Device authentication method using acoustic signal and computer apparatus therefor

Abstract

An embodiment of the present application provides a device authentication method using an acoustic signal and an apparatus therefor. The method may comprise the steps of: obtaining an acoustic reception signal by recording an acoustic signal transmitted from at least one device to be authenticated; extracting a first feature of the acoustic reception signal by inputting, into a pre-trained deep learning model, transformed data obtained by transforming the acoustic reception signal into the frequency domain; and comparing the first feature with a pre-registered reference feature to authenticate the device to be authenticated.

Description

Method for device authentication using acoustic signals and computer device for the same

[0001] The present application relates to a device authentication method using an acoustic signal and a computer device for the same.

[0002] Authentication technology for intelligent devices, particularly autonomous systems such as robots, drones, IoT devices, smart TVs, and smart speakers, is playing an increasingly important role today. This is because these devices are becoming more widely used in various industries and daily life, making it essential to ensure that only authorized and legitimate devices can perform specific tasks or access sensitive networks and environments. For example, robots used in factories, management devices in smart cities, and autonomous devices in the medical field each require a high level of reliability and security, and verification of device legitimacy is mandatory to ensure this. Through the authentication of legitimate devices, potential threats such as unauthorized intrusion, data corruption, and network disruption can be prevented, ultimately ensuring the stability and security of device operations.

[0003] Existing authentication methods include password input using a keyboard or mouse, biometric recognition, RFID technology, sensor signal comparison, and methods utilizing Physical Unclonable Functions (PUF). While these methods can be useful in specific environments and conditions, they share several common limitations.

[0004] First, many existing methods require physical input devices or additional hardware for intelligent devices. This not only increases the size and cost of the device but also causes issues such as changes to the physical structure or increased maintenance costs. In particular, for mobile devices, such additional hardware can increase weight and size, thereby reducing the device's efficiency.

[0005] Second, the authentication process requires direct user interaction or complex data exchange with a central server. This process consumes large-scale computing resources, which can overload servers and networks, and increases the risk of information loss or distortion during data transmission. This acts as a major factor in degrading the reliability and efficiency of the entire authentication process.

[0006] Meanwhile, an acoustic signal-based authentication method has been introduced as an existing authentication method. This method has the advantage of being able to operate non-contactually and, unlike other methods, does not require additional hardware. However, this acoustic signal authentication method also has limitations, such as a limited operating distance, user inconvenience, and the vulnerability of the signal to external interference depending on the environment.

[0007] Therefore, a new acoustic signal-based technology is required to overcome the limitations of such existing technologies and to identify and authenticate legitimate intelligent devices simply, efficiently, and accurately without additional hardware.

[0008] The purpose of this application is to provide a device authentication method using an acoustic signal and a computer device for the same.

[0009] According to an embodiment of the present application, a method for authenticating a device using an acoustic signal is provided. The method may include the steps of: recording an acoustic signal transmitted from at least one device to be authenticated to obtain an acoustic reception signal; inputting converted data obtained by converting the acoustic reception signal into the frequency domain into a pre-trained deep learning model to extract a first feature of the acoustic reception signal; and comparing the first feature with a pre-registered reference feature to perform authentication of the device to be authenticated.

[0010] Additionally, the method further includes the step of generating a frequency response curve (FRC) based on the magnitude ratio at each frequency between the acoustic reception signal and a predetermined design signal, and obtaining a second feature of the acoustic reception signal from the frequency response curve, and the step of performing authentication for the device to be authenticated can be performed by comparing the first feature and the second feature with the reference feature.

[0011] In addition, when each of the above-mentioned devices transmits the same design signal due to the unique physical characteristics of the acoustic output unit, the acoustic reception signal obtained from each of the above-mentioned devices may include at least one of the unique first characteristic and the second characteristic to enable identification of each of the above-mentioned devices.

[0012] In addition, the design signal is a high-frequency acoustic signal in the range of 19 kHz to 22 kHz, and the device subject to certification may transmit the design signal to perform certification.

[0013] In addition, the above design signal can be generated randomly so that phases do not overlap in order to minimize signal interference between each frequency component.

