A constant beamwidth beamforming method, system, device, and medium

Abstract

The application provides a constant beam width beam forming method, system, device and medium, relates to the technical field of beam forming, and the constant beam width beam forming method comprises the following steps: acquiring a received signal of a non-uniform symmetry microphone array; performing time domain and frequency domain based signal processing on the received signal to obtain a received signal model; sampling each microphone position in the microphone array to obtain an initial beam former weight; adjusting and normalizing each initial beam former weight in the microphone array to obtain a beam former weight; and outputting an estimated value corresponding to the incident direction of the incident signal source through a constant beam width beam former according to the received signal model and the beam former weight. The method, system, device and medium provided by the application can make the microphone array have the same spatial response to input signals of different frequencies within the main lobe width.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of beamforming, and particularly to a constant beamwidth beamforming method, system, device and medium. BACKGROUND

[0002] Beamforming technology based on microphone array is widely used to enhance the signal from a desired direction and suppress noise and interference signals from other directions. Deterministic and adaptive beamforming techniques have been widely studied and applied in radar, sonar, seismology, acoustic imaging, communication and many other fields. However, the beamwidth of the beam pattern of the traditional beamforming technology varies with frequency, such as delay-and-sum beamforming, which limits their application in wideband signals (e.g., voice communication). The basic method to solve this problem is to design a beamformer with constant beamwidth. Constant beamwidth beamforming refers to the fact that the main lobe width of the beam pattern remains relatively constant within the designed range after a wideband signal passes through a certain microphone sensor array, achieving consistent frequency in wideband beam.

[0003] Most of the existing beamformers with constant beamwidth are designed based on uniform linear array (ULA), but some are designed based on uniform circular array (UCA) and arbitrary geometry array. However, in practical applications, different shapes of geometry array will produce different beamformer performance, and the constant beamwidth beamforming technology using uniform symmetric linear array cannot maintain the required constant beamwidth in a lower frequency range. SUMMARY

[0004] In view of the above-mentioned shortcomings of the prior art, the purpose of the present application is to provide a constant beamwidth beamforming method, system, device and medium, which solves the problem that the constant beamwidth beamforming technology using uniform symmetric linear array cannot maintain the required constant beamwidth in a lower frequency range due to different shapes of geometry array producing different beamformer performance.

[0005] To achieve the above-mentioned purpose and other related purposes, the present application provides a constant beamwidth beamforming method, comprising: obtaining a received signal of a non-uniform symmetric microphone array; performing time-domain and frequency-domain based signal processing on the received signal to obtain a received signal model; sampling each microphone position in the microphone array to obtain an initial beamformer weight; adjusting and normalizing each initial beamformer weight in the microphone array to obtain a beamformer weight; and outputting an estimated value corresponding to the incident direction of the incident signal source through a constant beamwidth beamformer according to the received signal model and the beamformer weight.

[0006] In an embodiment of the present application, time-domain and frequency-domain based signal processing is performed on the received signal to obtain a received signal model, comprising: processing the received signal in a preset time domain to obtain a time-domain received signal; performing short-time Fourier transform on the time-domain received signal to obtain a frequency-domain received signal; generating a received signal model in the frequency domain according to the frequency-domain received signal; and rearranging the received signal model to obtain a vector form of the received signal model.

[0007] In an embodiment of the present application, the received signal model in the frequency domain is generated according to the frequency-domain received signal, comprising:

[0008] According to the non-uniform symmetric linear array with an odd number of microphones, the positions of the microphones are obtained

[0009] When n = 0, the positions of the microphones are obtained according to The frequency-domain representation X0(f) of the incident signal source x0(t) of interest corresponding to the microphone position x0 is obtained, where -N≤n≤N;

[0010] In the frequency-domain received signal at the n-th microphone in the non-uniform linear array, the output band noise signal Y n (f) in the frequency-domain representation in the frequency-domain received signal, the expected source signal X n (f), the additive noise signal V n (f) and the incident signal source X0(f) is generated, and a received signal model corresponding to the time frequency f in the frequency domain is obtained:

[0011] Wherein, n is the position index of the microphone, n satisfies 2n+1=M, M is the number of microphones in the non-uniform symmetric linear array; f is the time frequency and f>0; -j is the imaginary unit and j 2 =-1; c is the speed of sound; θ s is the incident direction of the source signal, and t is the time domain time.

[0012] In an embodiment of the present application, the received signal model is rearranged to obtain a vector form of the received signal model, comprising:

[0013] The received signal model is rearranged in a vector form, and the vector form of the received signal model is y(f)=x(f)+v(f)=d(f,θ s )X0(f)+v(f).

