An Identity-Based User-Controllable Cloud Storage Data Security Audit Method

By using a distributed public key generation and self-verification mechanism, the security and computational burden issues in cloud storage data auditing are resolved. This enables data owners to exercise self-control over the audit process and achieves provable security of TPA, making it suitable for environments with limited computing resources.

CN123122801APending Publication Date: 2026-11-17KUNMING UNIV OF SCI & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611167959.3
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-03
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing cloud storage data auditing solutions suffer from insufficient security, excessive computational burden, lack of self-verification mechanism for TPA credibility, and the risk of TPA data theft. In particular, it is difficult to construct a provably secure identity-based data auditing solution under the standard model.

Method used

The system uses a distributed public key generation center (D-PKG) to generate system parameters based on the Feldman verifiable secret sharing threshold protocol. Data owners generate authentication tags independently and verify them independently through the audit controller (AC). A random data blinding factor is introduced to protect data privacy. The TPA computation cost is constant and it supports resistance to multiple types of attacks.

Benefits of technology

It achieves provable security under the standard model, allows data owners to autonomously manage the audit process, reduces the computational burden of TPA, prevents TPA from stealing data, and is suitable for environments with limited computing resources.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122801A_ABST
    Figure CN123122801A_ABST
Patent Text Reader

Abstract

This invention discloses an identity-based, user-controllable cloud storage data security auditing method, belonging to the field of cloud storage data security technology. It includes the following steps: multiple key nodes perform threshold distributed key generation based on Feldman verifiable secret sharing to generate system parameters; each node collaboratively generates an identity private key for the data owner and supports blind extraction; the local audit controller generates control factors and their commitments in each round and binds them to authorization credentials; the data owner generates tags for file blocks and uploads them to the cloud; a third-party auditor initiates a challenge with credentials, and the cloud service provider introduces blinding factors to generate proof information and sends it to both the auditor and the controller, respectively for integrity verification and self-verification. This invention replaces a centralized key generation center with distributed threshold key generation, eliminating the risk of single-point failure of the master key and privacy leakage, achieving autonomous control and collusion detection in the audit process, and the verification cost is independent of the number of challenge blocks.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention belongs to the field of information security technology, specifically relating to an identity-based, user-controllable cloud storage data security auditing method. Background Technology

[0002] With the rapid development of cloud computing technology, more and more enterprises and individuals are storing data in the cloud. However, due to the lack of direct physical management of cloud data, data integrity verification has become a core issue of cloud storage security. Data auditing solutions effectively solve the problem of data owners being unable to conduct their own audits due to busy schedules or other reasons by authorizing third-party auditors (TPAs) to perform integrity verification on behalf of the data owners.

[0003] In Identity-Based Data Auditing (IBDA), Zhang et al., in their 2018 paper "Enabling efficient user revocation inident-based cloud storage auditing for shared big data" published in IEEE Transactions on Dependable and Secure Computing, proposed an identity-based cloud storage auditing scheme for shared big data, supporting efficient user revocation. However, its security proof relies on a random oracle model (ROM). Theoretically, however, provably secure schemes under the random oracle model are not absolutely secure in practical deployments. This is because it assumes that the hash function behaves completely randomly, but in reality, any specific hash implementation deviates from the ideal random function, potentially leading to exploitable vulnerabilities. Therefore, how to construct a provably secure IBDA scheme without relying on the standard random oracle model has become a pressing problem.

[0004] In their 2018 paper "Enabling identity-based integrity auditing and data sharing with sensitive information hiding for secure cloud storage" published in IEEE Transactions on Information Forensics and Security, Shen et al. pointed out that TPAs ​​(Third-Party Authorities) could potentially infer the original data content of the data owner by analyzing the proof information returned by the CSP (Content Sharing Service), leading to data privacy breaches. In reality, TPAs, as independent third-party organizations, have the motivation to actively steal data for commercial gain. Data owners should ensure that TPAs ​​only perform integrity verification functions and cannot access the data content.

[0005] In summary, the existing solutions have the following problems: (1) Public key generation centers (PKGs) generally adopt a centralized architecture, and the security of the entire system will collapse once the master key is leaked; (2) Data owners usually delegate the auditing task entirely to TPA, lacking an independent verification and control mechanism for audit parameters and audit results. Once TPA behaves abnormally or is compromised, data owners cannot independently verify the authenticity of the audit. (3) TPA may extract the data owner's data file through the proof information returned by CSP, causing data leakage. Existing solutions generally set TPA as a completely trusted entity and lack defense against TPA-CSP collusion. (4) The scheme generally relies on hash-to-point operations, the number of which increases linearly with the number of data blocks, resulting in excessive computational burden and making it unsuitable for computationally constrained environments such as mobile terminals and IoT devices.

