Privacy protection and verifiable continuous authentication method based on random forest

Abstract

The embodiment of the invention discloses a privacy protection and verifiable continuous authentication method based on a random forest. A specific embodiment of the method comprises the following steps: extracting a plurality of statistical features from an obtained initial user behavior data set to obtain a statistical feature set; generating a ciphertext feature representation according to the statistical feature set, and transmitting the ciphertext feature representation to a double-cloud authentication server; according to a pre-designed security comparison sub-protocol, performing security comparison on two pieces of to-be-compared ciphertext data in the ciphertext feature representation to obtain a ciphertext comparison result; according to the ciphertext sequence and the ciphertext comparison result, random forest training is carried out on the ciphertext feature representation in the double-cloud authentication server, and a random forest authentication model in a ciphertext environment is obtained; and continuously authenticating the user by using the random forest authentication model to generate an authentication result. According to the embodiment, resource consumption of the user during authentication is reduced, and privacy protection and verifiable continuous authentication are effectively realized.

Description

TECHNICAL FIELD

[0001] Embodiments of the present disclosure relate to the field of computer technology, and in particular, to a privacy-preserving and verifiable persistent authentication method based on random forest. BACKGROUND

[0002] The booming development of smart devices provides a series of services to improve daily life, including financial transactions, medical care, etc. However, this convenience has also raised some security issues. In order to be convenient, each user needs to store and process sensitive information on these devices, such as passwords, transaction records, biometric data, etc. In addition, the portability of smart devices makes them easy to lose, which may lead to data leakage. Therefore, it is urgent to provide an effective, reliable, and secure identity authentication mechanism for smart device users. Current authentication mechanisms mainly focus on one-time authentication (i.e., single sign-on). They can be divided into three types: knowledge-based (e.g., PIN and password), token-based (e.g., smart card and dynamic password token), and biometric-based authentication (e.g., fingerprint and facial recognition). Knowledge-based identity authentication is the most popular identity authentication method, but it is vulnerable to observational attacks such as shoulder surfing attacks and smudge attacks. In addition, users often use simple and easy-to-guess passwords for convenience, which makes it easy for imposters to access device content. As the dependence on smart devices increases, different applications often recommend users to use different passwords, which leads to an astonishing number of password and PIN requests required to access various data and services. For token-based authentication, physical tokens are easy to lose or damage, while electronic tokens usually depend on specific devices. Compared with knowledge-based and token-based authentication, biometric-based authentication provides enhanced convenience and security due to the strong association between biometric features and users. However, the implementation of biometric identity authentication often raises concerns about privacy and usability. More critically, once a user logs into a system through one-time identity authentication, they can continuously access resources until they log out. From user login to logout, the system often runs in a lower security state, which may expose sensitive resources to unauthorized users, causing significant vulnerabilities. Therefore, there is an urgent need for a persistent authentication method to improve the security of smart devices.

[0003] Unlike one-time identity authentication, continuous identity authentication requires continuous user identity monitoring to ensure that the user is always the owner during device usage. However, if the user needs to interact with the device every time for authentication, it will seriously interfere with the user's attention and affect the overall user experience. Obviously, the current explicit authentication method is not suitable for frequent authentication. Moreover, continuous authentication involves continuous monitoring and analysis, which requires certain computing and storage resources. These resource requirements may have a negative impact on the performance of smart devices. Therefore, continuous authentication should be both lightweight and implicit. Smart devices are equipped with various sensors that can capture user behavior in real time to implement implicit authentication technology. However, behavior-based identity authentication also faces challenges, especially in terms of privacy. Using behavior features for identity verification raises concerns about potential privacy violations due to their close relationship with personal privacy. In order to achieve privacy-protected continuous authentication, most existing solutions use feature vector similarity evaluation to determine user legitimacy. However, this solution is relatively simple and sacrifices accuracy to achieve continuous authentication. In addition, current privacy-protected continuous authentication methods are based on a semi-honest model and do not consider malicious behavior by the server. SUMMARY

[0004] This summary is provided to introduce a selection of concepts that are further described below in the detailed description. This summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used to limit the scope of the claimed subject matter.

