Paperless file borrowing identity verification method based on privacy protection Turing test

Abstract

The invention discloses a paperless file borrowing identity verification method based on a privacy protection Turing test, and particularly relates to the field of node security identity verification, comprising the steps of extracting a path state vector based on a borrowing request identifier, and generating a disturbance feature vector; constructing a disturbance binding type Turing task according to the disturbance feature vector, and generating a Turing verification interaction packet; and obtaining a response behavior vector and terminal state sampling data based on the Turing verification interaction packet, and generating a joint response feature vector. A disturbance feature vector analyzed by a path state vector is embedded in a Turing verification task, and a user response behavior is dynamically associated with terminal environment disturbance, so that an interaction task structure with a node binding capability is constructed; therefore, the problem that the Turing response result lacks authenticity verification of the original execution node and is easy to be forwarded and counterfeited by the relay is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of node security identity verification, and more specifically, to a paperless archive borrowing identity verification method based on a privacy-preserving Turing test. Background Art

[0002] In existing paperless archival lending systems, to reduce manual operation costs and improve remote access efficiency, identity authentication processes are increasingly being replaced by Turing tests to determine whether access requests originate from real users rather than automated programs. However, traditional Turing test authentication mechanisms typically assume that the request source and the response execution terminal are located on the same trusted node, and assume that intermediate nodes in the entire chain will not tamper with the interaction process.

[0003] In actual deployment environments, remote access paths often pass through multiple uncontrollable relay nodes, such as proxy servers, enterprise gateways, VPN forwarding points, or edge computing nodes. These nodes may hijack Turing test interaction content, cache authentication responses, or even replay verified Turing results of others without being detected, thereby fabricating the illusion of successful identity authentication.

[0004] In scenarios where a highly simulated attacker uses a forwarding or simulated terminal to perform a Turing test response, the traditional Turing mechanism cannot effectively verify the spatial execution location of the behavior and the terminal binding, thus posing a challenge to node security.

[0005] Therefore, in remote authentication with privacy protection goals, relying solely on the response data completed by the Turing test cannot ensure that the behavior originates from a real, non-hijacked terminal node. The core problem is that the Turing test results lack verifiability of the authenticity of the original execution node, resulting in the breach of the node security boundary, and then making the authentication path itself an attack channel for identity forgery. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a paperless archive borrowing identity verification method based on the privacy-preserving Turing test. By embedding the disturbance feature vector parsed by the path state vector in the Turing verification task and dynamically associating the user response behavior with the terminal environment disturbance, an interactive task structure with node binding capability is constructed, thereby solving the problem that the Turing response result lacks authenticity verification of the original execution node and is easily forged by relay forwarding.

[0007] To achieve the above-mentioned purpose, the present invention provides the following technical solution: a paperless archive borrowing identity verification method based on the privacy-preserving Turing test, comprising:

[0008] S1: Extract the path state vector based on the borrowing request identifier and generate a perturbation feature vector; construct a perturbation binding Turing task based on the perturbation feature vector and generate a Turing verification interaction package;

[0009] S2: Based on the Turing verification interaction package, the response behavior vector and terminal state sampling data are obtained to generate a joint response feature vector. The joint response feature vector is used to make a collaborative response judgment, output a consistent behavior vector, and construct a behavior execution trajectory graph to solve the disturbance behavior consistency feature index.

[0010] S3: Constructs a node perturbation vector sequence based on the behavior execution trajectory graph, compares it with the terminal perturbation registry, and outputs a perturbation matching confidence score. It confirms behavior consistency through synchronization verification, generates a trusted execution tag, and converts it into a node fingerprint hash summary.

[0011] S4: Generate a node-bound authentication token based on the node fingerprint hash summary and the borrowing request identifier, and establish an access permission mapping table; determine the legitimacy of the token through node consistency verification, and output the archive access permission or mark access exception events.

[0012] In a preferred embodiment, S1 further includes: extracting the user identity field and the access target field from the data generated by the archive borrowing request to generate a borrowing request identifier; extracting the request link path identifier and the execution timestamp from the borrowing request identifier to form a path state vector;

[0013] By parsing the path state vector, node disturbance related data items are extracted, and a disturbance feature vector is generated according to the construction requirements. The node disturbance related data items include cache response delay, voltage disturbance change and graphics frame rendering offset value;

[0014] By analyzing the perturbation response positions and noise distribution density with periodic consistency in the perturbation feature vector, a perturbation credible window is constructed. The perturbation indicators that meet the preset threshold in the perturbation credible window are embedded into the graphical test structure to form a perturbation-bound Turing task.

[0015] A graphical task structure is constructed through a perturbation-binding Turing task, wherein the graphical task structure includes a pseudo-behavior induction area, a true judgment area, and a path random perturbation area; the graphical task structure is combined with the perturbation-binding Turing task to generate a one-time interaction sequence, wherein the interaction sequence constitutes a Turing verification interaction package;

[0016] If the perturbation feature vector in the path state vector is lower than the preset execution confidence threshold, the construction process of the Turing verification interaction package is terminated, and the process returns to the path state vector extraction and is re-executed.

[0017] In a preferred embodiment, S2 further includes: guiding the user to complete pseudo-behavior selection and target judgment behavior through the deployed Turing verification interaction package, and outputting a response behavior vector; triggering periodic sampling of the current state of the terminal through Turing interaction, the data of the current state of the sampled terminal including image loading delay, thermal noise index and execution offset delay;

[0018] The response behavior vector and the terminal state sampling data are combined to form a behavior-environment mapping pair, and then converted into a joint response feature vector; it is identified whether the frequency of the behavior path node in the joint response feature vector and the synchronization rate of the terminal noise reach the preset collaborative response threshold; if the preset collaborative response threshold is reached, the joint response feature vector is marked as a consistent behavior vector; otherwise, it is recorded as an abnormal response behavior;

[0019] A behavior execution trajectory graph is established based on the consistent behavior vector. The nodes in the behavior execution trajectory graph include operation type identifiers and perturbation synchronization markers. The behavior labels embedded in the Turing test are matched with the behavior execution trajectory graph to solve the perturbation behavior consistency feature index.