[0006] Therefore, researching an identity-based data auditing scheme that possesses provable security under the standard model, supports data owners' autonomous control of the audit process, has constant TPA computational cost, and can effectively resist various types of attacks has significant theoretical and practical implications. Summary of the Invention

[0007] To address the aforementioned issues, this invention proposes an identity-based, user-controllable cloud storage data security auditing method. This invention offers provable security under a standard model, supports data owners in autonomously managing the audit process, and effectively resists various attacks, including external attackers forging tags, cloud service providers (CSPs) forging authentication information, and third-party auditors (TPAs) stealing data. It is suitable for cloud storage environments with limited computing resources.

[0008] The technical solution adopted in this invention is: an identity-based, user-controllable cloud storage data security auditing method, comprising the following steps: Step 1 System Initialization: by The D-PKG, a distributed public key generation center composed of 10 nodes, uses Feldman-based verifiable secret sharing. Threshold Protocol Collaborative Generation System Public Parameters Including bilinear pair mapping ,in for Additive group of order, for Factorial method group, and , , Three Generator; Master key No explicit refactoring required, parameters exposed ; Step 2 Key Extraction: The data owner selects a random identity blinding factor. Calculate blind identity And submit it to D-PKG; at least A D-PKG node uses a blinded identity Collaborative generation of blind private keys; the data owner deblinds the blind key locally, generating a private key and public key corresponding to their identity; Step 3 Audit Authorization: The data owner generates control factors through the local audit controller (AC). and its public commitments And generate a unique round identifier. Third-party auditor identifier and the validity period of the authorization certificate AC will Send to TPA and CSP, Stored locally, the TPA audit authorization is automatically revoked upon expiration; the authorization credential is generated by a dedicated hash function. Calculation generation, Hash functions independent of the proof information generation stage ; Step 4 Tag Generation: The data owner uses the private key obtained in Step 2 to divide the data file into blocks and generate an authentication tag for each data block. Then, the data block and its tag are uploaded to the CSP, and the public information required for verification is sent to the AC for safekeeping. Step 5 Challenge Generation: TPA with Authorization Certificate An audit challenge is launched against the CSP, and the CSP verifies the results before handling the challenge. the legitimacy of the challenge, and executes subsequent steps only after the verification is passed; otherwise, it rejects the challenge and notifies the AC; Step 6 Proof Generation and Verification: The CSP aggregates the tags of the challenged data blocks according to the challenge information, introduces a random data blinding factor to blind the aggregated data, and generates proof information and sends it to the TPA and the AC; the TPA performs verification according to the public identity information and submits the audit conclusion to the AC; Step 7 Autonomous Verification: The AC uses the locally stored control factor and the proof information to perform three independent verifications on the legitimacy of the TPA identity, the consistency of the proof information, and the audit result, then compares the verification result with the TPA conclusion. If they are consistent, the current round of audit is confirmed to be valid; if they are inconsistent, a re-audit is triggered and the abnormal behavior of the TPA is recorded.

[0009] Specifically, in the Step 1, each of the nodes of the D-PKG independently selects random polynomials and publishes the commitment values, the nodes mutually send shares to each other and verify the correctness thereof, and each node calculates its own joint share; the master public key is calculated by summing and publicizing the commitment values of all nodes; any or more nodes cooperatively generate the private key through Lagrange interpolation, and any fewer than nodes cannot obtain any information of .

[0010] Specifically, in the Step 2, at least D-PKG nodes cooperatively generate a private key pair and a public key pair for the data owner, satisfying

[0011]

[0012] wherein are two private key components of the data owner, is an auxiliary public parameter generated by the public key, and are jointly generated joint random numbers, and is a hash function.

[0013] Specifically, in the Step 3, the AC generates a control factor and a round identifier in each round, and calculates a public commitment , wherein is Elliptic curve commitment, publicly released, authorization certificate

[0014] For the validity period of the authorization, The filename of the data file. This is a collision-resistant hash function specifically designed for generating authorization credentials.

[0015] Specifically, in Step 4, the private key is used. , data file Divided into Block, F= ,set up Each block is randomly selected Calculate the authentication label:

[0016] in For the first Authentication tags for each data block; For the first The data values ​​of each data block; These are the two private key components for the data owner; For the group Two independent generators; For the first Random numbers corresponding to each data block This is a public commitment to the random number; For the first The filename of each data block is concatenated with its block index; , , These are the main auxiliary value, left auxiliary value, and right auxiliary value derived from three collision-resistant hash functions, respectively. For the first Each data block contains a complete tag pair; each block requires only 3 scalar multiplications, eliminating the need for hash-to-point operations.