[0005] Some embodiments of the present disclosure propose a privacy-protected and verifiable continuous authentication method based on random forests to solve the technical problems mentioned in the background section.

[0006] In a first aspect, some embodiments of the present disclosure provide a privacy-protected and verifiable continuous authentication method based on random forests, the method comprising: extracting a plurality of statistical features from an initial user behavior data set obtained to obtain a statistical feature set; generating a ciphertext feature representation according to the statistical feature set, and transmitting the ciphertext feature representation to a double-cloud authentication server; according to a pre-designed secure sorting sub-protocol, sorting the to-be-sorted data set in the ciphertext feature representation in ascending order according to the corresponding plaintext value to generate a ciphertext sequence; according to a pre-designed secure comparison sub-protocol, securely comparing two ciphertext data to be compared in the ciphertext feature representation to obtain a ciphertext comparison result; according to the ciphertext sequence and the ciphertext comparison result, training a random forest authentication model in the ciphertext environment according to the ciphertext feature representation in the double-cloud authentication server; and using the random forest authentication model to continuously authenticate the user to generate an authentication result.

[0007] In a second aspect, some embodiments of the present disclosure provide a privacy protection and verifiable persistent authentication device based on random forest, the device comprising: an acquisition unit configured to extract a plurality of statistical features from an acquired initial user behavior data set to obtain a statistical feature set; a generation unit configured to generate a ciphertext feature representation according to the statistical feature set, and transmit the ciphertext feature representation to a double-cloud authentication server; an ordering unit configured to sort a to-be-ordered data set in the ciphertext feature representation in ascending order according to a corresponding plaintext value according to a pre-designed secure ordering sub-protocol, to generate a ciphertext sequence; a comparison unit configured to securely compare two ciphertext data to be compared in the ciphertext feature representation according to a pre-designed secure comparison sub-protocol, to obtain a ciphertext comparison result; a training unit configured to train the ciphertext feature representation in the double-cloud authentication server according to the ciphertext sequence and the ciphertext comparison result, to obtain a random forest authentication model in a ciphertext environment; and an authentication unit configured to use the random forest authentication model to perform persistent authentication on a user, to generate an authentication result.

[0008] In a third aspect, some embodiments of the present disclosure provide an electronic device, comprising: one or more processors; a storage device having one or more programs stored thereon, when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any of the implementations of the first aspect.

[0009] In a fourth aspect, some embodiments of the present disclosure provide a computer readable medium having a computer program stored thereon, wherein the program is executed by a processor to implement the method described in any of the implementations of the first aspect.

[0010] The various embodiments disclosed herein have the following beneficial effects: The privacy-preserving and verifiable continuous authentication method based on random forests, as described in some embodiments of this disclosure, specifically aims to solve problems in existing authentication systems such as vulnerability to malicious attacks, easy leakage of user data, potential server theft of core model information, and high client authentication resource consumption. In the protocol, users encrypt and send their behavioral data to the authentication server. The authentication server can directly train a random forest model on the encrypted user behavioral data without decryption and use the trained encrypted model to perform user authentication. The protocol achieves full-process privacy protection for user data and the random forest model during the training and authentication phases. The server can use the random forest model for user authentication but cannot obtain the random forest's decision threshold or the user's behavioral data. This application assumes the authentication server is malicious, introduces a dual-cloud architecture and threshold Paillier encryption to improve the security of the authentication system, and adopts non-interactive authentication to reduce client computation and lower user resource consumption during authentication. Therefore, it can resist malicious attackers and prevent any leakage of behavioral data. The protocol allows the authentication server to authenticate users without knowing the user's input, i.e., behavioral feature data, while keeping the trained model confidential to all participants. The protocol reduces user involvement during both the training and authentication phases. For 28 behavioral features, each authentication requires only a 0.136s client-side computation latency, achieving both privacy protection and efficient authentication. This application effectively achieves privacy protection and verifiable continuous authentication, showing promising application prospects. It can achieve lossless authentication accuracy while ensuring privacy, demonstrating significant application value. Attached Figure Description