[0014] In addition, the deep learning model may be configured to include a recurrent neural network to extract the spatiotemporal relationship of the acoustic reception signal.

[0015] Additionally, the method may further include the steps of: causing a plurality of devices subject to certification registration to transmit the design signal and recording the transmitted acoustic signal to obtain a second acoustic reception signal; extracting the first feature and the second feature from the second acoustic reception signal, respectively; and registering the first feature and the second feature corresponding to the devices subject to certification registration as unique reference features of each of the devices subject to certification registration.

[0016] A computer program is provided according to an embodiment of the present application. The program may be stored on a recording medium to execute a method according to an embodiment of the present application.

[0017] According to an embodiment of the present application, a computer device for performing device authentication using an acoustic signal is provided. The device comprises at least one processor; and a memory for storing a program executable by the processor. By executing the program, the processor can record an acoustic signal transmitted from at least one device to be authenticated to obtain an acoustic reception signal, input the converted data obtained by converting the acoustic reception signal into the frequency domain into a pre-trained deep learning model to extract a first feature of the acoustic reception signal, and compare the first feature with a pre-registered reference feature to perform authentication of the device to be authenticated.

[0018] According to the embodiments of the present application, by utilizing a frequency response curve (FRC) and a deep learning-based feature extraction method together, stable authentication performance can be provided even with environmental noise or changes in distance from the device to be authenticated.

[0019] According to the embodiments of the present application, by performing authentication using FRC and deep learning-based features derived from the unique physical characteristics of the acoustic output unit of the device to be authenticated, the device to be authenticated can be uniquely identified without the provision of additional hardware.

[0020] According to the embodiments of the present application, by designing an authentication signal that is inaudible to human hearing using an inaudible high-frequency signal, it is possible to effectively reduce acoustic interference that may occur in the surrounding environment without causing inconvenience to the user.

[0021] The effects obtainable from the embodiments of the present application are not limited to those mentioned above, and other unmentioned effects will be clearly understood by those skilled in the art to which the present application pertains from the description below.

[0022] A brief description of each drawing is provided to help to better understand the drawings cited in this application.

[0023] FIG. 1 is a drawing for explaining an authentication system using an acoustic signal according to an embodiment of the present application.

[0024] FIG. 2 is a flowchart of a device authentication method using an acoustic signal according to an embodiment of the present application.

[0025] FIG. 3 is a flowchart of a device authentication method using an acoustic signal according to an embodiment of the present application.

[0026] FIGS. 4a and FIGS. 4b are drawings for exemplarily illustrating design signals for device authentication according to an embodiment of the present application.

[0027] FIG. 5 is a diagram illustrating an exemplary frequency domain conversion signal according to an embodiment of the present application.

[0028] FIG. 6 is a diagram illustrating an exemplary frequency response curve according to an embodiment of the present application.

[0029] FIG. 7 is a diagram illustrating, by way of example, a device authentication process using an acoustic signal according to an embodiment of the present application.

[0030] FIG. 8 is a diagram illustrating the performance of a device authentication method using an acoustic signal according to an embodiment of the present application.

[0031] FIG. 9 is a block diagram showing the configuration of a computer device for performing device authentication using an acoustic signal according to an embodiment of the present application.

[0032] The technical concept of the present application is subject to various modifications and may have various embodiments, and specific embodiments are illustrated in the drawings and described in detail. However, this is not intended to limit the technical concept of the present application to specific embodiments, and it should be understood that it includes all modifications, equivalents, and substitutions that fall within the scope of the technical concept of the present application.

[0033] In explaining the technical concept of the present application, detailed descriptions of related prior art are omitted if it is determined that such descriptions may unnecessarily obscure the essence of the present application.

[0034] The terms used herein are for describing embodiments and are not intended to limit or / or restrict the present application. Singular expressions include plural expressions unless the context clearly indicates otherwise. Additionally, numbers used herein (e.g., First, Second, etc.) are merely identifiers to distinguish one component from another.

[0035] In this specification, when it is stated that a part is connected to another part, this includes not only cases where they are directly connected, but also cases where they are indirectly connected with other components in between. Furthermore, when it is stated that a part includes a certain component, this means that, unless specifically stated otherwise, it does not exclude other components but may include additional components.