[0017] Specifically, in Step 5, the TPA holds the authorization certificate. Launch an audit challenge against the CSP, randomly selecting the number of data blocks to participate in the challenge. Select a random seed Challenge information The challenge is sent to the CSP, which then verifies it before processing it. Third-party auditor identification Is it in the license list? What is the license validity period? Whether it is within the validity period and the unique round identifier The system will check whether the device has been used before proceeding with the next steps.

[0018] Specifically, in Step 6, the CSP receives Afterwards, from Extract K Calculate the index of the challenged data block. With weight , , The loop number of the challenged block. The first one to be challenged One data block in the data file The original block number in the data; calculate the aggregation tag. With auxiliary aggregation value ,in , , , They are respectively the block number The corresponding authentication label, data block value, hash master and auxiliary value, and random number commitment; randomly select data blinding factor. λ ,calculate , , CSP will Simultaneously sent to TPA and AC, TPA calculates...

[0019]

[0020] And aggregate the hash values ​​of all challenged data blocks: left auxiliary value and right auxiliary value Verify the equation:

[0021] If the equation is true, output 1; otherwise, output 0. Submit the audit conclusion to AC. Blinding commitments, To blind aggregate data, For binding and The polymerization coefficient, , Each is a corresponding private key , The public key for verification, .

[0022] Specifically, in Step 7, the AC performs the following three verifications independently in sequence: Step 7.1, AC retrieves the current data from the local machine. corresponding Recalculate and The hash value in the system is compared. If they match, the TPA identity is deemed legitimate, and Step 7.2 is executed. If they do not match, the TPA identity is deemed illegitimate, the AC determines that this round of auditing is invalid, triggers a re-audit, and records the abnormal behavior of the TPA. Step 7.2, AC compares the data sent directly by CSP. Submitted with TPA If they are consistent, it is determined that TPA has not tampered with the proof information, and Step 7.3 continues; if they are inconsistent, it is determined that TPA has tampered with the proof information, AC determines that this round of audit is invalid, triggers a re-audit and records the abnormal behavior of TPA. Step 7.3, independently execute the same verification equation (5) as TPA to locally verify the audit results; If the AC independent verification results are consistent with the TPA conclusions, the audit is confirmed to be valid; otherwise, a re-audit is triggered and the TPA abnormal behavior is recorded.

[0023] The beneficial effects of this invention are: (1) Data owner: The tag generation stage requires 3 Subscalar multiplication (SM) without hash-to-point (HP) operations has a lower computational cost than existing comparable schemes.

[0024] (2) Cloud service provider: The proof information generation stage requires 2c SMs. The CSP has sufficient computing resources, so the overhead is acceptable.

[0025] (3) Third-party auditor: The verification phase of the proof information only requires 3 bilinear pairing operations (BP) and 2 SM operations, with a constant computational cost that does not increase with the number of data blocks c participating in the challenge. Taking c=10000 and c=200 as examples, the TPA calculation cost is about 62.693ms, while the TPA cost in the comparison scheme is as high as 4000ms or more.

[0026] (4) Audit Controller AC: The overhead of AC executing the Self-Verify algorithm is the same as that of TPA's Proof-Verify (3 BPs and 2 SMs, about 62.693ms). This overhead is executed locally by the data owner and does not affect the overall system throughput. Attached Figure Description

[0027] Figure 1 : Schematic diagram of the system model of this invention; Figure 2 : Schematic diagram of distributed PKG key generation protocol; Figure 3: Schematic diagram of audit authorization and self-verification mechanism; Figure 4 : Complete algorithm flowchart of this invention; Detailed Implementation

[0028] To describe the present invention in more detail and to facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings and embodiments. The embodiments in this section are for illustrative purposes and to facilitate understanding, and are not intended to limit the present invention.

[0029] Example 1: As Figure 1-4 As shown, the present invention: Innovation Point 1: Distributed Key Generation Protocol Based on Feldman VSS. This invention utilizes a Feldman Verifiable Secret Sharing (VSS)-based distributed key generation protocol. A threshold-based distributed key generation protocol replaces the traditional centralized key generator (PKG) architecture. Each key generation node independently selects a random polynomial and publishes its commitment value. The master key share is distributed and stored across the nodes using verifiable secret sharing, and only when the number of participating nodes reaches a threshold... Only then can the user's private key be generated collaboratively through Lagrange interpolation. This mechanism fundamentally eliminates the single point of failure risk caused by centralized storage of the master key, and the public nature of the commitment value allows any node to verify the correctness of the received share; at the same time, the protocol supports identity blind extraction, and D-PKG nodes cannot associate the private key with the real user's identity.

