Interference suppression method based on sparse dictionary in OTFS system, electronic device and program product

Abstract

The application provides a method for interference suppression based on a sparse dictionary in an OTFS system, an electronic device and a program product, and relates to the technical field of signal processing. The method comprises the following steps: acquiring an input signal carrying an interference signal; performing signal conversion on the input signal to obtain a received signal; reconstructing the interference signal based on a preset sparse dictionary strategy to obtain a reconstructed interference signal; and performing interference suppression on the input signal according to the received signal and the reconstructed interference signal to obtain a target signal from which the interference signal is removed. In this way, the interference signal is separated and reconstructed by the sparse dictionary, so that the interference signal in the OTFS system is reconstructed and removed with a small calculation complexity, and the problem that the traditional OTFS system interference suppression method cannot effectively suppress the interference signal in a complex communication environment with a low calculation complexity is improved.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and more specifically, to an interference suppression method, electronic device, and program product based on a sparse dictionary in an OTFS system. Background Technology

[0002] Future next-generation wireless communication systems need to provide high-frequency, low-efficiency, and high-reliability communication in highly mobile scenarios, such as high-speed rail, low Earth orbit (LEO) satellites, and unmanned aerial vehicles (UAVs). However, due to the high Doppler shift in these scenarios, traditional modulation and demodulation techniques such as Orthogonal Frequency Division Multiplexing (OFDM) are no longer suitable. Orthogonal Time-Frequency Space Modulation (OTFS) is a promising modulation technique to address these challenges. It maps all information symbols into a two-dimensional delay-Doppler (DD) domain, effectively combating the dynamic changes of time-varying multipath channels.

[0003] Similarly, these application scenarios also present numerous interferences, which have severe consequences for the signal processing of the OTFS system. Interference in wireless communication scenarios can be broadly categorized into four types: man-made interference (malicious interference), external interference, inter-system interference, and intra-system interference. These different types of interference have fundamental differences and all affect communication quality. Man-made interference refers to interference emitted by a hostile party using a specialized jammer, exhibiting high targeting and dynamic characteristics. External interference is caused by electromagnetic signal leakage from the natural environment and other devices, such as power lines and radiation from electronic equipment, and is generally related to the geographical environment. Inter-system interference refers to interference between wireless communication systems caused by similar operating frequency bands. Intra-system interference is caused by signals or operations of components within the system, which may affect the normal operation of the system, leading to performance degradation or functional failure.

[0004] Existing OTFS systems have limited anti-interference methods, with research focusing on the latter two types of interference, such as inter-user interference and inter-symbol interference (ISI), while neglecting external interference that is widespread in real-world communication environments and seriously affects signal detection performance. Only a few existing technologies have made some improvements to signal detection algorithms for narrowband noise interference, impulse noise interference, and DD domain impulse interference. However, these methods cannot effectively suppress interference signals in complex communication environments with low computational complexity, leading to severe signal distortion at the receiver and ultimately communication failure. Summary of the Invention

[0005] In view of this, the purpose of this application is to provide an interference suppression method, electronic device and program product based on sparse dictionary in OTFS system, which can improve the problem that the traditional interference suppression method of OTFS system cannot effectively suppress interference signals in complex communication environment with low computational complexity.

[0006] To achieve the above technical objectives, the technical solution adopted in this application is as follows:

[0007] In a first aspect, embodiments of this application provide an interference suppression method based on a sparse dictionary in an OTFS system, the method comprising:

[0008] Acquire input signals carrying interference signals;

[0009] The input signal is converted to obtain the received signal;

[0010] Based on a preset sparse dictionary strategy, the interference signal is reconstructed to obtain the reconstructed interference signal;

[0011] Based on the received signal and the reconstructed interference signal, interference suppression is performed on the input signal to obtain the target signal after removing the interference signal.

