Voiceprint authentication protection method and system based on SM2 algorithm

Abstract

The present invention relates to the field of smart home device control technology, and in particular to a voiceprint authentication protection method and system based on the SM2 algorithm, which collects the ambient temperature data and ambient volume data of smart home devices, and uses the SM3 algorithm to generate the pseudo-random large integer required for the digital signature authentication operation of the SM2 algorithm; collects user voice data, uses a pre-trained voiceprint feature model to extract the voiceprint feature vector corresponding to the user voice data, and uses the voiceprint feature vector to generate the public and private key pair required for the digital signature authentication operation of the SM2 algorithm, wherein the public key is transmitted to an FPGA chip for storage; for user voice data, based on the SM2 algorithm and calling the parallel processing operation unit in the FPGA chip to execute the digital signature authentication operation process. The present invention can avoid the threat of leakage of the user's voiceprint features during transmission, effectively protect the user's privacy, is flexible to use, and has efficient and convenient authentication, and is convenient for application in smart home authentication.

Description

Technical Field

[0001] The present invention relates to the field of smart home device control technology, and in particular to a voiceprint authentication protection method and system based on the SM2 algorithm. Background Art

[0002] With the advent of the Internet of Everything (IoE), smart homes are constantly evolving, encompassing both connected home appliance ecosystems like Xiaomi and individual appliances manufactured by various independent brands. While bringing convenience and peace of mind, the security and privacy of smart homes and appliances are drawing increasing attention. Some believe that smart home authentication and control have become a potential security threat, raising questions about whether they can truly meet theft prevention needs and whether the security of recorded data can be guaranteed. Traditionally, some smart home appliances, to demonstrate their responsiveness and high sensitivity, often skip authentication altogether, relying instead on voice control or signature verification based on certain biometrics to achieve security.

[0003] In the future, as people become increasingly dependent on smart technology, they will need to constantly authenticate their identities in their daily lives to protect their security and prevent privacy leaks or control by attackers. Using artificial intelligence to extract variable biometric features for personal identity authentication has become an effective measure, but this also brings the following problems: 1. The biometric features used for authentication are personal privacy, and without corresponding security measures during the AI ​​extraction process, they are prone to leakage. 2. In a large number of authentication activities, hardware authentication processing requires a lot of computing power. However, smaller mobile hardware often has insufficient processing speed due to hardware characteristics. 3. Smart furniture scenarios require high spatial flexibility for authentication. Common face and fingerprint authentication methods require authentication in fixed locations. Therefore, a dynamic and flexible security authentication solution is urgently needed to meet the needs of smart home device use. Summary of the Invention

[0004] To this end, the present invention provides a voiceprint authentication protection method and system based on the SM2 algorithm, which can avoid the threat of leakage of user voiceprint features during transmission, effectively protect the user's privacy, and is flexible, efficient and convenient.

[0005] According to the design scheme provided by the present invention, a voiceprint authentication protection method based on the SM2 algorithm is provided, comprising:

[0006] Collect ambient temperature and volume data of smart home devices, and use the SM3 algorithm to generate the pseudo-random large integer required for the digital signature authentication operation of the SM2 algorithm;

[0007] Collect user voice data, use the pre-trained voiceprint feature model to extract the voiceprint feature vector corresponding to the user voice data, and use the voiceprint feature vector to generate the public and private key pairs required for the SM2 algorithm digital signature authentication operation. The public key is transferred to the FPGA chip for storage;

[0008] For user voice data, the digital signature authentication operation process is performed based on the SM2 algorithm and the parallel processing unit in the FPGA chip is called.

[0009] As the voiceprint authentication protection method based on the SM2 algorithm of the present invention, further, the ambient temperature data and ambient volume data of the smart home device are collected, and the SM3 algorithm is used to generate the pseudo-random large integer required for the digital signature authentication operation of the SM2 algorithm, including:

[0010] First, set the environmental data collection cycle;

[0011] Next, use the set collection period to regularly collect ambient temperature data and ambient volume data;

[0012] Then, the collected ambient temperature data and ambient volume data are converted into digital form, and the pseudo-random large integer required for the multiplication operation in the digital signature of the SM2 algorithm is generated through the SM3 algorithm.

[0013] As the voiceprint authentication protection method based on the SM2 algorithm of the present invention, the voiceprint feature model pre-training process further includes:

[0014] A sample database is constructed by collecting sound data, and the speech data in the sample database is processed to obtain speech feature vectors;

[0015] The convolutional neural network is trained using the speech feature vector, and the trained convolutional neural network is used as a voiceprint feature model for extracting the voiceprint feature vector of the speech data.