[0020] In a preferred embodiment, S3 further includes: extracting disturbance parameters associated with each behavior node in the behavior execution trajectory graph, constructing a corresponding node disturbance vector sequence, performing difference regression comparison on the node disturbance vector sequence and a preset terminal disturbance registry, and outputting a disturbance matching confidence score;

[0021] Determine whether the perturbation matching confidence score exceeds the preset matching threshold. If so, construct a node execution confirmation result identifier; perform synchronization verification through the perturbation behavior consistency feature index and the sequence of perturbation indicators embedded in the Turing verification interaction package; if the synchronization verification is successful, merge the node execution confirmation result identifier and the behavior consistency label to form a trusted execution tag; if the perturbation matching confidence score does not meet the standard or the synchronization verification fails, output a pseudo execution identifier and terminate the borrowing process;

[0022] The trusted execution tag is converted into a node fingerprint hash digest, which serves as the binding basis for subsequent authentication token generation.

[0023] In a preferred embodiment, S4 further includes: forming a joint token input pair from the node fingerprint hash digest and the borrowing request identifier, performing a hash mixing operation on the joint token input pair to generate a node-bound authentication token, and mapping the node-bound authentication token with the user authority level, the request data confidentiality level, and the access time parameter to generate an access permission mapping table;

[0024] Determine whether the node fingerprint hash digest generated by the perturbation feature vector re-extracted by the current node is consistent with the digest recorded in the node-bound authentication token; if they are consistent, establish a data channel based on the access permission mapping table and output the corresponding file access permission; otherwise, destroy the node-bound authentication token and update the token status in the access permission mapping table to invalid;

[0025] If a subsequent node submits an expired or hash-inconsistent node binding authentication token again, the behavior will be marked as a pseudo-node access event and recorded in the access exception record.

[0026] In a preferred embodiment, S1 further includes:

[0027] Path state vector construction:

[0028]

[0029] Where V path is the path state vector; is the user identity field; To access the target field; μ req Identifies the borrowing request; Indicates extracting the request link path identifier x from the borrowing request identifier route and execution timestamp χ time ; Γ(·) is the borrowing request aggregation function;

[0030] Generation of perturbation eigenvectors:

[0031]

[0032] Where V dist is the perturbation eigenvector; ρ mem (t) represents the disturbance signal of cache response delay at time t; ρ volt (t) represents the voltage disturbance change response value at time t; ρ frame (t) represents the graphics frame rendering offset response value at time t; T is the total length of the sampling time;

[0033] The perturbation credible window extraction formula is expressed as:

[0034]

[0035] in is the perturbation credible window; ρ(t) represents the perturbation response strength at time t; ρ(s) represents the perturbation response strength at time s; is the second-order derivative of the disturbance signal; σ curve (t) is the perturbation curvature control function; ζ dense(t) is the noise distribution density function of the disturbance response; is the indicator function; ε is the preset disturbance stability threshold; Δt is the local disturbance observation window;

[0036] The spatial mapping of the disturbance index in the graph task structure is expressed as:

[0037]

[0038] in Pixel-space representation of graphics task structures; is the kth graphic area, which includes the pseudo-behavior induction area, the real judgment area and the path random perturbation area; δx k ,δy k is the coordinate perturbation offset generated when the perturbation point in the perturbation credible window is mapped to the graphic area; n is the total number of graphic interaction areas;

[0039]

[0040] in is a Turing-verified interaction package; ⊕ represents the synthetic operation of embedding the perturbation trusted window into the graph task structure; Λ exec (V dist ) is the execution confidence function; θ exec is the preset execution confidence threshold.

[0041] In a preferred embodiment, F is defined in S2 joint (t) is the joint response feature vector at time t;

[0042] F joint (t) = [α resp (t),β load (t),β noise (t),β devi (t)]

[0043] where α resp (t) is the response behavior vector value; β load (t) is the image loading delay; β noise (t) is the thermal noise index; β devi (t) is the execution offset delay;

[0044]

[0045] where Θ sync is a measure of the intensity of the collaborative response; The rate of change of user behavior rhythm; is the proportion of behavior offset to loading delay; cos(·) is the control function; is an exponential suppression function; T is the total length of the sampling time;

[0046]

[0047] Among them F valid is the consistent behavior vector; τ coh is a preset collaborative response threshold;

[0048]

[0049] in is the behavior execution trajectory diagram; θ i Identifies the type of operation i; is the ith disturbance synchronization mark; Op i Extract function for operation type; Corr i is the synchronization state evaluation function; N is the total number of trajectory behavior nodes;

[0050]

[0051] where Ξ match is the consistency feature index of the perturbation behavior; θ j is the jth behavior label in the behavior trajectory graph; τ j is the sequence of behavior labels embedded in the Turing task interaction package; is the jth disturbance synchronization mark; Indicates that the disturbance synchronization state is established.

[0052] In a preferred embodiment, in S3, a node perturbation vector sequence expression is constructed:

[0053]

[0054] Among them U trace is the node perturbation vector sequence; u j is the perturbation vector corresponding to the jth consistent behavior node; Ψ(·) is the perturbation parameter extraction function;

[0055] Regression comparison of the terminal disturbance registry based on component difference fitting:

[0056]

[0057] where μ match Assign confidence scores to perturbation matches; is the pre-stored disturbance vector of the corresponding behavior type in the terminal disturbance registration database; ||·|| is the vector bi-norm; ε is the disturbance noise robustness factor;

[0058]

[0059] where χexec Execute the confirmation result mark for the node; match is the preset disturbance matching threshold;

[0060]

[0061] where Ω sync is the consistency of disturbance synchronization; Embed the disturbance indicator sequence in the perturbation response window of the jth region in the Turing verification interaction package; is the indicator function;

[0062]

[0063] where χ trust Trusted execution mark; sync is the preset disturbance synchronization consistency threshold;

[0064]

[0065] in is the node fingerprint hash summary; Hash(·) is an irreversible hash function; || is the concatenation operator.

