Information traceability and classification method for same CANID (Controller Area Network Identity) signal and related equipment thereof

Abstract

The invention relates to the technical field of signal traceability and classification, and provides an information traceability and classification method of the same CANID signal and related equipment thereof. Physical layer signal acquisition is carried out through a synchronization time slot distribution table, so that acquired multi-source signals are preprocessed to generate a standardized signal data set, and then signal fingerprint feature extraction is carried out on the standardized signal data set to form a multi-dimensional feature matrix; and meanwhile, independent source signals are extracted from the multi-dimensional feature matrix through mixed signal blind source separation for evaluation, and finally, signal evaluation results are classified through a deep learning algorithm to obtain identity probability distribution of the equipment, so that the identity probability distribution of the equipment is obtained through data binding and a block chain evidence storage technology based on the probability distribution of the identity of the equipment. And the non-tampering and traceability of the equipment identity information are ensured. According to the method, through global time synchronization, signal fingerprint feature extraction, blind source separation and deep learning classification, accurate classification and traceability of multiple detection signals of the same CANID device are realized.

Description

Technical Field

[0001] This application relates to the technical field of signal traceability and classification, and in particular to a method for information traceability and classification of signals with the same CAN ID and related devices thereof. Background Art

[0002] Controller Area Network (CAN) communication is an efficient and reliable serial communication protocol commonly used in automobiles, industrial automation, and embedded systems. Especially in the in-vehicle key electronic components of electric vehicles, CAN communication plays a crucial role. During the production process of electric vehicles, multiple electronic modules need to generate and transmit CAN messages during the aging test. Since most modules use the same CAN ID, it is difficult to distinguish the messages of different modules on the same CAN bus.

[0003] To address the above problems, it is usually necessary to randomly assign different CAN IDs to devices again and set up dedicated communication hardware for each device (for example, a load with an adjustable CAN ID or a communication adapter module with multiple IO ports), so as to classify and trace signals of multi-source messages using the same CAN bus through different ID tags or control signal sources.

[0004] In the existing solutions, random CAN ID assignment requires users to manually distinguish different modules through the polling function, and setting up dedicated communication hardware for each device will increase the development and production costs. At the same time, the existing solutions have poor scalability and cannot adapt to the expansion of production scale. Summary of the Invention

[0005] In view of this, this application provides a method for information traceability and classification of signals with the same CAN ID and related devices thereof, so as to solve the problem of signal classification and traceability of multi-source signals with the same CAN ID.

[0006] The first aspect of this application provides a method for information traceability and classification of signals with the same CAN ID, and the method includes: Performing global time synchronization and time slot allocation on multiple CAN ID device networks to obtain a synchronized time slot allocation table; Performing physical layer signal acquisition according to the synchronized time slot allocation table, and preprocessing the obtained multi-source signals to obtain a standardized signal data set; Performing signal fingerprint feature extraction on the standardized signal data set to obtain a multi-dimensional feature matrix; Performing blind source separation of mixed signals on the multi-dimensional feature matrix to obtain an evaluation result of independent source signals; Perform deep learning classification on the evaluation results of the independent source signals to obtain the device identity probability distribution corresponding to each signal; Perform data binding and blockchain evidence storage on the device identity probability distribution to obtain an immutable traceability database.

[0007] In an optional implementation manner, the global time synchronization and time slot allocation for multiple CANID device networks to obtain a synchronized time slot allocation table includes: Broadcast synchronization signals to multiple CANID device networks according to a preset time protocol to obtain global time synchronization result data; Calculate the clock deviation of each CANID device according to the global time synchronization result data to obtain the calibration value of each CANID device; Divide the communication cycles of multiple CANID devices according to the calibration value, and dynamically calculate the time slot length of each CANID device according to the communication cycle, the number of CANID devices, and the payload and bit rate of a single-frame CAN message to obtain the dynamically adjusted time slot length; Optimize the time slot length to obtain a hash value, and allocate corresponding time slots to each CANID device according to the hash value to generate the synchronized time slot allocation table.

[0008] In an optional implementation manner, the physical layer signal acquisition according to the synchronized time slot allocation table and the preprocessing of the obtained multi-source signals to obtain a standardized signal data set includes: Perform real-time acquisition on each CANID device according to the synchronized time slot allocation table to obtain the original physical layer signals; Extract background noise from the original physical layer signals to obtain a dynamic threshold; Perform denoising and standardization processing on the original physical layer signals according to the dynamic threshold to obtain the standardized signal data set.

[0009] In an optional implementation manner, the signal fingerprint feature extraction for the standardized signal data set to obtain a multi-dimensional feature matrix includes: Perform time-domain analysis on the standardized signal data set to obtain time-domain eigenvalue; Perform message interval analysis on the standardized signal data set to obtain message interval statistical features; Perform fast Fourier transform on the standardized signal data set to obtain signal spectrum information, and calculate the dominant frequency and frequency band energy distribution of the signal spectrum information to obtain frequency-domain eigenvalue; Based on the time-domain eigenvalue and the frequency-domain eigenvalue, the packet interval statistical feature and the band energy obtained by calculating the band energy distribution are combined to obtain the multi-dimensional feature matrix.