[0011] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0012] Figure 1 This is a flowchart of some embodiments of the privacy-preserving and verifiable continuous authentication method based on random forest according to this disclosure;

[0013] Figure 2 This is a schematic diagram of the structure of some embodiments of the random forest-based privacy protection and verifiable continuous authentication apparatus according to the present disclosure;

[0014] Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure;

[0015] Figure 4This is an authentication method architecture diagram based on some embodiments of the privacy-preserving and verifiable continuous authentication method based on random forest disclosed herein;

[0016] Figure 5 This is a flowchart of a fully ordered sub-protocol based on some embodiments of a random forest-based privacy-preserving and verifiable continuous authentication method disclosed herein.

[0017] Figure 6 This is a flowchart of the optimal threshold determination process according to some embodiments of the random forest-based privacy-preserving and verifiable continuous authentication method disclosed herein. Detailed Implementation

[0018] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0019] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0020] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0021] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0022] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0023] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0024] Figure 1 A flow 100 is shown illustrating some embodiments of a random forest-based privacy-preserving and verifiable continuous authentication method according to this disclosure. This random forest-based privacy-preserving and verifiable continuous authentication method includes the following steps:

[0025] Step 101: Extract multiple statistical features from the obtained initial user behavior dataset to obtain a statistical feature set.

[0026] In some embodiments, the entity (e.g., a computing device) implementing the privacy protection and verifiable continuous authentication method based on random forest can extract multiple statistical features from the acquired initial user behavior dataset to obtain a statistical feature set.

[0027] Here, the aforementioned statistical features may include: time-domain features (e.g., mean, median, maximum, minimum, standard deviation, peak, lower quartile, upper quartile), frequency-domain features (e.g., HP1 energy, FHP2 energy, HP2 energy), morphological features (e.g., skewness, kurtosis), and information entropy.

[0028] As an example, the aforementioned executing entity can divide the time window... During the time window Within this framework, different statistical characteristics of the initial user behavior dataset are determined, resulting in a statistical feature set.

[0029] Step 102: Generate a ciphertext feature representation based on the above statistical feature set, and transmit the above ciphertext feature representation to the dual-cloud authentication server.

[0030] In some embodiments, the execution entity may generate a ciphertext feature representation based on the statistical feature set and transmit the ciphertext feature representation to the dual-cloud authentication server.

[0031] Optionally, the aforementioned executing entity can generate a ciphertext feature representation based on the aforementioned statistical feature set through the following steps, and transmit the ciphertext feature representation to the dual-cloud authentication server:

[0032] The first step is to preprocess the above statistical feature set to generate a processed statistical feature set.

[0033] As an example, the aforementioned executing entity can perform data cleaning on the aforementioned statistical feature set to generate a cleaned statistical feature set. Then, the cleaned statistical feature set is subjected to data standardization to generate a processed statistical feature set, which serves as the processed statistical feature set.

[0034] The second step is to discretize each processed statistical feature in the above processed statistical feature set in turn to obtain discretized statistical features.

[0035] As an example, the aforementioned executing entity can perform operations for each statistical feature data. k represents the first k. First, the user discretizes it using the following formula. Within the range. .in, It is the set of all discretized feature variables. yes One example, yes The minimum value, yes The maximum value. (The above) This is the discretization coefficient, which can be set to 16.

[0036] The third step is to determine the feature representation of the above discretized statistical features in order to generate feature representations, wherein the above feature representations include two forms.