[0066] In a preferred embodiment, in S4, the node disturbance behavior entropy is modeled and S is defined as node is the perturbation behavior path entropy;

[0067]

[0068] where p j represents the perturbation probability distribution at the jth node after the perturbation intensity is normalized; u k is the kth perturbation vector in the set of perturbation behavior consistency feature indices;

[0069]

[0070] in Bind authentication token to the node; Hash mix (·) is the hash mixing function;

[0071] Generate the ownership mapping tensor in S4:

[0072]

[0073] in is the value of the cell with index (i, j, k) in the access control tensor; is the i-th user authority level; is the jth data classification level; is the kth type access time window parameter; Δt kis the delay between the current system time and the generation time of the node-bound authentication token in the k-th access window;

[0074] Calculate the consistency difference of the current node disturbance spectrum:

[0075]

[0076] where φ j is the disturbance characteristic spectrum; D diff The difference between the current disturbance spectrum of the execution node and the disturbance spectrum of the authentication token bound to the node; is the spectrum feature of the current node in the jth disturbance dimension; is the spectrum feature recorded in the jth perturbation dimension when the node binding authentication token is generated; |·| is the absolute value;

[0077]

[0078] where χ grant is the final access decision result; τ D is the upper threshold of the disturbance spectrum difference; τ S is the lower threshold of the perturbation behavior path entropy; Represents a permission tensor consisting of permission level, data confidentiality level, and time window.

[0079] Technical effects and advantages of the present invention:

[0080] 1. By embedding the perturbation feature vector into the Turing verification task structure, the verification behavior is bound to the terminal perturbation environment, thereby ensuring that the Turing response result has node uniqueness and effectively preventing identity forgery under relay node hijacking;

[0081] 2. By integrating user behavior rhythm and terminal status sampling indicators through a joint response feature vector, we can achieve collaborative identification of behavior paths and physical environments, thereby improving the recognition accuracy of real-world behavioral interactions.

[0082] 3. By extracting the perturbation consistency index that meets the Turing task in the behavior execution trajectory graph, a precise mapping between the perturbation vector sequence and the preset registration information is established, thereby enhancing the authenticity verification between the execution behavior and the terminal state;

[0083] 4. By using perturbation behavior path entropy and spectrum consistency as dual filtering conditions for authentication token generation, the non-reusability of the perturbation structure is enhanced, effectively preventing pseudo-nodes from simulating perturbation patterns to obtain access tokens;

[0084] 5. By constructing a three-dimensional ownership mapping tensor to dynamically bind authentication tokens and permission parameters, the timeliness and context-aware control of access authorization conditions can be achieved, thereby improving the security adaptability and abuse prevention capabilities of access behaviors. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 It is a flowchart of the framework of the method steps of the present invention. DETAILED DESCRIPTION

[0086] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0087] Refer to the instruction manual Figure 1 A paperless archive borrowing identity verification method based on a privacy-preserving Turing test according to an embodiment of the present invention includes:

[0088] S1: Extract the path state vector based on the borrowing request identifier and generate a perturbation feature vector; construct a perturbation binding Turing task based on the perturbation feature vector and generate a Turing verification interaction package;

[0089] S2: Based on the Turing verification interaction package, the response behavior vector and terminal state sampling data are obtained to generate a joint response feature vector. The joint response feature vector is used to make a collaborative response judgment, output a consistent behavior vector, and construct a behavior execution trajectory graph to solve the disturbance behavior consistency feature index.

[0090] S3: Constructs a node perturbation vector sequence based on the behavior execution trajectory graph, compares it with the terminal perturbation registry, and outputs a perturbation matching confidence score. Behavior consistency is confirmed through synchronization verification, and a trusted execution tag is generated and converted into a node fingerprint hash summary.

[0091] S4: Generate a node-bound authentication token based on the node fingerprint hash summary and the borrowing request identifier, and establish an access permission mapping table; determine the legitimacy of the token through node consistency verification, output the archive access permission or mark access exception events.

[0092] S1 also includes: extracting the user identity field and the access target field from the data generated by the archive borrowing request to generate a borrowing request identifier; extracting the request link path identifier and the execution timestamp from the borrowing request identifier to form a path state vector;

[0093] By parsing the path state vector, node disturbance related data items are extracted, and a disturbance feature vector is generated according to the construction requirements. The node disturbance related data items include cache response delay, voltage disturbance change and graphics frame rendering offset value;

[0094] By analyzing the perturbation response positions and noise distribution density with periodic consistency in the perturbation feature vector, a perturbation credible window is constructed. The perturbation indicators that meet the preset threshold in the perturbation credible window are embedded into the graphical test structure to form a perturbation-bound Turing task.

[0095] A graphical task structure is constructed through a perturbation-binding Turing task, wherein the graphical task structure includes a pseudo-behavior induction area, a true judgment area, and a path random perturbation area; the graphical task structure is combined with the perturbation-binding Turing task to generate a one-time interaction sequence, wherein the interaction sequence constitutes a Turing verification interaction package;

[0096] If the perturbation feature vector in the path state vector is lower than the preset execution confidence threshold, the construction process of the Turing verification interaction package is terminated, and the process returns to the path state vector extraction and is re-executed.