[0010] In an alternative embodiment, the blind source separation of the mixed signals from the multi-dimensional feature matrix to obtain the independent source signal evaluation result includes: Model the mixed signals for the multi-dimensional feature matrix to obtain a mixed signal matrix; Perform independent component analysis on the mixed signal matrix to obtain an estimated separation matrix; Perform blind source separation processing on the mixed signal matrix according to the estimated separation matrix to obtain a source signal matrix; Calculate the eigenvalues and statistics of the source signal matrix to identify the characteristics of each CAN ID device signal, and summarize the identified characteristics to generate the independent source signal evaluation result.

[0011] In an alternative embodiment, the deep learning classification of the independent source signal evaluation result to obtain the device identity probability distribution corresponding to each signal includes: Extract the time-domain features and frequency-domain features of each signal in the independent source signal evaluation result to obtain a first data subset; Perform normalization processing on the first data subset to obtain a second data subset; Perform device identity authentication of the signals on the second data subset according to a preset convolutional neural network to obtain the device identity probability distribution corresponding to each signal.

[0012] In an alternative embodiment, the data binding and blockchain evidence storage of the device identity probability distribution to obtain an immutable traceability database includes: Perform data binding on the device identity probability distribution to obtain a data packet with device identity information; Perform hash processing and digital signature on the data packet to obtain a signed data packet; Construct a data block in the format of a blockchain with the signed data packet and the time stamp corresponding to the signed data packet, and add the data block to the blockchain; Verify the relevance between the current data block and the data blocks previously stored in the blockchain according to the hash chain structure in the blockchain to obtain an immutable traceability database.

[0013] The second aspect of the present application provides an information traceability and classification device for the same CAN ID signal, and the device includes: The time slot allocation module is used to perform global time synchronization and time slot allocation for multiple CAN ID device networks to obtain a synchronized time slot allocation table; The signal standard module is used to collect physical layer signals according to the synchronized time slot allocation table and preprocess the obtained multi-source signals to obtain a standardized signal data set; The feature matrix module is used to extract signal fingerprint features from the standardized signal data set to obtain a multi-dimensional feature matrix; The single-source evaluation module is used to perform blind source separation of mixed signals on the multi-dimensional feature matrix to obtain an independent source signal evaluation result; The probability distribution module is used to perform deep learning classification on the independent source signal evaluation result to obtain the device identity probability distribution corresponding to each signal; The traceability data module is used to perform data binding and blockchain evidence storage on the device identity probability distribution to obtain an immutable traceability database.

[0014] A third aspect of the present application provides an electronic device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it implements the steps of the information traceability and classification method for the same CAN ID signal as described above.

[0015] A fourth aspect of the present application provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the information traceability and classification method for the same CAN ID signal as described above.

[0016] In summary, the present application at least includes the following beneficial technical effects: 1. Through the extraction of the multi-dimensional feature matrix and the comprehensive analysis in the time domain and frequency domain, the unique fingerprint features of the signal are carefully captured. By combining features such as message interval, spectrum information, and band energy, the identification and classification of signals are made more accurate, providing rich information for the device identity verification of signals.

[0017] 2. Blind source separation of mixed signals effectively extracts independent source signals from multiple signal sources, avoiding signal confusion and interference. Through independent component analysis processing, the signal characteristics of each device can be accurately identified in a complex environment, ensuring the accuracy and independence of signal evaluation.

[0018] 3. Using deep learning methods and combining time domain and frequency domain features for device identity classification can achieve efficient and accurate device identity authentication. Description of the Drawings

[0019] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, the accompanying drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other accompanying drawings can also be obtained based on these drawings.

[0020] Figure 1 is a flowchart of a method for information traceability and classification of the same CAN ID signal provided by an embodiment of the present application; Figure 2 is a functional module diagram of a device for information traceability and classification of the same CAN ID signal provided by an embodiment of the present application; Figure 3 is a schematic structural diagram of an electronic device provided by an embodiment of the present application. Detailed implementation manners

[0021] The following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present application.

[0022] The method for information traceability and classification of the same CAN ID signal provided by the embodiment of the present application is executed by a device for information traceability and classification of the same CAN ID signal. Among them, the device for information traceability and classification of the same CAN ID signal includes, but is not limited to, an upper master device, a lower device (for connecting the parts to be detected and obtaining corresponding signals according to the upper master device), and a CAN communication line connecting the upper master device and the lower device. The following will illustrate the method for information traceability and classification of the same CAN ID signal provided by the embodiment of the present application from the perspective of the device for information traceability and classification of the same CAN ID signal in combination with the multi-source signal processing process of the same CAN ID in the aging chamber.

[0023] As Figure 1 shown, it is a flowchart of the method for information traceability and classification of the same CAN ID signal provided by the embodiment of the present application. The method for information traceability and classification of the same CAN ID signal provided by the embodiment of the present application includes the following steps.

[0024] Step S1: Perform global time synchronization and time slot allocation for multiple CAN ID device networks to obtain a synchronized time slot allocation table.