[0030] Innovation Point 2: A controllable audit mechanism with commitment binding and dual verification is implemented. This invention introduces an audit controller (AC) that runs locally on the data owner's end, which independently generates control factors in each round of auditing. and its elliptic curve commitment and will Embedded authorization certificate Simultaneously participating in the aggregation coefficient in the proof information The calculation. AC uses local secrets. right Hash commitments in Perform independent verification and obtain proof information through an independent channel. The verification algorithm is then run automatically. Even if the TPA is compromised or colludes with the CSP, it cannot detect the compromised TPA because the CSP cannot detect the compromised TPA. Attackers cannot forge authorization credentials that can be verified by the AC, which allows the AC to detect TPA-CSP collusion, thereby achieving user control over the entire audit process.

[0031] Innovation Point 3: Constant Computational Cost TPA Verification and Data Privacy Protection. During the proof verification phase, the computational cost required for TPA remains constant (3 bilinear pairing operations and 2 scalar multiplications), independent of the number of data blocks involved in the challenge. This significantly reduces the computational burden of TPA, making it suitable for environments with limited computational resources. Simultaneously, CSP introduces a randomized data blinding factor. By blinding aggregated data blocks, TPA aims to... Inverse solution It faces the challenge of discrete logarithm problems in elliptic curves, thus ensuring the privacy of data blocks.

[0032] I. System Model The system of this invention consists of five entities: a distributed public key generation center (D-PKG), a data owner (DO), a cloud service provider (CSP), a third-party auditor (TPA), and an audit controller (AC). The functions of each entity are as follows: (1) Distributed Public Key Generation Center (D-PKG): by It consists of several key generation nodes and employs Feldman Verifiable Secret Sharing (VSS) as its core. A threshold-based distributed key generation protocol manages the master key. Each node independently selects a random polynomial and publishes its commitment value for verification by other nodes; the system master key. The sum of the constant terms of the polynomials at all nodes, requiring no explicit reconstruction; arbitrary Each node can collaboratively generate a private key for the user through Lagrange interpolation, with any number of nodes less than [a certain value]. Each node is unable to obtain the master key information, thus eliminating the risk of single point of failure in centralized PKG.

[0033] (2) Audit Controller (AC): A lightweight component that runs locally by the data owner. It is responsible for managing the authorization and revocation of audit tasks, generating and distributing the basic random parameters of audit challenges, receiving audit conclusions from TPA and performing independent local reviews, thereby achieving autonomous control over the entire audit process.

[0034] (3) Data Owner (DO): Holds the identity ID and the corresponding private key, generates authentication tags for data blocks, uploads the data and tags to CSP, and authorizes TPA to perform audit tasks through AC.

[0035] (4) Cloud Service Provider (CSP): Responsible for storing data blocks and corresponding tags, generating proof information based on the challenge information of TPA, and ensuring the integrity of stored data.

[0036] (5) Third-party auditor (TPA): After being authorized by AC, the TPA periodically challenges the CSP and verifies the returned proof information, feeds back the audit results to the data owner, and is independently reviewed by AC.

[0037] System interaction process as follows Figure 1 As shown.

[0038] II. Algorithm Scheme The complete algorithm flow of this invention is as follows: Figure 4 As shown, the scheme consists of the following seven steps. For ease of verification, the following embodiments are all implemented using small prime numbers. Value calculus, in actual deployment It is a 256-bit prime number.

[0039] Step S1: System Initialization and Distributed Key Generation Input security parameters The Distributed Public Key Generation Center (D-PKG) is composed of... It consists of several key generation nodes, each operating through Feldman-based Verifiable Secret Sharing (VSS). Threshold distributed key generation protocol collaborative generation system public parameters. Each node ( Independent selection random polynomial Its constant term This serves as the node's secret contribution to the master key, and the commitment value is published. ( ).node via secure channel Send to node .node The correctness of the share is verified using the commitment value. After successful verification, each node calculates its own agreement with the master key. The joint share. The system master key is defined as... No explicit refactoring required, public parameters It can be publicly computed by all nodes. Based on the secret sharing property of Shamir, any Each node can be recovered or used via Lagrange interpolation. , and any less than The node cannot be obtained. Any information. Simultaneously select two pseudo-random functions. and and seven collision-resistant hash functions Public system parameters .