[0036] Furthermore, in this application, the term "or" is intended to mean an implicit "or" rather than an exclusive "or." That is, unless otherwise specified or evident from the context, "X uses A or B" is intended to mean one of the natural implicit substitutions. In other words, if X uses A; if X uses B; or if X uses both A and B, "X uses A or B" may apply to any of these cases. Additionally, the term "and / or" as used herein should be understood to refer to and include all possible combinations of one or more of the enumerated related configurations.

[0037] In addition, terms such as “~part,” “~device,” “~device,” and “~module” described in this application refer to a unit that processes at least one function or operation, and this can be implemented as hardware or software or a combination of hardware and software, such as a processor, microprocessor, microcontroller, CPU (Central Processing Unit), GPU (Graphics Processing Unit), APU (Accelerated Processor Unit), DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA (Field Programmable Gate Array), etc.

[0038] Furthermore, it is intended to clarify that the classification of the components in this application is merely based on the primary function each component is responsible for. That is, two or more components described below may be combined into a single component, or a single component may be divided into two or more components based on more subdivided functions. Additionally, each component described below may additionally perform some or all of the functions performed by other components in addition to its own primary function, and it is obvious that some of the primary functions performed by each component may be exclusively performed by other components.

[0039]

[0040] The method according to the embodiment of the present application may be performed on a personal computer, workstation, server computer device, etc., equipped with computing power, or on a separate device for this purpose.

[0041] Additionally, the method may be performed on one or more computing devices. For example, at least one step of the method according to an embodiment of the present application may be performed on a client device, and other steps may be performed on a server device. In this case, the client device and the server device may be connected via a network to transmit and receive computation results. Alternatively, the method may be performed by distributed computing technology.

[0042]

[0043] In this application, the term "artificial intelligence learning model" may be used interchangeably with "artificial intelligence model," "computational model," "machine learning model," etc. The artificial intelligence learning model may be trained by various algorithms, such as, for example, decision tree, random forest, Gaussian naive bayes, k-nearest neighbor, Ada Boost, support vector machine, voting, bagging, neural network, and deep learning. However, it is not limited thereto.

[0044] An artificial intelligence learning model can be trained using at least one of supervised learning, unsupervised learning, semi-supervised learning, or reinforcement learning. The training of an artificial intelligence learning model may be a process of applying knowledge to the model to perform a specific action.

[0045] When algorithms such as neural networks or deep learning are applied to an artificial intelligence learning model, the AI ​​learning model may be referred to as a network function. The term "network function" can be used interchangeably with "neural network." A neural network can generally be composed of a set of interconnected computational units referred to as nodes. These nodes may also be referred to as neurons. A neural network is composed of at least one node, and the nodes may be interconnected by one or more links.

[0046] Neural networks may include deep neural networks (DNNs). Deep neural networks may include convolutional neural networks (CNNs), recurrent neural networks (RNNs), autoencoders, restricted Boltzmann machines (RBMs), deep belief networks (DBNs), Q networks, U networks, Siamese networks, and Generative Adversarial Networks (GANs), but are not limited to these.

[0047]

[0048] Hereinafter, embodiments of the present application will be described in detail in turn.

[0049]

[0050] FIG. 1 is a drawing for explaining an authentication system using an acoustic signal according to an embodiment of the present application.

[0051] Referring to FIG. 1, the system includes a plurality of intelligent devices (10) and an authentication device (20), and the authentication device (20) can authenticate whether the intelligent device (10) corresponds to a legally registered device.

[0052] Here, the intelligent device (10) may include various types of devices such as drones, robots, etc., and each device may include an acoustic output device having unique physical characteristics.

[0053] During the authentication process, the intelligent device (10) transmits a specific acoustic signal in response to a request from the authentication device (20), and the authentication device (20) records the transmitted acoustic signal and can extract unique characteristics of the acoustic signal through processes such as deep learning-based analysis and frequency response curve (FRC) generation based on the above. Subsequently, the authentication device (20) can determine whether the intelligent device (10) is a legitimate device by comparing the extracted characteristics with pre-registered reference characteristics.