[0025] It should be understood that the preset time protocol adopted in the embodiments of this application is the IEEE-1588 Precision Time Protocol (hereinafter collectively referred to as the PTP protocol). In a device network composed of multiple lower-level devices, the upper-level master device serves as the master clock, and the lower-level devices serve as slave devices. The upper-level master device broadcasts a synchronization signal to send its own system time to all lower-level devices. Among them, each synchronization signal contains the timestamp information of the upper-level master device. After receiving the synchronization signal, the lower-level device records its own reception timestamp, and uses the timestamp information of the upper-level master device and its own reception timestamp as the global time synchronization result data.

[0026] After obtaining the global time synchronization result data, the lower-level device calculates the delay between the timestamp sent by the upper-level master device and its own reception timestamp through the following formula, so as to estimate its own clock deviation (i.e., calibration value).

[0027] Among them, is the timestamp sent by the upper-level master device, is the reception timestamp of the lower-level device, is the network transmission delay, is the clock deviation. After obtaining the calibration value, the lower-level device can perform clock correction through the compensation formula, so as to ensure that the information traceability and classification devices of the same CANID signal have consistent clocks after synchronization.

[0028] While the lower-level device is performing clock correction, the upper-level master device divides the communication cycle according to the master clock and the load situation in the device network, which helps to avoid different devices sending data at the same time, thus avoiding signal conflicts.

[0029] Specifically, the upper-level master device divides the total communication cycle according to the communication requirements (i.e., the sum of the communication cycles of all lower-level devices). The upper-level master device calculates the slot length of each lower-level device through the following formula according to the communication cycle and the total number of lower-level devices.

[0030] Among them, is the communication cycle, is the total number of lower-level devices, is the length of the effective payload of a single-frame CAN message, is the bit rate of the CAN bus. By calculating the slot length of each lower-level device, it is ensured that the lower-level device can exclusively occupy the CAN bus channel to transmit data within its specified slot.

[0031] To further optimize time slot allocation and ensure that there are no conflicts between subordinate devices, the master device uses the hash value of the subordinate device ID for time slot allocation. According to the ID information of the subordinate devices, the hash value of each subordinate device is calculated through a hash algorithm (e.g., SHA256). This hash value serves as the basis for time slot allocation. The master device sorts the time slots based on the hash values of the subordinate devices and assigns a unique time slot to each subordinate device, ensuring that each subordinate device can transmit data within its own time slot. At the same time, the master device generates a synchronized time slot allocation table based on the sorted time slots. The synchronized time slot allocation table contains the ID of each device and the corresponding time slot. After obtaining the synchronized time slot allocation table, the master device broadcasts the synchronized time slot allocation table to all subordinate devices, enabling the subordinate devices to transmit data based on the synchronized time slot allocation table.

[0032] Step S2: Collect physical layer signals according to the synchronized time slot allocation table and preprocess the obtained multi-source signals to obtain a standardized signal dataset.

[0033] During the aging detection of components, according to the synchronized time slot allocation table, the master device uses the subordinate devices to perform real-time acquisition on the device to be detected (i.e., the CANID device), thereby obtaining multiple raw physical layer signals with the same CANID. It should be understood that the raw physical layer signals include the timestamps collected by the subordinate devices. Since in step S1, the subordinate devices have completed clock calibration, the raw physical layer signals collected by the master device through the subordinate devices have completed time domain synchronization.

[0034] During the detection of multiple subordinate devices simultaneously, due to signal interference between subordinate devices and the influence of the electromagnetic environment, the raw physical layer signals often contain background noise. Before signal analysis, the master device must extract the background noise and set a dynamic noise threshold according to the statistical characteristics of the noise to denoise the raw physical layer signals.

[0035] The process of extracting background noise by the master device is usually carried out during the silent period when the device does not send data. At this time, the signals on the CAN bus only contain environmental noise and there is no effective data transmission. By sampling the signals during this period, the standard deviation of the background noise can be calculated, and then a dynamic threshold can be set for subsequent denoising processing. Specifically, the master device samples the signals during the silent period and calculates its standard deviation to obtain the distribution characteristics of the background noise. According to the statistical information of the background noise, a dynamic noise threshold (i.e., the dynamic threshold) can be set to filter the noise in the raw physical layer signals. The specific formula is as follows: Where is the standard deviation of the background noise. is the signal sample within the window W, used to estimate the local fluctuation of the original physical layer signal. is the dynamic noise threshold at the k-th sampling point.

[0036] After obtaining the dynamic noise threshold, the upper master device removes the random noise in the original physical layer signal and retains the effective signal components. The denoising process uses a dynamic threshold to distinguish noise and effective signals. When the amplitude of the original physical layer signal exceeds the set dynamic threshold, the original physical layer signal is considered an effective signal; otherwise, it is regarded as noise. The upper master device further standardizes the denoised original physical layer signal to eliminate the amplitude difference of the original physical layer signal. The standardization process adjusts the mean of the original physical layer signal to zero and the standard deviation to one, so that the standardized signal dataset (i.e., the dataset obtained after standardizing the original physical layer signal) can obtain a unified scale in subsequent analysis. The standardization formula is as follows: where, is the denoised original physical layer signal, is the mean of the denoised original physical layer signal, is the standard deviation of the denoised original physical layer signal.

[0037] Step S3: Extract signal fingerprint features from the standardized signal dataset to obtain a multi-dimensional feature matrix.

