] Retrieve the public parameter pp from the key generation center for the execution direction;

[0029] Choose a random number And calculate C = g r D = h r ,

[0030] The personal data of the data retrieval party is encrypted according to the following formula to form the corresponding vector ciphertext.

[0031] Furthermore, in step 4, the retrieval executor performs a decryption algorithm on the vector ciphertext based on the private key and public parameters, and retrieves the data from the data owner's private key. Specific operations include:

[0032] Based on the private key and the vector ciphertext, the retrieval executor first calculates the intermediate parameter E' according to the following formula, and finally calculates log g(E') to obtain the inner product of vector x and vector y.<x,y> This enables retrieval from the data owner's private key;

[0033]

[0034] Where E i x is the ciphertext parameter obtained from the encryption vector. i Let be an element in vector x, and C and D be random parameters in the ciphertext, s x ,t x These are the random parameters in the key.

[0035] Furthermore, if the inner product result<x,y> If the integers are within the range {0,…,L}, then the Pollard kangaroo method is used to optimize the decryption algorithm to achieve O(L) time complexity. 1 / 2 The decryption algorithm is completed in a time complexity of 1 / l, where l represents the upper limit of the range of the calculated result.

[0036] Furthermore, by pre-computing a scale of O(l) 1 / 3 The table optimizes the decryption algorithm, reducing its time complexity to O(l). 1 / 3 ).

[0037] Furthermore, the execution of the decryption algorithm is optimized according to the following optimization process:

[0038] According to the Euclidean distance formula, the Euclidean distance between two vectors x and y is calculated as follows:

[0039]

[0040] Since x and y belong to different entities, cross-calculation is not possible. Therefore, the vectors are transformed as follows:

[0041]

[0042] Therefore, based on the characteristics of inner product operations, we have:

[0043]

[0044] For squared terms and The high and low bits are split to make the bit length of the split item close to that of other elements in the vector. The specific optimization method is as follows:

[0045] First, let the square term Among them Ψ (h) and Ψ (l) Representing the squared terms respectively The high and low positions, and Φ (h) and Φ (l) Representing the squared terms respectively High and low positions;

[0046] The original vector is then represented as follows:

[0047]

[0048] Computation based on the arbitrary inner product function encryption (IPFE) scheme<x″,y″> Through calculation<x″,y″> The equation dist(x,y) =<x″,y″> It still holds true.

[0049] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0050] (1) In view of the problems of large computational overhead and large ciphertext size of existing fully homomorphic encryption algorithms, this invention proposes a data retrieval method based on inner product function encryption, thereby effectively reducing the computational overhead and the corresponding ciphertext size in the data retrieval process;

[0051] (2) This invention utilizes the concept of pre-computation to further optimize the performance of the data retrieval algorithm (i.e., the decryption algorithm), thereby significantly improving the efficiency of data retrieval.

[0052] (3) The method of this invention enables efficient and privacy-preserving data retrieval in multi-user scenarios. Specifically, it allows for data sharing, storage, and retrieval without leaking any private information, assuming the third-party cloud server is honest and curious. On one hand, the data owner uses the corresponding key instead of the data itself to upload it to the cloud server for outsourced storage and retrieval, thus saving data communication and storage costs while protecting data privacy. On the other hand, during data retrieval, the third-party cloud server can perform the retrieval without decrypting the data, and the data retrieval method based on data similarity measurement ensures efficient and lossless retrieval of encrypted data. Therefore, the method of this invention can ensure privacy and security during data retrieval in multi-user scenarios while supporting efficient and lossless data retrieval. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the data encryption and sharing phase.

[0054] Figure 2 This is a schematic diagram of the data retrieval stage;

[0055] Figure 3 This is a schematic diagram of a cloud-edge collaborative image escaping retrieval scenario.

[0056] Figure 4A comparison chart showing the time cost of each algorithm under different safety and optimization parameters;

[0057] Figure 5 This is a comparison chart of the time cost of each operation in the method of the present invention under different safety parameters;

[0058] Figure 6 This is a schematic diagram of the cryptographic parameter size evaluation of each operation in the method of the present invention under different vector dimensions;

[0059] Figure 7 This is a comparison chart of the encryption and decryption time overhead of the method of this invention and FHE under different vector dimensions;

[0060] Figure 8 A comparison chart showing the time overhead of each algorithm in host and Raspberry Pi environments. Detailed Implementation

[0061] To make the objectives, technical solutions, and advantages of this invention clearer, the following description, in conjunction with the accompanying drawings, will further illustrate the efficient data anonymization retrieval method based on inner product function encryption proposed in this invention.