[0097] S2 also includes: guiding the user to complete pseudo-behavior selection and target judgment behavior through the deployed Turing verification interaction package, and outputting a response behavior vector; triggering periodic sampling of the current state of the terminal through Turing interaction, and the data of the current state of the sampled terminal includes image loading delay, thermal noise index and execution offset delay;

[0098] The response behavior vector and the terminal state sampling data are combined to form a behavior-environment mapping pair, and then converted into a joint response feature vector; it is identified whether the frequency of the behavior path node in the joint response feature vector and the synchronization rate of the terminal noise reach the preset collaborative response threshold; if the preset collaborative response threshold is reached, the joint response feature vector is marked as a consistent behavior vector; otherwise, it is recorded as an abnormal response behavior;

[0099] A behavior execution trajectory graph is established based on the consistent behavior vector. The nodes in the behavior execution trajectory graph include operation type identifiers and perturbation synchronization markers. The behavior labels embedded in the Turing test are matched with the behavior execution trajectory graph to solve the perturbation behavior consistency feature index.

[0100] S3 also includes: extracting disturbance parameters associated with each behavior node in the behavior execution trajectory graph, constructing a corresponding node disturbance vector sequence, performing difference regression comparison on the node disturbance vector sequence and a preset terminal disturbance registry, and outputting a disturbance matching confidence score;

[0101] Determine whether the perturbation matching confidence score exceeds the preset matching threshold. If so, construct a node execution confirmation result identifier; perform synchronization verification through the perturbation behavior consistency feature index and the sequence of perturbation indicators embedded in the Turing verification interaction package; if the synchronization verification is successful, merge the node execution confirmation result identifier and the behavior consistency label to form a trusted execution tag; if the perturbation matching confidence score does not meet the standard or the synchronization verification fails, output a pseudo execution identifier and terminate the borrowing process;

[0102] The trusted execution tag is converted into a node fingerprint hash digest, which serves as the binding basis for subsequent authentication token generation.

[0103] S4 further includes: forming a joint token input pair from the node fingerprint hash digest and the borrowing request identifier, performing a hash mixing operation on the joint token input pair to generate a node-bound authentication token, and mapping the node-bound authentication token with the user permission level, the request data confidentiality level, and the access time parameter to generate an access permission mapping table;

[0104] Determine whether the node fingerprint hash digest generated by the perturbation feature vector re-extracted by the current node is consistent with the digest recorded in the node-bound authentication token; if they are consistent, establish a data channel based on the access permission mapping table and output the corresponding file access permission; otherwise, destroy the node-bound authentication token and update the token status in the access permission mapping table to invalid;

[0105] If a subsequent node submits an expired or hash-inconsistent node binding authentication token again, the behavior will be marked as a pseudo-node access event and recorded in the access exception record.

[0106] It should be noted that in the formula structure involved in this solution, dimensionless terms can serve as proportionality or structural adjustment factors. When combined with quantities with units, they only play a numerical scaling role and do not introduce new physical dimensions. Therefore, they will not change or confuse the overall unit system of expression. This combination of "dimensionless terms and units" can be understood as a composite structural expression commonly used in mathematical and physical modeling, conforming to the principle of dimensional consistency and having a clear physical interpretation basis.

[0107] Secondly, in the formula structure of this scheme, if multiple variables with different physical units are involved, including but not limited to time, mass or energy variables, their joint appearance is to express the collaborative modeling relationship of multiple physical mechanisms. Each variable can be formed into a unified structure through function mapping, ratio combination or normalization adjustment. The units and meanings are clear, and the overall expression conforms to the principle of dimensional consistency and the common formula of engineering modeling.

[0108] In this solution, any design constants, weights, adjustment factors, threshold parameters, and proportional coefficients are adjustable control parameters for different application environments. Their values ​​depend on the target device configuration, data input characteristics, and performance optimization goals. During the implementation phase, they are set within a reasonable range through model verification, performance constraints, or engineering calibration. Although these parameters do not have unique preset values, they have clear adjustment logic and calculation paths, and are part of the deterministic setting process in engineering implementation. The purpose of such setting is to ensure that the solution is both universally adaptable, reproducible, and operable, without affecting its technical clarity and feasibility.

[0109] Also included in S1:

[0110] Path state vector construction:

[0111]

[0112]

[0113] Where V path is the path state vector; is the user identity field; To access the target field; μ req Identifies the borrowing request; Indicates extracting the request link path identifier x from the borrowing request identifier route and execution timestamp χ time Γ(·) is the borrowing request aggregation function, which is used to form a unique identification key from the identity and the target;

[0114] Generation of perturbation eigenvectors:

[0115]

[0116] Where V dist is the perturbation eigenvector; ρ mem (t) represents the disturbance signal of cache response delay at time t; ρ volt (t) represents the voltage disturbance change response value at time t; ρ frame (t) represents the graphics frame rendering offset response value at time t; T is the total length of sampling time, which is used to define the disturbance observation interval;

[0117] The perturbation credible window extraction formula is expressed as:

[0118]

[0119] in is the perturbation credible window, which consists of time indices that satisfy the stable perturbation curvature and response density conditions; ρ(t) represents the perturbation response intensity at time t, and the perturbation response intensity can be ρ mem (t),ρ volt (t),ρ frame (t) is the dominant disturbance signal; ρ(s) represents the disturbance response intensity at time s; is the second-order derivative of the disturbance signal, which is used to measure the curvature or acceleration of the disturbance response. represents the disturbance response position; σ curve(t) is the perturbation curvature control function. In practical applications, the perturbation curvature control function represents the ability of the system to accept the curvature change of the perturbation; ζ dense (t) is the noise distribution density function of the disturbance response. The noise distribution density function of the disturbance response represents the cumulative duration that the disturbance signal and its surrounding disturbance values ​​maintain similar amplitudes within the current time window, and is used to measure the degree of local stable aggregation of the disturbance sequence. is the indicator function, in which if the condition in the brackets is met, it is 1, otherwise it is 0; ε is the preset disturbance stability threshold; Δt is the local disturbance observation window;

[0120] The spatial mapping of the disturbance index in the graph task structure is expressed as:

[0121]

[0122] in It is a pixel space representation of the graphics task structure, and the pixel space representation of the graphics task structure is used to form the final user interaction layer; is the kth graphic area, which includes the pseudo-behavior induction area, the real judgment area and the path random perturbation area; δx k ,δy k is the coordinate perturbation offset generated when the perturbation point in the perturbation credible window is mapped to the graphic area; n is the total number of graphic interaction areas;

[0123]

[0124] in is a Turing-verified interaction package; ⊕ represents the synthetic operation of embedding the perturbation trusted window into the graph task structure; Λ exec (V dist ) is the execution confidence function, which is used to determine whether the current disturbance feature vector is valid; θ exec is the preset execution confidence threshold, which represents the minimum acceptable disturbance structure strength;

[0125] If the confidence requirement is not met, the interaction package construction process is terminated and returns to the path state vector stage for re-extraction.