[0038] It should be understood that signal fingerprint feature extraction is to extract the feature information reflecting the CANID device and its working state from the signal data. The fingerprint features of the signal usually consist of time-domain features, frequency-domain features, message interval statistical features, etc.

[0039] The upper master device extracts the rise time and fall time of the signal from the standardized signal dataset. The rise time and fall time can reflect the waveform characteristics of the corresponding signal, especially the speed and stability when the signal changes. Among them, the rise time is the time for the signal to change from 10% to 90%, and the fall time is the time for the signal to change from 90% to 10%. The rise time and fall time, as time-domain features, can reflect the response speed and electrical characteristics of the CANID device drive circuit. For example, shorter rise and fall times indicate that the device has a higher response ability in signal transmission.

[0040] At the same time, the upper master device analyzes the time interval between every two adjacent signal messages in the standardized signal dataset, calculates the interval time of the signal messages, and analyzes all the message interval times to calculate their mean and variance. The specific calculation formulas are as follows: Among them, N is the number of signal message intervals, is the time interval of the k-th message, is the mean value of the signal message intervals, is the variance of the signal message intervals. By analyzing the mean value and variance of the message intervals, the timing stability of the CANID device when sending data can be identified. Abnormal interval times may indicate that there are faults or interferences in the device.

[0041] Meanwhile, the upper master device performs a fast Fourier transform on the standardized signal data set to obtain signal spectrum information. Among them, the signal spectrum information contains the frequency distribution information of the signal. Thus, based on the signal spectrum information, the upper master device calculates the dominant frequency and the frequency band energy distribution of the standardized signal data. The specific formulas are as follows: Among them, is the signal spectrum information, is the dominant frequency. The dominant frequency reflects the main frequency components in the signal and is a key feature of the signal in the frequency domain.

[0042] Among them, , are the start frequency and end frequency of the selected frequency band, is the energy within the frequency band. The selected frequency band is usually selected according to the operating frequency band of the device and can reflect the specific frequency domain characteristics of the device. Calculate the energy accumulation of the signal spectrum information within a specific frequency band to describe the energy distribution of the signal in different frequency bands.

[0043] Furthermore, the rise time and fall time characteristics obtained from the time domain analysis, the mean value and variance characteristics obtained from the message interval analysis, and the dominant frequency and frequency band energy characteristics obtained from the frequency domain analysis are combined to form a multi-dimensional feature matrix. Each row in the multi-dimensional feature matrix represents the complete feature data of a CANID device, including multiple features of the time domain, frequency domain, and message interval statistics. The dimension of the column vector is the number of feature dimensions.

[0044] Step S4: Perform blind source separation of mixed signals on the multi-dimensional feature matrix to obtain an independent source signal evaluation result.

[0045] On the CAN bus, the data of multiple lower devices are transmitted in the form of differential signals through a shared communication medium. Due to the number of lower devices and the diversity of signals, it is possible that the signals of multiple lower devices will be linearly superimposed on the CAN bus to form a mixed signal.

[0046] The upper master device models the multi-dimensional feature matrix to describe the linear mixing of signals between devices. Specifically, the signals of each lower device are independent source signals, and are combined through a mixing matrix to finally form the observed mixed signal. This mixing process can be described by the following mathematical model: where \(X\) is the mixed signal matrix, representing the signals observed on multiple acquisition channels (such as multiple ADC channels). \(A\) is the mixing matrix, which describes the linear relationship between the source signals and the mixed signals. \(S\) is the source signal matrix, and each column represents the independent signal of a CANID device. \(N\) is the noise matrix, representing the inevitable noise interference in the mixing process. And is the observed signal matrix, where \(M\) is the number of acquisition channels and \(T\) is the number of sampling points. is the mixing matrix, where \(N\) is the number of signal sources, usually equal to the number of lower devices. is the source signal matrix, and each column represents the signal of a CANID device.

[0047] Based on the established mixed signal matrix, the upper master device can perform independent component analysis, thereby estimating a separation matrix through a mathematical algorithm, and then recovering the source signals from the observed signals through the separation matrix. The specific formula for independent component analysis is as follows: where is a non-linear function used to measure non-Gaussianity. is the expectation operation, representing the statistical analysis of the signal. is the regularization parameter, used to constrain the orthogonality of the matrix \(W\) to prevent overfitting. is the Frobenius norm, representing the orthogonality constraint of the separation matrix. \(I\) is the identity matrix, ensuring that the separation matrix is orthogonal. By maximizing the non-Gaussianity of the source signals, the separation of the source signals is achieved.

[0048] The upper master device uses the estimated separation matrix to perform blind source separation processing on the mixed signal matrix, thereby obtaining the source signal matrix. The specific process can be represented by where is the estimated source signal matrix. By separating the independent signals of each CANID device from the mixed signals, an accurate data basis is provided for subsequent signal analysis and classification.

[0049] After the source signal separation is completed, the next step is to extract the features of the source signal matrix, including the calculation of the eigenvalues and statistics of the signal. The process of extracting features from the source signal matrix is similar to the signal fingerprint feature extraction, which will not be elaborated here. For details, please refer to the signal fingerprint feature extraction process in step S3. By extracting the features of the source signal matrix, an accurate independent source signal evaluation result can be provided for each CANID device signal.