[0016] As the voiceprint authentication protection method based on the SM2 algorithm of the present invention, further, the voice data in the sample database is processed to obtain the voice feature vector, including:

[0017] First, an audio frame sequence of speech data is obtained, and spectrogram information corresponding to each frame in the audio frame sequence is obtained through Fourier transform processing. The spectrogram information includes time domain information and frequency domain information of the frame;

[0018] Next, the time domain features and frequency domain features of the spectrum information are obtained;

[0019] Then, the speech feature vector is obtained by performing feature aggregation and vector embedding on the time domain features and frequency domain features.

[0020] The voiceprint authentication and protection method based on the SM2 algorithm of the present invention further includes:

[0021] A circuit logic configuration file is deployed in the FPGA chip, and the circuit logic configuration file is used to enable the FPGA chip logic circuit to provide a trusted execution environment for the SM2 algorithm parallel processing operation.

[0022] As the voiceprint authentication protection method based on the SM2 algorithm of the present invention, further, a circuit logic configuration file is deployed in the FPGA chip, including:

[0023] First, set the FPGA logic circuit design scheme and device type according to the business requirements of the SM2 algorithm, use the development tool to obtain the FPGA logic circuit, and use the simulation tool to perform functional simulation verification on the logic circuit;

[0024] Next, the logic circuit that meets the expectations of the FPGA logic circuit design solution is mapped to the device type structure resource and a logic netlist for configuration on the FPGA chip is generated;

[0025] Then, the FPGA chip is laid out and wired according to the logic netlist, the SM2 algorithm digital signature authentication operation is programmed into the chip and a parallel processing operation program is generated.

[0026] As a voiceprint authentication protection method based on the SM2 algorithm of the present invention, further, based on the SM2 algorithm and calling the parallel processing operation unit in the FPGA chip to perform the digital signature authentication operation process, the operation program in the parallel processing operation unit in the FPGA chip is used to respectively execute the point multiplication operation, cryptographic hash operation, point addition operation, multiplication operation, coordinate conversion operation, modular inverse operation and modular multiplication operation in the digital signature authentication of the SM2 algorithm.

[0027] Furthermore, the present invention also provides a voiceprint authentication and protection system based on the SM2 algorithm, comprising: a control terminal, a processing terminal connected to the control terminal, and a first acquisition terminal and a second acquisition terminal respectively connected to the processing terminal, wherein:

[0028] The first acquisition terminal is used to collect ambient temperature data and ambient sound data in real time, and transmit the ambient temperature data and ambient sound data to the processing terminal after analog-to-digital conversion;

[0029] The second collecting terminal is used to collect user voice data and transmit the voice data to the processing terminal;

[0030] The processing terminal is configured to use the ambient temperature data and the ambient sound data and generate a pseudo-random large integer required for the SM2 algorithm digital signature authentication operation through a pseudo-random number generator; input the user voice data into a trained voice feature model to extract a voiceprint feature vector of the user voice data, and use the voiceprint feature vector to generate a public-private key pair required for the SM2 algorithm digital signature authentication operation; and use an FPGA chip to execute the SM2 algorithm digital signature authentication operation based on the pseudo-random large integer and the public key and output a signature authentication result;

[0031] The control end controls the smart home devices based on the signature authentication results.

[0032] Beneficial effects of the present invention:

[0033] The present invention extracts the user's voiceprint features through an artificial intelligence model, generates a private key on the device side, and then combines it with the domestically produced, independently controllable SM2 algorithm to avoid transmitting the voiceprint features, thereby avoiding the threat of leakage of the user's voiceprint features during transmission and effectively protecting the user's privacy. FPGA hardware is used to execute the large-scale scalar multiplication operations of the SM2 algorithm, which has good flexibility and high performance. By combining it with the SM2 algorithm to implement a parallel processing architecture, the authentication speed is fast and can meet the needs of a large number of authentication activities. Compared with commonly used authentication methods such as face and fingerprint, the voiceprint authentication in this case can have higher spatial flexibility. The authenticator can authenticate anywhere their voice can be received, without having to go to a designated device for authentication. This provides the authenticator with extremely high convenience, versatility, and strong scalability, facilitating practical applications in smart home and other fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 Schematic diagram of the voiceprint authentication protection process based on the SM2 algorithm in the embodiment;

[0035] Figure 2 This is a schematic diagram of the FPGA structure design process in the embodiment;

[0036] Figure 3 This is a schematic diagram of the principle framework of the voiceprint authentication protection system in the embodiment;

[0037] Figure 4 Schematic diagram of the SM2 algorithm process in the embodiment. DETAILED DESCRIPTION

[0038] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention is further described in detail below with reference to the accompanying drawings and technical solutions.

