Method and device for determining whether the modulation method of the received signal matches the modulation method of the signal processing device
The method analyzes received signals using multidimensional vector rotation and division to enhance modulation classification accuracy, addressing the limitations of existing deep learning-based methods and improving system efficiency in non-cooperative communication scenarios.
Patent Information
- Application Number
- KR1020240085209
- Authority / Receiving Office
- KR · KR
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2024-06-28
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2044-06-28
AI Technical Summary
Existing automatic modulation classification methods using deep learning struggle to accurately determine the modulation scheme of a received signal when it is not included in the classification targets, leading to degraded performance and reduced system efficiency, especially in non-cooperative communication scenarios.
A method and apparatus that analyze the characteristics of a received signal by rotating and dividing input data at arbitrary angles, generating multidimensional vectors, and determining the number of divisions to identify if the modulation scheme matches the modulation scheme of a signal processing device, using statistical values and rotation matrices to enhance classification accuracy.
Enables reliable detection of new modulation schemes and improves overall system efficiency by accurately identifying signals from unknown sources, enhancing responsiveness in various communication environments.
Smart Images

Figure 112024070324736-PAT00006_ABST
Abstract
Description
Technology Field
[0001] The present invention relates to a method for determining whether the modulation scheme of a received signal matches the modulation scheme of a signal processing device, and to an apparatus for the same. Background Technology
[0002] The content described in this section merely provides background information regarding the present embodiment and does not constitute prior art.
[0003] Automatic modulation classification methods cannot effectively determine if the modulation scheme of a received signal is not included in the modulation schemes being considered, so it is necessary to classify it in advance.
[0004] Automatic modulation classification is a technology that identifies the modulation scheme used in a transmitted signal using only the received signal. In non-cooperative communication situations, information necessary for signal recovery, such as the communication specifications used by the transmitter, is not provided, so communication specifications must be estimated; in particular, the estimation of the modulation scheme is essential to extract symbols from the received signal.
[0005] While automatic modulation classification methods utilizing deep learning have been proposed, such existing automatic modulation classification methods suffer from a problem where classification performance decreases when the modulation method used at the transmitting end is not included in the modulation classification targets considered.
[0006] To overcome these limitations, there is a need for research and development on technology that can effectively determine whether the modulation scheme of a received signal falls within the modulation classification targets being considered prior to modulation classification. Prior art literature
[0007] Korean Patent Publication No. 10-2052103 (2019.11.28.) The problem to be solved
[0008] The objective of the present invention is to provide a method for determining whether the modulation method of a received signal matches the modulation method of a signal processing device and an apparatus for the same, which can analyze the characteristics of a received signal and enable reliable detection of signals including new modulation methods that existing deep learning-based methods cannot process through analysis by rotating and dividing at arbitrary angles, and can improve the overall efficiency and accuracy of the system.
[0009] Other unspecified objects of the present invention may be further considered to the extent that they can be easily inferred from the following detailed description and effects. means of solving the problem
[0010] A device for determining whether a modulation method of a received signal and a modulation method of a signal processing device match, according to an embodiment of the present invention for achieving the above-described purpose, comprises: a memory storing one or more programs for determining whether a modulation method of a received signal and a modulation method of a signal processing device match; and one or more processors that perform operations according to the one or more programs. The processor generates a symbol sequence using a modulation method used by the signal processing device, generates a first multidimensional vector based on the characteristics of the symbol sequence, generates a second multidimensional vector based on the characteristics of a received signal received from the outside, generates input data by combining the first multidimensional vector and the second multidimensional vector, determines an arbitrary axis in a multidimensional space corresponding to the input data, determines an arbitrary value among the values of the dimensions corresponding to the determined axis, divides the input data into a plurality of groups based on the determined value, repeats the division until only data obtained from the received signal remains, counts the number of divisions, and determines whether the received signal is a signal using a modulation method used by the signal processing device based on the number of divisions.
[0011] The processor is characterized by generating a reference signal by applying noise to the symbol sequence, calculating a first statistical value representing the characteristics of the reference signal, and generating the first multidimensional vector based on the first statistical value.
[0012] The processor is characterized by calculating a second statistical value representing the characteristics of the received signal and generating the second multidimensional vector based on the second statistical value.
[0013] The processor is characterized by rotating the input data at an arbitrary angle to redistribute it, and determining an arbitrary axis in a multidimensional space corresponding to the redistributed input data.
[0014] The above processor is characterized by redistributing the input data by rotating it by any angle within the range of 0 to 2π.
[0015] The above processor is characterized by rotating the input data by multiplying the input data by a rotation matrix.
[0016] The processor is characterized by determining that the modulation method of the received signal is one of at least one modulation method used by the signal processing device when the number of divisions exceeds a predetermined threshold value, and determining that the modulation method of the received signal is not one of at least one modulation method used by the signal processing device when the number of divisions is less than or equal to a predetermined threshold value.