[0050] Step S5: Perform deep learning classification on the independent source signal evaluation results to obtain the device identity probability distribution corresponding to each signal.

[0051] The upper master device summarizes the time-domain features and frequency-domain features of each signal from the obtained independent source signal evaluation results to obtain a first data subset. To ensure that the numerical features of different signals have the same scale and avoid the influence of data with larger eigenvalues on the training effect of the neural network, the upper master device must standardize the time-domain and frequency-domain features in the first data subset to ensure that all features have the same mean and variance and eliminate the scale differences between signal features. The conversion is specifically carried out through the following formula: where, is a certain feature of signal k, is the mean of feature and is the standard deviation of feature and is the standardized feature. Through the standardization process, it is ensured that all signal features are in a unified scale interval, avoiding the weight bias in the network training process.

[0052] The upper master device takes the standardized second data subset as the input and uses a convolutional neural network in deep learning for classification. At this stage, the convolutional neural network will extract the high-order features of the second data subset according to the pre-trained method and output the device identity probability distribution corresponding to each signal through the classification head.

[0053] Specifically, after inputting the standardized second data subset, feature extraction is performed through multiple convolutional layers. The convolutional layer captures the local patterns in the signal by sliding the convolutional kernel window. Each convolutional operation extracts information in different local regions of the signal and gradually obtains higher-level features. The output of the convolutional layer is a feature map for further classification. The convolutional operation can be expressed by the following formula: where, is the input second data subset, is the convolutional kernel weight, is the convolutional layer bias, is the output feature map of the convolutional layer. The features extracted by the convolutional layer will pass through the pooling layer for global average pooling, thereby converting the feature map into a feature vector of a fixed length. Then, through the fully connected layer, the feature vector is mapped into the classification space of the device identity, and the classification probability of the device is output. In the fully connected layer, the input feature vector will undergo a linear transformation through the weight matrix and the bias term. The formula for the linear transformation is as follows: Among them, is the weight matrix of the fully connected layer, is the bias term of the fully connected layer, is the output probability vector of the fully connected layer, is the input feature vector. Finally, after being processed by the Softmax function, the identity probability distribution of the device corresponding to each signal is output. The Softmax function converts the score of each category into a probability value, and the sum of these probability values is 1. The Softmax function can be expressed as the following formula: Among them, is the output score of the i-th category, is the probability of the i-th category. Finally, the network generates the corresponding device identity probability distribution according to the signal characteristics of the device.

[0054] Step S6: Perform data binding and blockchain evidence storage on the device identity probability distribution to obtain an immutable traceability database.

[0055] The upper-level master device binds the device identity probability distribution to the original CAN message data (i.e., the original physical layer signal) to generate a data packet with device identity information. Specifically, the device identity probability distribution is combined with the original physical layer signal to form a new data packet with device identity information. The data packet with device identity information includes, but is not limited to, the device identity classification probability distribution, the timestamp collected by the lower-level device, and the original physical layer signal. Further, the upper-level master device performs a hash process on the data packet with device identity information to generate a unique hash value. For example, an encryption hash algorithm such as SHA-256 is used to process the data packet to obtain the hash value. Since the same input data always produces the same hash value and the probability of different input data producing the same hash value is extremely low, the hash value can be used as the unique identifier of the data packet. And external devices cannot reverse-engineer the original data from the hash value, ensuring the security of the data packet. Thus, the upper-level master device encrypts the hash value using the private key of the device to generate a signature value. A legitimate device can verify the signature value using the public key of the device to ensure the validity of the signature and the integrity of the data. The digital signature ensures the authenticity of the source and the immutability of the data packet. Furthermore, the upper-level master device constructs a data block in the format of a blockchain with the signed data packet and the corresponding collection timestamp in the packet. Each data block includes, but is not limited to, the device identity probability distribution, the hash value of the original physical layer signal linking to the previous data block, the hash value of the current data packet, the signature value of the current data packet, the time identifier of the data block generation, and the hash value of the current data block for linking to the next data block. The upper-level master device adds the constructed data block to the blockchain. Among them, the blockchain adopts a decentralized distributed ledger technology to ensure the immutability and transparency of the data. Each node holds a complete copy of the blockchain, and any modification to the data requires network consensus to ensure the security of the data. To ensure the integrity and consistency of the data, the upper-level master device also matches the hash value of the current data block with the hash value of the previous data block to verify the relevance between the current data block and the data block previously stored in the blockchain. If they match, it means the data block has not been tampered with and the verification passes; if they do not match, it means the data block may have been tampered with and the verification fails. Any tampering with the data will result in a hash value mismatch, which can be detected in a timely manner. Thus, it is ensured that the verified blockchain is an immutable traceability database.

[0056] When a legitimate device obtains the data block in the blockchain from the information traceability and classification device of the same CANID signal in the embodiment of the present application, it can determine the identity of the CANID device of the original physical layer signal according to the probability ranking of each predicted device in the device identity probability distribution. Exemplarily, if the device identity probability distribution output by the model is 0.1, 0.85, 0.05, it is considered that the signal comes from the device with a probability of 0.85.