[0012] In conjunction with the first aspect, in some alternative implementations, acquiring the input signal carrying the interference signal includes:

[0013] Obtaining time-frequency domain signals:

[0014]

[0015] In the formula, X tf Represents a time-frequency domain signal. Let X represent the DFT matrix at point M, where j represents the imaginary unit, and X... dd This represents a time-delay-Doppler two-dimensional matrix of size M×N. Represents an N-point IDFT matrix;

[0016] Performing a Heisenberg transform on the time-frequency domain signal yields the time-domain signal:

[0017]

[0018] In the formula, S1 represents the time-domain signal. Represents the IDFT matrix of point M;

[0019] The time-domain signal is filtered to obtain the transmitted signal:

[0020]

[0021] In the formula, S2 represents the transmitted signal, and G tx This represents the first pulse shaping function;

[0022] Vectorization is performed on the matrix columns of the transmitted signal to obtain a vector representation of the transmitted signal:

[0023]

[0024] In the formula, x1 = vecX dd , Indicates the Kronecker product;

[0025] The input signal is determined based on the vector representation of the transmitted signal:

[0026] r = Hs + J + w

[0027] In the formula, r represents the input signal. P represents the number of paths, h i Let represent the channel gain of the i-th path, Π be the permutation matrix, represent the time delay, Δ be a diagonal matrix of dimension MN×MN, represent the Doppler offset, J represent the interference signal, and w represent the time-domain noise signal, where:

[0028]

[0029] Δ=diag[z 0 ,z 1 ,…,z MN-1 ]

[0030]

[0031] In the formula, j represents the imaginary unit.

[0032] In conjunction with the first aspect, in some optional embodiments, the interference signal is a single-tone interference, and the time-domain interference signal vector of the interference signal is represented as:

[0033]

[0034] In the formula, c∈(0,1,…,MN-1), A represents the interference amplitude, c represents the sampling time, and T s The sampling period is represented by f, and the carrier frequency is represented by f. Indicates a random initial phase. Represents the vector dimension.

[0035] In conjunction with the first aspect, in some optional embodiments, the input signal is converted to obtain a received signal, including:

[0036] The input signal is converted into vector form to obtain the input vector:

[0037]

[0038] in, This is the first equivalent channel matrix. For the interference signal vector, G is the noise vector. rx This represents the second pulse shaping function;

[0039] When the first pulse shaping function and the second pulse shaping function are ideal rectangular window functions, the first equivalent channel matrix is ​​converted to:

[0040]

[0041] Among them, I M Let M×M be the identity matrix, and substitute... The second equivalent channel matrix can be obtained:

[0042]

[0043] In the formula,

[0044] The interference signal vector is represented as The noise signal vector is represented as: Based on the second equivalent channel matrix, the interference signal vector, and the noise signal vector, the received signal is determined to be...

[0045] In conjunction with the first aspect, in some optional implementations, the interference signal is reconstructed based on a preset sparse dictionary strategy to obtain a reconstructed interference signal, including:

[0046] The interference signal is reconstructed using a pre-trained dictionary and sparse coefficient vector to obtain the reconstructed interference signal:

[0047] i c =Dc γ c

[0048] In the formula, i c D represents the reconstructed interference signal. c This represents a pre-trained dictionary, γ c This represents a pre-trained sparse coefficient vector.

[0049] In conjunction with the first aspect, in some alternative embodiments, the method further includes, before acquiring the input signal carrying the interference signal:

[0050] Obtain the preset interference signal vector representation;

[0051] Based on the preset interference signal vector representation, a sparse dictionary is constructed using a preset sparse dictionary learning strategy to obtain the pre-trained dictionary and the pre-trained sparse coefficient vector.

[0052] In conjunction with the first aspect, in some optional implementations, based on the preset interference signal vector representation, a sparse dictionary is constructed using a preset sparse dictionary learning strategy to obtain the pre-trained dictionary and the pre-trained sparse coefficient vector, including:

[0053] Based on the preset interference signal vector representation, construct a complete dictionary:

[0054]

[0055] In the formula, D represents a complete dictionary, and F N This represents the discrete Fourier transform matrix at N points. I M Let J be an M×M identity matrix, where J represents the time-domain interference signal vector representation corresponding to the preset interference signal vector representation.

[0056] Based on the complete dictionary, construct the sparse coefficient vector corresponding to the complete dictionary: γ = [γ1, γ2, ..., γ K ];

[0057] Obtain the preset received signal: in, x1 = vecX dd , X represents the second equivalent matrix corresponding to the preset received signal. dd Let n1 represent a time delay-Doppler two-dimensional matrix of size M×N, where n1 represents the noise signal vector representation corresponding to the preset received signal.

[0058] Based on the complete dictionary and the sparse coefficient vector corresponding to the complete dictionary, the k'-th atom in the complete dictionary is updated;

[0059] Specifically, updating the k'-th atom in the complete dictionary based on the complete dictionary and the corresponding sparse coefficient vector includes:

[0060] Perform matrix decomposition on the K-rank matrix Dγ:

[0061]

[0062]

[0063] Fix all atoms in the complete dictionary except k', and determine the sparsity of the complete dictionary:

[0064]

[0065] In the formula, Represents the residual;

[0066] Extract the corresponding γ from the residual k' From the non-zero sequence, we obtain the residual matrix.

