Multi-factor Internet of Things equipment authentication method and system

Abstract

The invention provides a multi-factor Internet of Things equipment authentication method and system, and the method comprises the steps: playing a registration audio through a loudspeaker of Internet of Things equipment in a registration stage, and collecting the played registration audio through a microphone of the Internet of Things equipment; acquiring spatial distance features of the loudspeaker and the microphone and FBank features and MFCC features corresponding to the registered audio as a sound fingerprint training authentication model; in the authentication stage, an authentication audio is played through a loudspeaker of the Internet of Things equipment, the played authentication audio is collected through a microphone of the Internet of Things equipment, spatial distance features of the loudspeaker and the microphone and FBank features and MFCC features corresponding to the authentication audio are obtained and serve as sound fingerprints to be input into the trained authentication model, and an authentication result is obtained. According to the invention, universal, safe, convenient and efficient verification of the legality of the Internet of Things equipment in the power distribution network environment can be realized.

Description

Technical Field

[0001] The present invention relates to the field of smart grids, and in particular to a multi-factor Internet of Things device authentication method and system. Background Art

[0002] The construction of smart grid is an important strategic deployment in my country's energy field. It will gradually promote the change of my country's power infrastructure production mode and is the development direction of the power industry. The Internet of Things technology has become an important technical means to promote the development of smart grid. However, the characteristics of the Internet of Things network, such as ubiquitous, comprehensive perception, reliable transmission, and intelligent processing, also make the distribution Internet of Things face greater security threats. One of the basic mechanisms to protect the security of the distribution Internet of Things is device legitimacy verification, which can verify the identity of the distribution Internet of Things entities and control their access. Therefore, in the face of security threats in the distribution Internet of Things environment, how to verify the legitimacy of the Internet of Things devices connected to the distribution network terminals has become an urgent problem to be solved.

[0003] With the continuous development of technology and the increase in security threats, traditional device authentication solutions, such as key authentication, MAC address filtering, IP address filtering, etc., lack effective protection solutions and response strategies in the face of new smart devices connected to the network and emerging malicious attacks against smart devices. At present, some novel device authentication schemes use physical layer characteristics to enhance device identity authentication. Physical layer device identification schemes can be roughly divided into three categories: software fingerprint-based, channel fingerprint-based, and hardware fingerprint-based. (1) In terms of software characteristics, many browser configuration information can be used to distinguish devices, such as user agent, installed fonts, plug-in information, benchmarks, etc. Software fingerprints are generated by the current configuration of the system, so software fingerprints are not static and are likely to change over time. In addition, software-based protocols cannot distinguish between different physical devices running the same software. (2) Channel fingerprint-based device authentication schemes usually use radio strength signal (RSS) and channel state information (CSI). Although RSS-based mechanisms have been intensively studied, their efficiency is still limited because RSS can only provide coarse-grained information and has a high correlation with transmission power and distance. CSI-based mechanisms are difficult to generalize because they rely on dedicated hardware and cannot be widely used on mobile devices. (3) Hardware fingerprinting schemes are designed based on reflections of hardware defects or the unique response of hardware to the transmission waveform, relying on some static source of characteristics. Many works are devoted to using small differences in the signals generated by hardware components to identify devices. For example, wireless network cards can be distinguished by using the characteristics of the RF signal emitted by the transmitter. However, these methods cannot be generalized to Internet tracking use. Data collected from accelerometers can also be used to distinguish users without active stimulation. Photos taken with a camera can also be distinguished by patterns and noise.