[0062] This embodiment proposes an efficient data anonymity retrieval method based on inner product function encryption, which mainly includes four steps, as follows:

[0063] Step 1: System Initialization This step is handled by the Key Generation Center (Key

[0064] The Generation Center (KGC) executes the process, with inputs including the security parameter λ and the maximum vector length. The output includes public parameters pp and the master key msk. Specifically, the key generation center KGC first uses the security parameter λ to generate the group parameters corresponding to the asymmetric group through a group generation algorithm: in Let q be a multiplicative cyclic group, and q be the order of the group. Next, the Key Generation Center (KGC) randomly selects generators from the group. Subsequently, for any Key Generation Center (KGC) selects random numbers And calculate Finally, the Key Generation Center (KGC) generates common parameters. and master key The Key Generation Center (KGC) broadcasts the public parameter pp publicly and stores the master key msk in a secure area.

[0065] Step 2: Key generation KeyGen(msk,x)→sk xThe key generation algorithm is executed by the Key Generation Center (KGC). This algorithm takes the master key msk and the user's facial feature data vector x as input, and calculates and outputs the corresponding private key sk. x Specifically, in order to generate vectors The corresponding key is calculated by the Key Generation Center (KGC) using the following formula:

[0066]

[0067] Step 3: Data Encryption Encrypt(pp,y) → C y The encryption algorithm is executed by the data retrieval party (or retrieval executor). The algorithm takes a common parameter pp and a data vector y to be encrypted as input, and outputs the corresponding vector ciphertext C. y Specifically, the user first selects a random number. And calculate C = g r D = h r , The corresponding ciphertext can be represented as After the ciphertext is generated, it is sent to the retrieval executor for data retrieval.

[0068] Step 4: Decrypt the data Decrypt(pp, sk x C y →<x,y> The decryption algorithm is executed by the retrieval executor. The algorithm takes into account the public parameter pp and the private key sk corresponding to the target vector as input. x and the ciphertext C to be retrieved y The algorithm outputs the inner product of vectors x and y.<x,y> Specifically, based on the private key sk x =(s x ,t x ) and vector ciphertext C y To perform the retrieval, the retrieval execution first calculates E' according to the following formula, and finally calculates log g(E') to obtain the inner product of vector x and vector u.<x,y> .

[0069]

[0070] The above decryption algorithm involves the calculation of a discrete logarithm. This calculation can be optimized from the following two perspectives:

[0071] 1) If the inner product result<x,y> For integers in a relatively small integer range {0,…,L}, the Pollard kangaroo method can be used for optimization, achieving O(L...) time complexity. 1 / 2 The calculation is completed in a time complexity of 10 ...

[0072] 2) Building upon the methods described above, the time complexity of the decryption algorithm can be further optimized through pre-computation. Specifically, this involves pre-computing a variable of size O(L...)... 1 / 3 The table can reduce the time complexity of the decryption process to O(L). 1 / 3 This is considered common sense in the field of computer science regarding optimizing computational complexity.

[0073] The specific optimization process is as follows:

[0074] According to the Euclidean distance formula, the Euclidean distance between two vectors x and y is calculated as follows:

[0075]

[0076] Since x and y belong to different entities, they cannot be cross-calculated. Therefore, the vector can be transformed as follows:

[0077]

[0078] Therefore, based on the characteristics of inner product operations, we have:

[0079]

[0080] Note that the above inner product operation involves the calculation of square terms (which are respectively) and The presence of squared terms significantly increases the time complexity of the proposed method. This is because encoding requires a sufficiently large number of bits to accommodate the potential range of computational results for each element. This range is determined by the bit length of the largest bit-length term, and the presence of squared terms causes this bit length to increase exponentially. To achieve optimal time complexity, the high and low bits of the squared terms can be split, making the bit length of the split terms closer to that of other elements in the vector. The specific optimization method is as follows:

[0081] First, let the square term Among them Ψ (h) and Ψ (l) Representing the squared terms respectively The high and low positions, and Φ (h) and Φ (1) Representing the squared terms respectively The high and low bits can then be used to further represent the original vector based on the vector transformation formula, as follows:

[0082]

[0083]

[0084] Through calculation<x″,y″> The equation dist(x,y) =<x″,y″> It still holds true, and calculations based on the IPFE scheme...<x″,y″> It has more advantages than direct calculation<x,y> Higher computational efficiency.