[0067] Perform singular value decomposition on the residual matrix:

[0068]

[0069] In the formula, U and V are unitary matrices, Σ is a diagonal matrix, and V T The apparatus matrix representing the unitary matrix V;

[0070] Determine V T The first column is used as the new atom d k' The product of the first row and first column of Σ and the first row of the singular matrix is ​​determined as the sparse coefficient vector corresponding to the new complete dictionary; expressed as:

[0071]

[0072] The k'-th atom in the complete dictionary is updated repeatedly based on the complete dictionary and the corresponding sparse coefficient vector until the sparsity of the complete dictionary satisfies the following convergence condition or the number of repetitions satisfies the preset maximum number of iterations:

[0073]

[0074] In the formula, U represents the preset sparsity threshold.

[0075] In conjunction with the first aspect, in some optional embodiments, interference suppression is performed on the input signal based on the received signal and the reconstructed interference signal to obtain a target signal after removing the interference signal, including:

[0076] Based on the reconstructed interference signal, the target signal is obtained by subtracting the reconstructed interference signal from the received signal:

[0077]

[0078] In the formula, y' represents the target signal, i c This represents the reconstructed interference signal.

[0079] Secondly, embodiments of this application also provide an electronic device, which includes a processor and a memory coupled to each other. The memory stores a computer program, and when the computer program is executed by the processor, the electronic device performs the above-described method.

[0080] Thirdly, embodiments of this application also provide a computer program product, including a computer program that implements the above-described method when executed by a processor.

[0081] The invention employing the above technical solution has the following advantages:

[0082] The technical solution provided in this application first acquires an input signal carrying interference signals and performs signal conversion on the input signal to obtain a received signal. Then, based on a preset sparse dictionary strategy, the interference signal is reconstructed to obtain a reconstructed interference signal. Finally, based on the received signal and the reconstructed interference signal, interference suppression is performed on the input signal to obtain the target signal after removing the interference signal. Thus, by separating and reconstructing the interference signal using a sparse dictionary, interference signals in the OTFS system can be reconstructed and removed with relatively low computational complexity, improving upon the problem that traditional OTFS system interference suppression methods cannot effectively suppress interference signals in complex communication environments with low computational complexity. Attached Figure Description

[0083] This application can be further illustrated by the non-limiting embodiments given in the accompanying drawings. It should be understood that the following drawings only illustrate some embodiments of this application and should not be considered as limiting the scope. For those skilled in the art, other related drawings can be obtained from these drawings without any inventive effort.

[0084] Figure 1 A structural block diagram of an electronic device provided in an embodiment of this application.

[0085] Figure 2 This is a structural block diagram of a communication connection system provided in an embodiment of this application.

[0086] Figure 3This is one of the flowcharts illustrating the interference suppression method based on a sparse dictionary in the OTFS system provided in this application embodiment.

[0087] Figure 4 This is the second schematic diagram of the interference suppression method based on sparse dictionaries in the OTFS system provided in the embodiments of this application.

[0088] Figure 5 The figure shows the bit error rate performance under different interference signal suppression strategies proposed in this application.

[0089] Icons: 100 - Electronic device; 101 - Processor; 102 - Memory. Detailed Implementation

[0090] The present application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that similar or identical parts are referred to by the same reference numerals in the drawings or description. Implementations not shown or described in the drawings are forms known to those skilled in the art. In the description of this application, terms such as "first" and "second" are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0091] Please refer to Figure 1 This application provides an electronic device 100 that may include a processor 101 and a memory 102. The memory 102 stores a computer program, which, when executed by the processor 101, enables the electronic device 100 to perform corresponding steps in the interference suppression method based on a sparse dictionary in the OTFS system described below.

[0092] In this embodiment, refer to Figure 2 The electronic device 100 can be a receiver in a communication connection system. This communication connection system may also include a signal transmitter and a jamming transmitter (used to simulate various types of interference signals; in practical applications, this device may be omitted when the signal transmitted by the transmitter carries interference signals). The electronic device 100 is used to acquire an input signal carrying interference signals and perform signal conversion on the input signal to obtain a received signal. Then, based on a preset sparse dictionary strategy, the interference signal is reconstructed to obtain a reconstructed interference signal. Finally, based on the received signal and the reconstructed interference signal, interference suppression is performed on the input signal to obtain the target signal after removing the interference signal.