[0040] In this embodiment, the number of key generation nodes is determined. 3. Threshold Each of the three nodes independently selects a random polynomial once. , , and made a public commitment The system master key is the sum of the constant terms of each polynomial. (mold ), system public key At this point, any two nodes can collaboratively generate a private key for the user through Lagrange interpolation, and no single node can know this key. This eliminates any information, thereby removing the risk of a single point of failure in a centralized PKG.

[0041] Step S2: Key Extraction Data owners submit identity verification. , any Each D-PKG node collaboratively generates a private key for the data owner via a threshold key generation protocol (supporting identity blinding extraction: the data owner can select a random identity blinding factor). Calculate blind identity Submitted to D-PKG, node uses Collaborative generation of blind private keys, with local deblinding by the data owner, prevents nodes from associating the private key with their real identity. (Assume the participating...) The set of nodes is The key extraction process is as follows: (1) Each node in the process collaboratively generates joint random numbers. and Each participating node Independently select random numbers And publish the corresponding elliptic curve points. Calculate. , No single node can know and The complete value.

[0042] (2) Each node Calculate the local private key share using its own joint share and Lagrange coefficients, and send the local share to the data owner via a secure channel. The data owner then merges these shares to obtain the final private key share.

[0043]

[0044] (3) Output the complete private key Throughout the entire process, no single node could obtain the complete master key. Or the complete private key.

[0045] In this embodiment, the hospital submits its identity verification. Then, the two D-PKG nodes collaboratively generate joint random numbers. , ,correspond , ;Pick , The private key components are calculated using equations (1) and (2). , (Mod. 23), final private key If an identity needs to be concealed, the hospital can obtain a blinding factor. ,by Submitted, D-PKG node only However, it is impossible to associate the private key with the real identity.

[0046] Step S3: Audit Authorization The data owner generates independent control factors for this round of auditing through the locally running audit controller (AC). and unique round identifier Simultaneously calculate public commitments (in for Elliptic curve commitment (publicly released), calculation license.

[0047] in A collision-resistant hash function specifically designed for authorization credentials, used in the verification of proof information. They are not interchangeable, ensuring domain separation and security. The AC, via a securely authenticated channel, will... Send them to TPA and CSP respectively, and Securely stored in the local audit log, not publicly disclosed. The AC records the validity period of audit tasks and... The TPA audit authorization will be automatically revoked upon expiration. Each Only corresponds to one round of audit tasks (by...) (Unique binding), TPA cannot initiate challenges repeatedly without authorization within the validity period.

[0048] In this embodiment, the audit controller AC is the control factor for this round of auditing. Unique round identifier (Referring to "R2026-01"), calculate the public commitment. The authorization certificate is generated by equation (3). Control factor The TPA authorization is stored only in the AC local audit log and is not publicly disclosed. It will be automatically revoked upon expiration. Ensure that the voucher corresponds only to this round of audit to prevent TPA from launching repeated challenges within the validity period.

[0049] Step S4: Tag Generation and Upload Data owners use private keys Authentication tags are generated for each data block using scalar multiplication, eliminating the need for point-to-point hashing. The data file... Divided into fixed lengths Block, i.e. ,set up ( For each data block Random selection Calculate random auxiliary points and hash auxiliary value Then, the authentication label is calculated: Each data block requires a total of 3 scalar multiplication (SM) operations. Total of data blocks This SM operation requires no hashing to the point. After the tags are generated, Upload to CSP, Send it to AC for safekeeping.

[0050] In this embodiment, the file Divided into Data blocks, retrieve block values , , , Random numbers are taken from each block. , , , Three derived hashes , , Taking the first data block as an example, its authentication tag is calculated using equation (4). Each data block requires only 3 scalar multiplications, with no hash-to-point operations throughout.

[0051] Step S5: Challenge Generation TPA holds a valid authorization certificate. TPA launched an audit challenge against the CSP. The number of data blocks selected for the challenge was randomly chosen. Randomly select seeds Constructing challenge information Send the challenge to the CSP. The CSP will first verify the challenge upon receiving it. Legality: Check Is it in the authorization list? Check Is it within the validity period? Check. Check if it has not been used before (to prevent replay); if the verification is successful, proceed with the proof information generation step; otherwise, reject the challenge request and notify AC.

[0052] In this embodiment, TPA is determined by the number of challenge blocks. and random seeds , After pseudo-random function , The selected data blocks are determined to be blocks 1 and 3, with corresponding weights. , CSP verifies the challenge upon receiving it. (examine Within the authorization list, Not expired (If it is not used), the certificate information will be generated after the verification is passed.