[0004] Therefore, the above device authentication methods cannot be applied to the distribution network environment due to various defects. Therefore, it is necessary to explore and research new distribution network environment equipment authentication technologies to achieve universal, safe, convenient and efficient identity verification of IoT devices in the distribution network, and form the legitimacy verification capability of IoT devices in the distribution network. Summary of the invention

[0005] Technical problem to be solved by the present invention: In view of the above-mentioned problems in the prior art, a multi-factor IoT device authentication method and system are provided, which utilize acoustic characteristics to capture the manufacturing defects of the built-in microphone-speaker system of the device and the spatial distance to distinguish whether it is a legitimate device, and extract the acoustic fingerprint for device authentication, so as to realize universal, safe, convenient and efficient verification of the legitimacy of devices on terminal devices and IoT devices in the distribution network environment.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: A multi-factor IoT device authentication method comprises the following steps: Play the registered audio through the speaker of the IoT device and collect the played registered audio through the microphone of the IoT device, perform signal processing on the registered audio collected by each microphone to obtain the corresponding FBank features and MFCC features, and calculate the spatial distance features between the speaker and microphone of the IoT device according to the registered audio collected by each microphone, and use the spatial distance features and the FBank features and MFCC features corresponding to the registered audio collected by each microphone as the sound fingerprint training authentication model; The authentication audio is played through the speaker of the IoT device and collected through the microphone of the IoT device. Signal processing is performed on the authentication audio collected by each microphone to obtain the corresponding FBank features and MFCC features. At the same time, the spatial distance features of the speaker and microphone of the IoT device are calculated according to the authentication audio collected by each microphone. The spatial distance features and the FBank features and MFCC features corresponding to the authentication audio collected by each microphone are input into the trained authentication model as sound fingerprints to obtain the authentication result.

[0007] Furthermore, before playing the registration audio through the speaker of the Internet of Things device and before playing the authentication audio through the speaker of the Internet of Things device, it also includes: collecting environmental sounds through the microphone of the Internet of Things device.

[0008] Furthermore, the signal processing of the registration audio collected by each microphone and the signal processing of the authentication audio collected by each microphone include: The sound signal collected by the current microphone is denoised, and the denoised sound signal is pre-emphasized, framed and windowed. The frequency domain signal is obtained by fast Fourier transform, and the frequency domain spectrum is obtained. The Mel spectrum is obtained by using a filter group on the frequency domain spectrum, and the obtained Mel spectrum is paired to calculate the FBank feature, and the FBank feature is discrete cosine transformed to obtain the MFCC feature.

[0009] Furthermore, the IoT device is provided with two microphones, and when the spatial distance feature of the speaker and microphone of the IoT device is calculated based on the registered audio collected by each microphone, and the spatial distance feature of the speaker and microphone of the IoT device is calculated based on the authentication audio collected by each microphone, the spatial distance feature expression is as follows:

[0010] in, and They represent the sound distortion produced by the first and second microphones of the IoT device, and Represents the spatial distance between the speaker and two microphones of the IoT device, It represents the attenuation of sound during propagation. It is the distance and frequency function.

[0011] Furthermore, when the spatial distance features and the FBank features and MFCC features corresponding to the registered audio collected by each microphone are used as sound fingerprints to train the authentication model, the sound fingerprints are specifically used as sample features to train the unsupervised learning vector machine to obtain the authentication model.

[0012] The present invention also proposes a multi-factor IoT device authentication system, comprising two IoT devices connected in pairs, a verification server, and a distribution network terminal device, wherein: The IoT device is used to send a registration request or an authentication request to the inspection server, and after receiving the instruction from the inspection server, generates a corresponding registration audio or an authentication audio, plays the registration audio or the authentication audio through a speaker, and collects the played registration audio or the authentication audio through a microphone and sends it to the inspection server; The verification server is used to send a corresponding registration audio generation instruction to the IoT device after receiving a registration request from the IoT device, so as to play the registration audio through the speaker of the IoT device and collect the played registration audio through the microphone of the IoT device, and then obtain the registration audio collected by each microphone of the IoT device and perform signal processing to obtain the corresponding FBank features and MFCC features, and at the same time calculate the spatial distance features of the speaker and microphone of the IoT device according to the registration audio collected by each microphone, and use the spatial distance features and the FBank features and MFCC features corresponding to the registration audio collected by each microphone as a sound fingerprint training authentication model; The verification server is also used to send a corresponding authentication audio generation instruction to the IoT device after receiving an authentication request from the IoT device, so as to play the authentication audio through the speaker of the IoT device and collect the played authentication audio through the microphone of the IoT device, and then obtain the authentication audio collected by each microphone of the IoT device and perform signal processing to obtain the corresponding FBank features and MFCC features, and at the same time calculate the spatial distance features of the speaker and microphone of the IoT device according to the authentication audio collected by each microphone, and input the spatial distance features and the FBank features and MFCC features corresponding to the authentication audio collected by each microphone into the trained authentication model as sound fingerprints to obtain the authentication result and send it to the distribution network terminal device; The distribution network terminal device is used to obtain and analyze the authentication result. If the authentication result of the Internet of Things device is a legal device, the Internet of Things device is allowed to access, otherwise the Internet of Things device is blocked from accessing.