[0093] In this embodiment, the processor 101 can be an integrated circuit chip with signal processing capabilities. The processor 101 can be a general-purpose processor. For example, the processor 101 can be a central processing unit (CPU), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, or discrete hardware components, capable of implementing or executing the methods, steps, and logic block diagrams disclosed in the embodiments of this application.

[0094] The memory 102 can be, but is not limited to, random access memory, read-only memory, programmable read-only memory, erasable programmable read-only memory, electrically erasable programmable read-only memory, etc. In this embodiment, the memory 102 can be used to store input signals, received signals, preset sparse dictionary strategies, reconstructed interference signals, target signals, etc. Of course, the memory 102 can also be used to store programs, which the processor 101 executes after receiving an execution instruction.

[0095] Understandable Figure 1 or Figure 2 The electronic device 100 shown is only a schematic diagram; the electronic device 100 may also include components that are more... Figure 1 or Figure 2 More components are shown. Figure 1 or Figure 2 The components shown can be implemented using hardware, software, or a combination thereof.

[0096] Please refer to Figure 3 This application also provides an interference suppression method based on a sparse dictionary in an OTFS system, which can be applied to the aforementioned electronic device 100 and executed or implemented by the electronic device 100. The interference suppression method based on a sparse dictionary in the OTFS system may include the following steps:

[0097] Step 210: Acquire the input signal carrying the interference signal;

[0098] Step 220: Perform signal conversion on the input signal to obtain the received signal;

[0099] Step 230: Based on a preset sparse dictionary strategy, the interference signal is reconstructed to obtain the reconstructed interference signal;

[0100] Step 240: Based on the received signal and the reconstructed interference signal, perform interference suppression on the input signal to obtain the target signal after removing the interference signal.

[0101] In the above implementation, the input signal carrying the interference signal is first acquired and then converted to obtain the received signal. Then, based on a preset sparse dictionary strategy, the interference signal is reconstructed to obtain the reconstructed interference signal. Finally, based on the received signal and the reconstructed interference signal, interference suppression is applied to the input signal to obtain the target signal after removing the interference signal. Thus, by separating and reconstructing the interference signal using a sparse dictionary, interference signals in the OTFS system can be reconstructed and removed with relatively low computational complexity. This improves upon the problem that traditional OTFS interference suppression methods cannot effectively suppress interference signals in complex communication environments with low computational complexity.

[0102] Understandably, in practical applications, the target signal after removing interference signals is usually used in highly mobile scenarios to locate transmitters mounted on mobile devices (such as high-speed trains, low Earth orbit satellites, drones, etc.).

[0103] The following is a detailed explanation of each step in the interference suppression method based on sparse dictionaries in the OTFS system:

[0104] In step 210, acquiring the input signal carrying the interference signal may include:

[0105] Obtain the time-frequency domain signal transmitted by the aforementioned signal transmitter:

[0106]

[0107] In the formula, X tf Represents a time-frequency domain signal. Let X represent the DFT matrix at point M, where j represents the imaginary unit, and X... dd This represents a time-delay-Doppler two-dimensional matrix of size M×N. Represents an N-point IDFT (Inverse Discrete Fourier Transform) matrix;

[0108] Then the time-frequency domain signal X tf Performing the Heisenberg transform, we obtain the time-domain signal:

[0109]

[0110] In the formula, S1 represents the time-domain signal. Represents the IDFT matrix of point M;

[0111] Then, the first pulse shaping function G is used. tx The transmitted signal is obtained by performing pulse shaping filtering on the time-domain signal at the transmitter end:

[0112]

[0113] In the formula, S2 represents the transmitted signal, and G tx This represents the first pulse shaping function, which is also the pulse shaping matrix at the transmitting end;

[0114] Vectorization is performed on the matrix columns of the transmitted signal S2 to obtain a vector representation of the transmitted signal with dimension MN×1:

[0115]

[0116] In the formula, x1 = vecX dd , Indicates the Kronecker product;

[0117] Based on the vector representation of the transmitted signal, the input signal at the receiving end (i.e., at the receiver) is determined:

[0118] r = Hs + J + w

[0119] In the formula, r represents the input signal, and the channel response matrix in the time domain is... P represents the number of paths (i.e., the signal has P propagation paths in total), h i Let Π represent the channel gain of the i-th path, Π be the permutation matrix, and Π represent the time delay. This represents the permutation matrix corresponding to the i-th path (i.e., the time delay corresponding to the i-th path). Let l represent the diagonal matrix corresponding to the i-th path (i.e., the Doppler offset corresponding to the i-th path). i k i Let represent the time delay tap (index) and Doppler tap (index) of the i-th path, respectively; Δ is a diagonal matrix of dimension MN×MN, representing the Doppler offset frequency; J represents the interference signal; and w represents the time-domain noise signal, where:

[0120]

[0121] Δ=diag[z 0 ,z 1 ,…,z MN-1 ]

[0122]

[0123] In the formula, j represents the imaginary unit.

[0124] In this embodiment, the input signal r received at the receiving end is converted into a time-frequency domain signal through shaping filtering and Wigner transform, and then transformed into a time-delay-Doppler domain signal through SFFT. Where R = vec -1 r.

[0125] In this embodiment, the interference signal can be a single-tone interference simulated by an interference transmitter, and the time-domain interference signal vector of the interference signal can be expressed as:

[0126]

[0127] In the formula, c∈(0,1,…,MN-1), A represents the interference amplitude, c represents the sampling time, and T s The sampling period is represented by f, and the carrier frequency is represented by f. Indicates a random initial phase. The dimension of the vector is MN×1.

[0128] Understandably, in practical applications, interference signals can also be any form of signal other than single-tone interference. Converting them into a time-domain signal vector representation and substituting it into the technical solution of this application can also achieve the removal of interference signals.

[0129] In step 220, the input signal is converted to obtain the received signal, which may include:

[0130] The input signal is converted into vector form to obtain the input vector:

[0131]

[0132] in, This is the first equivalent channel matrix. For the interference signal vector, G is the noise vector. rx This represents the second pulse shaping function, which is also the pulse shaping matrix at the receiving end;

[0133] When the first pulse shaping function and the second pulse shaping function are ideal rectangular window functions, the first equivalent channel matrix is ​​converted to:

[0134]

[0135] Among them, I M Let M×M be the identity matrix, and substitute... The second equivalent channel matrix can be obtained:

[0136]

[0137] In the formula,

[0138] In this embodiment, a matrix is ​​used. From the block cycle characteristics, we can know that P i The p-th and q-th elements of (0≤p,q≤MN-1) are:

[0139]

[0140] Where m∈[0,M-1] represents the frequency domain index, l i Let represent the delay index of the i-th path, where n∈[0,M-1] represents the time-domain index, and [·] M This is the modulo function.

[0141] Similarly, using matrices From the characteristics of block cycles, we can know that:

[0142]

[0143] Where 0≤p,q≤MN-1, m′=qn′M.

[0144] Let T i =P i Q i Then we have:

[0145]

[0146] in:

[0147]

[0148] Additionally, the interference signal vector can be represented as The noise signal vector is represented as: Based on the second equivalent channel matrix, the interference signal vector, and the noise signal vector, the received signal is determined to be...

[0149] In step 230, the interference signal is reconstructed based on a preset sparse dictionary strategy to obtain the reconstructed interference signal, which may include:

[0150] The interference signal is reconstructed using a pre-trained dictionary and sparse coefficient vector to obtain the reconstructed interference signal:

[0151] i c =D c γ c

[0152] In the formula, i cD represents the reconstructed interference signal. c This represents a pre-trained dictionary, γ c This represents a pre-trained sparse coefficient vector.

[0153] In step 240, based on the received signal and the reconstructed interference signal, interference suppression is performed on the input signal to obtain the target signal after removing the interference signal, which may include:

[0154] Based on the reconstructed interference signal, the target signal is obtained by subtracting the reconstructed interference signal from the received signal:

[0155]

[0156] In the formula, y' represents the target signal, i c This represents the reconstructed interference signal.

[0157] Let ε = i1 - i c ,but:

[0158]

[0159] In the formula, ε represents the disturbance estimation error.

[0160] As an optional implementation, prior to step 210, the method may further include:

[0161] Obtain the preset interference signal vector representation;

[0162] Based on the preset interference signal vector representation, a sparse dictionary is constructed using a preset sparse dictionary learning strategy to obtain the pre-trained dictionary and the pre-trained sparse coefficient vector.