[0085] The workflow of the method proposed in this embodiment is divided into two stages: encrypted data sharing and encrypted data retrieval.

[0086] Figure 1 This demonstrates the execution sequence of each step in the encrypted data sharing phase:

[0087] Step 101: The data owner extracts the features of the personal data and generates the corresponding feature vector.

[0088] Step 102: The data owner sends the personal data feature vector to the key generation center.

[0089] Step 103: The key generation center executes the key generation algorithm to generate a corresponding key for the data owner's data vector. This key is also used as a tag for the data.

[0090] Step 104: The key generation center returns the generated data tag (key) to the data owner.

[0091] Step 105: The data owner uploads the data tag (key) to the cloud service provider for subsequent data retrieval.

[0092] Figure 2 This demonstrates the execution order of each step in the encrypted data retrieval phase:

[0093] Step 201: Retrieve the information of the data retrieval party collected by the executor (such as facial information).

[0094] Step 202: Retrieve the public parameters used for encrypting data from the execution direction key generation center.

[0095] Step 203: The retrieval execution party extracts the features of the collected user information and generates the corresponding feature vector.

[0096] Step 204: The retrieval executor executes the encryption algorithm and encrypts the extracted data feature vector using common parameters.

[0097] Step 205: The retrieval component of the retrieval executor executes the decryption algorithm and uses the encrypted feature vector to search for tags uploaded by the data owner.

[0098] like Figure 3As shown, in a cloud-edge collaborative image data anonymity retrieval scenario, users can register by sending the key corresponding to the feature vector of their facial data to the cloud server. The cloud server then sends the key corresponding to the user's facial data to the associated edge node based on the correspondence between the user and the edge node. When a user needs to access a node, the node's facial information acquisition device first collects the user's facial features, then encrypts them, and finally sends the encrypted ciphertext to the edge node's data retrieval component for retrieval. Based on the retrieval results, it is determined whether the user has access rights to the current node.

[0099] To address this scenario, the proposed method (IPFE-IR) was systematically tested. The experimental environment was set up as follows: The method in this embodiment is implemented based on the PBC function library and C++-14. The server configuration used in the test was as follows: CPU parameters: "2.40GHz Intel Xeon Silver 4314CPU", 32GB of RAM; Raspberry Pi configuration: CPU parameters: "2.4GHz Broadcom BCM2712 Cortex-A76 (Arm v8)", 8GB of RAM. Both devices ran on Ubuntu 22.04LTS system. To prevent runtime errors from affecting the accuracy of the experimental results, each algorithm in this embodiment was executed 10,000 times and the average time cost was calculated. Since the Setup algorithm is only executed once during system initialization, its time cost has a negligible impact on the overall system performance; therefore, the time cost of this algorithm is not included in the following experimental tests.

[0100] (1) Algorithm Optimization Evaluation

[0101] Figure 4 The time cost of each algorithm in this embodiment is given. The optimization process of the computational characteristics of this embodiment is as follows:

[0102]

[0103] The optimization process focuses on reducing the size of the element with the longest bit length in the vector. Therefore, the implementation process needs to balance the relationship between the element length boundary that the vector can accommodate and the vector scaling scale. During the experiment, it was found that the highest bit length corresponding to the extracted image feature vector is 10 bits. Therefore, the bit length boundary can be set to 9 for scaling, i.e., k = 9 in the other two vector transformation formulas (corresponding to Bound in the figure). According to the experimental results in the figure, in tests with multiple vector dimension sizes under the same security parameters, as the Bound value increases, the time overhead of key generation and encryption algorithms is not significantly different, but the time overhead of decryption algorithms increases significantly. This indicates that the optimization in this embodiment is effective; by selecting a smaller Bound value, the time overhead of decryption operations can be effectively reduced. Furthermore, this embodiment also evaluates different security parameters (…). Figure 4 (a)} Figure 4 (b) Figure 4 The security parameter corresponding to (c) is 1024. Figure 4 (d) in Figure 4 e) Figure 4 The (f) in the equation corresponds to the security parameter 2048. Figure 4 (g) in Figure 4 (h) in Figure 4 (i) in the figure corresponds to the performance of each algorithm under a security parameter of 4096. According to the results in the figure, as the size of the security parameter increases, the time cost of each algorithm increases, but the performance difference between different bounds remains unchanged.