[0013] The present invention also proposes a multi-factor IoT device authentication device, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of any one of the multi-factor IoT device authentication methods. The present invention also provides a computer-readable storage medium having a computer program stored thereon, which implements the steps of any one of the multi-factor IoT device authentication methods when executed by a processor. The present invention also proposes a computer program product, comprising a computer program, which implements the steps of any one of the multi-factor IoT device authentication methods when executed by a processor.

[0014] Compared with the prior art, the advantages of the present invention are: (1) The present invention utilizes the built-in microphone and speaker of the IoT device to complete device authentication without introducing additional hardware or sensors. The present invention can be applied to devices that lack dedicated hardware or sensors in device authentication between distribution network terminal devices and device configurations.

[0015] (2) The present invention utilizes the spatial distance feature from the speaker to the microphone of the IoT device as an additional factor for device authentication, which can resist replay attacks and injection attacks and improve the security of device authentication. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The figure is a flow chart of a method according to an embodiment of the present invention.

[0017] Figure 2 Schematic diagram of the interaction between power grid terminal equipment and IoT devices.

[0018] Figure 3 Schematic diagram of the sound propagation path of IoT devices. DETAILED DESCRIPTION

[0019] The present invention is further described below in conjunction with the accompanying drawings and specific preferred embodiments, but the protection scope of the present invention is not limited thereby.

[0020] Embodiment 1 With the widespread use of audio communications, many existing IoT devices have built-in speakers and microphones. During the manufacturing process, due to differences in the hardware manufacturing process and other uncontrollable factors, the analog circuits of these components have unique defects and characteristics. Through existing research, these defects make each microphone and speaker unique. And the spatial distance from the speaker to the microphone of the IoT device is fixed. The audio characteristics and the spatial distance characteristics from the speaker to the microphone can be obtained by recording the speaker-microphone system response audio and performing power spectrum calculation on the recorded audio stream. These acoustic characteristics can be used to capture the manufacturing defects and spatial distance of the built-in microphone-speaker system of the device to distinguish whether it is a legitimate device, and extract the acoustic fingerprint for device authentication.

[0021] Based on the above characteristics of IoT devices, this embodiment proposes a multi-factor IoT device authentication method, which can realize universal, safe, convenient and efficient verification of the legitimacy of terminal devices and IoT devices in a distribution network environment, including the following steps: Registration stage, such as Figure 1 As shown, the following steps are included: S101) playing a registration audio through a speaker of the IoT device and collecting the played registration audio through a microphone of the IoT device; S102) performing signal processing on the registered audio collected by each microphone to obtain corresponding FBank features and MFCC features; S103) Calculate the spatial distance features of the speaker and microphone of the IoT device based on the registered audio collected by each microphone, and use the spatial distance features and the FBank features and MFCC features corresponding to the registered audio collected by each microphone as a sound fingerprint training authentication model; Authentication stage, such as Figure 1 As shown, the following steps are included: S201) playing authentication audio through a speaker of the IoT device and collecting the played authentication audio through a microphone of the IoT device; S202) performing signal processing on the authentication audio collected by each microphone to obtain corresponding FBank features and MFCC features; S203) Calculate the spatial distance features between the speaker and microphone of the IoT device based on the authentication audio collected by each microphone, and input the spatial distance features and the FBank features and MFCC features corresponding to the authentication audio collected by each microphone as sound fingerprints into the trained authentication model to obtain the authentication result.