[0163] In this embodiment, to facilitate the handling of interference signals, the received signal received in the DD domain can be represented as:

[0164] In this embodiment, based on the preset interference signal vector representation, a sparse dictionary is constructed using a preset sparse dictionary learning strategy to obtain the pre-trained dictionary and the pre-trained sparse coefficient vector, which may include:

[0165] Based on the preset interference signal vector representation, construct a complete dictionary:

[0166]

[0167] In the formula, D represents a complete dictionary, and F NI represents the N-point Discrete Fourier Transform (DFT) matrix. M Let J be an M×M identity matrix, where J represents the time-domain interference signal vector representation corresponding to the preset interference signal vector representation.

[0168] Based on the complete dictionary, construct the sparse coefficient vector corresponding to the complete dictionary: γ = [γ1, γ2, ..., γ K ];

[0169] The preset received signal can then be represented as: in, x1 = vecX dd , X represents the second equivalent matrix corresponding to the preset received signal. dd Let n1 represent a time delay-Doppler two-dimensional matrix of size M×N, where n1 represents the noise signal vector representation corresponding to the preset received signal.

[0170] Based on the complete dictionary and the sparse coefficient vector corresponding to the complete dictionary, the k'-th atom in the complete dictionary is updated;

[0171] The updating of the k'-th atom in the complete dictionary based on the complete dictionary and the corresponding sparse coefficient vector can include:

[0172] The K-rank matrix Dγ is decomposed into:

[0173]

[0174]

[0175] Fix all atoms in the complete dictionary except k', and determine the sparsity of the complete dictionary:

[0176]

[0177] In the formula, Represents the residual;

[0178] Understandably, the residuals cannot be directly solved here; directly solving for γ will yield a different result. k' It may not be sparse. Therefore, it is necessary to extract the corresponding γ from the residual. k' Extracting the non-zero sequences yields the residual matrix.

[0179] Then, singular value decomposition (SVD) is performed on the residual matrix:

[0180]

[0181] In the formula, Σ is a diagonal matrix, and V T The apparatus matrix representing the unitary matrix V;

[0182] During the atomic update phase, V is determined. T The first column is used as the new atom d k' The product of the first row and first column of Σ and the first row of the singular matrix is ​​determined as the sparse coefficient vector corresponding to the new complete dictionary; expressed as:

[0183]

[0184] Finally, the k'-th atom in the complete dictionary is updated repeatedly based on the complete dictionary and the corresponding sparse coefficient vector, until the sparsity of the complete dictionary satisfies the following convergence condition or the number of repetitions satisfies the preset maximum number of iterations:

[0185]

[0186] In the formula, U represents the preset sparsity threshold.

[0187] Understandably, in practical applications, multiple transmission tests of the input signal can be conducted, and interference signals can be added to the input signal during the transmission process to obtain multiple input signals as learning samples. The learning samples are then repeated in step 220 to obtain the corresponding interference signal vector representation and received signal, which are then used as the preset interference signal vector representation and preset received signal, respectively. That is, the parameters required in the sparse dictionary learning process can be the input signal obtained in steps 210 and 220 and the received signal obtained through signal conversion. For the sake of brevity and logical coherence, the formula parameters involved in the sparse dictionary learning process are not distinguished from the formula parameters in steps 210 and 220.

[0188] Understandable, refer to Figure 4 This application proposes an interference suppression method based on a sparse dictionary in an OTFS system, using single-tone interference as an example of the interference signal. After Wigner transform and SFFT at the receiver, the interference signal is reconstructed in the DD domain using a sparse dictionary and effectively suppressed. The specific steps include: constructing an atomic model and a complete dictionary of the interference signal in the DD domain; updating the dictionary and sparse coefficient vector using singular value decomposition; reconstructing the interference signal in the DD domain based on the learned dictionary and sparse coefficient vector; and subtracting the reconstructed interference signal from the initially received signal at the receiver to achieve effective suppression of the interference signal.

[0189] Reference Figure 5 , Figure 5In the diagram, the horizontal axis SNR represents the signal-to-noise ratio (SNR), and the vertical axis represents the bit error rate (BER) performance of signal detection.

[0190] The line “MMSE-OFDM” is a graph showing the change of BER as a function of SNR under the condition of interference (interference not suppressed) using OFDM modulation technology and MMSE detector;

[0191] The line “MMSE-OTFS” is a graph showing the change of BER as a function of SNR under the condition of interference (interference not suppressed) using OTFS modulation technology and MMSE detector;

[0192] The line “MMSE-OTFS-SDL” represents the curve of BER versus SNR after using OTFS modulation technology and the proposed scheme to suppress interference in the presence of interference, and then using an MMSE detector for signal detection.