[0054] This authentication system applies a method that utilizes the unique acoustic output characteristics of the device (10) for a specifically designed signal, thereby having high accuracy and robustness, and can operate stably in various environments.

[0055] The configuration of the system illustrated in FIG. 1 is exemplary, and various configurations may be applied according to embodiments of the present application.

[0056]

[0057] FIG. 2 is a flowchart of a device authentication method using an acoustic signal according to an embodiment of the present application.

[0058] In step S210, the authentication device (20) can obtain an acoustic reception signal by recording an acoustic signal transmitted from at least one authentication target device (10).

[0059] As described above, the authentication device (20) can identify the device to be authenticated (10) and perform authentication based on the acoustic signal output by the device to be authenticated (10). This utilizes the unique physical characteristics of the acoustic output device (i.e., speaker) of the device to be authenticated. That is, even if the device to be authenticated (10) transmits the same design signal, the acoustic output section of each device generates minute distortion or deformation, and such distortion can act as a characteristic that can uniquely identify the device to be authenticated. In other words, the acoustic reception signal obtained from each device to be authenticated (10) may include unique characteristics (first characteristics and / or second characteristics) that enable individual identification of the device to be authenticated (10) by means of the unique physical characteristics of the acoustic output device. Through this, the acoustic signal enables highly reliable authentication based on the physical characteristics of the device. To perform such authentication, the characteristics of the acoustic signals transmitted from the acoustic output devices of multiple devices may be stored in a database in advance by the authentication device (20).

[0060] For example, when a device subject to authentication (10) transmits an authentication request within a specific area requiring authentication, the authentication device (20) can transmit information regarding an acoustic signal designed for authentication (i.e., a design signal) to the device subject to authentication (10). Based on this information, the device subject to authentication (10) transmits the design signal through its acoustic output device, and in this process, the unique physical characteristics of the device are reflected in the design signal. The authentication device (20) can record this to generate an acoustic reception signal and proceed with a subsequent authentication procedure.

[0061] In an embodiment, the device to be authenticated (10) may have prior information regarding an already designed acoustic signal. In this case, the device to be authenticated can output the design signal along with the authentication request without needing to receive separate design signal information.

[0062] In step S220, the authentication device (20) can input the converted data, which is converted from the acoustic reception signal into the frequency domain, into a pre-trained deep learning model to extract the first feature of the acoustic reception signal.

[0063] Here, frequency domain transformation is the process of converting an acoustic reception signal from the time domain to the frequency domain, and can be performed using a Fourier transform technique such as the Short-Time Fourier Transform (STFT), as illustrated in Fig. 5, for example. This transformed data reveals the unique frequency patterns and relationships of the acoustic signal and can be used to analyze the unique physical characteristics of the device to be authenticated.

[0064] A pre-trained deep learning model can effectively analyze the pattern of the acoustic reception signal based on frequency domain transformed data and extract a first feature that reflects the unique acoustic output characteristics of the device to be authenticated (i.e., the acoustic output device of the device to be authenticated (10)).

[0065] In an embodiment, the deep learning model may be configured to include a Recurrent Neural Network (RNN) structure. Through this, the deep learning model is designed to learn the spatiotemporal relationships of acoustic signals, thereby enabling the derivation of a first feature that is robust against noise or environmental changes.

[0066] In step S230, the authentication device (20) generates a frequency response curve (FRC) based on the magnitude ratio at each frequency between the acoustic reception signal and a predetermined design signal, and can obtain a second feature of the acoustic reception signal from the frequency response curve.

[0067] Here, the design signal may refer to a signal in a specific frequency band that the authentication device (20) pre-defines to perform authentication. For example, the design signal may consist of an inaudible high-frequency acoustic signal in the range of 19 kHz to 22 kHz, and may be designed so that the phase of each frequency component is randomized to minimize signal interference. This design signal can effectively reflect the physical characteristics of the device to be authenticated without causing discomfort to the user by using frequencies outside the range of human hearing.