[0014] According to the vector form of the received signal model y(f)=x(f)+v(f)=d(f,θ s )X0(f)+v(f), the band noise signal vector y(f)=[Y -N(f),..., Y0(f),..., Y N (f)] T , the desired source signal vector x(f) = [X -N (f),..., X0(f),..., X N (f)] T , the additive noise signal vector v(f) = [V -N (f),..., V0(f),..., V N (f)] T and a steering vector of length 2n+1

[0015] where T is a permutation operator.

[0016] In an embodiment of the present application, sampling is performed at each microphone position in the microphone array to obtain initial beamformer weights, comprising:

[0017] Sampling is performed at different microphone positions in the microphone array by using a continuous Kaiser window to obtain initial beamformer weights

[0018] where x n represents the position of a microphone in the array, and x n satisfies the formula n is the position index of the microphone, and n satisfies -N≤n≤N; J0 is a first-order modified Bessel function of the first kind; β is a window shape factor, and β represents an adjustable parameter; and W is a window support.

[0019] In an embodiment of the present application, each initial beamformer weight in the microphone array is adjusted and normalized to obtain a beamformer weight, comprising:

[0020] Using trapezoidal integration technology, the initial beamformer weights w k (x n ) of the outermost microphone and the microphone inside the outermost microphone are respectively multiplied by the corresponding microphone spacing Δn to obtain an intermediate value of the adjusted beamformer weight;

[0021] The intermediate value of the beamformer weight is normalized to obtain a beamformer weight

[0022] where the microphone spacing Δn of the outermost microphone satisfies Δn = x N -x N-1 , |n| = N, and the microphone spacing Δn of the microphone inside the outermost microphone satisfies

[0023] In an embodiment of the present application, an estimation corresponding to an incident direction of an incident signal source is obtained by a constant beamwidth beamformer output according to a received signal model and a beamformer weight, comprising:

[0024] obtaining an incident direction θ of an incident signal source x0(t) s ;

[0025] obtaining an incident direction θ of an incident signal source x0(t) s , the output of the constant beamwidth beamformer is as an estimation;

[0026] wherein h(f) is a normalized beamformer weight with a length of 2N+1; * is a complex conjugate operator; H is a conjugate transpose operator; Z(f) is an estimation of an expected source signal X0(f) in a frequency domain representation; d(f, θ) is a steering vector with a length of 2n+1; Y n (f) is an output noise signal in a frequency domain representation; X0(f) is an expected signal source in a frequency domain representation; and y(f) is a noise signal vector in a frequency domain representation.

[0027] To achieve the above object and other related objects, the present application further provides a constant beamwidth beamforming system, comprising: a receiving module, configured to obtain a received signal of a non-uniform symmetric microphone array; a processing module, configured to perform time domain and frequency domain based signal processing on the received signal to obtain a received signal model; a sampling module, configured to sample each microphone position in the microphone array to obtain an initial beamformer weight; an adjusting module, configured to adjust and normalize each initial beamformer weight in the microphone array to obtain a beamformer weight; and an output module, configured to obtain an estimation corresponding to an incident direction of an incident signal source by a constant beamwidth beamformer output according to the received signal model and the beamformer weight.

[0028] To achieve the above object and other related objects, the present application further provides an electronic device, comprising: one or more processors; and a storage device configured to store one or more programs, which, when executed by the one or more processors, cause the electronic device to implement the constant beamwidth beamforming method.

[0029] To achieve the above object and other related objects, the present application further provides a computer readable storage medium, which stores a computer program, and the computer program, when executed by a processor of a computer, causes the computer to perform the constant beamwidth beamforming method.

[0030] As above, the constant beamwidth beamforming method, system, device and medium provided by the present application has the following beneficial effects: by using the trapezoidal integration technique based on the non-uniform symmetric linear array, each beamformer weight is multiplied by the microphone spacing under the non-uniform array, so that at low frequencies, the inner microphone sensors with smaller spacing are ineffective due to the smaller weight assigned to them, and the outer sensors with larger spacing are assigned a larger weight, and by utilizing the symmetric property of the non-uniform linear array, the array has the same spatial response to input signals of different frequencies within the main lobe width, effectively expanding the low frequency range that maintains the required constant beamwidth, thereby achieving the maintenance of constant beamwidth at lower frequencies by controlling the window parameter (window shape factor), i.e., achieving the maintenance of constant beamwidth in a larger frequency range, and improving directivity at low frequencies. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 A flowchart of the constant beamwidth beamforming method provided by an embodiment of the present application.

[0032] Figure 2 A position diagram of a non-uniform symmetric linear array used by an embodiment of the present application is shown.

[0033] Figure 3 A constant beamwidth beam pattern based on the prior art method of a uniform symmetric linear array is shown.

[0034] Figure 4 A constant beamwidth beam pattern based on the non-uniform symmetric microphone array method provided by an embodiment of the present application is shown.