[0037] The two forms mentioned above are the vector form for the registration phase and the vector form for the authentication phase, respectively.

[0038] As an example, the aforementioned executing entity can provide the aforementioned discretized statistical characteristics. The data after discretization is represented in the registration phase as follows: The authentication phase is represented as... In the above, m represents the plaintext value. a represents an element, i represents the i-th sample, and j represents the j-th feature.

[0039] The fourth step is to encrypt each matrix element in the above feature matrix to generate encrypted feature elements, obtain ciphertext feature representation, and transmit the above ciphertext feature representation to the dual-cloud authentication server.

[0040] As an example, the aforementioned execution entity can use Paillier encryption. For the characteristic matrix Encrypt each element in and represent the ciphertext features Submitted to the dual-cloud authentication server. The 'm' above represents the plaintext value. The above... This is a ciphertext feature representation. The 'c' above represents a single ciphertext element within the ciphertext feature representation. Wherein, =(N,g) is the Paillier public key. r represents an integer used to ensure semantic security (i.e., the same plaintext will produce different ciphertexts each time it is encrypted, preventing frequency analysis attacks).

[0041] Step 103: According to the pre-designed secure sorting sub-protocol, the dataset to be sorted in the above-mentioned ciphertext feature representation is sorted in ascending order according to the corresponding plaintext values ​​to generate a ciphertext sequence.

[0042] In some embodiments, the execution entity may sort the dataset to be sorted in the ciphertext feature representation in ascending order according to the corresponding plaintext values, based on a pre-designed secure sorting sub-protocol, to generate a ciphertext sequence.

[0043] As an example, the aforementioned execution entity can use a secure sorting protocol to sort the Paillier ciphertext data to be sorted in ascending order according to the corresponding plaintext values. The data to be sorted is defined as a tuple. ,in The sorted ciphertext sequence is represented as , where for any ,satisfy .

[0044] Optionally, the aforementioned execution entity can generate a ciphertext sequence by sorting the dataset to be sorted in the ciphertext feature representation in ascending order according to the corresponding plaintext values ​​based on a pre-designed secure sorting sub-protocol through the following steps:

[0045] The first step is to determine the first unsorted data in the unsorted dataset in the above encrypted feature representation as the pivot element.

[0046] As an example, the aforementioned execution entity can choose As a reference element .

[0047] The second step is to generate the sorted left subvector and the sorted right subvector based on the dataset to be sorted and the pre-designed secure sorting sub-protocol.

[0048] As an example, the aforementioned execution entity can invoke the Security Comparison Protocol (SDCP) for collaborative computation: ,if ,but ,otherwise Identification based on comparison results , The original vector is divided into two new subvectors. The left subvector... , including all that meet elements right subvector Includes all that satisfy elements The same process (steps 1-3) is recursively applied to the left subvector. and right subvector Each recursive call produces a sorted subvector. and .

[0049] The third step is to combine the above-mentioned base element, the above-mentioned sorted left sub-vector, and the above-mentioned sorted right sub-vector to obtain the ciphertext sequence.

[0050] As an example, the aforementioned executing entity can... and The combination yields the sorted vector: The detailed flowchart is as follows: Figure 5 As shown.

[0051] Step 104: According to the pre-designed security comparison sub-protocol, perform a security comparison on the two ciphertext data to be compared in the above ciphertext feature representation to obtain the ciphertext comparison result.

[0052] In some embodiments, the aforementioned execution entity may perform a secure comparison of two ciphertext data to be compared in the aforementioned ciphertext feature representation according to a pre-designed secure comparison sub-protocol, and obtain a ciphertext comparison result.

[0053] As an example, the aforementioned execution entity can compare two Paillier ciphertexts. and The Security Comparison Protocol (SDCP) determines the numerical relationship of its corresponding plaintext, that is ( or The security comparison protocol is denoted as The details are as follows:

[0054] Step 1, random selection Publish authentication key , ,in , It refers to the number of servers.