[0039] The embodiment of the present invention, see Figure 1 As shown, a voiceprint authentication protection method based on the SM2 algorithm is provided, including:

[0040] S101, collecting ambient temperature data and ambient volume data of smart home devices, and using the SM3 algorithm to generate a pseudo-random large integer required for the digital signature authentication operation of the SM2 algorithm;

[0041] S102. Collect user voice data, use a pre-trained voiceprint feature model to extract the voiceprint feature vector corresponding to the user voice data, and use the voiceprint feature vector to generate the public and private key pairs required for the SM2 algorithm digital signature authentication operation, wherein the public key is transmitted to the FPGA chip for storage;

[0042] S103: For the user voice data, based on the SM2 algorithm and calling the parallel processing unit in the FPGA chip, a digital signature authentication operation process is performed.

[0043] The authenticator can perform authentication anywhere his voice can be received, without having to go to a designated device for authentication. This gives the authenticator extremely high convenience, versatility, and strong scalability. It generates SM2 algorithm-related parameters through environmental data and user voice data, and executes related digital signature verification operations in parallel in the FPGA chip. The authentication speed is fast and can meet the needs of a large number of authentication activities.

[0044] As a preferred embodiment, further, the ambient temperature data and ambient volume data of the smart home device are collected, and the SM3 algorithm is used to generate the pseudo-random large integer required for the digital signature authentication operation of the SM2 algorithm, including:

[0045] First, set the environmental data collection cycle;

[0046] Next, use the set collection period to regularly collect ambient temperature data and ambient volume data;

[0047] Then, the collected ambient temperature data and ambient volume data are converted into digital form, and the pseudo-random large integer required for the multiplication operation in the digital signature of the SM2 algorithm is generated through the SM3 algorithm.

[0048] It should be noted that the environmental data collection period can be set according to the application and user needs. For example, it can be set to collect data every 1 second, or the collection interval can be extended to save energy. There are also various settings, such as in the smart home device app interface or smart home device interaction panel.

[0049] In this embodiment, the voiceprint feature model pre-training process further includes:

[0050] A sample database is constructed by collecting sound data, and the speech data in the sample database is processed to obtain speech feature vectors;

[0051] The convolutional neural network is trained using the speech feature vector, and the trained convolutional neural network is used as a voiceprint feature model for extracting the voiceprint feature vector of the speech data.

[0052] By constructing a CNN convolutional neural network and using sample data to train and optimize the CNN convolutional neural network, the trained and optimized network model is used as a voiceprint feature model for extracting voiceprint feature vectors. The model can be embedded in smart terminals to meet the needs of intelligent control, combining artificial intelligence with intelligent control, improving the efficiency and accuracy of speech recognition, and enhancing the convenience of using intelligent control equipment.

[0053] Furthermore, in the embodiment of this case, the step of obtaining a speech feature vector by processing the speech data in the sample database can be designed to include:

[0054] First, an audio frame sequence of speech data is obtained, and spectrogram information corresponding to each frame in the audio frame sequence is obtained through Fourier transform processing. The spectrogram information includes time domain information and frequency domain information of the frame;

[0055] Next, the time domain features and frequency domain features of the spectrum information are obtained;

[0056] Then, the speech feature vector is obtained by performing feature aggregation and vector embedding on the time domain features and frequency domain features.

[0057] Original speech is a time series signal of indefinite length and is not suitable as direct input to a machine learning algorithm. In this application, the collected speech time domain signal is preprocessed and converted into a speech feature vector to reduce the difficulty of model training and improve the efficiency of model training.

[0058] As a preferred embodiment, the solution of this case further includes: deploying a circuit logic configuration file in the FPGA chip, and the circuit logic configuration file is used to enable the FPGA chip logic circuit to provide a trusted execution environment for the parallel processing operation of the SM2 algorithm.