[0126] Define F in S2 joint (t) is the joint response feature vector at time t;

[0127] F joint (t) = [α resp (t),β load (t),β noise (t),β devi (t)]

[0128] where α resp(t) is the response behavior vector value, which includes the behavior judgment response value at the interaction time point t in the Turing verification interaction package, such as the click path or judgment option; β load (t) is the image loading delay, which indicates the time required for the terminal to fully render the graphics content during the task loading process; β noise (t) is the thermal noise index, which indicates the degree of thermal stability disturbance during CPU operation; β devi (t) is the execution offset delay, which represents the offset time of the behavioral input relative to the preset Turing task rhythm;

[0129]

[0130] where Θ sync The synergistic response strength metric is used to measure the degree of temporal dynamic coupling between behavior and terminal state; The user behavior rhythm change rate is used to measure the consistency of behavior; is the ratio of behavior offset to loading delay, which reflects the degree of synchronization between the behavior and the terminal loading state. cos(·) is the control function used to suppress synchronization failure caused by excessive offset. is an exponential suppression function, which is used to penalize the response mismatch caused by excessive terminal thermal noise; T is the total length of the sampling time;

[0131]

[0132] Among them F valid is the consistent behavior vector, which indicates the effective behavior state that the joint response characteristics meet the synchronization criterion; τ coh is the preset collaborative response threshold; if the judgment fails, an empty set is output and recorded as an abnormal response;

[0133]

[0134] in is the behavior execution trajectory graph, which consists of a set of nodes; θ i is the i-th operation type identifier, which indicates the specific behavior category, such as click, drag, judgment, etc. is the i-th disturbance synchronization mark, which indicates whether the marking behavior is aligned with the node disturbance synchronization area; Op i Extraction function for the operation type, which comes from the consistent behavior vector F valid The interactive behavior term in Corri is a synchronization state evaluation function, which is used to determine whether the response behavior falls within the corresponding section of the disturbance trust window; Where N is the total number of trajectory behavior nodes;

[0135]

[0136] where Ξ match is the perturbation behavior consistency feature index, which represents the set of node indices in the behavior trajectory graph that are consistent with the Turing task behavior label and the perturbation synchronization is established; θ j is the jth behavior label in the behavior trajectory graph; τ j is the sequence of behavior labels embedded in the Turing task interaction package; is the jth disturbance synchronization mark; Indicates that the disturbance synchronization state is established; the set of disturbance behavior consistency feature indexes is used for subsequent trusted execution path judgment or behavior authenticity confirmation.

[0137] In S3, construct the node perturbation vector sequence expression:

[0138]

[0139] Among them U trace is the node perturbation vector sequence, which represents the set of perturbation signal vectors extracted from the behavior execution trajectory graph; j∈Ξ match Indicates that only the consistent feature index of the perturbation behavior is extracted; u j is the disturbance vector corresponding to the jth consistent behavior node; Ψ(·) is the disturbance parameter extraction function, which is used to extract disturbance quantities such as cache response and image jitter from task execution according to the operation type;

[0140] Regression comparison of the terminal disturbance registry based on component difference fitting:

[0141]

[0142] where μ match is the perturbation matching confidence score, which is used to reflect the consistency between the node perturbation feature and the registered terminal; is the pre-stored disturbance vector corresponding to the behavior type in the terminal disturbance registry; ||·|| is the vector bi-norm; ε is the disturbance noise robustness factor, which is used to prevent division by zero in the above formula and in practical applications; the above formula performs nonlinear normalized comparison based on the similarity of disturbance features, combining directional differences and energy differences;

[0143]

[0144] where χ exec λ is the result mark of the node execution confirmation, where 1 is confirmation and 0 is denial; match is the preset disturbance matching threshold; if the disturbance matching confidence is higher than the threshold, it is considered to be executed by the original terminal and is assigned a value of 1; otherwise, it is considered to be suspicious execution and is assigned a value of 0;

[0145]

[0146] where Ω sync is the perturbation synchronization consistency, and the value range of the perturbation synchronization consistency is [0, 1]; Embed the disturbance indicator sequence in the perturbation response window of the jth region in the Turing verification interaction package; is the indicator function, in Ω sync The indicator function in the formula means that if u j If it is located in the embedded perturbation indicator window in the chart area to which it belongs, it returns 1, otherwise it returns 0;

[0147]

[0148] where χ trust is the trusted execution flag, a trusted execution flag of 1 indicates trusted, and a trusted execution flag of 0 indicates pseudo execution; κ sync A preset perturbation synchronization consistency threshold is used; a trusted execution tag is generated when both the perturbation configuration confidence score and the perturbation synchronization conditions are met;

[0149]

[0150] in is the node fingerprint hash summary; Hash(·) is an irreversible hash function, including SHA256; || is a concatenation operator, which is used to combine the perturbation vector sequence, behavioral feature index, and trusted tag into a unique input; the node fingerprint hash summary serves as the node signature bound to the authentication token to ensure that it cannot be forged.