[0022] The following is a detailed description of each step.

[0023] The method of this embodiment is applied to Figure 2 The system shown includes an IoT device and a verification server. In step S101, the IoT device sends a registration request and generates audio for registration after receiving an instruction from the verification server, and uses a built-in speaker to play the registration audio. At the same time, the built-in microphone is used to collect the played registration audio as response audio and store it in the storage unit of the device, and then sent to the verification server.

[0024] In this embodiment, the built-in microphone is used to sense and collect the response audio, and a mixed sound signal including the ambient sound and the response audio is obtained. Therefore, before playing the registration audio through the speaker of the IoT device, this embodiment also collects the ambient sound through the microphone of the IoT device, thereby facilitating the denoising operation in the subsequent steps.

[0025] In step S102 of this embodiment, the inspection server performs signal processing on the collected response audio, removes noise and calculates FBank features and MFCC features. Specifically, the collected mixed sound signal is divided into noise and signal, and the MMSE algorithm is used for sound noise reduction. The noise-reduced sound signal is pre-emphasized, framed and windowed, and then the frequency domain signal is obtained through fast Fourier transform, the frequency domain spectrum is obtained, and a filter group is used on the frequency domain spectrum to obtain the Mel spectrum, and the FBank features and MFCC features corresponding to each microphone are calculated. Therefore, when the registered audio collected by each microphone is signal processed, the following steps are included: S1021) De-noise the sound signal collected by the current microphone. Specifically, the sound signal collected by the current microphone is divided into two parts, noise and signal, according to the time length. Then, the MMSE algorithm is used to de-noise the signal part according to the noise part. The MMSE algorithm is an algorithm that performs noise reduction processing by estimating the statistical characteristics of the signal, thereby improving the quality and clarity of the signal. The noise part is estimated, and the statistical characteristics of the noise are used to de-noise the signal part, thereby reducing the influence of the noise on the signal and obtaining the de-noised signal. The de-noising principle of the MMSE algorithm is well known to those skilled in the art and will not be described in detail in this embodiment.

[0026] S1022) Pre-emphasize, frame and window the de-noised sound signal, then perform fast Fourier transform to obtain the frequency domain signal, obtain the frequency domain spectrum, and use the filter group to obtain the Mel spectrum. Specifically, the frequency domain spectrum is passed through a group of Mel-scale triangular filters, and the spectrum is smoothed by the filter to eliminate the effect of harmonics and highlight the resonance peak of the original speech. The output within the frequency range of each filter is used as the weight, and the corresponding energy of the corresponding frequency in the frequency domain spectrum is weighted and summed to obtain the Mel spectrum. The expression is as follows:

[0027] in, represents the first of a set of Mel-scale triangular filters filters, Indicates the Fourier transform Points, represents the center frequency of the filter, is the Fourier transform The output obtained by filtering the points is represents the Fourier transform After calculation, we get outputs.

[0028] S1023) The obtained Mel spectrum is paired and calculated to obtain the Mel filter bank coefficient, that is, FBank feature, which is expressed as follows:

[0029] in, represents the first of a set of Mel-scale triangular filters filters, Indicates Frame data, Indicates Frame data List, Indicates Frame data passes through the The Mel spectrum value obtained by the filter.

[0030] S1024) Perform discrete cosine transform (DCT) on the FBank feature to obtain the Mel cepstral coefficient, i.e., MFCC feature, which is expressed as follows:

[0031] in, represents the first of a set of Mel-scale triangular filters filters, represents the total number of a set of Mel-scale triangular filters, Indicates Frame data, Indicates Frame data List.