[0017] A method performed in a device comprising a memory storing one or more programs for determining whether a modulation method of a received signal and a modulation method of a signal processing device match, according to an embodiment of the present invention for achieving the above-described purpose, and one or more processors performing operations according to said one or more programs, may include: generating a symbol sequence using a modulation method used by the signal processing device, generating a first multidimensional vector based on the characteristics of said symbol sequence, generating a second multidimensional vector based on the characteristics of a received signal received from the outside, and generating input data by combining the first multidimensional vector and the second multidimensional vector; determining an arbitrary axis in a multidimensional space corresponding to said input data, determining an arbitrary value among the values of the dimensions corresponding to said axis, and dividing said input data into a plurality of groups based on said determined value; repeating said division until only data obtained from said received signal remains and counting said divisions; and determining whether said received signal is a signal using a modulation method used by said signal processing device based on said divisions.
[0018] The step of generating the input data is characterized by generating a reference signal by applying noise to the symbol sequence, calculating a first statistical value representing the characteristics of the reference signal, and generating the first multidimensional vector based on the first statistical value.
[0019] The step of generating the input data is characterized by calculating a second statistical value representing the characteristics of the received signal and generating the second multidimensional vector based on the second statistical value.
[0020] The step of dividing the input data into a plurality of groups is characterized by rotating the input data at an arbitrary angle to redistribute it, and determining an arbitrary axis in a multidimensional space corresponding to the redistributed input data.
[0021] The step of determining whether the received signal is a signal using a modulation method used by the signal processing device based on the number of divisions is characterized by determining that the modulation method of the received signal is one of at least one modulation method used by the signal processing device when the number of divisions exceeds a predetermined threshold value, and determining that the modulation method of the received signal is not one of at least one modulation method used by the signal processing device when the number of divisions is less than or equal to a predetermined threshold value.
[0022] A computer program according to one embodiment of the present invention for achieving the above-described purpose is stored in a computer-readable recording medium and executes on a computer any one of the methods for determining whether the modulation method of the received signal and the modulation method of the signal processing device match. Effects of the invention
[0023] As described above, according to one embodiment of the present invention, by applying a method for determining whether the modulation scheme of a received signal matches the modulation scheme of a signal processing device and a device for such determination, the characteristics of the received signal are analyzed, and reliable detection is possible even for signals including new modulation schemes that existing deep learning-based methods cannot process through analysis by rotating and dividing at arbitrary angles, thereby improving the overall efficiency and accuracy of the system. In particular, it enhances the responsiveness to unexpected signals or signals received from unknown signal sources, enabling stable and efficient signal processing in various communication environments.
[0024] Even if an effect is not explicitly mentioned herein, the effects and potential effects described in the following specification expected by the technical features of the present invention are treated as described in the specification of the present invention. Brief explanation of the drawing
[0025] FIG. 1 is a diagram illustrating the configuration of a device for determining whether a modulation method of a received signal matches a modulation method of a signal processing device according to one embodiment of the present invention. FIG. 2 is a flowchart illustrating a method for determining whether a modulation method of a received signal matches a modulation method of a signal processing device according to an embodiment of the present invention. Figure 3 is a table showing the theoretical values of higher-order cumulants calculated for signals generated using BPSK (Binary Phase Shift Keying), QPSK (Quadrature Phase Shift Keying), 8PSK (8-ary Phase Shift Keying), 16QAM (16-ary Quadrature Amplitude Modulation), and 64QAM (64-ary Quadrature Amplitude Modulation) modulation methods. FIG. 4 is a diagram showing the distribution of the average number of divisions obtained by applying a method for determining whether the modulation method of a received signal and the modulation method of a signal processing device match according to an embodiment of the present invention for each modulation method. FIG. 5 is a diagram showing an ROC curve for calculating AUROC, which is a performance indicator of a method for determining whether the modulation method of a received signal matches the modulation method of a signal processing device according to an embodiment of the present invention. Specific details for implementing the invention
[0026] Hereinafter, embodiments of the present invention will be described in detail with reference to the attached drawings. The advantages and features of the present invention, and the methods for achieving them, will become clear by referring to the embodiments described below in detail together with the attached drawings. However, the present invention is not limited to the embodiments disclosed below but may be implemented in various different forms. These embodiments are provided merely to ensure that the disclosure of the present invention is complete and to fully inform those skilled in the art of the scope of the invention, and the present invention is defined only by the scope of the claims. Unless otherwise defined, all terms used in this specification (including technical and scientific terms) may be used in a meaning that is commonly understood by those skilled in the art to which the present invention belongs. Furthermore, terms defined in commonly used dictionaries are not to be interpreted ideally or excessively unless explicitly and specifically defined otherwise.
[0027] The terms used in this application are used merely to describe specific embodiments and are not intended to limit the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, terms such as “have,” “may have,” “include,” or “may include” are intended to indicate the existence of the features, numbers, steps, actions, components, parts, or combinations thereof described in the specification, and should be understood as not precluding the existence or addition of one or more other features, numbers, steps, actions, components, parts, or combinations thereof. Terms including ordinal numbers, such as “second,” “first,” etc., may be used to describe various components, but said components are not limited by said terms.
[0028] The above terms are used solely for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the second component may be named the first component, and similarly, the first component may be named the second component. The term "and / or" includes a combination of a plurality of related described items or any of a plurality of related described items.