[0057] This application is applied to the field of signal traceability and classification technology. By synchronizing the time slot allocation table, physical layer signal acquisition is performed, and then the collected multi-source signals are preprocessed to generate a standardized signal data set. Furthermore, signal fingerprint feature extraction is performed on the standardized signal data set to form a multi-dimensional feature matrix. At the same time, independent source signals are extracted from the multi-dimensional feature matrix through mixed signal blind source separation for evaluation. Finally, a deep learning algorithm is used to classify the signal evaluation results to obtain the identity probability distribution of the device. Based on the probability distribution of the device identity, through data binding and blockchain evidence storage technology, the immutability and traceability of the device identity information are ensured. This application realizes the accurate classification and traceability of a multi-source signal set of detection signals from multiple devices with the same CAN ID through global time synchronization, signal fingerprint feature extraction, blind source separation, and deep learning classification.

[0058] As Figure 2 shown, it is a functional module diagram of an information traceability and classification device for signals with the same CAN ID provided by an embodiment of this application.

[0059] In some embodiments, the information traceability and classification device 2 for signals with the same CAN ID may include multiple functional modules composed of computer program segments. The computer programs of each program segment in the information traceability and classification device 2 for signals with the same CAN ID can be stored in the memory of the server and executed by at least one processor to execute (see details in Figure 1 the description) the functions of the information traceability and classification method for signals with the same CAN ID.

[0060] In this embodiment, according to the functions it executes, the information traceability and classification device 2 for signals with the same CAN ID can be divided into multiple functional modules. The functional modules may include: a time slot allocation module 21, a signal standard module 22, a feature matrix module 23, an independent source evaluation module 24, a probability distribution module 25, and a traceability data module 26. The module referred to in the present invention means a series of computer program segments that can be executed by at least one processor and can complete fixed functions, and are stored in the memory. In this embodiment, the functions of each module will be described in detail in subsequent embodiments.

[0061] The time slot allocation module 21 is used to perform global time synchronization and time slot allocation for multiple CAN ID device networks to obtain a synchronized time slot allocation table.

[0062] In an alternative embodiment, the time slot allocation module 21 is specifically used for: Performing synchronous signal broadcasting on multiple CAN ID device networks according to a preset time protocol to obtain global time synchronization result data; Calculate the clock deviation of each CAN ID device according to the global time synchronization result data to obtain the calibration value of each CAN ID device; Divide the communication cycles of multiple CAN ID devices according to the calibration value, and dynamically calculate the time slot length of each CAN ID device according to the communication cycle, the number of CAN ID devices, and the payload and bit rate of a single-frame CAN message to obtain the dynamically adjusted time slot length; Optimize the time slot length to obtain a hash value, and allocate corresponding time slots to each CAN ID device according to the hash value to generate the synchronized time slot allocation table.

[0063] The signal standard module 22 is used to collect physical layer signals according to the synchronized time slot allocation table, and preprocess the obtained multi-source signals to obtain a standardized signal data set.

[0064] In an optional implementation manner, the signal standard module 22 is specifically used for: Perform real-time collection on each CAN ID device according to the synchronized time slot allocation table to obtain the original physical layer signal; Extract the background noise of the original physical layer signal to obtain a dynamic threshold; Perform denoising and standardization processing on the original physical layer signal according to the dynamic threshold to obtain the standardized signal data set.

[0065] The feature matrix module 23 is used to extract signal fingerprint features from the standardized signal data set to obtain a multi-dimensional feature matrix.

[0066] In an optional implementation manner, the feature matrix module 23 is specifically used for: Perform time-domain analysis on the standardized signal data set to obtain time-domain feature values; Perform message interval analysis on the standardized signal data set to obtain message interval statistical features; Perform fast Fourier transform on the standardized signal data set to obtain signal spectrum information, and calculate the dominant frequency and frequency band energy distribution of the signal spectrum information to obtain frequency-domain feature values; According to the time-domain feature values and the frequency-domain feature values, merge the message interval statistical features and the frequency band energy obtained by calculating the frequency band energy distribution to obtain the multi-dimensional feature matrix.

[0067] The single-source evaluation module 24 is used to perform blind source separation of mixed signals on the multi-dimensional feature matrix to obtain an independent source signal evaluation result.

[0068] In an optional implementation manner, the single-source evaluation module 24 is specifically configured to: Perform mixed-signal modeling on the multi-dimensional feature matrix to obtain a mixed-signal matrix; Perform independent component analysis on the mixed-signal matrix to obtain an estimated separation matrix; Perform blind source separation processing on the mixed-signal matrix according to the estimated separation matrix to obtain a source signal matrix; Calculate eigenvalues and statistics of the source signal matrix to identify the characteristics of each CANID device signal, and summarize the identified characteristics to generate the single-source signal evaluation result.

[0069] The probability distribution module 25 is configured to perform deep learning classification on the single-source signal evaluation result to obtain the device identity probability distribution corresponding to each signal.

[0070] In an optional implementation manner, the probability distribution module 25 is specifically configured to: Extract the time-domain characteristics and frequency-domain characteristics of each signal in the single-source signal evaluation result to obtain a first data subset; Perform normalization processing on the first data subset to obtain a second data subset; Perform device identity authentication on the second data subset according to a preset convolutional neural network to obtain the device identity probability distribution corresponding to each signal.