[0032] Mel filter bank coefficients (FBank) and Mel cepstral coefficients (MFCC) are two parameters extracted in the Mel scale frequency domain. The Mel scale describes the nonlinear characteristics of human ear frequency. Compared with the normal frequency mechanism, the Mel value is closer to the human ear's hearing mechanism. Its relationship with frequency can be approximately expressed by the following formula:

[0033] in, is the frequency of the sound signal, in Hz.

[0034] In step S103 of this embodiment, when calculating the spatial distance characteristics of the speaker and microphone of the Internet of Things device based on the registered audio collected by each microphone, the parameters related to the spatial distance are extracted using mathematical methods based on the constructed sound propagation model of the noise-reduced sound signal to obtain the spatial distance characteristics.

[0035] Specifically, when the built-in speaker is used to play audio and two built-in microphones are used to collect audio, the sound propagation paths of the two channels are as follows: Figure 3 As shown, build the following model:

[0036]

[0037] in and Represent the sound signals of the first and second channels respectively. Indicates the authenticated audio being played. Indicates the sound distortion produced by the speaker, and Indicates the sound distortion produced by the first and second microphones, It is the distance and frequency The propagation of sound waves follows a power law attenuation, so It represents the attenuation of sound during propagation. and Indicates the spatial distance between the speaker and the two microphones.

[0038] From the above sound propagation expression, we can obtain the coefficient related to the spatial distance between the IoT device speaker and microphone, that is, , expressed as follows:

[0039] in and Indicates the sound distortion produced by the first and second microphones, and Indicates the spatial distance between the speaker and the two microphones.

[0040] In step S103 of this embodiment, when the spatial distance feature and the FBank feature and MFCC feature corresponding to the registered audio collected by each microphone are used as the sound fingerprint training authentication model, the sound fingerprint is specifically used as a sample feature to train the unsupervised learning vector machine to obtain a classifier model as the authentication model. How to use sample features to train an unsupervised learning vector machine is well known to those skilled in the art and will not be described in detail in this embodiment.

[0041] It should be noted that due to the influence of sound echo, there will be slight differences in the sound signals collected under different circumstances. In order to avoid the influence of sound echo, this embodiment collects response audio, i.e., registration audio, multiple times during the registration phase, and then obtains the corresponding FBank features, MFCC features, and spatial distance features through steps S102 and S103, so that a large number of samples can be used to train the unsupervised vector machine, enhance the model recognition capability, and reduce the error caused by sound echo.

[0042] Step S201 of this embodiment is similar to step S101. After the IoT device issues an authentication request and receives an instruction from the verification server, it generates an audio for authentication, and uses a built-in speaker to play the authentication audio. At the same time, it uses a built-in microphone to collect the played authentication audio as a response audio and stores it in the storage unit of the device, and then sends it to the verification server. Before playing the authentication audio through the speaker of the IoT device, the microphone of the IoT device also collects the ambient sound, so as to facilitate the denoising operation in the subsequent steps.

[0043] Step S202 of this embodiment is the same as step S102, which also performs signal processing on the collected response audio, divides the collected mixed sound signal into noise and signal, uses the MMSE algorithm to perform sound noise reduction, pre-emphasizes, frames and windows the noise-reduced sound signal, obtains the frequency domain signal through fast Fourier transform, obtains the frequency domain spectrum, uses a filter group on the frequency domain spectrum, obtains the Mel spectrum, and calculates the FBank features and MFCC features of the channels of the two microphones. No further details are given here.

[0044] The calculation of the spatial distance characteristics of the speaker and microphone of the IoT device based on the authentication audio collected by each microphone in step S203 of this embodiment is the same as the calculation of the spatial distance characteristics of the speaker and microphone of the IoT device based on the registration audio collected by each microphone in step S102, and also uses the sound propagation model of two channels to calculate and extract parameters related to the spatial distance between the speaker and the microphone to obtain the spatial distance characteristics of the device speaker and the microphone. No further details are given here.