[0059] Set the FPGA logic circuit design plan and device type according to the SM2 algorithm business requirements, use the development tool to obtain the FPGA logic circuit, and use the simulation tool to perform functional simulation verification on the logic circuit; map the logic circuit that meets the expectations of the FPGA logic circuit design plan to the device type structure resource and generate a logic netlist for configuration on the FPGA chip; layout and route the FPGA chip according to the logic netlist, program the SM2 algorithm digital signature authentication operation into the chip and generate a parallel processing operation program.

[0060] See also Figure 2As shown, an FPGA is used as the hardware carrier for the SM2 algorithm. The appropriate design scheme and device type are selected by evaluating operating speed, device resources, cost, and wiring routability. A top-down design approach is used to layer the design until EDA component libraries can be directly used. After the circuit design is completed, a dedicated simulation tool is used to perform functional simulation of the design to verify that the circuit meets the design requirements. During synthesis optimization, the high-level description is converted into a low-level description. The standard delay file generated by synthesis is back-annotated into the synthesis simulation model to estimate the impact of gate delays. An implementation tool is used to map the logic to the FPGA resources, determine the optimal layout of the logic, select routing channels that link the logic to the input and output functions, and generate the corresponding files. After layout and routing, a timing simulation is performed, during which the layout and routing delay file is back-annotated into the design. Generate the data file (bit data stream file) to be used, and then download the programming data to the FPGA to use the operation program in the parallel processing operation unit in the FPGA chip to respectively execute the point multiplication operation, cryptographic hash operation, point addition operation, multiplication operation, coordinate conversion operation, modular inverse operation and modular multiplication operation in the SM2 algorithm digital signature authentication.

[0061] Furthermore, based on the above method, an embodiment of the present invention also provides a voiceprint authentication and protection system based on the SM2 algorithm, comprising: a control end, a processing terminal connected to the control end, and a first acquisition end and a second acquisition end respectively connected to the processing terminal, wherein:

[0062] The first acquisition terminal is used to collect ambient temperature data and ambient sound data in real time, and transmit the ambient temperature data and ambient sound data to the processing terminal after analog-to-digital conversion;

[0063] The second collecting terminal is used to collect user voice data and transmit the voice data to the processing terminal;

[0064] The processing terminal is configured to use the ambient temperature data and the ambient sound data and generate a pseudo-random large integer required for the SM2 algorithm digital signature authentication operation through a pseudo-random number generator; input the user voice data into a trained voice feature model to extract a voiceprint feature vector of the user voice data, and use the voiceprint feature vector to generate a public-private key pair required for the SM2 algorithm digital signature authentication operation; and use an FPGA chip to execute the SM2 algorithm digital signature authentication operation based on the pseudo-random large integer and the public key and output a signature authentication result;

[0065] The control end controls the smart home devices based on the signature authentication results.

[0066] It is important to note that the dot multiplication module, as the core of the entire SM2 algorithm, can affect the implementation rate of the entire algorithm. The large-scale calculation processes in the SM2 algorithm are configured as parallel processing operations in the FPGA structure. For example, the hash algorithm, dot multiplication, dot addition, multiplication, modular inverse, and modular multiplication operations are all implemented by FPGA logic circuits. By using parallel processing operations in the FPGA structure to obtain voiceprint authentication results, the calculation efficiency is improved and the calculation time is reduced.

[0067] The coordinate transformation operation can be from affine coordinates to Jacobian coordinates. The modular inverse operation can be performed using the Montgomery algorithm combined with Fermat's theorem. The point multiplication operation can be performed using a Jacobi weighted projective coordinate system, and the result of the calculation is finally converted to an affine coordinate system. The hashing algorithm can use the SM3 cryptographic hashing algorithm to compress the message during the SM2 signing and verification process.

[0068] See also Figure 3 and 4 As shown, in the embodiment of this case, the processing terminal can use Raspberry Pi 4B to coordinate the collaborative work of various devices. The collaborative devices may include temperature sensors, sound sensors, analog-to-digital converters and microphones. The sound data collected by the microphone is converted into voiceprint features through the artificial intelligence model in the Raspberry Pi 4B.

[0069] In FPGA design, the compiled program can be added to the elinx project file and a test file can be created for the program. The algorithm can be simulated and tested using ModelSim simulation software. After verifying the ModelSim simulation results, the compiled code can be burned into the EQ6HL130 FPGA board. Connecting the burned FPGA board to the Raspberry Pi completes the simulation by running multiple sets of data.

