A lightweight user authentication method and system based on device fingerprinting

Abstract

This invention discloses a lightweight user authentication method and system based on device fingerprints, belonging to the field of power Internet of Things (IoT) technology. The lightweight user authentication method based on device fingerprints of this invention constructs a trace generation model, a clustering coupling model, anomaly calculation model, and a user authentication model. It uses the traces of Static Random Access Memory (SRAM) as device fingerprints; simultaneously, it clusters the data, selecting objects deviating from their cluster centers as trace objects, and uses a local outlier factor to measure the outliers of the trace objects; then, it obtains a ranking of user detection based on the outlier factor, removing users with high outliers, thereby avoiding false detections and achieving lightweight, legitimate user authentication. This method offers high user authentication efficiency and good flexibility, enabling IoT nodes to respond quickly. It is suitable for scenarios requiring rapid processing of a large number of requests, and the solution is scientific, reasonable, and feasible.

Description

TECHNICAL FIELD

[0001] The present application relates to a lightweight user authentication method and system based on device fingerprints, belonging to the technical field of power Internet of Things. BACKGROUND

[0002] Under the promotion of carbon emission reduction targets, the penetration rate of distributed energy in China has steadily increased. A large number of Internet of Things nodes are introduced into the new power grid dominated by distributed new energy to collect data on the operation and energy use of distributed energy access, transmission, distribution, and end users. The interactive business instruction interaction and information data transmission brought about by the collection process mostly use wireless public network transmission, and these interaction data are easily attacked physically, causing data tampering.

[0003] Further, Chinese patent (publication (announcement) number: CN110620820A) provides a ubiquitous power Internet of Things intelligent management system, including a device terminal and a management terminal; the device terminal is provided with an information collection module and an information sending module; the information collection module is used to collect Internet of Things data of the device terminal; the information sending module is used to send the Internet of Things data to each node in the blockchain network for legality judgment of the Internet of Things data, and if the Internet of Things data is legal, the Internet of Things data is written into the management terminal and the blockchain respectively. By sending the Internet of Things data to each node in the blockchain network for legality judgment of the Internet of Things data, if the Internet of Things data is legal, the Internet of Things data is written into the management terminal and the blockchain respectively to prevent data loss or changes, prevent the traditional structured center mode from causing power grid data management to be paralyzed and affecting people's normal life if the terminal fails, and ensure normal operation of the power grid.

[0004] The above uses each node in the blockchain network to judge the legality of the Internet of Things data, but the Internet of Things nodes have low processing and storage capabilities, and usually have low-bandwidth communication channels, so the above node processing scheme has large node resource overhead, making it impossible for the Internet of Things nodes to respond quickly, and thus the above scheme cannot be applied to scenarios that require fast processing of a large number of requests, resulting in low user authentication efficiency and poor flexibility.

[0005] The information disclosed in this background technology is only used to understand the background of the present application concept, and therefore it can include information that does not constitute prior art. SUMMARY

[0006] In view of the above problems or one of the above problems, a first object of the present application is to provide a lightweight user authentication method based on device fingerprints, which obtains a lightweight legal user authentication by constructing a trace generation model, a clustering coupling model, an anomaly calculation model and a user authentication model, has high user authentication efficiency, has good flexibility, enables the Internet of Things node to respond quickly, is suitable for a scene requiring quick processing of a large number of requests, and is particularly suitable for a low-bandwidth communication channel with low processing and storage capacity, and is scientific, reasonable and feasible.

[0007] In view of the above problems or one of the above problems, a second object of the present application is to provide a lightweight user authentication method and system based on device fingerprints, which characterizes device fingerprints by a static random access memory (SRAM) trace and uses the SRAM trace as a unique identifier to accurately distinguish devices, uses a clustering algorithm to cluster data, selects objects deviating from the clustering center as an anomaly candidate set, uses a local outlier factor to measure outliers in the anomaly candidate set, improves the outlier overlap problem of the local outlier factor method, obtains a user detection ranking through the outlier factor, uses descending order of the ranking to eliminate users with high outliers, thereby avoiding false detection and obtaining a lightweight legal user authentication.

[0008] In view of the above problems or one of the above problems, a third object of the present application is to provide a lightweight user authentication system based on device fingerprints, which realizes lightweight legal user authentication by setting an acquisition layer, an access layer and an application layer, has high authentication efficiency, has good flexibility, enables the Internet of Things node to respond quickly, is suitable for a scene requiring quick processing of a large number of requests, and is particularly suitable for a low-bandwidth communication channel with low processing and storage capacity, and is scientific, reasonable and feasible.