[0045] The trained authentication model in step S203 of this embodiment is the classifier model obtained by training the unsupervised vector machine using a large number of samples in step S103. When the spatial distance feature and the FBank feature and MFCC feature corresponding to the authentication audio collected by each microphone are input into the trained authentication model as sound fingerprints, the three features are integrated into the device sound fingerprint, which is input into the trained classifier model for classification, and the classification result is obtained to determine whether it is a legal device or an abnormal device. When judging a legal device or an abnormal device based on the classification result, this embodiment uses test accuracy as a verification method and sets an accuracy threshold. When the test accuracy of the classifier model reaches the threshold, the authentication is passed and the device is identified as a legal device, otherwise the authentication fails.

[0046] Embodiment 2 This embodiment proposes a multi-factor IoT device authentication system, including two IoT devices connected in pairs, a verification server, and a distribution network terminal device, wherein: The IoT device is used to send a registration request or an authentication request to the inspection server, and after receiving the instruction from the inspection server, generates a corresponding registration audio or an authentication audio, plays the registration audio or the authentication audio through a speaker, and collects the played registration audio or the authentication audio through a microphone and sends it to the inspection server; The verification server is used to send a corresponding registration audio generation instruction to the IoT device after receiving a registration request from the IoT device, so as to play the registration audio through the speaker of the IoT device and collect the played registration audio through the microphone of the IoT device, and then obtain the registration audio collected by each microphone of the IoT device and perform signal processing to obtain the corresponding FBank features and MFCC features, and at the same time calculate the spatial distance features of the speaker and microphone of the IoT device according to the registration audio collected by each microphone, and use the spatial distance features and the FBank features and MFCC features corresponding to the registration audio collected by each microphone as a sound fingerprint training authentication model; The verification server is also used to send a corresponding authentication audio generation instruction to the IoT device after receiving an authentication request from the IoT device, so as to play the authentication audio through the speaker of the IoT device and collect the played authentication audio through the microphone of the IoT device, and then obtain the authentication audio collected by each microphone of the IoT device and perform signal processing to obtain the corresponding FBank features and MFCC features, and at the same time calculate the spatial distance features of the speaker and microphone of the IoT device according to the authentication audio collected by each microphone, and input the spatial distance features and the FBank features and MFCC features corresponding to the authentication audio collected by each microphone into the trained authentication model as sound fingerprints to obtain the authentication result and send it to the distribution network terminal device; The distribution network terminal device is used to obtain and analyze the authentication result. If the authentication result of the Internet of Things device is a legal device, the Internet of Things device is allowed to access, otherwise the Internet of Things device is blocked from accessing.

[0047] Embodiment 3 This embodiment provides a multi-factor IoT device authentication device, including a memory, a processor, and a computer program stored in the memory, and the processor executes the computer program to implement the steps of the multi-factor IoT device authentication method described in Embodiment 1. This embodiment further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the multi-factor IoT device authentication method described in the first embodiment are implemented. This embodiment also proposes a computer program product, including a computer program, which, when executed by a processor, implements the steps of the multi-factor IoT device authentication method described in the first embodiment.

[0048] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A multi-factor IoT device authentication method, characterized in that: The following steps are involved: Play the registered audio through the speaker of the IoT device and collect the played registered audio through the microphone of the IoT device, perform signal processing on the registered audio collected by each microphone to obtain the corresponding FBank features and MFCC features, and calculate the spatial distance features between the speaker and microphone of the IoT device according to the registered audio collected by each microphone, and use the spatial distance features and the FBank features and MFCC features corresponding to the registered audio collected by each microphone as the sound fingerprint training authentication model; The authentication audio is played through the speaker of the IoT device and collected through the microphone of the IoT device. Signal processing is performed on the authentication audio collected by each microphone to obtain the corresponding FBank features and MFCC features. At the same time, the spatial distance features of the speaker and microphone of the IoT device are calculated according to the authentication audio collected by each microphone. The spatial distance features and the FBank features and MFCC features corresponding to the authentication audio collected by each microphone are input into the trained authentication model as sound fingerprints to obtain the authentication result.