[0070] A large amount of sound data is collected in advance to establish a sample database; a convolutional neural network is built in the Raspberry Pi 4B and trained through the sample database to obtain a voiceprint feature model; the temperature sensor can be set to collect ambient temperature data every 1 second, and the sound sensor can collect ambient volume data every 1 second; the analog-to-digital converter processes the temperature and volume data and passes it to the SM3 algorithm in the Raspberry Pi 4B to calculate and generate a pseudo-random large integer.

[0071] The microphone collects human voice data and passes it to the voiceprint feature model trained in the Raspberry Pi 4B to extract the corresponding person's voiceprint feature vector and generate the public and private key pair of SM2; the private key is deleted and the public key is transferred to the FPGA for storage; when the user's voice data is transmitted, the data before and after encryption and decryption are compared using the generated private key and the public key stored in the FPGA. If they are consistent, the authentication is passed, otherwise the authentication fails.

[0072] When the Raspberry Pi is actually working, various components such as the CPU, GPU, and memory jointly generate heat, causing the temperature to fluctuate continuously. The actual temperature can be collected through the DS18B20 temperature sensor to obtain the actual temperature value. When the authenticator performs actual authentication, the volume of the authentication sound is random and difficult to control according to actual conditions. The authentication sound is collected and processed by the sound sensor, and the volume is digitized and output. By reading the corresponding dictionary value, the three can be combined to generate a pseudo-random large integer required for the multiplication operation in the subsequent authentication algorithm in the Raspberry Pi using the national secret SM3 algorithm.

[0073] Since the Raspberry Pi only has an audio output port but no audio input, it's impossible to record sound signals. Therefore, a USB external sound card is used as the audio input port. Install the corresponding recording program and its corresponding UI to facilitate volume adjustment. Once the program is installed, plug the USB external sound card into the Raspberry Pi's USB 3.0 port, then plug the 3.5mm jack of the microphone into the 3.5mm port on the USB external sound card to start recording. The collected user voice is automatically generated into a WAV file and temporarily stored on the backend.

[0074] Using Python's PyTorch library, a convolutional neural network model was built, and a sample database was established. The convolutional neural network model in the Raspberry Pi 4B was trained using the speech feature vectors in the sample database to obtain a voiceprint feature model; the speech data to be recognized collected by the microphone was preprocessed to obtain a speech feature vector; the preprocessed speech feature vector was input into the voiceprint feature model to obtain a voiceprint feature vector.

[0075] The voiceprint feature vector and pseudo-random large integer are used as the private key in the SM2 algorithm to participate in the digital signature. After the voice password entered by the user is signed, it is authenticated by the public key stored in the FPGA. If the authentication is successful, the ESP8266 chip as the control end will enable the user to connect via WIFI to control the smart device. Compared with commonly used authentication methods such as face and fingerprint, it has higher spatial flexibility. The authenticator can authenticate anywhere where his voice can be received, without having to go to the designated device for authentication, which gives the authenticator extremely high convenience.

[0076] Unless otherwise specifically stated, the relative steps, numerical expressions and values ​​of the components and steps set forth in these embodiments do not limit the scope of the present invention.

[0077] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the systems disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0078] The units and method steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the components and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. A person of ordinary skill in the art may use different methods to implement the described functions for each specific application, but such implementation is not considered to be beyond the scope of the present invention.

[0079] Those skilled in the art will appreciate that all or part of the steps in the above method can be performed by a program instructing related hardware. The program can be stored in a computer-readable storage medium, such as a read-only memory, a magnetic disk, or an optical disk. Alternatively, all or part of the steps in the above embodiment can be implemented using one or more integrated circuits. Accordingly, each module / unit in the above embodiment can be implemented in the form of hardware or software functional modules. The present invention is not limited to any specific combination of hardware and software.

[0080] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present invention, which are used to illustrate the technical solutions of the present invention, rather than to limit them. The scope of protection of the present invention is not limited thereto. Although the present invention has been described in detail with reference to the above-described embodiments, those skilled in the art should understand that any person skilled in the art can modify or easily conceive of changes to the technical solutions described in the above-described embodiments within the technical scope disclosed by the present invention, or replace some of the technical features therein with equivalents. Such modifications, changes, or replacements do not deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be subject to the scope of protection of the claims.