[0104] (2) Data retrieval performance evaluation

[0105] To accurately evaluate the performance of the method in this embodiment in real-world scenarios, the performance overhead of different operations in this embodiment is tested. To test the optimal performance of the method in this embodiment, a Bound=9 setting is selected, with the data vector length ranging from [2...]. 4 ,2 10 Under the condition of [unclear], the time cost of each operation in the method of this embodiment was tested. This embodiment includes tag generation (i.e., KeyGen algorithm); image acquisition and processing (including Encrypt algorithm and image feature extraction); and retrieval operations (including Decrypt algorithm and inner product result comparison operation). The selection of elliptic curve parameters in this experiment refers to the "secp224r1" curve recommended by the NIST standard. Table 1 shows the specific performance test results of each operation in the method of this embodiment when the security parameter λ = 2048. According to the test results in the table, as the data vector dimension increases, the time cost of each operation increases linearly with the vector dimension. This experiment further tested the time cost of each operation in the method of this embodiment under different security parameters, and the test results are as follows: Figure 5 As shown in the figure, the experimental results show that as the safety parameters increase (the safety of the scheme increases), the execution time of each operation increases linearly, which is in line with the general rule.

[0106] Table 1 shows the time overhead test results for each operation in the proposed method when the safety parameter is 2048 (unit: milliseconds).

[0107]

[0108] (3) Performance comparison with similar methods

[0109] Before making comparisons, we first analyze similar anonymity retrieval methods. Privacy-preserving data retrieval schemes based on homomorphic encryption are limited by the ciphertext expansion and high computational complexity inherent in homomorphic encryption, typically resulting in significant computational, communication, and storage overhead. Secure multi-party computation involves frequent interactions between multiple parties, leading to high communication complexity in the protocol itself, causing related anonymity retrieval methods to suffer from high storage overhead and high communication latency. While data feature digest-based schemes do not introduce excessive computational, communication, or storage overhead, the hash collision problem they present can potentially impact the accuracy of data retrieval.

[0110] In contrast, the data retrieval technique based on inner product function encryption in this embodiment uses a key corresponding to the data feature vector instead of the vector itself for sharing and retrieval. This not only avoids the heavy overhead of storage and communication for each participant in the system but also effectively ensures the accuracy of data retrieval, thus achieving a good trade-off between performance and privacy protection. Since methods based on data feature summarization and secure multi-party computation have significant performance defects, they are not suitable for data retrieval in cloud-edge collaborative scenarios. Therefore, this experiment only compares the performance with methods based on homomorphic encryption.

[0111] Specifically, this embodiment compares and analyzes the method with the stealth retrieval technique based on fully homomorphic encryption (FHE) (CHEON JH, KIM A, KIM M, et al. Homomorphic encryption for arithmetic of approximate numbers [C] / / International Conference on the Theory and Applications of Cryptology and Information Security. Hong Kong, China, 2017:409-437) from two aspects: cryptographic parameter scale and algorithm performance. First, regarding the cryptographic parameter scale, this experiment compares the size of the relevant cryptographic parameters (including public parameters, keys, ciphertext, or secret parameters) involved in the data retrieval process of the two techniques. This experiment is tested under data vectors of different dimensions to evaluate the communication and storage overhead of the two techniques in the stealth retrieval process of data vectors of different sizes. Since the security of the fully homomorphic encryption algorithm depends on mathematical problems in high-dimensional linear spaces (such as the LWE problem), its ciphertext must be large enough to accommodate the increasing noise during the calculation process and ensure security. Considering the security of the scheme, this experiment selected a minimum vector dimension of 4096 when evaluating the fully homomorphic encryption technology. All data vectors with actual dimensions less than 4096 were processed as 4096. This experiment compared two stealth retrieval technologies from three stages: initialization, registration, and retrieval. The statistical security parameter length of the fully homomorphic encryption technology was set to 128 bits, while the security parameter length of the proposed method was set to 2048 bits, with a Bound of 9. The experiment tested data with vector dimensions of 4096, 8192, and 16384. Table 2 shows the comparison results of the cryptographic parameter size of the two technologies under different vector dimensions. In the initialization stage, both technologies require passing common parameters for system initialization. In the user registration stage, the fully homomorphic encryption requires passing the parameters used for calculation (including reline and gal in the table) and the ciphertext of the data feature vector, while the proposed method only requires passing the key corresponding to the data feature vector. In the retrieval stage, both technologies require passing the ciphertext for retrieval. Experimental results show that the proposed method has a larger common parameter and ciphertext size than fully homomorphic encryption, while the parameter size during the registration phase is much smaller. In a single data retrieval process, the common parameters and the ciphertext used for retrieval only need to be transmitted / stored once, while the parameter size during the registration phase is related to the number of registered users. Therefore, in large-scale data retrieval scenarios, compared to fully homomorphic encryption, the method in this embodiment significantly reduces the communication and storage overhead between the central server and edge nodes.