[0029] In this specification, identification symbols (e.g., a, b, c, etc.) for each step are used for convenience of explanation and do not indicate the order of the steps; the steps may occur differently from the specified order unless the context clearly indicates a specific order. That is, the steps may occur in the same order as specified, may be performed substantially simultaneously, or may be performed in the reverse order.
[0030] Existing automatic modulation classification methods utilizing deep learning have a problem in that they lack an effective mechanism to determine when a received signal's modulation scheme is not included in the classification targets being considered. Consequently, it is difficult to quickly and accurately determine the signal's modulation scheme, leading to significantly degraded classification performance and reduced overall system efficiency. Furthermore, since these methods are designed to be optimized only for signals of specific modulation schemes, it is difficult to effectively classify unexpected new signals.
[0031] In this invention, it is determined whether the modulation method of a received signal is included in the modulation classification targets being considered, using only the received signal without prior information. This improves modulation classification performance and overall system efficiency, and strengthens the ability to respond to the reception of unexpected signals.
[0032] Various embodiments of a method for determining whether a modulation method of a received signal matches a modulation method of a signal processing device and a device for such a method, according to the present invention, will be described in detail below with reference to the attached drawings.
[0033] The embodiments described in this specification may be applied to automatic modulation classification devices and signal information collection or analysis equipment.
[0034] In the present invention, a signal processing device is a device used to analyze, convert, classify electrical signals, or extract desired information, and encompasses a modulation classification device, a signal information collection device, and a signal analysis device. The modulation classification device identifies the modulation method of a received signal and analyzes the corresponding information, the signal information collection device gathers useful information from various signals, and the signal analysis device analyzes the collected signals in detail to extract specific patterns or characteristics. Such devices play an essential role in various fields, such as communications, radar, satellites, and information security.
[0035] FIG. 1 is a diagram illustrating the configuration of a device for determining whether a modulation method of a received signal matches a modulation method of a signal processing device according to one embodiment of the present invention.
[0036] A device (100) for determining whether the modulation method of a received signal matches the modulation method of a signal processing device includes at least one processor (110), a computer-readable storage medium (120), and a communication bus (150).
[0037] The processor (110) can be controlled to operate as a device (100) for determining whether the modulation method of a received signal matches the modulation method of a signal processing device. For example, the processor (110) can execute one or more programs (121) stored in a computer-readable storage medium (120). One or more programs (121) may include one or more computer-executable instructions, and the computer-executable instructions may be configured to cause the device (100) for determining whether the modulation method of a received signal matches the modulation method of a signal processing device to perform operations according to an exemplary embodiment when executed by the processor (110).
[0038] A computer-readable storage medium (120) is configured to store computer-executable instructions or program code, program data and / or other suitable forms of information. Computer-executable instructions or program code, program data and / or other suitable forms of information may also be provided through an input / output interface (130) or a communication interface (140). A program (121) stored in the computer-readable storage medium (120) includes a set of instructions executable by a processor (110). In one embodiment, the computer-readable storage medium (120) may be a memory (volatile memory such as random access memory, non-volatile memory, or a suitable combination thereof), one or more magnetic disk storage devices, optical disk storage devices, flash memory devices, or other forms of storage media that can store desired information and are accessed by a device (100) that determines whether the modulation method of a received signal matches the modulation method of a signal processing device, or a suitable combination thereof.
[0039] The communication bus (150) interconnects various other components of the device (100) for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, including a processor (110) and a computer-readable storage medium (120).
[0040] The device (100) for determining whether the modulation method of a received signal matches the modulation method of a signal processing device may also include one or more input / output interfaces (130) and one or more communication interfaces (140) that provide interfaces for one or more input / output devices. The input / output interface (130) and the communication interface (140) are connected to a communication bus (150). An input / output device (not shown) may be connected to other components of the device (100) for determining whether the modulation method of a received signal matches the modulation method of a signal processing device through the input / output interface (130).
[0041] The processor (110) generates a symbol sequence using a modulation method used by a signal processing device, generates a first multidimensional vector based on the characteristics of the symbol sequence, generates a second multidimensional vector based on the characteristics of a received signal received from the outside, generates input data by combining the first multidimensional vector and the second multidimensional vector, determines an arbitrary axis in a multidimensional space corresponding to the input data, determines an arbitrary value among the values of the dimensions corresponding to the determined axis, divides the input data into a plurality of groups based on the determined value, repeats the division until only the data obtained from the received signal remains, counts the number of divisions, and determines whether the received signal is a signal using the modulation method used by the signal processing device based on the number of divisions.
[0042] The processor (110) preferably generates a plurality of first multidimensional vectors, but is not necessarily limited thereto.
[0043] The process of generating a symbol sequence using the modulation method employed by a signal processing device may involve determining the type of modulation method and generating a sequence of modulated symbols accordingly. For example, if the signal processing device uses the Binary Phase Shift Keying (BPSK) modulation method, data bits are converted into two different phases (0 degrees and 180 degrees) to form a symbol sequence. If the Quadrature Phase Shift Keying (QPSK) modulation method is used, data bits are converted into four different phases (0 degrees, 90 degrees, 180 degrees, and 270 degrees) to form a symbol sequence. The symbol sequence generated in this way is arranged continuously over time, and each symbol represents a specific signal waveform according to the modulation method.