[0193] The line “MP-OTFS” is the curve of BER as a function of SNR when interference is present (interference is not suppressed), using OTFS modulation technology and a message passing (MP) detector for signal detection.

[0194] The line “MP-OTFS-SDL” represents the curve of BER versus SNR when OTFS modulation is used and the proposed scheme is applied to suppress interference, and a message passing (MP) detector is used for signal detection under interference conditions.

[0195] Depend on Figure 5 It is evident that, without suppressing interference signals, the BER performance of the MP detector is slightly better than that of the MMSE detector. After suppressing interference signals using the technical solution of this application, both detectors achieve significant performance improvements, and the traditional MMSE detector also achieves better BER performance than the MP detector, reaching 10 when the SNR is 30dB. -3 Magnitude.

[0196] It should be noted that those skilled in the art will understand that, for the sake of convenience and brevity, the specific working process of the above-mentioned electronic device 100 can be referred to the corresponding process of each step in the aforementioned method, and will not be elaborated further here.

[0197] This application also provides a computer program product, including a computer program that, when executed by processor 101, implements the above-described interference suppression method based on a sparse dictionary in the OTFS system.

[0198] Based on the above description of the embodiments, those skilled in the art can clearly understand that this application can be implemented by hardware or by using software plus necessary general-purpose hardware platforms. Based on this understanding, the technical solution of this application can be embodied in the form of a software product. This software product can be stored in a non-volatile storage medium (such as CD-ROM, USB flash drive, mobile hard drive, etc.) and includes several instructions to cause a computer device (such as a personal computer, server, or network device, etc.) to execute the methods described in the various implementation scenarios of this application.

[0199] In summary, this application provides an interference suppression method, electronic device 100, and program product based on a sparse dictionary in an OTFS system. In this technical solution, an input signal carrying interference signals is first acquired and then converted to obtain a received signal. Then, based on a preset sparse dictionary strategy, the interference signal is reconstructed to obtain a reconstructed interference signal. Finally, based on the received signal and the reconstructed interference signal, interference suppression is performed on the input signal to obtain the target signal after removing the interference signal. Thus, by separating and reconstructing the interference signal using a sparse dictionary, interference signals in the OTFS system can be reconstructed and removed with relatively low computational complexity, improving upon the problem that traditional OTFS system interference suppression methods cannot effectively suppress interference signals in complex communication environments with low computational complexity.

[0200] In the embodiments provided in this application, it should be understood that the disclosed methods can also be implemented in other ways. The method embodiments described above are merely illustrative. For example, the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of this application. In this regard, each block in a flowchart or block diagram may represent a module, program segment, or part of code, which includes one or more executable instructions for implementing a specified logical function. It should also be noted that each block in a block diagram and / or flowchart, and combinations of blocks in block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or action, or using a combination of dedicated hardware and computer instructions. Furthermore, the functional modules in the various embodiments of this application can be integrated together to form an independent part, or each module can exist independently, or two or more modules can be integrated to form an independent part.

[0201] The above description is merely an embodiment of this application and is not intended to limit the scope of protection of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of protection of this application.