[0055] The second step, and Calculate separately ;

[0056] The third step, Perform partial decryption operation ,calculate and Then calculate .in, It is an output range of The hash function, finally, calculate Finally Send to . Performing similar operations yields , ,and and will Send to .

[0057] Step 4 calculate , And verify Similarly, verify If the equation does not hold true, then malicious operation has occurred, and the protocol is invalid. Among these, It is an output range of The hash function.

[0058] Step 5 Aggregate the shares to obtain And calculate if ,but ;if This means The detailed flowchart is as follows: Figure 6 As shown.

[0059] Step 105: Based on the above ciphertext sequence and the above ciphertext comparison results, perform random forest training on the ciphertext feature representation in the above dual-cloud authentication server to obtain the random forest authentication model in the ciphertext environment.

[0060] In some embodiments, the execution entity may train a random forest on the ciphertext feature representation in the dual-cloud authentication server based on the ciphertext sequence and the ciphertext comparison result to obtain a random forest authentication model in the ciphertext environment.

[0061] Here, the dual-cloud authentication server achieves lossless authentication accuracy while ensuring the privacy of user data and random forest model based on the sub-protocol. The dual-cloud authentication server trains a privacy-preserving random forest on the statistical features of encrypted user behavior data to obtain a continuous authentication model under ciphertext. This ensures that the user data and random forest model thresholds are privacy-preserving during the training phase. The server stores the authentication model under ciphertext, but cannot obtain the decision threshold of the random forest or the user's behavior data.

[0062] As an example, the aforementioned execution entity can perform operations for each data dimension. Dual-server collaborative execution For each Execute on both servers one at a time To eliminate duplicates and identify all potential thresholds, denoted as Dual-server collaborative computing Comparison matrices:

[0063] Then, in each decision tree of the random forest, from Random selection Each feature is denoted as . ,in Additionally, the server from the first Randomly selecting a threshold from the dimensional features yields ,in Next, the server evaluates candidate thresholds for the selected data dimensions on the current node, calculates the Gini coefficient, and determines the threshold and feature dimension that maximizes the Gini gain G.

[0064] Next, the candidate threshold evaluation step is recursively performed on the left and right child nodes until a predefined stopping condition is met. At this point, the node becomes a leaf node, and the decision tree model in the encrypted environment is obtained. Finally, the random forest model is obtained. The model parameters trained by the server in encrypted form are still in encrypted form. The authentication server can only store the authentication model in encrypted form and cannot obtain the decision thresholds of the random forest or the user's behavior data.

[0065] Step 106: Use the above random forest authentication model to continuously authenticate users in order to generate authentication results.

[0066] In some embodiments, the aforementioned implementing entity may use the aforementioned random forest authentication model to continuously authenticate users in order to generate authentication results.

[0067] The server uses a random forest authentication model under encrypted data to achieve continuous authentication with privacy protection. The protocol only outputs the authentication result and does not disclose other information. If the result is 1, the authentication is successful; otherwise, the service is terminated.

[0068] Optionally, the aforementioned implementing entity can continuously authenticate users using the random forest authentication model described above through the following steps to generate authentication results:

[0069] The first step is to collect raw behavioral data within the preset authentication time window.

[0070] As an example, the aforementioned implementing entity can [do something] at a specific time. Inside, raw behavioral data is collected.

[0071] The second step is to extract features from the above-mentioned raw behavioral data to generate plaintext feature vectors.

[0072] As an example, the aforementioned executing entity can extract feature vectors from the aforementioned raw behavioral data. That is, the plaintext feature vector. .

[0073] The third step is to encrypt the plaintext feature vector to generate an encrypted feature vector.

[0074] As an example, the aforementioned execution entity can encrypt each element in the plaintext feature vector using the following formula: This is used to generate encrypted feature vectors.

[0075] The fourth step is to send the aforementioned encrypted feature vector to the dual-cloud authentication server.