[0044] The processor (110) can generate a reference signal by applying noise to the symbol sequence, calculate a first statistical value representing the characteristics of the reference signal, and generate the first multidimensional vector based on the first statistical value.
[0045] The processor (110) can calculate a second statistical value representing the characteristics of the received signal and generate the second multidimensional vector based on the second statistical value.
[0046] Applying noise refers to the process of modifying a signal by artificially adding noise to the original signal. This can be used to simulate and analyze the effects of noise that may occur when a signal is transmitted in a real-world environment.
[0047] The first statistical value may be a statistical indicator representing the characteristics of a reference signal generated by applying noise to a symbol sequence generated by a modulation method. The first statistical value includes higher-order cumulant values of the reference signal and can be used to analyze the characteristics of the signal. For example, the second moment (C) of the reference signal 20), 4th cumulant (C 40 ), 4th-order cumulant including complex conjugate (C 42 ), 4th cumulant using only real components (C 40I ) values may correspond to the first statistical value.
[0048] The second statistical value may be a statistical indicator representing the characteristics of a signal received from an external source. The second statistical value includes statistical characteristic values such as higher-order cumulants and can be used to analyze the characteristics of the received signal. For example, the second moment (C) of the received signal 20 ), 4th cumulant (C 40 ), 4th-order cumulant including complex conjugate (C 42 ), 4th cumulant using only real components (C 40I ) values may correspond to the second statistical value.
[0049] The first multidimensional vector is derived from the characteristics of a symbol sequence generated by the processor based on a modulation scheme. First, a symbol sequence is generated using a modulation scheme, and noise is added to the sequence to generate a reference signal. Subsequently, a first statistical value representing the characteristics of the reference signal is calculated, and a first multidimensional vector is formed based on this. This vector primarily includes high-order statistical values such as the cumulant and reflects the signal's energy, asymmetry, kurtosis, and phase information. For example, the second moment (C) of the received signal 20 ), 4th cumulant (C 40 ), 4th-order cumulant including complex conjugate (C 42 ), 4th cumulant using real components (C 40I ) values are included.
[0050] The second multidimensional vector is derived from the characteristics of a signal received from the outside. A second statistical value representing the characteristics of the received signal is calculated. A second multidimensional vector is formed based on the second statistical value, and this vector also includes higher-order statistical values such as cumulants. To accurately represent the characteristics of the received signal, the second multidimensional vector includes the second moment (C) of the received signal. 20 ), 4th cumulant (C 40 ), 4th-order cumulant including complex conjugate (C 42 ), 4th cumulant using real components (C 40I Includes ) values.
[0051] Generating input data by combining a first multidimensional vector and a second multidimensional vector may involve combining the two vectors to form a single multidimensional data set. This combined input data may be in the form of combining each dimension of the first multidimensional vector and the second multidimensional vector. More specifically, during the combination process, each element of the first multidimensional vector and each element of the second multidimensional vector may be merged to create a single new vector.
[0052] There may be multiple first multidimensional vectors, and the combination process may be a process of stacking four-dimensional vectors vertically. Specifically, the input data constructed by combining the first multidimensional vector and the second multidimensional vector may take the form of an Nx4 matrix. Here, N may represent the total number of the first multidimensional vector and the second multidimensional vector.
[0053] Determining an arbitrary axis in a multidimensional space corresponding to input data may refer to the process of selecting a specific axis to analyze input data composed of a combined first multidimensional vector and a second multidimensional vector in a multidimensional space. In a multidimensional space, each dimension represents the characteristics or variables of the data, and determining an arbitrary axis in such a space is used to divide or transform the data in a specific direction. For example, when rotating or dividing data by selecting a specific axis in a four-dimensional space, the distribution or characteristics of the data can be analyzed based on this axis, and the selected axis can be used as a reference point in subsequent data processing.
[0054] Dividing input data into multiple groups based on a random value selected from the values of the dimensions corresponding to a determined axis can refer to the process of randomly selecting one of the values along an axis of a selected multidimensional space and dividing the data into several groups based on that value. For example, if the input data consists of a 4-dimensional vector, a specific axis (e.g., the first dimension) is selected, and the data is divided into two groups based on a random value (e.g., the median) among the values of this axis. This criterion value is determined based on a specific characteristic of the data, and dividing the data in this way can result in each group forming a data set distinguished according to the value of the corresponding axis. This division process can be performed repeatedly.
[0055] Repeating the partitioning process until only data obtained from the received signal remains can be interpreted as a process of repeatedly partitioning the input data to ultimately retain only data possessing characteristics that match the received signal. For example, the input data is initially divided into two groups based on an arbitrary axis, and among these partitioned groups, the group containing more data matching the characteristics of the received signal is selected. Subsequently, the selected group is partitioned again based on a different axis, and in this process, the group containing more data with characteristics matching the received signal is again selected. This partitioning process is repeated until only data satisfying the characteristics of the received signal remains.
[0056] The processor (110) can rotate the input data at an arbitrary angle to redistribute it and determine an arbitrary axis in a multidimensional space corresponding to the redistributed input data.