[0009] To achieve one of the above objects, a first technical solution of the present application is as follows:

[0010] A lightweight user authentication method based on device fingerprints, comprising the following steps:

[0011] In a first step, a trace sample of a static random access memory (SRAM) is collected by a pre-constructed trace generation model, a trace dataset is constructed, and the trace dataset is used to characterize device fingerprints;

[0012] In a second step, a clustering coupling model is used to cluster the trace dataset, and one or more trace objects deviating from the clustering center are obtained;

[0013] In a third step, an anomaly calculation model is used to measure the anomaly information of the one or more trace objects based on a local outlier factor;

[0014] Fourthly, through the pre-constructed user authentication model, the authentication of the lightweight user is completed based on the abnormal information.

[0015] Through continuous exploration and tests, the application constructs a trace generation model, a clustering coupling model, an abnormality calculation model and a user authentication model, takes the trace of a static random access memory (SRAM) as a device fingerprint, i.e., a unique identifier of the device, accurately distinguishes the device, clusters data, selects an object deviating from the clustering center as a trace object, measures the abnormal value of the trace object by using a local outlier factor, improves the abnormal value overlap problem of the local outlier factor method, obtains the ranking of user detection according to the abnormal value factor, eliminates the user with a high abnormal value by using the descending order of the ranking, avoids false detection, and obtains the lightweight legal user authentication, which is high in user authentication efficiency and good in flexibility.

[0016] Further, the static random access memory (SRAM) exists in the on-board controller of the Internet of Things node in the collection layer and is used to store collected data, and each node has the static random access memory, so the trace of the static random access memory (SRAM) can be taken as a device fingerprint feature, i.e., a unique identifier, to accurately distinguish the device.

[0017] Further, the local outlier factor (LOF) can be used to detect abnormal values, but cannot handle overlapping abnormal values. Therefore, the application introduces a clustering coupling model to cluster, analyzes the static random access memory (SRAM) trace of the user by using the clustering coupling model, lists the user far from the cluster center as an abnormal candidate, calculates the abnormality degree of the abnormal candidate by using the local outlier factor, and authenticates the legal user.

[0018] As a preferred technical measure:

[0019] In the first step, the method for constructing the trace data set by the trace generation model is as follows:

[0020] Step 11: Determine the static random access memory (SRAM) of all user devices in the collection layer to be authenticated.

[0021] Step 12: Collect the trace samples of the power-on static random access memory (SRAM) from each static random access memory by turning on and off the power supply.

[0022] Step 13: Assemble the trace samples to obtain the trace data set.

[0023] As a preferred technical measure:

[0024] In the second step, the method for clustering the trace dataset using a clustering coupling model is as follows:

[0025] Step 21: Obtain the trace dataset and initialize the clustering parameters;

[0026] Step 22: Use the Gaussian mixture algorithm to process the trace dataset and obtain the corresponding trace clusters;

[0027] Step 23: Based on trace clustering, calculate the number of trace clusters in the trace dataset;

[0028] Step 24: Calculate the probability that each trace sample belongs to each trace cluster;

[0029] Step 25: Maximize the probability that a trace sample belongs to a certain trace cluster, and assign the trace sample to a certain trace cluster;

[0030] Step 26: Iterate through steps 23, 24, and 25 until all trace samples are clustered.

[0031] As a preferred technical measure:

[0032] In the second step, the method for obtaining trace objects that deviate from their cluster centers using the cluster coupling model is as follows:

[0033] S21: After clustering is completed, calculate the cluster center for any trace cluster;

[0034] S22: Calculate the distance from each trace sample in the trace cluster to its respective cluster center to obtain the deviation data;

[0035] S23: Based on the deviation data, obtain one or more trace objects that deviate from their cluster centers;

[0036] S24: Aggregate one or more trace samples that are off-center from the cluster to form a candidate set of trace anomaly data.

[0037] As a preferred technical measure:

[0038] The formula for calculating the distance from each trace sample to its respective cluster center is as follows:

[0039]

[0040] Where, x j The cluster center corresponding to the trace sample. For trace samples, This represents the distance from the trace sample to its respective cluster center;

[0041] Or / and, the method to obtain trace objects that deviate from their cluster centers is as follows:

[0042] The method for calculating the trace anomaly dataset A for all trace clusters is as follows:

[0043] To satisfy The trace samples are taken as trace objects that deviate from their cluster centers and included in the trace anomaly dataset A;

[0044] Among them, |D j | is the size of the j-th cluster, |D| is the trace dataset, and δ is the value of n based on the local outlier factor. The deviation is the triple standard deviation from the j-th cluster.