[0053] Step S6: Generation and Verification of Proof Information CSP generates proof information based on the challenge information: it uses a pseudo-random function to calculate the index of the challenged data block. With weight ; Calculate aggregate tags and auxiliary aggregation value ;from Extract commitment Randomly select data blinding factor ,calculate , , .because The existence of TPA observations The aggregated values, after being randomly blinded, cannot be used to reverse-engineer the original data block by solving a system of linear equations. (Equivalent to the discrete logarithm problem). CSP will prove the information. Simultaneously send to TPA and AC (dual-send mechanism), and... Forward it to AC to ensure that AC can independently obtain complete audit data.

[0054] TPA received Verification is then performed. First, the identity aggregation public key is calculated:

[0055]

[0056] Then calculate the weighted aggregate hash value of the challenged data block (here). , This is a weighted aggregation of the challenged data blocks, distinct from the block-by-block calculation in Step 4. , ):

[0057]

[0058] Finally, verify the equation:

[0059] If the equation is true, output 1 and pass the conclusion along with the equation. Submit to AC, otherwise output 0, indicating that data integrity has been compromised and notify AC immediately.

[0060] In this embodiment, CSP calculates the actual aggregated data. (Modulus 23), take the data blinding factor. have to polymerization coefficient After blinding , aggregated , Proof information middle TPA calculation , and aggregate hash 、 Substitute this into the verification equation (5). Let... 、 Then the left side of the equation The product of the three terms on the right-hand side, after simplification, is also equal to... If both ends are strictly equal, output 1 to confirm data integrity. As a counterexample, if CSP loses the third block of data and uses a forged value... Impersonation is impossible because the authentication label cannot be forged (forgery requires possession of the private key). , ), aggregate tags Still The blinded aggregation value becomes Auxiliary aggregation value The coefficient becomes (Right now (degenerates to infinity), substituting into equation (5), the right side becomes If the left side is not equal to the right side, output 0, indicating that the tampering has been effectively detected.

[0061] Step S7: Self-verification After receiving the audit conclusions submitted by the TPA, the AC first performs three independent verifications, and then makes a conclusion determination and handles subsequent anomalies based on the verification results, as follows: (1) AC retrieves the current data from the local audit log. Corresponding control factors Recalculate , with TPA submission The hash value is compared with the data. If they do not match, it indicates that the TPA used forged credentials, and the AC immediately rejects the audit conclusion and reports the anomaly.

[0062] (2) AC compares the data sent directly by CSP. Submitted with TPA Are they consistent? If they are inconsistent, it indicates that the TPA has tampered with the proof information, the AC record is abnormal, and it was sent using CSP. Further verification will be conducted to ensure accuracy.

[0063] (3) AC sends directly using CSP and the challenge parameters for CSP forwarding Combined with local storage Execute the same verification equation (5) as Step 6 independently to obtain the independent verification result.

[0064] If the AC independent verification results are consistent with the TPA conclusions, the current audit is confirmed to be valid; if they are inconsistent, a re-audit is triggered and the TPA abnormal behavior is recorded.

[0065] In this embodiment, AC uses a local control factor. Recalculate and The hash commitment comparison is used to verify the legitimacy of the TPA identity, and to compare it with the direct CSP issuance. Transferred to TPA The consistency is used to detect whether the TPA has tampered with the proof information, and then obtained using an independent channel. and Combined with locally saved If the verification equation (5) is executed independently and the results of the triple verification are consistent, the audit is confirmed to be valid; if a certain TPA is continuously If this conclusion does not conform to AC's verification, AC will automatically revoke its audit authorization and select another auditor from the trust list. Since neither TPA nor CSP possesses control factors... Even if the two collude, they cannot forge a credential commitment that can pass AC verification. Therefore, collusion will inevitably be detected by the AC, enabling data owners to exercise autonomous control over the audit process.

[0066] III. Security Analysis This invention provides security proofs against three types of adversaries under the standard model: Theorem 1 (Resistance to Type I Adversaries): Under the Standard Model, if the CDH problem is difficult, then the solution of this invention is resistant to external attackers (Type I adversaries). The forgery attack is secure. This is proven by... The advantages of this reduction are applied to the solution of the CDH problem.

[0067] Theorem 2 (Resistance to Type II Adversaries): Under the Standard Model, if and If there are two collision-resistant hash functions, then the solution of this invention is effective against CSP (Type II adversaries). Deception attacks using this method are safe. In the missing data block If valid proof information is constructed under the given circumstances, then it can be deduced that... or The collision contradicts the collision resistance assumption.