[0057] That is, according to one embodiment of the present invention, the processor (110) generates a first multidimensional vector, generates a second multidimensional vector based on the characteristics of a received signal received from the outside, generates input data by combining the first multidimensional vector and the second multidimensional vector, generates redistributed input data by rotating the input data at an arbitrary angle, determines an arbitrary axis in a multidimensional space corresponding to the redistributed input data, determines an arbitrary value among the values of the dimension corresponding to the determined axis, divides the redistributed input data into a plurality of groups based on the determined value, repeats the division until only the data obtained from the received signal remains, counts the number of divisions, and determines whether the received signal is a signal using a modulation method used by the signal processing device based on the number of divisions.
[0058] The process of the processor (110) rotating input data at an arbitrary angle may mean a process of creating a new distribution by rotating the data in a multidimensional space where the input data exists. This rotation is performed using a rotation matrix and includes transforming the data at various angles to analyze the structure or pattern of the data. The rotated data may have a form different from the original distribution. For example, if input data is rotated in a four-dimensional space, the relative position and distribution of the data change according to the angle of rotation. This is referred to as redistribution in the present invention.
[0059] The processor (110) can redistribute the input data by rotating it by any angle within the range of 0 to 2π.
[0060] The above processor can rotate the input data by multiplying the input data by a rotation matrix.
[0061] The processor (110) may determine that if the number of divisions exceeds a predetermined threshold value, the modulation method of the received signal is one of at least one modulation method used by the signal processing device, and if the number of divisions is less than or equal to a predetermined threshold value, the modulation method of the received signal is not one of at least one modulation method used by the signal processing device.
[0062] In this invention, the threshold value is a numerical value serving as a criterion for determining the modulation method of a received signal, and is used to classify the signal as normal or abnormal based on the number of divisions. The threshold value is set to distinguish between the modulation method used by the signal processing device and that which is not. The number of divisions is recorded until the input data is divided multiple times and only data matching the characteristics of the received signal remains; if this number of divisions exceeds the threshold value, the received signal is determined to be a normal modulation method used by the signal processing device. Conversely, if the number of divisions is less than or equal to the threshold value, the received signal is considered to be an abnormal modulation method not used by the signal processing device.
[0063] The threshold is determined based on the distribution of the number of splits for pre-examined modulation schemes. The average number of splits obtained from multiple experiments on normal modulation schemes is analyzed, and an appropriate threshold is set based on this value. This can be adjusted depending on the usage environment and the applied modulation scheme, and the performance of modulation scheme classification can be optimized by setting the optimal threshold. If the threshold is too low, there is a higher probability of misidentifying normal modulation schemes as abnormal, while conversely, if it is too high, there is a higher probability of misidentifying abnormal modulation schemes as normal.
[0064] The threshold determination method can be based on statistical analysis and experimental data. First, multiple experiments are conducted for each modulation scheme used by the signal processing unit, and the average number of divisions obtained for each scheme is recorded. To this end, multiple signals are generated under various signal-to-noise ratio (SNR) conditions, their higher-order cumulants are calculated and converted into multidimensional vectors, and then this data is analyzed through a division process. Sufficient data samples are collected for each modulation scheme to calculate the average number of divisions, and their distribution is analyzed.
[0065] For example, the average and standard deviation of the number of divisions are calculated through 1,000 experiments each for BPSK and QPSK modulation schemes. A threshold is set based on this collected data on the number of divisions. Based on the average number of divisions and standard deviation of normal modulation schemes, the threshold is usually determined by adding or subtracting a constant multiple of the standard deviation from the average value. For example, if the average number of divisions for BPSK modulation is 25 and the standard deviation is 3, the threshold can be set to 34, which is the average value plus three times the standard deviation. This ensures that the number of divisions for normal modulation schemes mostly falls below the threshold, thereby enabling the identification of normal modulation schemes.
[0066] FIG. 2 is a flowchart illustrating a method for determining whether a modulation method of a received signal matches a modulation method of a signal processing device according to an embodiment of the present invention.
[0067] A method for determining whether the modulation scheme of a received signal matches the modulation scheme of a signal processing device may be performed in a device comprising a memory storing one or more programs for determining whether the modulation scheme of a received signal matches the modulation scheme of a signal processing device and one or more processors performing operations according to one or more programs. For example, a method for determining whether the modulation scheme of a received signal matches the modulation scheme of a signal processing device may be performed by a device for determining whether the modulation scheme of a received signal matches the modulation scheme of a signal processing device as described through FIG. 1.
[0068] In step S210, a symbol sequence is generated using a modulation method utilized by a signal processing device, a first multidimensional vector is generated based on the characteristics of the symbol sequence, a second multidimensional vector is generated based on the characteristics of a received signal received from the outside, and input data can be generated by combining the first multidimensional vector and the second multidimensional vector.
[0069] In step S220, an arbitrary axis in a multidimensional space corresponding to the input data is determined, and an arbitrary value is determined among the values of the dimensions corresponding to the determined axis, and the input data can be divided into a plurality of groups based on the determined value.
[0070] In step S230, the division can be repeated until only the data obtained from the received signal remains, and the number of divisions can be counted.