[0045] As a preferred technical measure:

[0046] In the third step, the method for measuring abnormal information using an anomaly calculation model is as follows:

[0047] Step 31: For a trace sample x in the trace dataset, calculate the n nearest neighbors N of the trace sample x in the trace dataset. n (x ) And the nth nearest distance dist between trace samples in the trace dataset and trace sample y. n (y);

[0048] Step 32: Based on the nth nearest distance dist n Rd(x,y) is used to calculate the reachability distance between two trace samples x and y in the trace dataset. The formula is as follows:

[0049] Rd(x,y)=max{dist(x,y),dist n (y)};

[0050] Where dist(x,y) represents the Euclidean distance between two trace samples x and y;

[0051] Step 33: Calculate the local reachability density ρ of trace sample x based on the reachability distance Rd(x,y). n (x), its calculation formula is as follows:

[0052]

[0053] Where n is the nth nearest neighbor of x. This represents the reciprocal of the average distance between trace sample x and trace sample y when a trace sample y is selected from the nearest neighbors of trace sample x.

[0054] Step 34: Based on local reachability density ρ n(x), calculate the ratio of the local reachability density (LOF) of trace sample x to the local reachability density (LOF) of its nearest neighbor trace sample y for a specific n. n (x), used to measure the outlier degree of x among its nearest neighbors, is calculated as follows:

[0055]

[0056] Where, ρ n (y) represents the local reachability density of trace sample y and a selected trace sample among its nearest neighbors;

[0057] Step 35: Calculate the Locally Accessible Density Ratio (LOF). n (x) is taken as an outlier of the trace sample x, and multiple outliers are arranged in order to obtain the outlier information.

[0058] As a preferred technical measure:

[0059] In the fourth step, the method for authenticating lightweight users using the user authentication model is as follows:

[0060] Step 41: Sort the trace samples in the trace anomaly dataset according to the size of the outlier values ​​in the anomaly information to obtain user ranking data;

[0061] Step 42: Based on user ranking data, remove users with high average rankings to obtain authorized user authentication results.

[0062] To achieve one of the above objectives, the second technical solution of the present invention is as follows:

[0063] A lightweight user authentication method based on device fingerprints includes the following:

[0064] Trace samples are collected from static random access memory (SRAM) to construct a trace dataset, which is used to characterize device fingerprints. The trace dataset is clustered, and trace objects that deviate from their cluster centers are selected as trace anomaly candidate sets.

[0065] Based on the local outlier factor, outliers in the candidate set of trace anomalies are measured;

[0066] Authentication for lightweight users is completed based on outliers.

[0067] This invention uses SRAM traces to represent device fingerprints and serves as unique identifiers for accurate device differentiation. Simultaneously, it employs clustering algorithms to cluster data, selecting objects deviating from their cluster centers as anomaly candidates. Local outlier factors are used to measure outliers within these candidate sets, improving upon the outlier overlap problem inherent in the local outlier factor method. Furthermore, user detection rankings are obtained through outlier factors, and users with high outliers are removed using descending order of ranking, thus avoiding false detections and achieving lightweight legitimate user authentication.

[0068] To achieve one of the above objectives, the third technical solution of the present invention is as follows:

[0069] A lightweight user authentication system based on device fingerprints, employing the aforementioned lightweight user authentication method based on device fingerprints, comprises a data acquisition layer, an access layer, and an application layer.

[0070] The acquisition layer is used to collect node data from distributed energy sources, transmission network equipment, distribution network equipment, and end users. It deploys IoT nodes and trace extraction modules.

[0071] The access layer is equipped with 4G and / or 5G base stations for transmitting node data;

[0072] The application layer is the power grid business cloud, which is used to process node data. Its deployment includes an anomaly calculation module based on clustering coupling model and outlier detection, as well as a user authentication module that can eliminate abnormal users.

[0073] The node data collected by the IoT nodes is accessed to the power grid business cloud via 4G or / and 5G base stations for anomaly detection and abnormal user removal, thereby achieving lightweight user authentication.

[0074] This invention achieves lightweight, legitimate user authentication by setting up a collection layer, an access layer, and an application layer. The authentication is highly efficient and flexible, enabling IoT nodes to respond quickly. It is suitable for scenarios that require rapid processing of a large number of requests, and is particularly suitable for low-bandwidth communication channels with limited processing and storage capabilities. The solution is scientific, reasonable, and feasible.