[0076] The fifth step involves determining that the aforementioned dual-cloud authentication server has received the aforementioned encrypted feature vector, and then using the aforementioned random forest authentication model to continuously authenticate the user in order to generate an authentication result.

[0077] As an example, the aforementioned execution entity could, for each decision tree, have the authentication server store the model containing the threshold in ciphertext and the feature index in plaintext. This enables dual-server collaborative computation.

[0078] .if The server continues to evaluate the comparison results of the right subtree. Otherwise, it calculates the comparison results of the left subtree until the current node is identified as a leaf node, at which point the leaf node's label is returned as the prediction result. The final authentication result is determined based on the vote. If the result is 1, the service continues. Otherwise, a warning "authentication failed" is issued, the service terminates, and the authentication process continues as follows. Figure 4 As shown.

[0079] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a privacy-preserving and verifiable continuous authentication device based on random forest. These device embodiments are similar to... Figure 1 Corresponding to the method embodiments shown, this random forest-based privacy protection and verifiable continuous authentication device can be specifically applied to various electronic devices.

[0080] like Figure 2 As shown, some embodiments of the privacy protection and verifiable continuous authentication device 200 based on random forest include: an acquisition unit 201, a generation unit 202, a sorting unit 203, a comparison unit 204, a training unit 205, and an authentication unit 206. The acquisition unit 201 is configured to extract multiple statistical features from the acquired initial user behavior dataset to obtain a statistical feature set; the generation unit 202 is configured to generate a ciphertext feature representation based on the statistical feature set and transmit the ciphertext feature representation to the dual-cloud authentication server; the sorting unit 203 is configured to sort the dataset to be sorted in the ciphertext feature representation in ascending order according to the corresponding plaintext values ​​according to a pre-designed secure sorting sub-protocol to generate a ciphertext sequence; the comparison unit 204 is configured to perform a secure comparison of two ciphertext data to be compared in the ciphertext feature representation according to a pre-designed secure comparison sub-protocol to obtain a ciphertext comparison result; the training unit 205 is configured to train a random forest on the ciphertext feature representation in the dual-cloud authentication server based on the ciphertext sequence and the ciphertext comparison result to obtain a random forest authentication model in a ciphertext environment; and the authentication unit 206 is configured to continuously authenticate users using the random forest authentication model to generate authentication results.

[0081] It is understandable that the units described in the random forest-based privacy protection and verifiable continuous authentication device 200 are related to the reference Figure 1 The steps in the described method correspond accordingly. Therefore, the operations, features, and beneficial effects described above for the method also apply to the random forest-based privacy protection and verifiable continuous authentication device 200 and the units contained therein, and will not be repeated here.

[0082] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device 300 (e.g., a computing device) suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0083] like Figure 3 As shown, the electronic device 300 may include a processing unit 301 (e.g., a central processing unit, a graphics processor, etc.), which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0084] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0085] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0086] It should be noted that, in some embodiments of this disclosure, the computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0087] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0088] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: extract multiple statistical features from an acquired initial user behavior dataset to obtain a statistical feature set; generate a ciphertext feature representation based on the statistical feature set, and transmit the ciphertext feature representation to a dual-cloud authentication server; sort the dataset to be sorted in the ciphertext feature representation in ascending order according to the corresponding plaintext values ​​according to a pre-designed secure sorting sub-protocol to generate a ciphertext sequence; perform a secure comparison of two ciphertext data to be compared in the ciphertext feature representation according to a pre-designed secure comparison sub-protocol to obtain a ciphertext comparison result; train a random forest on the ciphertext feature representation in the dual-cloud authentication server based on the ciphertext sequence and the ciphertext comparison result to obtain a random forest authentication model in a ciphertext environment; and continuously authenticate users using the random forest authentication model to generate authentication results.