[0068] Meanwhile, the Frequency Response Curve (FRC) is a visual representation of the frequency-specific magnitude changes and distortions that occur as an acoustic signal passes through the acoustic output device of the device to be certified, and can reflect the unique acoustic output characteristics of the device to be certified (i.e., the acoustic output device of the device to be certified (10)). The frequency response curve is derived by comparing the magnitudes of the design signal and the acoustic reception signal at each frequency. Specifically, as shown in FIG. 6, the magnitude ratio at each frequency can be calculated to represent the degree of signal change at each frequency in the form of a curve.

[0069] The authentication device (20) can derive a second feature of the acoustic reception signal by analyzing the frequency response curve. The second feature may include the pattern of the FRC, the relationship between frequencies, the change in the magnitude of a specific frequency band, etc.

[0070] In step S240, the authentication device (20) can perform authentication of the device to be authenticated (10) by comparing the first feature and the second feature with a pre-registered reference feature.

[0071] For example, the authentication device (20) may compare the first feature and the second feature with the reference feature, respectively, and then determine the legality of the device to be authenticated (10) by combining the results of the comparison of the two features. Additionally, for example, the authentication device (20) may compare the entire feature, which is a combination of the first feature and the second feature, with the reference feature, and then determine the legality of the device to be authenticated (10) based on this.

[0072] In the embodiment, the comparison between features can be performed through cosine similarity. Cosine similarity can quantitatively measure the similarity between two features by calculating the cosine value of the angle formed by two vectors in a vector space. At this time, the closer the calculated cosine value is to 1, the more similar the directions of the two vectors are, and this can indicate that the device to be authenticated is legitimate. Accordingly, the authentication device (20) can calculate the cosine similarity between each of the first feature and the second feature and the reference feature, and determine the legitimacy of the device to be authenticated (10) through comparison with a predetermined threshold value.

[0073] The first feature is that it can extract spatiotemporal relationships and unique patterns of acoustic signals by learning features that are strong against noise and environmental changes through a deep learning model, and the second feature is that it can express the unique physical characteristics of the device (10) to be certified based on the ratio of magnitudes at each frequency between the design signal and the acoustic reception signal.

[0074] The method (100) according to the embodiment of the present application utilizes both of these two features to provide higher accuracy and robustness, and enables reliable authentication to be performed in various environments. That is, errors that may occur due to noise, environmental changes, signal distortion, etc., can be effectively reduced, and in particular, even if one feature shows slight variation, the other feature can compensate for this to perform more stable authentication.

[0075] Meanwhile, although FIG. 2 describes that authentication of the device (10) to be authenticated is performed using both the first feature and the second feature, according to the embodiment, the authentication device (20) may be implemented to perform authentication based on one of the first feature and the second feature.

[0076] Additionally, the method (200) illustrated in FIG. 2 is exemplary, and various configurations may be applied according to embodiments of the present application.

[0077]

[0078] FIG. 3 is a flowchart of a method for certifying a device using an acoustic signal according to an embodiment of the present application. More specifically, FIG. 3 is a flowchart illustrating a process of performing prior registration for a legitimate device by registering the acoustic output characteristics of a plurality of devices subject to certification registration as reference characteristics.

[0079] In step S310, the authentication device (20) can cause a plurality of authentication registration target devices to transmit a design signal and record the transmitted acoustic signal to obtain a second acoustic reception signal.

[0080] In step S320, the authentication device (20) can extract the first feature and the second feature, respectively, from the second acoustic reception signal.

[0081] Here, the first feature refers to a feature extracted using a deep learning model as in step S220 of FIG. 2, and the second feature may refer to a feature extracted based on a frequency response curve as in step S230 of FIG. 2.

[0082] In step S330, the authentication device (20) may register the first feature and the second feature corresponding to the device subject to authentication registration as unique reference features for each device subject to authentication registration. Registration of the reference features may be performed by matching information about the device subject to authentication registration (device ID, etc.) with information about the reference features and storing them in a database.

[0083] The registered reference feature is used as a comparison standard in the authentication process described above with reference to FIG. 2, and the legitimacy can be verified by determining whether the acoustic signal transmitted by the device to be authenticated (10) matches the reference feature stored in the database.

[0084] Through steps S310 to S330, multiple devices subject to certification registration can be registered as legitimate devices based on their respective unique acoustic output characteristics.