[0071] In step S240, based on the number of divisions, it can be determined whether the received signal is a signal using the modulation method used by the signal processing device.
[0072] The processor may repeat steps S220 and S230 multiple times to calculate a number of divisions, and calculate the average value of the calculated number of divisions, and determine whether the received signal is a signal using the modulation method used by the signal processing device based on the average value. That is, the processor may determine whether the received signal is a signal using the modulation method used by the signal processing device based on whether the average value exceeds a threshold value.
[0073] The step of generating the input data is characterized by generating a reference signal by applying noise to the symbol sequence, calculating a first statistical value representing the characteristics of the reference signal, and generating the first multidimensional vector based on the first statistical value.
[0074] The step of generating the input data is characterized by calculating a second statistical value representing the characteristics of the received signal and generating the second multidimensional vector based on the second statistical value.
[0075] The step of dividing the input data into a plurality of groups is characterized by rotating the input data at an arbitrary angle to redistribute it, and determining an arbitrary axis in a multidimensional space corresponding to the redistributed input data.
[0076] The step of determining whether the received signal is a signal using a modulation method used by the signal processing device based on the number of divisions is characterized by determining that the modulation method of the received signal is one of at least one modulation method used by the signal processing device when the number of divisions exceeds a predetermined threshold value, and determining that the modulation method of the received signal is not one of at least one modulation method used by the signal processing device when the number of divisions is less than or equal to a predetermined threshold value.
[0077] The symbols received from the signal that has passed through the additive white Gaussian noise channel are as shown in Equation 1 below.
[0078]
[0079] Here, y[n] is the nth received symbol, x[n] is the nth transmitted symbol, and w[n] is additive white Gaussian noise. y[n] has different statistical characteristics depending on the modulation method. A statistical value that can effectively represent these signal characteristics is the cumulant, which can be calculated as shown in Equation 2 below.
[0080]
[0081] Here, E[·] is the expectation operator, cum[·] is the cumulant operator, and y * [n] is the complex conjugate of y[n], y I [n] is a component of the in-hoc relationship of y[n]. C 20 , C 40 , C 42 , C 40I are higher-order cumulant values representing the statistical characteristics of the received signal. C 20is the second moment of the received signal y[n], representing the signal energy. C 40 ε is the fourth-order cumulant of the received signal y[n], a value representing the signal's asymmetry and kurtosis. C 42 is the fourth-order cumulant containing the complex conjugate of the received signal y[n], reflecting the phase information of the signal. C 40I is a fourth-order cumulant using only the real components of the received signal y[n], representing the asymmetry and kurtosis of the real part of the signal.
[0082] For random variables A, B, C, and D with a mean of 0, the 4th cumulant can be calculated as shown in Equation 3 below.
[0083]
[0084] Figure 3 is a table showing the theoretical values of higher-order cumulants calculated for signals generated using BPSK (Binary Phase Shift Keying), QPSK (Quadrature Phase Shift Keying), 8PSK (8-ary Phase Shift Keying), 16QAM (16-ary Quadrature Amplitude Modulation), and 64QAM (64-ary Quadrature Amplitude Modulation) modulation methods.
[0085] In this invention, outlier detection using an unsupervised modulation method utilizing RIF (Rotated Isolation Forest) is performed using a higher-order cumulant as input.
[0086] First, a symbol sequence is generated using the target modulation method, and then noise is added to randomly generate a signal and calculate the higher-order cumulant of each signal. By repeating this process, multiple 4-dimensional vectors can be constructed with the four types of calculated higher-order cumulants as elements. At this time, the vectors calculated from each modulation method are called generated data, and the generated data will be closely distributed around the theoretical cumulant values of Fig. 3.
[0087] Afterwards, a 4-dimensional vector is generated from the received signal r[k] in the same way, with four types of higher-order cumulants as elements, and combined with the generated data calculated earlier. The generated data and the vector calculated from the received signal can be collectively referred to as input data.
[0088] To detect anomalies in the modulation method, the input data is first rotated and redistributed by an arbitrary angle. To rotate the input data by an arbitrary angle, six rotation matrices as shown in Equation 4 below are multiplied to the input data. Here, θ1, θ2, θ3, θ4, θ5, and θ6 are arbitrary angles between [0, 2π].
[0089]
[0090] Here, R1, R2, R3, R4, R5, and R6 represent the rotation matrices available to the processor.
[0091] Since the rotated input data is a set of 4-dimensional vectors, one of the four axes is arbitrarily selected, and an arbitrary value is chosen as a dividing boundary within the data range of that axis. Through this process, the rotated input data is divided based on the selected dividing boundary to leave only the data containing the vector calculated from the received signal. Subsequently, the division process is repeated until only the vector calculated from the received signal remains, and the number of divisions is recorded.
[0092] As described above, the process of recording the number of divisions after redistributing input data through rotation is repeated many times to calculate the average number of divisions, and the average number of divisions serves as a criterion for detecting outliers. If the modulation method of the received signal is one of the target modulation methods, the average number of divisions will be large. On the other hand, if the modulation method of the received signal is an untarget modulation method, the average number of divisions will be small.