[0071] The traceability data module 26 is configured to perform data binding and blockchain evidence preservation on the device identity probability distribution to obtain an immutable traceability database.

[0072] In an optional implementation manner, the traceability data module 26 is specifically configured to: Perform data binding on the device identity probability distribution to obtain a data packet with device identity information; Perform hash processing and digital signature on the data packet to obtain a signed data packet; Construct a data block in the format of a blockchain with the signed data packet and the timestamp corresponding to the signed data packet, and add the data block to the blockchain; Verify the relevance between the current data block and the data blocks previously stored in the blockchain according to the hash chain structure in the blockchain to obtain an immutable traceability database.

[0073] It should be understood that the various variations and specific embodiments of the methods provided in the above embodiments are equally applicable to the information traceability and classification device for the same CANID signal in this embodiment. Through the detailed description of the information traceability and classification method for the same CANID signal above, those skilled in the art can clearly know the implementation method of the information traceability and classification device for the same CANID signal in this embodiment. For the sake of brevity of the specification, it will not be elaborated here.

[0074] As Figure 3 shown, it is a schematic structural diagram of an electronic device provided by an embodiment of the present application.

[0075] In a preferred embodiment of the present invention, the electronic device 3 may include, but is not limited to: a memory 31, at least one processor 32, and at least one communication bus 33.

[0076] Those skilled in the art should understand that Figure 3 the structure of the electronic device 3 shown does not constitute a limitation on the embodiments of the present invention. The electronic device 3 may also include more or fewer other hardware or software than shown, or different component arrangements.

[0077] In some embodiments, the electronic device 3 is a device capable of automatically performing numerical calculations and / or information processing according to pre-set or stored instructions. Its hardware includes, but is not limited to, microprocessors, application-specific integrated circuits, programmable gate arrays, digital signal processors, and embedded devices, etc.

[0078] It should be noted that the electronic device 3 is only an example. Other existing or future possible electronic products that can be adapted to the present application should also be included within the protection scope of the present application and are included herein by reference.

[0079] In some embodiments, a computer program is stored in the memory 31. When the computer program is executed by the at least one processor 32, all or some of the steps in the information tracing and classification method of the same CAN ID signal as described above are implemented. The memory 31 includes a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electrically-erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CD-ROM), or other optical disc memories, magnetic disk memories, tape memories, or any other computer-readable medium capable of carrying or storing data. Further, the computer-readable storage medium mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system, application programs required for at least one function, and the like.

[0080] In some embodiments, the at least one processor 32 is the control core (Control Unit) of the electronic device 3, connecting various components of the entire electronic device 3 through various interfaces and circuits. By running or executing the programs or modules stored in the memory 31, and by calling the data stored in the memory 31, various functions of the electronic device 3 are executed and data is processed. For example, when the at least one processor 32 executes the computer program stored in the memory 31, all or some of the steps in the information tracing and classification method of the same CAN ID signal in the embodiments of the present application are implemented; or all or some of the functions of the information tracing and classification device of the same CAN ID signal are implemented. The at least one processor 32 can be composed of integrated circuits. For example, it can be composed of a single packaged integrated circuit, or can be composed of multiple integrated circuits with the same or different functions packaged, including a combination of one or more central processing units (CPUs), microprocessors, digital processing chips, graphics processors, and various control chips.

[0081] In some embodiments, the at least one communication bus 33 is configured to enable connection communication between the memory 31, the at least one processor 32, and the like. Although not shown, the electronic device 3 may further include a power source (such as a battery) for powering each component. Preferably, the power source may be logically connected to the at least one processor 32 through a power management device, so as to implement functions such as management of charging, discharging, and power consumption management through the power management device. The power source may further include any components such as one or more DC or AC power sources, a recharge device, a power failure detection circuit, a power converter or inverter, a power status indicator, etc. The electronic device 3 may further include various sensors, a Bluetooth module, a Wi-Fi module, etc., which will not be elaborated herein.

[0082] The integrated units implemented in the form of software functional modules as described above may be stored in a computer-readable storage medium. The above-mentioned software functional modules are stored in a storage medium and include several instructions for causing an electronic device (which may be a personal computer, an electronic device, or a network device, etc.) or a processor to execute a part of the methods described in various embodiments of the present application.

[0083] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods may be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and there may be other division methods in actual implementation.

[0084] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical units. They may be located in one place or distributed to multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0085] The above are all preferred embodiments of the present application. Without limiting the protection scope of the present application accordingly, therefore, any equivalent changes made according to the structure, shape, and principle of the present application shall be covered within the protection scope of the present application.

Claims

1. A method for tracing and classifying information of the same CANID signal, characterized in that: The method comprises: Perform global time synchronization and time slot allocation for multiple CANID device networks to obtain a synchronized time slot allocation table; Performing physical layer signal acquisition according to the synchronization time slot allocation table, and preprocessing the acquired multi-source signals to obtain a standardized signal data set; Extracting signal fingerprint features from the standardized signal data set to obtain a multi-dimensional feature matrix; Performing mixed signal blind source separation on the multidimensional feature matrix to obtain an independent source signal evaluation result; Performing deep learning classification on the independent source signal evaluation results to obtain a device identity probability distribution corresponding to each signal; The probability distribution of the device identity is data-bound and stored in the blockchain to obtain an unalterable traceability database.