[0151] In S4, the node perturbation behavior entropy is modeled and S is defined node is the perturbation behavior path entropy, which is used to characterize the discreteness and uniqueness of the perturbation feature vector distribution;

[0152]

[0153] where p j represents the perturbation probability distribution at the jth node after the perturbation intensity is normalized; u k is the kth perturbation vector in the set of perturbation behavior consistency feature indexes; in addition, log2(p j) represents the self-information of the perturbation intensity at the jth node; by introducing the information entropy of the behavioral perturbation sequence for token generation discrimination, it makes it difficult to reuse or forge highly discrete perturbation paths;

[0154]

[0155] in Bind authentication token to the node; Hash mix (·) is a hash mixing function, which means that multiple input fields are mixed and encoded and hashed into a unique output;

[0156] Generate the ownership mapping tensor in S4:

[0157]

[0158] in is the value of the cell indexed by (i, j, k) in the access control tensor, which represents the access feasibility score of the user under different combinations of permissions, confidentiality levels, and time; The i-th user authority level includes but is not limited to administrator, visitor, etc. The data classification level is Class j, which includes but is not limited to ordinary, confidential, top secret, etc. is the kth access time window parameter, which is used to control the access validity period; Δt k is the delay between the current system time and the generation time of the node-bound authentication token under the k-th access window; In addition, in the above formula An exponential decay factor representing the effectiveness of access over time;

[0159] Calculate the consistency difference of the current node disturbance spectrum:

[0160]

[0161] where φ j is the disturbance characteristic spectrum, which uses Fourier transform FFT to obtain the node disturbance periodicity information; D diff The difference between the current disturbance spectrum of the execution node and the disturbance spectrum of the authentication token bound to the node; is the spectrum feature of the current node in the jth disturbance dimension; The spectrum characteristics recorded in the jth perturbation dimension when the node binding authentication token is generated; |·| is the absolute value, which indicates the spectrum difference;

[0162]

[0163] where χ grantis the final access decision result. A final access decision result of 1 indicates that access is allowed, and a final access decision result of 0 indicates that access is denied; τ D is the upper threshold of the disturbance spectrum difference; τ S is the lower threshold of the perturbation behavior path entropy; Represents a permission tensor consisting of permission level, data confidentiality level, and time window. If the above formula satisfies the spectrum similarity, perturbation entropy complexity, and corresponding permission authorization, the access is considered valid.

[0164] It should be noted that the core purpose of this solution is to construct a dynamic authentication method that does not rely on external identity carriers and can deeply integrate interactive behavior and environmental perturbation responses in the context of the failure of traditional identity authentication methods and highly untrustworthy terminal environments. By coupling the Turing test mechanism with the perturbation behavior path, a dynamic token generation logic with difficult-to-forge behavior and node uniqueness is formed, fundamentally ensuring that the initiator of the borrowing request is credible, the path is authentic, and the response is complete.

[0165] The solution uses the borrowing request identifier as the only entry point. By extracting the request link path identifier and execution timestamp, it constructs a path state vector. It then further analyzes the node perturbation-related data items in the path state vector to generate a perturbation feature vector. The perturbation feature vector not only includes basic parameters such as cache response latency, voltage perturbation variation, and graphics frame rendering offset, but also extracts a perturbation trust window by analyzing its periodic consistency and noise distribution density. This perturbation trust window is the basis for constructing perturbation-bound Turing tasks, ensuring that the behavioral judgments contained in the Turing tasks are unique to the terminal-side environmental response.

[0166] On this basis, the graphical task structure is constructed with three parts: a pseudo-behavior induction area, a true judgment area, and a path random perturbation area. By embedding perturbation indicators that meet the preset threshold in the perturbation trust window and synthesizing a one-time interaction sequence, a Turing-verified interaction package is formed. This interaction package induces and guides user behavior and periodically samples terminal status data during execution, including image loading delay, thermal noise index, and execution offset delay. Finally, the response behavior vector is combined with the terminal status sampling data to generate a joint response feature vector. Based on the behavior path node frequency and the terminal noise synchronization rate, it is determined whether the collaborative response threshold has been reached.

[0167] If the response meets the threshold requirements, a consistent behavior vector is constructed and further formed into a behavior execution trajectory graph. Each node in the graph contains an operation type identifier and a perturbation synchronization marker. The behavior label sequence pre-embedded in the Turing test is matched with the trajectory graph to solve the perturbation behavior consistency feature index, which is used to locate which interactive behaviors meet the dual conditions of preset label matching and environmental perturbation synchronization. This feature index is then used as a key reference point to extract the corresponding perturbation parameters and construct a node perturbation vector sequence.

[0168] The perturbation vector sequence is then compared with the terminal perturbation registry through differential regression, and a perturbation matching confidence score is output to determine whether the behavior is indeed initiated by the bound terminal. Synchronicity verification is performed by combining the perturbation behavior consistency feature index with the perturbation indicator sequence embedded in the Turing interaction package. Only when the perturbation matching score reaches the preset threshold and the synchronization verification is established is a trusted execution tag generated. This tag is then quantized using the perturbation behavior path entropy to form the final node fingerprint hash summary, which serves as the unique binding benchmark for generating the authentication token.

[0169] During authentication token generation, the node fingerprint hash digest is combined with the borrowing request identifier to form a joint token input, which is then processed through a hash mixing function to form a node-bound authentication token. To ensure dynamic access control and multi-dimensional authorization, the authentication token must also be mapped to a three-dimensional ownership mapping tensor constructed from user permission level, request data confidentiality level, and access time parameters. Access control no longer relies on a static table structure, but instead relies on a decaying authorization structure formed by the three elements of permission, content, and time, ensuring that the token authorization status is automatically updated over time and with the request context.