[0068] Theorem 3 (Resistance to Type III Adversaries): Under the Standard Model, the present invention's solution is effective against TPAs ​​(Type III adversaries). Data extraction attacks using TPA are safe. TPA attempts to extract data from proof information. When extracting data blocks, two paths are deemed infeasible: Path 1: due to Protected by the difficulty of discrete logarithm ( TPA cannot solve this problem. Therefore, it is impossible to obtain from Recover the data block; Path 2: Solving the data block directly from the label also boils down to the discrete logarithm problem.

[0069] Theorem 4 (Resisting Type IV Adversaries, TPA-CSP Collusion): Under the Standard Model, if the discrete logarithm problem is difficult and... Since all are collision-resistant hash functions, the solution of this invention is effective against TPA-CSP collusion attackers (Class IV adversaries). It is safe. Even Simultaneous control of TPA and CSP, due to control factors Saved only locally on AC. Unable to know In cases where authorization certificates that can be verified by AC are forged (its hash commitment) ( , At the same time, it was bound With public commitment AC obtains proof information through an independent channel. and with local secrets Independently run the verification algorithm: on the one hand, verify middle On the one hand, it verifies the correctness of the commitment, and on the other hand, it checks the aggregation coefficient. The binding in Whether it is consistent with AC's public commitments, any objections... or Any alteration will break the verification equation and be detected by AC. Therefore, it can be determined according to... The difficulty of discrete logarithms and Its impact resistance will The advantage is reduced to negligible quantities, structurally enabling provable detection of TPA-CSP collusion attacks.

[0070] User controllability guarantee: control factor It is stored only in the local AC of the data owner, and is unknown to both TPA and CSP. The value; Elliptic curve commitments exposed by AC and bound to the grant certificate. hash commitment ( , and polymerization coefficient Within the calculations. Even if the TPA and CSP collude (a Category IV adversary), due to the inability to obtain... Attackers cannot forge authorization credentials that can be verified by the AC; the AC obtains proof information through an independent channel. and with local secrets The independently running verification algorithm can detect any collusive attacks, thus ensuring that data owners have complete and provable autonomy over the audit process.

[0071] The above description is merely a specific embodiment of the present invention to facilitate understanding by those skilled in the art. However, the embodiments of the present invention are not limited to those described above. Those skilled in the art can make improvements or modifications based on the present invention, and all improvements or modifications utilizing the concept of the present invention are considered to be within the scope of protection of the present invention.

Claims

1. A user-controllable cloud storage data security auditing method based on identity, characterized in that, Includes the following steps: Step 1 System Initialization: by The D-PKG, a distributed public key generation center composed of 10 nodes, uses Feldman-based verifiable secret sharing. Threshold Protocol Collaborative Generation System Public Parameters ; Step 2 Key Extraction: The data owner selects a random identity blinding factor. Calculate blind identity And submit it to D-PKG; at least A D-PKG node uses a blinded identity Collaborative generation of blind private keys; the data owner deblinds the blind key locally, generating a private key and public key corresponding to their identity; Step 3 Audit Authorization: The data owner generates control factors through the local audit controller (AC). and its public commitments And generate a unique round identifier. Third-party auditor identifier and the validity period of the authorization certificate AC will Send to TPA and CSP, Saved locally; Step 4 Tag Generation: The data owner uses the private key obtained in Step 2 to divide the data file into blocks and generate an authentication tag for each data block. Then, the data block and its tag are uploaded to the CSP, and the public information required for verification is sent to the AC for safekeeping. Step 5 Challenge Generation: TPA with Authorization Certificate An audit challenge is launched against the CSP, and the CSP verifies the results before handling the challenge. The legitimacy of the challenge is verified before proceeding to the next step; otherwise, the challenge is rejected and the AC is notified. Step 6: Proof Generation and Verification: CSP aggregates the labels of the challenged data blocks based on the challenge information and introduces a random data blinding factor to blind the aggregated data, generating proof information. And send it to TPA and AC; TPA will then process the information based on the publicly available identity information. Perform the verification and submit the audit findings to the AC; Step 7 Self-verification: AC uses locally stored control factors and supporting information The legitimacy of the TPA's identity, the consistency of the supporting information, and the audit results are verified independently in three ways. The verification results are compared with the TPA's conclusions. If they are consistent, the audit is confirmed to be valid. If they are inconsistent, a re-audit is triggered and the abnormal behavior of the TPA is recorded.