[0085] The method (300) illustrated in FIG. 3 is exemplary, and various configurations may be applied according to embodiments of the present application.

[0086]

[0087] FIGS. 4a and 4b are drawings illustrating an exemplary design signal for device authentication according to an embodiment of the present application. Specifically, FIG. 4a illustrates a design signal for authentication designed randomly such that the phases of each frequency component do not overlap, and FIG. 4b illustrates a signal implemented such that the phases of at least some frequency components overlap.

[0088] Referring to Fig. 4a, the design signal for certification can be designed by randomizing the phase of each frequency component. This allows for the derivation of a stable and robust frequency response curve (FRC) by effectively reducing signal interference in specific frequency bands and minimizing signal distortion. This random phase design can more accurately reflect physical distortions occurring in the interaction between the design signal and the acoustic output device, and can provide a basis for reliably identifying the unique characteristics of the device to be certified.

[0089] On the other hand, as shown in Fig. 4b, when a signal with superimposed phases of frequency components is applied, the likelihood of interference between frequency components increases, and a phenomenon in which the signal is amplified or canceled out in a specific frequency band may occur. This can degrade the stability of the frequency response curve and may not accurately reflect the physical characteristics of the device to be certified.

[0090] Therefore, in this application, the reliability and accuracy of authentication were improved by using a randomized design signal such that the phases of each frequency component do not overlap, as shown in FIG. 4a.

[0091]

[0092] FIG. 7 is a diagram illustrating, by way of example, a device authentication process using an acoustic signal according to an embodiment of the present application.

[0093] First, an acoustic signal transmitted from the device to be authenticated (10) is received and recorded by the authentication device (20) and converted into an acoustic reception signal.

[0094] The recorded acoustic reception signal is converted into frequency domain data through a Fourier transform (e.g., Short-Time Fourier Transform, STFT). The Fourier transform converts the signal in the time domain into the frequency domain to generate data containing magnitude and phase information for each frequency. The data converted into the frequency domain is input into a deep learning model (710) to extract the first feature of the acoustic signal.

[0095] At the same time, the authentication device (20) generates a frequency response curve (FRC) by calculating the magnitude ratio at each frequency between the acoustic reception signal and the predefined design signal. The generated frequency response curve is processed through the FRC feature extraction unit (720) to derive a second feature.

[0096] The extracted first and second features are transmitted to the judgment unit (730) and compared with reference features stored in the database (DB) in advance. In this process, quantitative comparison techniques such as cosine similarity may be used to evaluate the similarity between the features. The judgment unit (730) synthesizes the comparison results of the two features to determine the legitimacy of the device (10) to be certified and finally outputs the certification result.

[0097]

[0098] FIG. 8 is a diagram illustrating the performance of a device authentication method using an acoustic signal according to an embodiment of the present application.

[0099] For performance testing, seven different audio output devices were used. Specifically, two Google Nest Audio speakers (N-Audio 1, 2), two IKEA SYMFONISK speakers (SYMFONISK 1, 2), a Samsung Smart TV (SamsungTV), a Sonos speaker (SonosSB), and a Harman Kardon speaker (HAKA) were registered as legitimate devices according to the method of Fig. 3, and subsequently, performance tests were performed on the certification results using the method of Fig. 2.

[0100] Referring to FIG. 8, test results show that the method according to the embodiment of the present application exhibits high authentication performance in terms of F1-score, Error-rate, Precision, and Recall for all acoustic output devices.

[0101]

[0102] FIG. 9 is a block diagram showing the configuration of a computer device for performing device authentication using an acoustic signal according to an embodiment of the present application.

[0103] Referring to FIG. 9, the computer device (900) may include a communication unit (910), an input unit (920), a memory (930), and a processor (940). The computer device (900) may be the authentication device (20) described above with reference to FIG. 1.

[0104] The communication unit (910) can receive or transmit data from inside or outside. The communication unit (910) may include a wired or wireless communication unit. If the communication unit (910) includes a wired communication unit, the communication unit (910) may include one or more components that enable communication through a Local Area Network (LAN), a Wide Area Network (WAN), a Value Added Network (VAN), a mobile radio communication network, a satellite communication network, and combinations thereof. Additionally, if the communication unit (910) includes a wireless communication unit, the communication unit (910) may transmit or receive data or signals wirelessly using cellular communication, a wireless LAN (e.g., Wi-Fi), etc. In an embodiment, the communication unit (910) may transmit or receive data or signals to and from an external device or an external server under the control of a processor (940).