[0112] Furthermore, to further illustrate the advantages of the proposed method over fully homomorphic encryption when processing smaller-dimensional image data, this experiment tested the cryptographic parameter scale of the proposed method when processing smaller-dimensional image data. The experimental results are as follows: Figure 6 As shown in the figure. This experiment tests the IPFE-IR technology from three stages: initialization, registration, and retrieval. Bound=9 was set, and security parameters were selected as 1024, 1536, 2048, 2560, 3072, and 4096. The experimental results show that for the common parameters in the initialization stage, the size of the common parameters increases linearly with the increase of security parameters and vector dimension. In the user registration stage, since the password corresponding to the user's image feature vector is only related to the security parameters, the key size remains unchanged regardless of the vector dimension under the same security parameter settings. For the image retrieval stage, the ciphertext size also increases with the increase of security parameters and vector dimension. Compared with fully homomorphic encryption technology, the method in this embodiment provides a flexible data processing approach, possessing a smaller cryptographic parameter scale when processing smaller-dimensional data vectors, thus resulting in less communication and storage overhead.

[0113] Table 2 Comparison of cryptographic parameter sizes between the method in this embodiment and FHE under different vector dimensions (unit: KB)

[0114]

[0115]

[0116] In the comparative analysis of algorithm performance, since the KeyGen algorithm is executed only once, this experiment does not involve a performance comparison of that algorithm. The main focus is on comparing the time overhead of the encryption and decryption algorithms in the two technologies. This experiment is conducted with a prime field size of 2^24 and a security parameter λ = 2048. The selected test vector size ranges from {2^24} to {2^24}. 4 -2 10 The test results are as follows: Figure 7 As shown in the figure. Experimental results show that, under different test vector dimensions, although the encryption time of the method in this embodiment is longer, its decryption time is much shorter than that of the fully homomorphic encryption method. In a single image retrieval process, the encryption operation for the retrieved image only needs to be performed once, while the number of data retrievals (corresponding to the decryption of both) is proportional to the size of the image database. Therefore, the overall overhead of the retrieval process is proportional to the decryption overhead of a single operation in related schemes. Thus, the method in this embodiment has a significantly higher retrieval efficiency than fully homomorphic encryption schemes with equivalent security in large-scale image database retrieval scenarios, demonstrating a clear performance advantage.

[0117] (4) Performance evaluation on terminal devices

[0118] To further demonstrate the broad applicability of the method in this embodiment, the performance of each algorithm of this embodiment was tested on a Raspberry Pi device. To comprehensively evaluate the performance of the method in this embodiment, this experiment tested the time overhead of each algorithm on the Raspberry Pi device when the Bound value was 9, 10, and 11. This experiment used different security parameters (1024, 2048, and 4096) and test data vector sizes (ranging from 2). 4 -2 10 The test was conducted below. The test results are as follows: Figure 8 As shown in the figure. The test results show that even on a low-performance Raspberry Pi device, the algorithms in this embodiment do not suffer significant performance loss. For 2... 10 The time cost of a single query for the dimensional data vector is no more than 0.2 seconds. Furthermore, by comparing test results with different Bound values, it can be seen that the method in this embodiment is also effective on Raspberry Pi devices, indicating that the method in this embodiment has good performance in device environments with varying performance levels.