[0035] Figure 5 A comparison diagram of the constant beamwidth frequency range based on the non-uniform symmetric microphone array method and the uniform symmetric linear array method provided by an embodiment of the present application is shown.

[0036] Figure 6 A comparison diagram of the directivity index based on the non-uniform symmetric microphone array method and the uniform symmetric linear array method provided by an embodiment of the present application is shown.

[0037] Figure 7 A structural block diagram of the constant beamwidth beamforming system provided by an embodiment of the present application is shown.

[0038] Figure 8 A structural diagram of an electronic device according to an embodiment of the present application is shown.

[0039] Element Number Description

[0040] Electronic device 1; constant beamwidth beamforming system 11; memory 12; processor 13; receiving module 111; processing module 112; sampling module 113; adjusting module 114; output module 115. DETAILED DESCRIPTION

[0041] The present application is herein described, by way of example only, with the assistance of the accompanying drawings wherein:

[0042] Needless to say, the drawings provided in the following embodiments are only schematic and are intended to provide a general understanding of the present application. In actual implementation, the shapes, numbers and sizes of the components are not drawn according to the actual implementation, and the shapes, numbers and proportions of the components can be arbitrarily changed, and the layout of the components can be more complex.

[0043] In the following description, a large number of details are discussed in order to provide a more thorough explanation of embodiments of the application, however, it will be readily apparent to those skilled in the art that the embodiments of the application can be practiced without these specific details, in other embodiments, well-known structures and devices are shown in block diagram form, rather than in detail in order to avoid obscuring the embodiments of the application.

[0044] Referring to Figure 1 The present application provides a constant beamwidth beamforming method, comprising:

[0045] Step S10: obtaining a received signal of a non-uniform symmetric microphone array;

[0046] Step S20: performing time-domain and frequency-domain based signal processing on the received signal to obtain a received signal model;

[0047] Step S30: sampling each microphone position in the microphone array to obtain an initial beamformer weight;

[0048] Step S40: adjusting and normalizing each initial beamformer weight in the microphone array to obtain a beamformer weight;

[0049] Step S50: outputting an estimated value corresponding to the incident direction of the incident signal source through a constant beamwidth beamformer according to the received signal model and the beamformer weight.

[0050] It can be found by the above steps that, in order to obtain a constant beam width, according to the symmetry property of the non-uniform linear array, the received signal of the non-uniform symmetric microphone array is obtained, the received signal is processed in time domain and frequency domain to obtain a frequency domain received signal, and then a received signal model is generated according to the frequency domain received signal. Because some microphones in the symmetric linear array are closely spaced, the initial beamformer weight can be obtained by sampling at each microphone position in the microphone array, and the initial beamformer weight is adjusted so that the required constant beam width can still be maintained at low frequencies, and then normalized to obtain the final constant beam width beamformer weight, so as to reduce the influence of the closely spaced central sensors on the constant beam width. Finally, the estimated value corresponding to the incident direction of the incident signal source is calculated by the constant beam width beamformer according to the received signal model and the beamformer weight. By the above method, at low frequencies, the inner microphone sensors with small spacing are ineffective due to the small weight assigned to them, and the outer sensors with large spacing are assigned a larger weight, effectively expanding the low frequency range that maintains the required constant beam width, and achieving constant beam width at lower frequencies.

[0051] Figure 1 A flowchart of the constant beam width beamforming method in an example embodiment of the present application is shown, including steps S10-S60. The technical solutions of the present application will be described in detail below with reference to the accompanying drawings. Figure 1

[0052] First, step S10 is performed to obtain the received signal of the non-uniform symmetric microphone array.

[0053] When the received signal is obtained, it corresponds to the position of each microphone in the non-uniform symmetric microphone array. In the present application, the non-uniform symmetric linear array has an odd number of microphones, and the number of microphones in the array satisfies M=2n+1. Moreover, because the microphones are arranged in a symmetric array, the phase center of the array is frequency invariant, so the position of the microphone can be represented as where x n is the position of the microphone, and n is the position index of the microphone. Because the microphone array has symmetry, the range of n is constrained to -N≤n≤N.

[0054] Next, step S20 is performed to process the received signal based on time domain and frequency domain signal processing to obtain a received signal model.

[0055] ​After obtaining the received signal of the non-uniform symmetric microphone array, further conversion of the received signal in time domain and frequency domain is needed to obtain a received signal model for calculating the estimated value of the constant beamwidth beamformer output.

[0056] In step S20, time domain and frequency domain based signal processing is performed on the received signal to obtain a received signal model, including:

[0057] Step S201: processing the received signal in a preset time domain to obtain a time domain received signal;

[0058] Step S202: performing short-time Fourier transform on the time domain received signal to obtain a frequency domain received signal;

[0059] Step S203: generating a received signal model in the frequency domain according to the frequency domain received signal;

[0060] Step S204: rearranging the received signal model to obtain a vector form of the received signal model.