[0075] As a preferred technical measure:

[0076] The trace extraction module is used to extract the static random access memory (SRAM) traces of all user equipment in the acquisition layer. By turning the power on and off, it collects trace samples from each SRAM that is powered on.

[0077] Or / and, the anomaly calculation module is used to cluster all trace samples using a clustering coupling model, select trace objects that deviate from their cluster centers as trace anomaly candidate sets, and use the local outlier factor to measure the outliers in the trace anomaly candidate set.

[0078] Or / and, the user authentication module is used to obtain the user ranking at the time of detection by outlier detection. If the average ranking of the detected users is always at the top, the user corresponding to the outlier is removed, and the user authentication is completed.

[0079] Or / and, node data includes at least operational data and energy consumption data;

[0080] Or / and, IoT nodes include smart devices and / or smart meters.

[0081] To achieve one of the above objectives, the fourth technical solution of the present invention is as follows:

[0082] An electronic device comprising:

[0083] One or more processors;

[0084] Storage device for storing one or more programs;

[0085] When the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned lightweight user authentication method based on device fingerprints.

[0086] To achieve one of the above objectives, the fifth technical solution of the present invention is as follows:

[0087] A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the aforementioned lightweight user authentication method based on device fingerprints.

[0088] Compared with existing technical solutions, the present invention has the following beneficial effects:

[0089] Through continuous exploration and experimentation, this invention constructs a trace generation model, a clustering coupling model, anomaly calculation model, and a user authentication model. It uses the traces of Static Random Access Memory (SRAM) as device fingerprints, i.e., unique identifiers for each device, to accurately distinguish them. Simultaneously, it clusters the data, selecting objects deviating from their cluster centers as trace objects. The local outlier factor is used to measure the outliers of these trace objects, improving upon the outlier overlap problem in the local outlier factor method. Furthermore, user detection rankings are obtained based on the outlier factor, and users with high outliers are removed using descending order, thus avoiding false detections and achieving lightweight, efficient, and flexible user authentication. Therefore, this invention's solution has low node resource overhead, enabling rapid response from IoT nodes. It is suitable for scenarios requiring rapid processing of large numbers of requests, and is particularly suitable for low-bandwidth communication channels with limited processing and storage capabilities. The solution is scientific, reasonable, and feasible.

[0090] Furthermore, this invention uses SRAM traces to represent device fingerprints and serves as unique identifiers to accurately distinguish devices. Simultaneously, it employs clustering algorithms to cluster data, selecting objects deviating from their cluster centers as anomaly candidate sets. Local outlier factors are used to measure outliers in these candidate sets, improving upon the outlier overlap problem in the local outlier factor method. User detection rankings are obtained through outlier factors, and users with high outliers are removed using descending order of ranking, thus avoiding false detections and achieving lightweight legitimate user authentication.

[0091] Furthermore, this invention achieves lightweight legitimate user authentication by setting up a collection layer, an access layer, and an application layer. The authentication is highly efficient and flexible, enabling IoT nodes to respond quickly. It is suitable for scenarios that require rapid processing of a large number of requests, and is particularly suitable for low-bandwidth communication channels with limited processing and storage capabilities. The solution is scientific, reasonable, and feasible. Attached Figure Description

[0092] Figure 1 This is a flowchart illustrating a lightweight user authentication method based on device fingerprinting according to the present invention.

[0093] Figure 2 This is a schematic diagram of a novel power grid architecture that introduces legitimate user authentication in this invention;

[0094] Figure 3 This is another flowchart illustrating the lightweight user authentication method based on device fingerprints according to the present invention.

[0095] Figure 4 This is a schematic diagram of a process for calculating LOF according to the present invention. Detailed Implementation

[0096] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0097] Conversely, this invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of the invention as defined in the claims. Furthermore, to provide a better understanding of the invention, certain specific details are described in detail below. However, those skilled in the art will fully understand the invention even without these detailed descriptions.

[0098] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the invention. The term "or / and" as used herein includes any and all combinations of one or more of the associated listed items.

[0099] like Figure 1 As shown, this is the first specific embodiment of the lightweight user authentication method based on device fingerprinting of the present invention:

[0100] A lightweight user authentication method based on device fingerprints includes the following steps:

[0101] The first step is to collect trace samples from static random access memory (SRAM) using the previously constructed trace generation model, and construct a trace dataset to characterize the device fingerprint.

[0102] The second step is to use the previously constructed clustering coupling model to cluster the trace dataset, resulting in one or more trace objects that deviate from their cluster centers.