[0093] Based on this feature, a specific threshold is set, and data is classified as normal if the average number of divisions is greater than the threshold, and as abnormal if it is less. In this case, the threshold can be determined according to the usage environment based on the average number of divisions of the modulation schemes under consideration that were previously investigated.
[0094] Deep learning-based modulation classification methods suffer from a problem where classification performance is significantly degraded and the overall efficiency of the system decreases when receiving a signal with a modulation method that was not previously considered as input. This invention solves this problem by providing a mechanism capable of determining whether the modulation method of a received signal is included in the classification target. This improves modulation classification performance, enhances system efficiency, and enables effective response even when receiving unexpected signals. Additionally, this invention facilitates the classification and collection of signals received from unknown signal sources, allowing them to be utilized for signal information analysis.
[0095] The following description is an example of the application of the present invention.
[0096] FIG. 4 is a diagram showing the distribution of the average number of divisions obtained by applying a method for determining whether the modulation method of a received signal and the modulation method of a signal processing device match according to an embodiment of the present invention for each modulation method.
[0097] FIG. 5 is a diagram showing an ROC curve for calculating AUROC, which is a performance indicator of a method for determining whether the modulation method of a received signal matches the modulation method of a signal processing device according to an embodiment of the present invention.
[0098] The modulation schemes considered in the modulation classification are set as BPSK, QPSK, and 8PSK, and these are defined as normal modulation schemes (i.e., modulation schemes used by the signal processing device). Conversely, 16QAM and 64QAM, which are not considered in the modulation classification, are defined as abnormal modulation schemes (i.e., modulation schemes not used by the signal processing device). The signal's Eb / N o It considers 1dB intervals from -5dB to 10dB.
[0099] E considered in each normal modulation scheme b / N o For each, 1,000 signal vectors containing 5,000 symbols are generated. For each generated signal vector, a higher-order cumulant |C 20 |, |C 40 | |C 42 |, |C 40I Calculate | to generate a 4-dimensional vector containing it as an element. At this time, the 48,000 4-dimensional high-order cumulant vectors generated by normal modulation are defined as generated data.
[0100] In addition to this, E that considers the received signal to determine whether it is an outlier in all modulation schemes b / N o For each, 200 4-dimensional high-order cumulant vectors are generated in the same way as above.
[0101] Data generated to determine whether they are outliers are added one by one to the previously generated data and used as input to the RIF model. The input data is rotated at an arbitrary angle and redistributed, and random splitting is applied to calculate the number of splits required 100 times until only the data made into a received signal remains, and the average number of splits is calculated by averaging these values. The average number of splits is represented as a histogram as shown in Figure 4.
[0102] In Figure 4, it can be seen that the average number of divisions of a signal generated by a normal modulation method is greater than the average number of divisions of a signal generated by an abnormal modulation method. In this embodiment, a threshold value of 20 is selected to determine whether a received signal is included in the modulation method being considered, and a signal having an average number of divisions smaller than that value is classified as an abnormal modulation method, while a signal having an average number of divisions larger than that value is determined as a normal modulation method.
[0103] In the case of binary classification, such as determining whether a value is outlier, performance varies depending on the threshold value. Therefore, the Area Under Receiver Operating Characteristic curve (AUROC), which is not affected by the threshold value, was used as a performance indicator for the described embodiment and is shown in Fig. 5.
[0104] The graph shown in Fig. 5 is a Receiver Operating Characteristic (ROC) curve illustrating the changes in False Positive Rate (FPR) and True Positive Rate (TPR) measured while gradually adjusting the threshold value. The area under the curve, known as AUROC, has a value between 0 and 1, and a higher AUROC value indicates that the two distributions being compared are well separated. The fact that the AUROC of the example is 0.995, which is very close to 1, demonstrates that the performance of the present invention is excellent.
[0105] Although FIG. 3 describes each process as being executed sequentially, this is merely an illustrative description, and a person skilled in the art may modify and adapt it in various ways without departing from the essential characteristics of the embodiment of the present invention, such as changing the order described in FIG. 3, executing one or more processes in parallel, or adding other processes.
[0106] The present application also provides a computer storage medium. Program instructions are stored in the computer storage medium, and when the program instructions are executed by a processor, a method for determining whether the modulation method of the received signal described above matches the modulation method of the signal processing device is realized.
[0107] A computer storage medium according to one embodiment of the present invention may be a U disk, SD card, PD optical drive, mobile hard disk, large capacity floppy drive, flash memory, multimedia memory card, server, etc., but is not necessarily limited thereto.
[0108] Although it is described that all components constituting the embodiments of the present invention described above are combined or operate in combination, the present invention is not necessarily limited to such embodiments. That is, within the scope of the purpose of the present invention, all such components may be selectively combined in one or more ways to operate. Furthermore, while all such components may each be implemented as a single independent piece of hardware, they may also be implemented as a computer program having a program module that performs some or all of the combined functions in one or more pieces of hardware by selectively combining some or all of the components. Additionally, such a computer program may be stored on a computer-readable media such as a USB memory, CD disk, or flash memory, and read and executed by a computer to implement the embodiments of the present invention. Magnetic recording media, optical recording media, etc., may be included as recording media for the computer program.