2. The identity-based user-controllable cloud storage data security auditing method according to claim 1, characterized in that, In the described Step 1, the public parameters include a bilinear pairing mapping , where is an -order additive group, is an -order multiplicative group, and , , three generators; the master secret key does not need to be explicitly reconstructed, the public parameters , the nodes of the D-PKG each independently select a -degree random polynomial and publish a commitment value, nodes mutually send shares and verify their correctness, and each node calculates its own joint share; the master public key is calculated publicly by accumulating the commitment values of all nodes; any or more nodes cooperatively generate a private key through Lagrange interpolation, and any fewer than nodes cannot obtain any information of .

3. The identity-based user-controllable cloud storage data security auditing method according to claim 1, characterized in that, In Step 2, at least Each D-PKG node collaborates to generate a private key pair for the data owner. public key pair ,satisfy ; ; in These are the two private key components for the data owner. Auxiliary public parameters generated for the public key, and For jointly generated random numbers, This is a hash function.

4. The identity-based user-controllable cloud storage data security auditing method according to claim 1, characterized in that, In Step 3, AC generates control factors in each round. and round identification Calculate public commitment ,in for Elliptic curve commitment, publicly released, authorization certificate ; For the validity period of the authorization, The filename of the data file. This is a collision-resistant hash function specifically designed for generating authorization credentials.

5. The identity-based user-controllable cloud storage data security auditing method according to claim 1, characterized in that, In Step 4, the private key is used. , data file Divided into Block, F= ,set up Each block is randomly selected Calculate the authentication label: ; in For the first Authentication tags for each data block; For the first The data values ​​of each data block; These are the two private key components for the data owner; For the group Two independent generators; For the first Random numbers corresponding to each data block This is a public commitment to the random number; For the first The filename of each data block is concatenated with its block index; , , These are the main auxiliary value, left auxiliary value, and right auxiliary value derived from three collision-resistant hash functions, respectively. For the first Each data block contains a complete tag pair; each block requires only 3 scalar multiplications, eliminating the need for hash-to-point operations.

6. The identity-based user-controllable cloud storage data security auditing method according to claim 1, characterized in that, In Step 5, the TPA holds the authorization certificate. Launch an audit challenge against the CSP, randomly selecting the number of data blocks to participate in the challenge. Select a random seed Challenge information The challenge is sent to the CSP, which then verifies it before processing it. Third-party auditor identification Is it in the license list? What is the license validity period? Whether it is within the validity period and the unique round identifier The system will check whether the device has been used before proceeding with the next steps.

7. The identity-based user-controllable cloud storage data security auditing method according to claim 1, characterized in that, In Step 6, the CSP receives Afterwards, from Extract K Calculate the index of the challenged data block. With weight , , The loop number of the challenged block. The first one to be challenged One data block in the data file The original block number in the data; Calculate aggregate tags With auxiliary aggregation value ,in , , , They are respectively the block number The corresponding authentication label, data block value, hash master and auxiliary value, and random number commitment; randomly select data blinding factor. λ ,calculate , , CSP will Simultaneously sent to TPA and AC, TPA calculates... ; ; And aggregate the hash values ​​of all challenged data blocks: left auxiliary value and right auxiliary value Verify the equation: ; If the equation is true, output 1; otherwise, output 0. Submit the audit conclusion to AC. Blinding commitments, To blind aggregate data, For binding and The polymerization coefficient, , Each is a corresponding private key , The public key for verification, .

8. The identity-based user-controllable cloud storage data security auditing method according to claim 7, characterized in that, In Step 7, the AC performs the following three verifications independently in sequence: Step 7.1, AC retrieves the current data from the local machine. corresponding Recalculate and The hash value in the system is compared. If they match, the TPA identity is deemed legitimate, and Step 7.2 is executed. If they do not match, the TPA identity is deemed illegitimate, the AC determines that this round of auditing is invalid, triggers a re-audit, and records the abnormal behavior of the TPA. Step 7.2, AC compares the data sent directly by CSP. Submitted with TPA If they are consistent, it is determined that TPA has not tampered with the proof information, and Step 7.3 continues; if they are inconsistent, it is determined that TPA has tampered with the proof information, AC determines that this round of audit is invalid, triggers a re-audit and records the abnormal behavior of TPA. Step 7.3, independently execute the same verification equation (5) as TPA to locally verify the audit results; If the AC independent verification results are consistent with the TPA conclusions, the audit is confirmed to be valid; otherwise, a re-audit is triggered and the TPA abnormal behavior is recorded.