[0103] The third step is to use the previously constructed anomaly calculation model to measure the anomaly information of one or more trace objects based on the local outlier factor.

[0104] The fourth step involves authenticating lightweight users based on the previously constructed user authentication model and anomaly information.

[0105] A second specific embodiment of the lightweight user authentication method based on device fingerprinting of the present invention:

[0106] A lightweight user authentication method based on device fingerprints includes the following steps:

[0107] The first step involves collecting trace samples from Static Random Access Memory (SRAM) using the previously constructed trace generation model to construct a trace dataset for characterizing device fingerprints.

[0108] The second step is to use the previously constructed clustering coupling model to cluster the trace dataset and select trace objects that deviate from their cluster centers as trace anomaly candidate sets.

[0109] The third step is to use the previously constructed anomaly calculation model to measure outliers in the trace anomaly candidate set based on the local outlier factor.

[0110] The fourth step is to complete the authentication of lightweight users based on outliers using the user authentication model built in advance.

[0111] A third specific embodiment of the lightweight user authentication method based on device fingerprinting of the present invention:

[0112] A lightweight user authentication method based on device fingerprints includes the following:

[0113] This invention collects SRAM trace samples from IoT devices to form a dataset D. A clustering coupling model is used to cluster the data, and objects deviating from their cluster centers are selected as anomaly candidate set A. A local outlier factor is used to measure outliers in A, obtaining their ranking at each detection time. If the average ranking of the detected outliers is consistently among the top (top 3% of the entire dataset), the outlier is removed, thus obtaining user authentication. This invention utilizes the SRAM traces of IoT nodes as unique and legitimate device identifiers, and employs lightweight clustering and local outlier detection to authenticate legitimate users. The solution is scientific, reasonable, and ingeniously conceived.

[0114] Static Random Access Memory (SRAM) resides in the onboard controller of IoT nodes in the acquisition layer and is used to store the acquired data. The SRAM structure is divided into a data portion, a high-variance stack, and a low-variance stack. The mean and variance of the high-variance stack of a device differ significantly, and this difference can be used as a device fingerprint to uniquely identify the device. The SRAM trace represents the size of the high-variance stack, which consists of bytes. These bytes are converted to their corresponding integer values ​​and scaled down by a factor of 255 to the range [0,1].

[0115] The method for selecting objects that deviate from their cluster centers is as follows:

[0116] S1. First calculate the cluster center x. j ;

[0117] S2. Next, calculate the distance from the trace samples to their respective cluster centers, using the following formula:

[0118]

[0119] S3. Calculate the triple standard deviation from the j-th cluster.

[0120] S4. Classify those that meet the following criteria into the abnormal candidate set A.

[0121]

[0122] Among them, |D j | is the size of the j-th cluster, |D| is the size of the entire dataset D, and δ is a parameter, which is 0.05 in this embodiment.

[0123] Because IoT nodes typically have limited processing and storage capabilities and low-bandwidth communication channels, the use of security solutions such as cryptographic public-key encryption algorithms is restricted. In new power grid architectures that introduce legitimate user authentication (such as...),... Figure 2 As shown in the figure, each node has a static random access memory (SRAM), so the trace of the SRAM is used as the fingerprint feature of the device.

[0124] Meanwhile, while the local outlier factor (LOF) can be used to detect outliers, it cannot handle overlapping outliers. Therefore, this invention introduces a clustering coupling model to cluster users. This model is used to analyze the user's static random access memory (SRAM) traces, listing users far from the cluster center as outlier candidates. The LOF is then used to calculate the outlier degree of these candidates, thereby authenticating legitimate users.

[0125] like Figure 3 As shown, this is the fourth specific embodiment of the lightweight user authentication method based on device fingerprints of the present invention:

[0126] A lightweight user authentication method based on device fingerprints requires collecting Static Random Access Memory (SRAM) trace samples from each IoT node in the novel power grid acquisition layer as device fingerprints. A lightweight user authentication method based on device fingerprints is then established to authenticate legitimate users. The specific implementation steps are as follows:

[0127] Step 1: Obtain the trace dataset D of the static random access memory (SRAM) and initialize the relevant parameters n and δ;

[0128] Step 2: Obtain the number of clusters and the corresponding clusters from the Gaussian Mixture Algorithm clustering. The method for obtaining this information includes the following:

[0129] The obtained device fingerprints (trace dataset D) of IoT nodes are divided into k clusters, which are k Gaussian distributions. The k Gaussian distributions are mixed and superimposed to form a Gaussian mixture distribution (cluster). Each Gaussian distribution has variance, mean and external mixing coefficient.