[0109] The foregoing description is merely an illustrative explanation of the technical concept of the present invention, and those skilled in the art to which the present invention pertains will be able to make various modifications, changes, and substitutions within the scope of the essential characteristics of the present invention. Accordingly, the embodiments disclosed in the present invention and the accompanying drawings are intended to explain, not limit, the technical concept of the present invention, and the scope of the technical concept of the present invention is not limited by such embodiments and accompanying drawings. The scope of protection of the present invention shall be interpreted by the claims below, and all technical concepts within an equivalent scope shall be interpreted as being included within the scope of rights of the present invention.
Claims
Claim 1 A device for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, comprising: a memory storing one or more programs for determining whether the modulation method of a received signal matches the modulation method of a signal processing device; and one or more processors performing operations according to the one or more programs; wherein the processor generates a symbol sequence using the modulation method used by the signal processing device, generates a first multidimensional vector based on the characteristics of the symbol sequence, generates a second multidimensional vector based on the characteristics of a received signal received from the outside, generates input data by combining the first multidimensional vector and the second multidimensional vector, determines an arbitrary axis in a multidimensional space corresponding to the input data, determines an arbitrary value among the values of the dimensions corresponding to the determined axis, divides the input data into a plurality of groups based on the determined value, repeats the division until only the data obtained from the received signal remains and counts the number of divisions, and determines whether the received signal is a signal using the modulation method used by the signal processing device based on the number of divisions. Claim 2 A device for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein, in claim 1, the processor generates a reference signal by applying noise to the symbol sequence, calculates a first statistical value representing the characteristics of the reference signal, and generates the first multidimensional vector based on the first statistical value. Claim 3 A device for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein, in claim 1, the processor calculates a second statistical value representing the characteristics of the received signal and generates the second multidimensional vector based on the second statistical value. Claim 4 A device for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein, in claim 1, the processor rotates the input data by an arbitrary angle and redistributes it, and determines an arbitrary axis in a multidimensional space corresponding to the redistributed input data. Claim 5 A device for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein, in claim 4, the processor rotates the input data by any angle within the range of 0 to 2π and redistributes it. Claim 6 A device for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein, in paragraph 4, the processor rotates the input data by multiplying the input data by a rotation matrix. Claim 7 A device for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein, in claim 1, the processor determines that the modulation method of the received signal is one of at least one modulation method used by the signal processing device when the number of divisions exceeds a predetermined threshold value, and determines that the modulation method of the received signal is not one of at least one modulation method used by the signal processing device when the number of divisions is less than or equal to a predetermined threshold value. Claim 8 A method for determining whether a modulation method of a received signal matches a modulation method of a signal processing device, comprising a memory storing one or more programs for determining whether a modulation method of a received signal matches a modulation method of a signal processing device, and one or more processors performing operations according to said one or more programs, wherein the method comprises: generating a symbol sequence using a modulation method used by the signal processing device, generating a first multidimensional vector based on the characteristics of said symbol sequence, generating a second multidimensional vector based on the characteristics of a received signal received from the outside, and generating input data by combining the first multidimensional vector and the second multidimensional vector; determining an arbitrary axis in a multidimensional space corresponding to the input data, determining an arbitrary value among the values of the dimensions corresponding to said axis, and dividing the input data into a plurality of groups based on said determined value; repeating said division until only data obtained from said received signal remains and counting the number of divisions; and determining whether said received signal is a signal using a modulation method used by the signal processing device based on said number of divisions. Claim 9 A method for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein the step of generating the input data comprises applying noise to the symbol sequence to generate a reference signal, calculating a first statistical value representing the characteristics of the reference signal, and generating the first multidimensional vector based on the first statistical value. Claim 10 A method for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein the step of generating the input data comprises calculating a second statistical value representing the characteristics of the received signal and generating the second multidimensional vector based on the second statistical value. Claim 11 A method for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein, in claim 8, the step of dividing the input data into a plurality of groups is characterized by rotating the input data at an arbitrary angle to redistribute it and determining an arbitrary axis in a multidimensional space corresponding to the redistributed input data. Claim 12 A method for determining whether the modulation method of a received signal matches the modulation method of a signal processing device, wherein, in claim 8, the step of determining whether the received signal is a signal using a modulation method used by the signal processing device based on the number of divisions is characterized by determining that the modulation method of the received signal is one of at least one modulation method used by the signal processing device when the number of divisions exceeds a predetermined threshold value, and determining that the modulation method of the received signal is not one of at least one modulation method used by the signal processing device when the number of divisions is less than or equal to a predetermined threshold value. Claim 13 A computer program stored on a computer-readable recording medium for executing on a computer a method for determining whether the modulation method of a received signal and the modulation method of a signal processing device described in any one of paragraphs 8 through 12 match.
Citation Information
Patent Citations
Communication device and transmission mode selectingmethod
KR1020020093964A
Symbol detection method for MIMO system
KR1020130143416A
Method and apparatus for providing modulation recognition in mobile communication system
KR1020140032190A
Classification method of modulation based on deep learning
KR1020180079050A
Automatic modulation recognition method with feature selection for uncertain fading channels
KR1020190031832A