Claims

1. A method for encrypted data sharing, characterized in that: Includes the following steps: Data owners extract features from their personal data and generate corresponding personal data feature vectors; Data owners send their personal data feature vectors to the key generation center; The key generation center executes a key generation algorithm to generate a corresponding private key for the data owner's personal data feature vector; The key generation center returns the generated private key to the data owner; The data owner uploads the private key to the cloud service provider; The key generation algorithm includes the following steps: The key generation center obtains the security parameter λ and the maximum vector length l; Using a security parameter λ, the group parameters corresponding to the asymmetric group are generated using a group generation algorithm: in It is a multiplicative cyclic group, and q is the order of the group; The key generation center randomly selects generators from the group. For any i ∈ {1,…,l}, the key generation center selects a random number. And calculate Key generation center generates public parameters and master key The key generation center will broadcast the public parameter pp publicly and store the master key msk in a secure area; Assuming the data owner's personal data feature vector is x, the key generation center generates the vector according to the following formula. Corresponding private key:

2. A highly efficient data anonymity retrieval method based on inner product function encryption, characterized in that: Includes the following steps: Step 1: The key generation center uses a data encryption and sharing method to generate public parameters, master keys, and private keys for the personal data of different data owners; and sends the generated private keys to the corresponding data owners. Step 2: Retrieve the personal data of the data collection party; Step 3: Obtain public parameters from the execution direction key generation center, and encrypt the personal data of the data retrieval party based on these public parameters to form vector ciphertext; Step 4: The retrieval executor performs a decryption algorithm on the vector ciphertext based on the private key and public parameters, and retrieves the data from the data owner's private key; The data encryption and sharing method described in claim 1 is the data encryption and sharing method described in claim 1. In step 3, the retrieval execution direction key generation center obtains public parameters and encrypts the personal data of the data retrieval party based on these public parameters. Specific operations include: Retrieve the public parameter pp from the key generation center for the execution direction; Choose a random number And calculate C = g r D = h r , The personal data of the data retrieval party is encrypted according to the following formula to form the corresponding vector ciphertext C. y =(C,D,E1,…,E) l ).

3. The efficient data anonymity retrieval method based on inner product function encryption according to claim 2, characterized in that: In step 4, the retrieval executor performs a decryption algorithm on the vector ciphertext based on the private key and public parameters, and retrieves the data from the data owner's private key. Specific operations include: Based on the private key and the vector ciphertext, the retrieval executor first calculates the intermediate parameter E' according to the following formula, and finally calculates log g(E') to obtain the inner product of vector x and vector y.<x,y> This enables retrieval from the data owner's private key; Among them, E i x is the ciphertext parameter obtained from the encryption vector. i Let be an element in vector x, and C and D be random parameters in the ciphertext, s x ,t x These are the random parameters in the key.

4. The efficient data anonymity retrieval method based on inner product function encryption according to claim 3, characterized in that: If the inner product result<x,y> If the integers are within the range {0,…,L}, then the Pollard kangaroo method is used to optimize the decryption algorithm to achieve O(L) time complexity. 1 / 2 The decryption algorithm is completed in a time complexity of 1 / L, where L represents the upper limit of the range of the calculated result.

5. The efficient data anonymity retrieval method based on inner product function encryption according to claim 3, characterized in that: By pre-calculating a size of O(L) 1 / 3 The table optimizes the decryption algorithm, reducing its time complexity to O(L). 1 / 3 ).

6. The efficient data anonymity retrieval method based on inner product function encryption according to claim 3, characterized in that: The decryption algorithm is optimized using the following process: According to the Euclidean distance formula, the Euclidean distance between two vectors x and y is calculated as follows: Since x and y belong to different entities, cross-calculation is not possible. Therefore, the vectors are transformed as follows: Therefore, based on the characteristics of inner product operations, we have: For squared terms and The high and low bits are split to make the bit length of the split item close to that of other elements in the vector. The specific optimization method is as follows: First, let the square term Among them Ψ (h) and Ψ (l) Representing the squared terms respectively The high and low positions, and Φ (h) and Φ (l) Representing the squared terms respectively High and low positions; The original vector is then represented as follows: Calculation based on arbitrary inner product function encryption scheme<x″,y″> Through calculation<x″,y″> The equation dist(x,y) =<x″,y″> It still holds true.

Images