[0130] Step 3: Calculate the number of Gaussian models (clusters); Step 4: Calculate the probability that each trace sample belongs to each cluster, using the following method:

[0131] The expectation-maximization algorithm is used to obtain 3k parameters, and Bayes' theorem is used to obtain the posterior probability of each trace sample belonging to a certain Gaussian distribution (cluster).

[0132] Step 5: Maximize the probability that a trace sample belongs to a certain cluster. The specific method is as follows:

[0133] The Bayesian formula is used to calculate the posterior probability of each node data (trace sample) belonging to the 1st, 2nd, ... up to the kth Gaussian distribution. The node data is assigned to the corresponding cluster with the largest posterior probability.

[0134] Step 6: Iterate through steps 3 and 4 until convergence;

[0135] Step 7: For any cluster, calculate the cluster center and the distance from each object in the cluster to its respective cluster center;

[0136] Step 8: Obtain the candidate set A of anomalous data consisting of objects that are off-center;

[0137] Step 9: Calculate the candidate set of outlier data for all clusters;

[0138] Step 10: Calculate the Local Outlier Factor (LOF);

[0139] Step 11: Obtain the ranking of the candidate set of outlier data based on LOF;

[0140] Step 12: Remove users with high average rankings and obtain the authentication results of authorized users. The authorized users refer to legitimate IoT nodes that are connected to the power grid and recognized by the power grid.

[0141] This invention also provides a novel power grid system for lightweight legitimate user authentication based on SRAM trace samples from IoT devices, utilizing clustering coupling models and local outlier factor measurement. The system mainly includes a data acquisition layer, which deploys IoT nodes (smart devices, smart meters, etc.) to collect data on distributed energy access, transmission networks, distribution networks, and the operation and energy consumption of end users. At the access layer, the collected data is accessed via 4G and 5G base stations to the application layer's novel power grid service cloud.

[0142] In this embodiment, the specific implementation steps for calculating LOF are as follows (e.g. Figure 4 ):

[0143] S1: The node establishes a communication connection with the base station;

[0144] S2: IoT nodes use a Gaussian model to obtain clusters;

[0145] S3: For any cluster, calculate the cluster center x. j And each object to its respective cluster center x j distance Calculate the triple standard deviation from the j-th cluster.

[0146] S4: For those that satisfy The object is included in the abnormal dataset A;

[0147] Among them, |D j | is the size of the j-th cluster, |D| is the trace dataset, and δ is the value taken from n in LOF;

[0148] S5: Calculate A for all clusters;

[0149] S6: For an object x in dataset D, calculate the n nearest neighbors N of x in D. n (x), and the nth nearest distance dist between the object in D and y. n (y);

[0150] S7: Calculate the reachability distance Rd(x,y) between two objects x and y in dataset D. The calculation formula is as follows:

[0151] Rd(x,y)=max{dist(x,y),dist n (y)};

[0152] Here, dist(x,y) represents the Euclidean distance between two objects x and y.

[0153] S8: Calculate the local reachability density ρ of x. n (x), its calculation formula is as follows:

[0154]

[0155] Where n is the nth nearest neighbor of x. This represents the reciprocal of the average distance that x and y can reach when an object y is selected from the nearest neighbors of x.

[0156] S9: Calculate the ratio of the local reachability density (LOF) of x to that of its nearest neighbor y for a given n. n (x), used to measure the outlier degree of x among its nearest neighbors, is calculated as follows:

[0157]

[0158] Where, ρ n(y) represents the local reachability density of y and a selected object among its nearest neighbors.

[0159] A first specific embodiment of the lightweight user authentication system based on device fingerprinting of the present invention:

[0160] A lightweight user authentication system based on device fingerprints, employing the aforementioned lightweight user authentication method based on device fingerprints, comprises a data acquisition layer, an access layer, and an application layer.

[0161] The acquisition layer is used to collect node data from distributed energy sources, transmission network equipment, distribution network equipment, and end users. It deploys IoT nodes and trace extraction modules.

[0162] The access layer is equipped with 4G and / or 5G base stations for transmitting node data;

[0163] The application layer is the power grid business cloud, which is used to process node data. Its deployment includes an anomaly calculation module based on clustering coupling model and outlier detection, as well as a user authentication module that can eliminate abnormal users.

[0164] The node data collected by the IoT nodes is accessed to the power grid business cloud via 4G or / and 5G base stations for anomaly detection and abnormal user removal, thereby achieving lightweight user authentication.