[0170] During the actual access phase, the current node's perturbation feature vector and behavior consistency index are re-extracted, and the spectrum structure of the current perturbation behavior is generated. The perturbation spectrum difference is calculated by performing a component-by-component difference analysis on this spectrum feature and the spectrum recorded when the binding authentication token was generated. The final determination of whether access is authorized no longer depends on a single comparison value, but is based on the combined results of three conditions: whether the perturbation spectrum difference is below an upper threshold, whether the perturbation path entropy is above a lower threshold, and whether the authentication token is still in a valid position in the mapping tensor.

[0171] If any of the above three conditions is not met, the access request will be rejected, and the system will mark the current authentication token as invalid or abnormally used, and write it into the access exception record; this structure realizes end-to-end, non-forgeable, non-migratable, and non-relayable real identity verification of paperless archive borrowing behavior through a combination of multi-stage, multi-dimensional, and multi-source data security verification logic; in this way, even if the terminal device encounters counterfeit, replay, or behavior forwarding attacks, as long as there is a deviation in any layer of the disturbance structure, path behavior, spectrum morphology, or access mapping, authorization can be denied, thereby completing a high-trust identity determination of the three-dimensional coupling of terminal, behavior, and request without relying on external physical media or traditional static authentication factors.

[0172] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A paperless archive borrowing identity verification method based on the privacy-preserving Turing test, characterized in that: include: S1: Extract the path state vector based on the borrowing request identifier and generate the disturbance feature vector; Construct a perturbation-bound Turing task based on the perturbation feature vector and generate a Turing verification interaction package; S2: Obtain the response behavior vector and terminal state sampling data based on the Turing verification interaction package to generate a joint response feature vector; Perform collaborative response judgment on the joint response feature vector, output the consistent behavior vector and construct the behavior execution trajectory map, and solve the disturbance behavior consistency feature index; S3: Constructs a node perturbation vector sequence based on the behavior execution trajectory graph, compares it with the terminal perturbation registry, and outputs a perturbation matching confidence score. It confirms behavior consistency through synchronization verification, generates a trusted execution tag, and converts it into a node fingerprint hash summary. S4: Generate a node-bound authentication token based on the node fingerprint hash summary and the borrowing request identifier, and establish an access permission mapping table; determine the legitimacy of the token through node consistency verification, and output the archive access permission or mark access exception events.

2. The paperless archive borrowing identity verification method based on the privacy-preserving Turing test according to claim 1 is characterized by: S1 also includes: extracting the user identity field and the access target field from the data generated by the archive borrowing request to generate a borrowing request identifier; extracting the request link path identifier and the execution timestamp from the borrowing request identifier to form a path state vector; By parsing the path state vector, node disturbance related data items are extracted, and a disturbance feature vector is generated according to the construction requirements. The node disturbance related data items include cache response delay, voltage disturbance change and graphics frame rendering offset value; By analyzing the perturbation response positions and noise distribution density with periodic consistency in the perturbation feature vector, a perturbation credible window is constructed. The perturbation indicators that meet the preset threshold in the perturbation credible window are embedded into the graphical test structure to form a perturbation-bound Turing task. A graphical task structure is constructed through a perturbation-binding Turing task, wherein the graphical task structure includes a pseudo-behavior induction area, a true judgment area, and a path random perturbation area; the graphical task structure is combined with the perturbation-binding Turing task to generate a one-time interaction sequence, wherein the interaction sequence constitutes a Turing verification interaction package; If the perturbation feature vector in the path state vector is lower than the preset execution confidence threshold, the construction process of the Turing verification interaction package is terminated, and the process returns to the path state vector extraction and is re-executed.

3. The paperless archive borrowing identity verification method based on the privacy-preserving Turing test according to claim 2 is characterized by: S2 also includes: guiding the user to complete pseudo-behavior selection and target judgment behavior through the deployed Turing verification interaction package, and outputting a response behavior vector; triggering periodic sampling of the current state of the terminal through Turing interaction, and the data of the current state of the sampled terminal includes image loading delay, thermal noise index and execution offset delay; The response behavior vector and the terminal state sampling data are combined to form a behavior-environment mapping pair, and then converted into a joint response feature vector; it is identified whether the frequency of the behavior path node in the joint response feature vector and the synchronization rate of the terminal noise reach the preset collaborative response threshold; if the preset collaborative response threshold is reached, the joint response feature vector is marked as a consistent behavior vector; otherwise, it is recorded as an abnormal response behavior; A behavior execution trajectory graph is established based on the consistent behavior vector. The nodes in the behavior execution trajectory graph include operation type identifiers and perturbation synchronization markers. The behavior labels embedded in the Turing test are matched with the behavior execution trajectory graph to solve the perturbation behavior consistency feature index.

4. The paperless archive borrowing identity verification method based on the privacy-preserving Turing test according to claim 3 is characterized by: S3 also includes: extracting disturbance parameters associated with each behavior node in the behavior execution trajectory graph, constructing a corresponding node disturbance vector sequence, performing difference regression comparison on the node disturbance vector sequence and a preset terminal disturbance registry, and outputting a disturbance matching confidence score; Determine whether the perturbation matching confidence score exceeds the preset matching threshold. If so, construct a node execution confirmation result identifier; perform synchronization verification through the perturbation behavior consistency feature index and the sequence of perturbation indicators embedded in the Turing verification interaction package; if the synchronization verification is successful, merge the node execution confirmation result identifier and the behavior consistency label to form a trusted execution tag; if the perturbation matching confidence score does not meet the standard or the synchronization verification fails, output a pseudo execution identifier and terminate the borrowing process; Convert the trusted execution token into a node fingerprint hash digest, which serves as the binding basis for subsequent authentication token generation.