[0105] The input unit (920) can receive various user commands through external operation. To this end, the input unit (920) may include or be connected to one or more input devices. For example, the input unit (920) may receive user commands by being connected to an interface for various inputs, such as a keypad or a mouse. To this end, the input unit (920) may include an interface such as a Thunderbolt as well as a USB port. Additionally, the input unit (920) may include or be combined with various input devices, such as a touchscreen or a button, to receive external user commands. In an embodiment, the input unit (920) may further include an audio receiver that receives an external audio signal and converts it into data. The audio receiver may be implemented through a speaker device.

[0106] The memory (930) can store programs and / or program instructions for the operation of the processor (940) and can temporarily or permanently store input / output data. Specifically, the memory (930) can store various data, programs (one or more instructions), applications, software, instructions, code, etc. for driving and controlling the processor (940). For example, the memory (930) may include at least one type of storage medium among flash memory type, hard disk type, multimedia card micro type, card type memory (e.g., SD or XD memory, etc.), RAM, SRAM, ROM, EEPROM, PROM, magnetic memory, magnetic disk, and optical disk.

[0107] In an embodiment, the memory (930) may store instructions for implementing at least one module, learning model, etc. for processing a method according to the embodiments of the present application.

[0108] The processor (940) can control the overall operation of the computer device (900). The processor (940) can execute one or more programs or software stored in memory (930). The processor (940) may refer to a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), or a dedicated processor (940) on which the methods according to embodiments of the present application are performed.

[0109] In an embodiment, the processor (940) can implement the FRC feature extraction unit (720) and the judgment unit (730) described above with reference to FIG. 7 by executing one or more programs or software stored in memory (930).

[0110]

[0111] In an embodiment, the processor (940) can obtain an acoustic reception signal by recording an acoustic signal transmitted from at least one device to be authenticated, and input the converted data obtained by converting the acoustic reception signal into the frequency domain into a pre-trained deep learning model to extract a first feature of the acoustic reception signal, and perform authentication of the device to be authenticated by comparing the first feature with a pre-registered reference feature.

[0112] In an embodiment, the processor (940) can generate a frequency response curve (FRC) based on the magnitude ratio at each frequency between an acoustic reception signal and a predetermined design signal, obtain a second feature of the acoustic reception signal from the frequency response curve, and perform authentication of the device to be authenticated by comparing the first feature and the second feature with a reference feature.

[0113] Here, the design signal is a high-frequency acoustic signal in the range of 19 kHz to 22 kHz, and the device subject to certification may transmit the design signal to perform certification. In addition, the design signal may be generated randomly so that phases do not overlap in order to minimize signal interference between each frequency component. Furthermore, the deep learning model may be configured to include a Recurrent Neural Network to extract spatiotemporal relationships of the received acoustic signal.

[0114] In an embodiment, the processor (940) causes a plurality of devices subject to certification registration to transmit a design signal, records the transmitted acoustic signal to obtain a second acoustic reception signal, extracts a first feature and a second feature from the second acoustic reception signal, and registers the first feature and the second feature corresponding to each device subject to certification registration as unique reference features of each device subject to certification registration.

[0115] The configuration of the computer device (900) shown in FIG. 9 is exemplary, and various configurations may be applied according to embodiments of the present application.

[0116]

[0117] The method according to an embodiment of the present application may be implemented in the form of program instructions that can be executed through various computer means and recorded on a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, etc., either alone or in combination. The program instructions recorded on the medium may be those specifically designed and configured for the present application or may be those known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes; optical recording media such as CD-ROMs and DVDs; magneto-optical media such as floptical disks; and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include machine code, such as that generated by a compiler, as well as high-level language code that can be executed by a computer using an interpreter, etc.

[0118] Additionally, the method according to the disclosed embodiments may be provided by being included in a computer program product. The computer program product may be traded between a seller and a buyer as a product.