[0165] A second specific embodiment of the lightweight user authentication system based on device fingerprints of the present invention:

[0166] A lightweight user authentication system based on device fingerprinting mainly includes three parts: a static random access memory (SRAM) trace extraction module, an anomaly calculation module based on clustering coupling model and outlier detection, and a user authentication module for removing abnormal users.

[0167] Static Random Access Memory (SRAM) Trace Extraction Module: Completes the extraction of SRAM traces from all user devices in the acquisition layer. By turning the power on and off, it manually collects SRAM trace samples from each device upon power-on.

[0168] An anomaly calculation module based on cluster coupling model and outlier detection: It performs clustering of data using cluster coupling model, selects objects that deviate from their cluster centers as anomaly candidate set A, and uses the local outlier factor to measure outliers in A.

[0169] User authentication module: Utilize the above outlier detection to obtain the ranking at the time of detection. If the average ranking of the detection is always at the top, remove the outlier and obtain user authentication.

[0170] An embodiment of a device applying the method of the present invention:

[0171] An electronic device comprising:

[0172] One or more processors;

[0173] Storage device for storing one or more programs;

[0174] When the one or more programs are executed by the one or more processors, the one or more processors implement the aforementioned lightweight user authentication method based on device fingerprints.

[0175] An embodiment of a computer medium applying the method of the present invention:

[0176] A computer-readable storage medium having a computer program stored thereon that, when executed by a processor, implements the aforementioned lightweight user authentication method based on device fingerprints.

[0177] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0178] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0179] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0180] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0181] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A lightweight user authentication method based on device fingerprint, characterized in that it comprises the following steps: Step 1: Collecting trace samples of static random access memory (SRAM) to construct a trace dataset for representing device fingerprint by using a pre-constructed trace generation model; The method for constructing a trace dataset by using a trace generation model is as follows: Step 11: Determining the static random access memory (SRAM) of all user devices in the collection layer to be authenticated; Step 12: Collecting the trace samples of the power-on static random access memory (SRAM) from each static random access memory by turning on and off the power supply; Step 13: Assembling the trace samples to obtain a trace dataset; Step 2: Clustering the trace dataset by using a pre-constructed clustering coupling model to obtain one or more trace objects deviating from the cluster center; Step 3: Measuring the abnormal information of one or more trace objects based on a local outlier factor by using a pre-constructed abnormality calculation model; Step 4: Completing the authentication of lightweight users based on the abnormal information by using a pre-constructed user authentication model; The method for completing the authentication of lightweight users by using a user authentication model is as follows: Step 41: Sorting the trace samples in the trace abnormal dataset according to the size of the abnormal value in the abnormal information to obtain user ranking data; Step 42: Based on the user ranking data, eliminating the users with high average ranking to obtain the authorized user authentication result. 2.The lightweight user authentication method based on device fingerprint according to claim 1, characterized in that the method for clustering the trace dataset by using a clustering coupling model in the second step is as follows: Step 21: Obtaining the trace dataset and initializing the clustering parameters; Step 22: Processing the trace dataset by using a Gaussian mixture algorithm to obtain the corresponding trace clusters; Step 23: Calculating the number of trace clusters of the trace dataset based on the trace clusters; Step 24: Calculating the probability of each trace sample belonging to each trace cluster; Step 25: Maximizing the probability of a trace sample belonging to a certain trace cluster to divide the trace sample into a certain trace cluster; Step 26: Iterating steps 23, 24 and 25 until all trace samples are clustered. 3.The lightweight user authentication method based on device fingerprint according to claim 1, characterized in that the method for obtaining trace objects deviating from the cluster center by using a clustering coupling model in the second step is as follows: S21: After clustering, calculating the cluster center for any trace cluster; S22: Calculating the distance of each trace sample in the trace cluster to the respective cluster center to obtain deviation data; S23: Obtaining one or more trace objects deviating from the cluster center based on the deviation data; S24: Converging one or more trace samples deviating from the cluster center to form a trace abnormal data candidate set. 4.The lightweight user authentication method based on device fingerprint according to claim 3, characterized in that the formula for calculating the distance of a trace sample to the respective cluster center is as follows: Or / and, the method for obtaining trace objects deviating from the cluster center is as follows: ​ ​ ​ wherein x j is the cluster center corresponding to the trace sample, is the trace sample, is the distance of the trace sample to the respective cluster center; ​ An abnormal trace data set A is calculated for all trace clusters, and the method is as follows: to meet the trace sample as a trace object deviating from its cluster center, and is included in the trace anomaly dataset A; where |D| is the size of the jth cluster, |D| is the trace data set, and δ is a value taken from the local outlier factor, j |is the size of the jth cluster, |D| is the trace data set, and δ is a value taken from the local outlier factor, is three standard deviations away from the jth cluster.