2. The multi-factor IoT device authentication method according to claim 1, characterized in that: Before playing the registration audio through the speaker of the IoT device, and before playing the authentication audio through the speaker of the IoT device, it also includes: collecting environmental sounds through the microphone of the IoT device.

3. The multi-factor IoT device authentication method according to claim 1, characterized in that: The signal processing for the registration audio collected by each microphone and the signal processing for the authentication audio collected by each microphone include: The sound signal collected by the current microphone is denoised, and the denoised sound signal is pre-emphasized, framed and windowed. The frequency domain signal is obtained by fast Fourier transform, and the frequency domain spectrum is obtained. The Mel spectrum is obtained by using a filter group on the frequency domain spectrum, and the obtained Mel spectrum is paired to calculate the FBank feature, and the FBank feature is discrete cosine transformed to obtain the MFCC feature.

4. The multi-factor IoT device authentication method according to claim 1, characterized in that: The IoT device is provided with two microphones. When the spatial distance feature of the speaker and microphone of the IoT device is calculated based on the registered audio collected by each microphone, and the spatial distance feature of the speaker and microphone of the IoT device is calculated based on the authentication audio collected by each microphone, the spatial distance feature expression is as follows: in, and They represent the sound distortion produced by the first and second microphones of the IoT device, and Represents the spatial distance between the speaker and two microphones of the IoT device, It represents the attenuation of sound during propagation. It is the distance and frequency function.

5. The multi-factor IoT device authentication method according to claim 1, characterized in that: When the spatial distance features and the FBank features and MFCC features corresponding to the registered audio collected by each microphone are used as sound fingerprints to train the authentication model, the sound fingerprints are specifically used as sample features to train the unsupervised learning vector machine to obtain the authentication model.

6. A multi-factor IoT device authentication system, characterized in that: It includes two-to-two connected IoT devices, inspection servers, and distribution network terminal devices, among which: The IoT device is used to send a registration request or an authentication request to the inspection server, and after receiving the instruction from the inspection server, generates a corresponding registration audio or an authentication audio, plays the registration audio or the authentication audio through a speaker, and collects the played registration audio or the authentication audio through a microphone and sends it to the inspection server; The verification server is used to send a corresponding registration audio generation instruction to the IoT device after receiving a registration request from the IoT device, so as to play the registration audio through the speaker of the IoT device and collect the played registration audio through the microphone of the IoT device, and then obtain the registration audio collected by each microphone of the IoT device and perform signal processing to obtain the corresponding FBank features and MFCC features, and at the same time calculate the spatial distance features of the speaker and microphone of the IoT device according to the registration audio collected by each microphone, and use the spatial distance features and the FBank features and MFCC features corresponding to the registration audio collected by each microphone as a sound fingerprint training authentication model; The verification server is also used to send a corresponding authentication audio generation instruction to the IoT device after receiving an authentication request from the IoT device, so as to play the authentication audio through the speaker of the IoT device and collect the played authentication audio through the microphone of the IoT device, and then obtain the authentication audio collected by each microphone of the IoT device and perform signal processing to obtain the corresponding FBank features and MFCC features, and at the same time calculate the spatial distance features of the speaker and microphone of the IoT device according to the authentication audio collected by each microphone, and input the spatial distance features and the FBank features and MFCC features corresponding to the authentication audio collected by each microphone into the trained authentication model as sound fingerprints to obtain the authentication result and send it to the distribution network terminal device; The distribution network terminal device is used to obtain and analyze the authentication result. If the authentication result of the Internet of Things device is a legal device, the Internet of Things device is allowed to access, otherwise the Internet of Things device is blocked from accessing.

7. A multi-factor IoT device authentication device, comprising a memory, a processor, and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the multi-factor Internet of Things device authentication method according to any one of claims 1 to 5.

8. 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 multi-factor IoT device authentication method described in any one of claims 1 to 5 are implemented.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the multi-factor IoT device authentication method described in any one of claims 1 to 5 are implemented.

Images