[0119] A computer program product may include a software program and a computer-readable storage medium on which the software program is stored. For example, a computer program product may include a product in the form of a software program (e.g., a downloadable app) that is electronically distributed through a manufacturer of an electronic device or an electronic market (e.g., Google Play Store, App Store). For electronic distribution, at least a portion of the software program may be stored on a storage medium or temporarily created. In this case, the storage medium may be a server of the manufacturer, a server of the electronic market, or a storage medium of a relay server that temporarily stores the software program.

[0120] A computer program product may include a storage medium of a server or a storage medium of a client device in a system composed of a server and a client device. Alternatively, if there is a third device (e.g., a smartphone) that communicates with the server or the client device, the computer program product may include a storage medium of the third device. Alternatively, the computer program product may include the S / W program itself that is transmitted from the server to the client device or the third device, or transmitted from the third device to the client device.

[0121] In this case, one of the server, the client device, and the third device may execute the computer program product to perform the method according to the disclosed embodiments. Alternatively, two or more of the server, the client device, and the third device may execute the computer program product to perform the method according to the disclosed embodiments in a distributed manner.

[0122] For example, a server (e.g., a cloud server or an artificial intelligence server, etc.) can execute a computer program product stored on the server to control a client device connected to the server in communication to perform the method according to the disclosed embodiments.

[0123]

[0124] Although the embodiments have been described in detail above, the scope of the present application is not limited thereto, and various modifications and improvements by those skilled in the art using the basic concept of the present application as defined in the following claims also fall within the scope of the present application.

Claims

1. As a device authentication method using acoustic signals, A step of obtaining an acoustic reception signal by recording an acoustic signal transmitted from at least one authentication target device; A step of extracting a first feature of the acoustic reception signal by inputting the converted data, obtained by converting the acoustic reception signal into the frequency domain, into a pre-trained deep learning model; and A method comprising the step of performing authentication of the device to be authenticated by comparing the first feature with a pre-registered reference feature.

2. In Paragraph 1, The method further includes the step of generating a frequency response curve (FRC) based on the magnitude ratio at each frequency between the acoustic reception signal and a predetermined design signal, and obtaining a second feature of the acoustic reception signal from the frequency response curve. The step of performing authentication for the above-mentioned authentication target device is, A method performed by comparing the first feature and the second feature with the reference feature.

3. In Paragraph 2, A method in which, when each of the above-mentioned devices for authentication transmits the same design signal due to the unique physical characteristics of the acoustic output section, the acoustic reception signal obtained from each of the above-mentioned devices for authentication includes at least one of the unique first characteristic and the second characteristic so as to enable identification of each of the above-mentioned devices for authentication.

4. In Paragraph 2, The above design signal is a high-frequency acoustic signal in the range of 19 kHz to 22 kHz, and A method in which the above-mentioned device subject to authentication transmits the above-mentioned design signal to perform authentication.

5. In Paragraph 4, A method in which the above design signal is randomly generated so that the phases do not overlap in order to minimize signal interference between each frequency component.

6. In Paragraph 2, A method in which the deep learning model is configured to include a recurrent neural network to extract the spatiotemporal relationship of the acoustic reception signal.

7. In Paragraph 2, A step of causing a plurality of certified registered devices to transmit the design signal, and recording the transmitted acoustic signal to obtain a second acoustic reception signal; A step of extracting the first feature and the second feature, respectively, from the second acoustic reception signal; and A method further comprising the step of registering the first feature and the second feature corresponding to the device subject to certification registration as unique reference features of each of the devices subject to certification registration.

8. A computer program stored on a recording medium to execute a method according to any one of paragraphs 1 through 7.

9. A computer device for performing device authentication using acoustic signals, At least one processor; and It includes memory for storing a program executable by the above processor, and A device comprising: a processor that, by executing the program, records an acoustic signal transmitted from at least one device subject to authentication to obtain an acoustic reception signal, inputs converted data obtained by converting the acoustic reception signal into the frequency domain into a pre-trained deep learning model to extract a first feature of the acoustic reception signal, and compares the first feature with a pre-registered reference feature to perform authentication of the device subject to authentication.

Images