5. The lightweight user authentication method based on device fingerprints according to claim 1, characterized in that: In the third step, the method for measuring abnormal information by using an abnormal calculation model is as follows: Step 31 : For a trace sample x in the trace data set, compute the n nearest neighbors N n (x) of trace sample x in the trace data set, and the n-th nearest distance dist n (y) between trace sample x and trace sample y in the trace data set. Step 32: Calculate the reachable distance Rd(x, y) of two trace samples x, y in the trace dataset according to the nth nearest distance distn(x, y) n (y), the calculation formula is as follows: Rd(x, y) = max {dist(x, y), dist n (y)}; where dist(x, y) represents the Euclidean distance between two trace samples x and y. Step 33: Calculate the local reachable density p of the trace sample x according to the reachable distance Rd(x, y), which is calculated as follows: n (x), which is calculated as follows: where n is the nth nearest neighbor of x, represents the inverse of the average reachable distance of trace sample x from a trace sample y selected from the nearest neighbors of trace sample x. Step 34: Local reachable density p based on n (x), the ratio of the local reachable density of the trace sample x and the trace sample y corresponding to n in its nearest neighbor n (x), which measures the outlying degree of x and its nearest neighbor, and its calculation formula is as follows: where p n (y) represents the local reachable density of a selected one of the trace samples y and its nearest neighbors. Step 35: Calculate the local out-of-core density ratio LOF n (x) as an outlier of the trace sample x, and arrange the plurality of outliers in order to obtain the outlier information.

6. A lightweight user authentication method based on device fingerprints, characterized in that: It includes the following contents: Collecting trace samples of static random access memory (SRAM) to construct a trace data set for representing device fingerprints; The method for constructing a trace data set is as follows: Step 11: Determine the static random access memory (SRAM) of all user devices in the collection layer to be authenticated; Step 12: Collect the power-on static random access memory (SRAM) trace samples from each static random access memory (SRAM) by turning on and off the power supply; Step 13: Assemble the trace samples to obtain a trace data set; Perform clustering processing on the trace data set, and select the trace objects deviating from the clustering center as the trace abnormal candidate set; Based on the local outlier factor, the abnormal values in the trace abnormal candidate set are measured; Based on the abnormal values, the authentication of lightweight users is completed; The method for completing the authentication of lightweight users is as follows: Step 41: Sort the trace samples in the trace abnormal data set according to the size of the abnormal values in the abnormal information, and obtain user ranking data; Step 42: Based on the user ranking data, remove the users with high average ranking to obtain the authorized user authentication result.

7. A lightweight user authentication system based on device fingerprints, characterized in that: Any one of the lightweight user authentication methods based on device fingerprints according to claims 1-6 is used, which includes a collection layer, an access layer, and an application layer; The collection layer is used to collect node data of distributed energy, transmission network equipment, distribution network equipment, and terminal users, and deploy Internet of Things nodes and trace extraction modules; The access layer is provided with a 4G or / and 5G base station for transmitting node data; The application layer is a power grid service cloud for processing node data, which deploys an abnormal calculation module based on a clustering coupling model and an outlier detection, and a user authentication module capable of removing abnormal users; The node data collected by the Internet of Things nodes is accessed to the power grid service cloud through the 4G or / and 5G base station, and the abnormal value detection and abnormal user removal are performed to realize the authentication of lightweight users.

8. The lightweight user authentication system based on device fingerprints according to claim 7, characterized in that: The trace extraction module is used to extract the static random access memory (SRAM) traces of all user devices in the collection layer, and collect the power-on static random access memory (SRAM) trace samples from each static random access memory (SRAM) by turning on and off the power supply; Or / and, the abnormal calculation module is used to cluster all trace samples by using a clustering coupling model, select trace objects deviating from the clustering center as a trace abnormal candidate set, and measure the abnormal values in the trace abnormal candidate set by using a local outlier factor. Or / and, a user authentication module is used to obtain the user ranking at the detection time by using the outlier detection, if the average ranking of the detection is always in the front, the user corresponding to the outlier is eliminated, and the authentication of the user is completed; Or / and, the node data at least includes operation data and energy consumption data; Or / and, the Internet of Things node includes a smart device or / and a smart meter.

Images