2. The information tracing and classification method of the same CANID signal according to claim 1 is characterized in that: The performing global time synchronization and time slot allocation on multiple CANID device networks to obtain a synchronized time slot allocation table comprises: Broadcast synchronization signals to multiple CANID device networks according to a preset time protocol to obtain global time synchronization result data; Calculate the clock deviation of each CANID device according to the global time synchronization result data to obtain the calibration value of each CANID device; Dividing the communication cycle of the plurality of CANID devices according to the calibration value, and dynamically calculating the time slot length of each CANID device according to the communication cycle, the number of CANID devices, and the effective load and bit rate of a single-frame CAN message to obtain a dynamically adjusted time slot length; The time slot length is optimized to obtain a hash value, and a corresponding time slot is allocated to each CANID device according to the hash value to generate the synchronization time slot allocation table.

3. The information tracing and classification method of the same CANID signal according to claim 1 is characterized in that: The step of collecting physical layer signals according to the synchronization time slot allocation table and preprocessing the acquired multi-source signals to obtain a standardized signal data set includes: Perform real-time acquisition of each CANID device according to the synchronization time slot allocation table to obtain the original physical layer signal; Performing background noise extraction on the original physical layer signal to obtain a dynamic threshold; The original physical layer signal is denoised and standardized according to the dynamic threshold to obtain the standardized signal data set.

4. The information tracing and classification method of the same CANID signal according to claim 1 is characterized in that: The extracting signal fingerprint features from the standardized signal data set to obtain a multi-dimensional feature matrix comprises: Performing time domain analysis on the standardized signal data set to obtain time domain feature values; Performing message interval analysis on the standardized signal data set to obtain message interval statistical characteristics; Performing a fast Fourier transform on the standardized signal data set to obtain signal spectrum information, and performing dominant frequency and band energy distribution calculation on the signal spectrum information to obtain frequency domain eigenvalues; According to the time domain eigenvalues ​​and the frequency domain eigenvalues, the message interval statistical characteristics and the frequency band energy obtained by the frequency band energy distribution calculation are combined to obtain the multi-dimensional feature matrix.

5. The information tracing and classification method of the same CANID signal according to claim 1 is characterized in that: The performing mixed signal blind source separation on the multidimensional feature matrix to obtain an independent source signal evaluation result comprises: Performing mixed signal modeling on the multidimensional feature matrix to obtain a mixed signal matrix; Performing independent component analysis on the mixed signal matrix to obtain an estimated separation matrix; Performing blind source separation processing on the mixed signal matrix according to the estimated separation matrix to obtain a source signal matrix; The eigenvalues ​​and statistics of the source signal matrix are calculated to identify the characteristics of each CANID device signal, and the identified characteristics are summarized to generate the independent source signal evaluation result.

6. The information tracing and classification method of the same CANID signal according to claim 1 is characterized in that: The performing deep learning classification on the independent source signal evaluation results to obtain the device identity probability distribution corresponding to each signal includes: Extracting the time domain features and the frequency domain features of each signal in the independent source signal evaluation result to obtain a first data subset; performing standardization processing on the first data subset to obtain a second data subset; The device identity of the signal of the second data subset is authenticated according to a preset convolutional neural network to obtain a device identity probability distribution corresponding to each signal.

7. The information tracing and classification method of the same CANID signal according to claim 1 is characterized in that: The data binding and blockchain storage of the device identity probability distribution to obtain an unalterable traceability database includes: Performing data binding on the device identity probability distribution to obtain a data packet with device identity information; Performing hash processing and digital signing on the data packet to obtain a data packet with a signature; Constructing a data block from the data packet with the signature and the timestamp corresponding to the data packet with the signature in a blockchain format, and adding the data block to the blockchain; According to the hash chain structure in the blockchain, the association between the current data block and the data block previously stored in the blockchain is verified to obtain an unalterable traceability database.

8. An information tracing and classification device for the same CANID signal, characterized in that: The device comprises: A time slot allocation module, used for performing global time synchronization and time slot allocation on multiple CANID device networks to obtain a synchronized time slot allocation table; A signal standard module, used for collecting physical layer signals according to the synchronization time slot allocation table, and preprocessing the acquired multi-source signals to obtain a standardized signal data set; A feature matrix module, used for extracting signal fingerprint features from the standardized signal data set to obtain a multi-dimensional feature matrix; An independent source evaluation module is used to perform mixed signal blind source separation on the multidimensional feature matrix to obtain an independent source signal evaluation result; A probability distribution module, used for performing deep learning classification on the independent source signal evaluation results to obtain a device identity probability distribution corresponding to each signal; The traceability data module is used to bind the device identity probability distribution to the data and store it in the blockchain to obtain an unalterable traceability database.

9. An electronic device, characterized in that: The electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the information tracing and classification method for the same CANID signal according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the information tracing and classification method for the same CANID signal according to any one of claims 1 to 7 are implemented.

Images