[0089] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0090] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0091] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0092] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A privacy-preserving and verifiable continuous authentication method based on random forest, characterized in that, include: Multiple statistical features are extracted from the initial user behavior dataset to obtain a statistical feature set; Based on the statistical feature set, a ciphertext feature representation is generated, and the ciphertext feature representation is transmitted to the dual-cloud authentication server; According to a pre-designed secure sorting sub-protocol, the dataset to be sorted in the ciphertext feature representation is sorted in ascending order according to the corresponding plaintext values ​​to generate a ciphertext sequence; According to a pre-designed secure comparison sub-protocol, a secure comparison is performed on the two ciphertext data to be compared in the ciphertext feature representation to obtain the ciphertext comparison result. Based on the ciphertext sequence and the ciphertext comparison result, a random forest is trained on the ciphertext feature representation in the dual-cloud authentication server to obtain a random forest authentication model in the ciphertext environment. The random forest authentication model is used to continuously authenticate users in order to generate authentication results.

2. The method according to claim 1, characterized in that, The step of generating a ciphertext feature representation based on the statistical feature set and transmitting the ciphertext feature representation to the dual-cloud authentication server includes: The statistical feature set is preprocessed to generate a processed statistical feature set; Each processed statistical feature in the processed statistical feature set is sequentially discretized to obtain discretized statistical features; The discretized statistical features are subjected to feature representation determination to generate feature representations, wherein the feature representations include two forms; Each element in the feature representation is encrypted to generate an encrypted feature element, resulting in a ciphertext feature representation, which is then transmitted to the dual-cloud authentication server.

3. The method according to claim 1, characterized in that, The step of sorting the dataset to be sorted in the ciphertext feature representation according to the corresponding plaintext values ​​in ascending order based on the pre-designed secure sorting sub-protocol to generate a ciphertext sequence includes: The first unsorted data in the unsorted dataset in the encrypted feature representation is determined as the pivot element; Based on the dataset to be sorted and the pre-designed secure sorting sub-protocol, generate the sorted left sub-vector and the sorted right sub-vector; The base element, the sorted left subvector, and the sorted right subvector are combined to obtain the ciphertext sequence.

4. The method according to claim 1, characterized in that, The random forest authentication model is used to continuously authenticate users and generate authentication results, including: Collect raw behavioral data within the preset authentication time window; Feature extraction is performed on the original behavioral data to generate plaintext feature vectors; The plaintext feature vector is encrypted to generate an encrypted feature vector; The encrypted feature vector is sent to the dual-cloud authentication server; In response to determining that the dual-cloud authentication server has received the encrypted feature vector, the random forest authentication model is used to continuously authenticate the user in order to generate an authentication result.

5. A privacy-preserving and verifiable continuous authentication device based on random forest, characterized in that, include: The acquisition unit is configured to extract multiple statistical features from the acquired initial user behavior dataset to obtain a statistical feature set; The generation unit is configured to generate a ciphertext feature representation based on the statistical feature set, and to transmit the ciphertext feature representation to the dual-cloud authentication server. The sorting unit is configured to sort the dataset to be sorted in the ciphertext feature representation in ascending order according to the corresponding plaintext values, according to a pre-designed secure sorting sub-protocol, so as to generate a ciphertext sequence. The comparison unit is configured to perform a secure comparison of two ciphertext data to be compared in the ciphertext feature representation according to a pre-designed secure comparison sub-protocol, and obtain a ciphertext comparison result. The training unit is configured to train a random forest on the ciphertext feature representation in the dual-cloud authentication server based on the ciphertext sequence and the ciphertext comparison result, so as to obtain a random forest authentication model in the ciphertext environment. An authentication unit is configured to continuously authenticate users using the random forest authentication model to generate authentication results.

6. An electronic device, characterized in that, include: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 4.

7. A computer-readable medium, characterized in that, It stores a computer program thereon, wherein the computer program, when executed by a processor, implements the method as described in any one of claims 1 to 4.

Images