Claims

1. A method for interference mitigation based on sparse dictionary in OTFS system, characterized in that, The method comprises: acquiring a preset interference signal vector representation; constructing a sparse dictionary according to the preset interference signal vector representation through a preset sparse dictionary learning strategy, to obtain a pre-trained dictionary and a pre-trained sparse coefficient vector; acquiring an input signal carrying an interference signal; performing signal conversion on the input signal to obtain a received signal; reconstructing the interference signal based on a preset sparse dictionary strategy to obtain a reconstructed interference signal; performing interference suppression on the input signal according to the received signal and the reconstructed interference signal to obtain a target signal after removing the interference signal; the interference signal is a single-tone interference, and a time-domain interference signal vector representation of the interference signal is: wherein A denotes the interference amplitude, c denotes the sampling instant, T s denotes the sampling period, f denotes the carrier frequency, denotes the random initial phase, denotes the vector dimension; constructing a complete dictionary according to the preset interference signal vector representation; updating a k'th atom in the complete dictionary according to the complete dictionary and a sparse coefficient vector corresponding to the complete dictionary; In the formula, D represents a complete dictionary, F N represents an N-point discrete Fourier transform matrix, I M is a unit matrix with a dimension of MxM, J represents a preset interference signal vector, and represents a corresponding time-domain interference signal vector Based on the complete dictionary, construct the sparse coefficient vector corresponding to the complete dictionary: γ = [γ1, γ2, ..., γ K ]; acquiring a preset received signal: wherein, x1 = vecX dd , represents a second equivalent matrix corresponding to the preset received signal, X dd represents a time-delay-Doppler two-dimensional matrix with a size of MxN, and n1 represents a noise signal vector corresponding to the preset received signal wherein updating the k'th atom in the complete dictionary according to the complete dictionary and the sparse coefficient vector corresponding to the complete dictionary comprises: performing matrix decomposition on a K-rank matrix Dγ; fixing atoms other than the k'th atom in the complete dictionary and determining a sparsity of the complete dictionary; performing singular value decomposition on the residual matrix; In the formula, represents a residual error; extracting non-zero sequences of the corresponding gamma k' from the residual, resulting in a residual matrix repeating the updating of the k'th atom in the complete dictionary according to the complete dictionary and the sparse coefficient vector corresponding to the complete dictionary until the sparsity of the complete dictionary meets a convergence condition or a repetition number meets a preset maximum iteration number: In the formula, U and V are unitary matrices, and Σ is a diagonal matrix. V T The device matrix represents the unitary matrix V. determine the first column of V T as a new atom d k' , determine the product of the first row and first column of ∑ and the first row of the singular matrix as a new sparse coefficient vector corresponding to the complete dictionary pair; denoted as: wherein U represents a preset sparsity threshold. acquiring an input signal carrying an interference signal comprises:

2. The method of claim 1, wherein, acquiring a time-frequency domain signal; performing a Heisenberg transformation on the time-frequency domain signal to obtain a time-domain signal; wherein X tf denotes a time-frequency domain signal, denotes a DFT matrix of M points, j denotes the imaginary unit, X dd denotes a time-delay-Doppler two-dimensional matrix of size M x N, denotes an IDFT matrix of N points; filtering the time-domain signal to obtain a sending signal; In the formula, S1 represents a time domain signal, represents an IDFT matrix of M points; vectorizing the matrix column in the sending signal to obtain a vector representation of the sending signal; In the formula, S2 represents a transmission signal, G tx represents a first pulse shaping function; determining the input signal according to the vector representation of the sending signal: where x1= vecX dd , denotes the Kronecker product; r = Hs + J + w wherein j represents an imaginary unit. where r represents an input signal, P represents the number of paths, h i represents the channel gain of the i-th path, Π is a permutation matrix, represents the time delay, Δ is a diagonal matrix of dimension MN x MN, represents the Doppler frequency offset, J represents the interference signal, and w represents the time-domain noise signal, where: Δ = diag[z 0 ,z 1 ,…,z MN-1 ] performing signal conversion on the input signal to obtain a received signal comprises:

3. The method of claim 2, wherein, converting the input signal into a vector form to obtain an input vector; when the first pulse shaping function and the second pulse shaping function are ideal rectangular window functions, converting the first equivalent channel matrix into: wherein is a first equivalent channel matrix, is an interference signal vector, is a noise vector, G rx denotes a second pulse-shaping function; reconstructing the interference signal based on a preset sparse dictionary strategy to obtain a reconstructed interference signal comprises: where I M is the M x M identity matrix, and substituting the second equivalent channel matrix is obtained: In the formulae, representing the interference signal as representing the noise signal as determining the received signal as 4. The method of claim 3, wherein, reconstructing the interference signal using the pre-trained dictionary and the sparse coefficient vector to obtain a reconstructed interference signal: performing interference suppression on the input signal according to the received signal and the reconstructed interference signal to obtain a target signal after removing the interference signal comprises: i c = D c γ c In the formula, i c represents the reconstructed interference signal, D c represents a pre-trained dictionary, γ c represents a pre-trained sparse coefficient vector.

5. The method of claim 4, wherein, subtracting the reconstructed interference signal from the received signal according to the reconstructed interference signal to obtain the target signal: ​ where y' represents the target signal, i c represents the reconstructed interference signal.

6. An electronic device, comprising: The electronic device comprises a processor and a memory coupled to each other, and the memory stores a computer program, which, when executed by the processor, causes the electronic device to perform the method of any one of claims 1-5.

7. A computer program product, characterised in that, A computer program which, when executed by a processor, implements the method of any one of claims 1-5.

Images