Claims

1. A voiceprint authentication protection method based on SM2 algorithm, characterized in that: Include: Collect the ambient temperature data and ambient volume data of smart home devices, and use the SM3 algorithm to generate the pseudo-random large integer required for the digital signature authentication operation of the SM2 algorithm; Collect user voice data, use a pre-trained voiceprint feature model to extract a voiceprint feature vector corresponding to the user voice data, and use the voiceprint feature vector to generate a public-private key pair required for the digital signature authentication operation of the SM2 algorithm, wherein the public key is transmitted to the FPGA chip for storage; wherein the voiceprint feature model pre-training process includes: constructing a sample database by collecting voice data, and processing the voice data in the sample database to obtain a voice feature vector; using the voice feature vector to train a convolutional neural network, and using the trained convolutional neural network as a voiceprint feature model for extracting the voiceprint feature vector of the voice data; and, processing the voice data in the sample database to obtain a voice feature vector, including: first, obtaining an audio frame sequence of the voice data, and obtaining the spectrum information corresponding to each frame in the audio frame sequence through Fourier transform processing, the spectrum information including the time domain information and frequency domain information of the frame; then, obtaining the time domain features and frequency domain features of the spectrum information; then, obtaining the voice feature vector by performing feature aggregation and vector embedding on the time domain features and the frequency domain features; A circuit logic configuration file is deployed in an FPGA chip, and the circuit logic configuration file is used to enable the FPGA chip logic circuit to provide a trusted execution environment for the parallel processing operation of the SM2 algorithm; for user voice data, the operation program in the parallel processing operation unit in the FPGA chip is used to respectively execute the point multiplication operation, cryptographic hash operation, point addition operation, multiplication operation, coordinate conversion operation, modular inverse operation and modular multiplication operation in the digital signature authentication of the SM2 algorithm, so as to perform the digital signature authentication operation process based on the SM2 algorithm and calling the parallel processing operation unit in the FPGA chip; wherein, the circuit logic configuration file is deployed in the FPGA chip, including: first, setting the FPGA logic circuit design scheme and device type according to the business requirements of the SM2 algorithm, using the development tool to obtain the FPGA logic circuit, and performing functional simulation verification on the logic circuit through the simulation tool; then, mapping the logic circuit that meets the expectations of the FPGA logic circuit design scheme to the device type structure resource and generating a logic netlist for configuration on the FPGA chip; then, laying out and routing the FPGA chip according to the logic netlist, performing chip programming for the digital signature authentication operation of the SM2 algorithm and generating a parallel processing operation program.

2. The voiceprint authentication protection method based on the SM2 algorithm according to claim 1 is characterized in that: Collect the ambient temperature data and ambient volume data of smart home devices, and use the SM3 algorithm to generate the pseudo-random large integer required for the digital signature authentication operation of the SM2 algorithm, including: First, set the environmental data collection cycle; Next, the set collection period is used to regularly collect the ambient temperature data and ambient volume data; Then, analog-to-digital conversion is performed on the collected ambient temperature data and ambient volume data, and the pseudo-random large integer required for the multiplication operation in the digital signature in the SM2 algorithm is generated through the SM3 algorithm.

3. A voiceprint authentication protection system based on SM2 algorithm, characterized in that: The method according to claim 1 is implemented, comprising: a control end, a processing terminal connected to the control end, and a first collection end and a second collection end respectively connected to the processing terminal, wherein: The first acquisition terminal is used to collect ambient temperature data and ambient sound data in real time, and transmit the ambient temperature data and ambient sound data to the processing terminal after analog-to-digital conversion; The second collecting terminal is used to collect user voice data and transmit the voice data to the processing terminal; A processing terminal is used to use the ambient temperature data and the ambient sound data and generate a pseudo-random large integer required for the SM2 algorithm digital signature authentication operation through a pseudo-random number generator; input the user voice data into the trained sound feature model to extract the voiceprint feature vector of the user voice data, and use the voiceprint feature vector to generate the public and private key pair required for the SM2 algorithm digital signature authentication operation; and use the FPGA chip to perform the SM2 algorithm digital signature authentication operation based on the pseudo-random large integer and the public key and output the signature authentication result; The control end controls the smart home devices based on the signature authentication results.

4. An electronic device, characterized in that: It includes one or more FPGA processors, communication interfaces, memories and communication buses, wherein the FPGA processors, communication interfaces and memories communicate with each other via the communication buses; Memory, used to store computer programs; A processor, configured to execute a program stored in a memory and implement the method steps described in any one of claims 1 to 2 when the program is executed.

5. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions enable the computer to execute the method steps according to any one of claims 1 to 2.

Images