5. The paperless archive borrowing identity verification method based on the privacy-preserving Turing test according to claim 4 is characterized by: S4 further includes: forming a joint token input pair from the node fingerprint hash digest and the borrowing request identifier, performing a hash mixing operation on the joint token input pair to generate a node-bound authentication token, and mapping the node-bound authentication token with the user permission level, the request data confidentiality level, and the access time parameter to generate an access permission mapping table; Determine whether the node fingerprint hash digest generated by the perturbation feature vector re-extracted by the current node is consistent with the digest recorded in the node-bound authentication token; if they are consistent, establish a data channel based on the access permission mapping table and output the corresponding file access permission; otherwise, destroy the node-bound authentication token and update the token status in the access permission mapping table to invalid; If a subsequent node submits an expired or hash-inconsistent node binding authentication token again, the behavior will be marked as a pseudo-node access event and recorded in the access exception record.

6. The paperless archive borrowing identity verification method based on the privacy-preserving Turing test according to claim 5 is characterized by: Also included in S1: Path state vector construction: Where V path is the path state vector; is the user identity field; To access the target field; μ req Identifies the borrowing request; Indicates extracting the request link path identifier x from the borrowing request identifier route and execution timestamp χ time ; Γ(·) is the borrowing request aggregation function; Generation of perturbation eigenvectors: Where V dist is the perturbation eigenvector; ρ mem (t) represents the disturbance signal of cache response delay at time t; ρ volt (t) represents the voltage disturbance change response value at time t; ρ frame (t) represents the graphics frame rendering offset response value at time t; T is the total length of the sampling time; The perturbation credible window extraction formula is expressed as: in is the perturbation credible window; ρ(t) represents the perturbation response strength at time t; ρ(s) represents the perturbation response strength at time s; is the second-order derivative of the disturbance signal; σ curve (t) is the perturbation curvature control function; ζ dense (t) is the noise distribution density function of the disturbance response; is the indicator function; ε is the preset disturbance stability threshold; Δt is the local disturbance observation window; The spatial mapping of the disturbance index in the graph task structure is expressed as: in Pixel-space representation of graphics task structures; is the kth graphic area, which includes the pseudo-behavior induction area, the real judgment area and the path random perturbation area; δx k ,δy k is the coordinate perturbation offset generated when the perturbation point in the perturbation credible window is mapped to the graphic area; n is the total number of graphic interaction areas; in is a Turing-verified interaction package; ⊕ represents the synthetic operation of embedding the perturbation trusted window into the graph task structure; Λ exec (V dist ) is the execution confidence function; θ exec is the preset execution confidence threshold.

7. The paperless archive borrowing identity verification method based on the privacy-preserving Turing test according to claim 6 is characterized by: Define F in S2 joint (t) is the joint response feature vector at time t; F joint (t)=[a resp (t),β load (t),β noise (t),β devi (t)] where α resp (t) is the response behavior vector value; β load (t) is the image loading delay; β noise (t) is the thermal noise index; β devi (t) is the execution offset delay; where Θ sync is a measure of the intensity of the collaborative response; The rate of change of user behavior rhythm; is the proportion of behavior offset to loading delay; cos(·) is the control function; is an exponential suppression function; T is the total length of the sampling time; Among them F valid is the consistent behavior vector; τ coh is a preset collaborative response threshold; in is the behavior execution trajectory diagram; θ i Identifies the type of operation i; is the ith disturbance synchronization mark; Op i Extract function for operation type; Corr i is the synchronization state evaluation function; N is the total number of trajectory behavior nodes; where Ξ match is the consistency feature index of the perturbation behavior; θ j is the jth behavior label in the behavior trajectory graph; τ j is the sequence of behavior labels embedded in the Turing task interaction package; is the jth disturbance synchronization mark; Indicates that the disturbance synchronization state is established.

8. The paperless archive borrowing identity verification method based on the privacy-preserving Turing test according to claim 7 is characterized by: In S3, construct the node perturbation vector sequence expression: Among them U trace is the node perturbation vector sequence; u j is the perturbation vector corresponding to the jth consistent behavior node; Ψ(·) is the perturbation parameter extraction function; Regression comparison of the terminal disturbance registry based on component difference fitting: where μ match Assign confidence scores to perturbation matches; is the pre-stored disturbance vector of the corresponding behavior type in the terminal disturbance registration database; ||·|| is the vector bi-norm; ε is the disturbance noise robustness factor; where χ exec Execute the confirmation result mark for the node; match is the preset disturbance matching threshold; where Ω sync is the consistency of disturbance synchronization; Embed the disturbance indicator sequence in the perturbation response window of the jth region in the Turing verification interaction package; is the indicator function; where χ trust Trusted execution mark; sync is the preset disturbance synchronization consistency threshold; in is the node fingerprint hash summary; Hash(·) is an irreversible hash function; || is the concatenation operator.

9. The paperless archive borrowing identity verification method based on the privacy-preserving Turing test according to claim 8 is characterized by: In S4, the node perturbation behavior entropy is modeled and S is defined node is the perturbation behavior path entropy; where p j represents the perturbation probability distribution at the jth node after the perturbation intensity is normalized; u k is the kth perturbation vector in the set of perturbation behavior consistency feature indices; in Bind authentication token to the node; Hash mix (·) is the hash mixing function; Generate the ownership mapping tensor in S4: in is the value of the cell with index (i, j, k) in the access control tensor; is the i-th user authority level; is the jth data classification level; is the kth type access time window parameter; Δt k is the delay between the current system time and the generation time of the node-bound authentication token in the k-th access window; Calculate the consistency difference of the current node disturbance spectrum: where φ j is the disturbance characteristic spectrum; D diff The difference between the current disturbance spectrum of the execution node and the disturbance spectrum of the authentication token bound to the node; is the spectrum feature of the current node in the jth disturbance dimension; is the spectrum feature recorded in the jth perturbation dimension when the node binding authentication token is generated; |·| is the absolute value; where χ grant is the final access decision result; τ D is the upper threshold of the disturbance spectrum difference; τ S is the lower threshold of the perturbation behavior path entropy; Represents a permission tensor consisting of permission level, data confidentiality level, and time window.

Images