[0061] In this embodiment, when processing the received signal, first, the received signal of the non-uniform symmetric linear microphone array is processed in a preset time domain to generate a time domain received signal. Since short-time Fourier transform can analyze signals in both time domain and frequency domain to effectively capture the time-varying characteristics of signals, the frequency domain received signal is obtained by performing short-time Fourier transform on the time domain received signal. Then, the frequency domain received signal is converted into a received signal model in the frequency domain, and by rearranging the received signal model, a vector form corresponding to the received signal model can be obtained. The time domain describes the relationship between physical signals and time, and the time domain waveform of the signal can represent the change of the signal with time.

[0062] In step S203, a received signal model in the frequency domain is generated according to the frequency domain received signal, including:

[0063] Step S2031: obtaining the positions of the microphones according to the non-uniform symmetric linear array with an odd number of microphones

[0064] Step S2032: when n = 0, obtaining the frequency domain representation X0(f) of the incident signal source x0(t) of interest corresponding to the microphone position x0 according to the positions of the microphones

[0065] Step S2033: in the frequency domain received signal at the nth microphone in the non-uniform linear array, obtaining the frequency domain representation of the output noise signal Y n ​(f), the desired source signal X n (f), the additive noise signal V n (f) and the incident signal source X0(f), a received signal model corresponding to the time frequency f in the frequency domain is generated:

[0066] wherein n is the position index of the microphone, n satisfies 2n+1=M, M is the number of microphones in the non-uniform symmetric linear array; f is the time frequency and f>0; -j is the imaginary unit and has j 2 =-1; c is the sound speed; θ s is the incident direction of the source signal, and t is the time domain time.

[0067] In the embodiment, after obtaining the received signal in the frequency domain, the position of the microphone is obtained according to the non-uniform symmetric linear array Further, the frequency domain representation X0(f) of the incident signal source x0(t) of interest corresponding to the microphone position x0 is obtained when n=0. Further, the output band noise signal Y n (f), the desired source signal X n (f), the additive noise signal V n (f), so that the received signal model corresponding to the time frequency f in the frequency domain wherein, considering the far-field sound source case, the incident signal source x0(t) of interest is a plane wave propagating at the sound speed (i.e. c=340m / s) in an anechoic environment.

[0068] In step S204, the received signal model is rearranged to obtain a vector form of the received signal model, including:

[0069] Step S2041: the received signal model is arranged in a vector form to obtain a vector form of the received signal model y(f)=x(f)+v(f)=d(f,θ s )X0(f)+v(f);

[0070] Step S2042: according to the vector form of the received signal model y(f)=x(f)+v(f)=d(f,θ s )X0(f)+v(f), a band noise signal vector y(f)=[Y -N (f),…,Y0(f),…,Y N (f)] T , the desired source signal vector x(f)=[X -N (f),…,X0(f),…,X N (f)] T , the additive noise signal vector v(f)=[V-N (f),…,V0(f),…,V N (f)] T and the steering vector of length 2n+1

[0071] where T is a permutation operator.

[0072] In the embodiment, after obtaining the received signal model , the signal model can be rearranged into a vector form, i.e., y(f) = x(f) + v(f) = d(f, θ s )X0(f) + v(f), and further, y(f) = [Y -N (f),…,Y0(f),…,Y N (f)] T , x(f) = [X -N (f),…,X0(f),…,X N (f)] T , v(f) = [V -N (f),…,V0(f),…,V N (f)] T and where y(f) is a noisy signal vector, x(f) is a desired source signal vector, v(f) is an additive noise signal vector, and d(f, θ s ) is a steering vector of length 2n+1.

[0073] Then, step S30 is performed to sample each microphone position in the microphone array to obtain initial beamformer weights.

[0074] Before determining the beamformer weights of the microphone array, each microphone position in the microphone array also needs to be sampled to obtain the initial beamformer weights in advance.

[0075] In step S30, each microphone position in the microphone array is sampled to obtain the initial beamformer weights, including:

[0076] The different microphone positions in the microphone array are sampled by a continuous Kaiser window to obtain the initial beamformer weights

[0077] where x n represents the position of the microphone in the array, and x n satisfies the formula n is the position index of the microphone, and n satisfies -N≤n≤N; J0 is a first-order modified Bessel function of zero; β is a window shape factor, and β represents an adjustable parameter; and W is a window support.

[0078] In the embodiment, for the non-uniform symmetric linear microphone array, when sampling at each microphone position in the microphone array, the sampling can be performed at different microphone positions using a continuous Kaiser window function to obtain initial beamformer weights wherein the position x of the microphone in the array needs to be utilized n The aforementioned formula should be satisfied n is the position index of the microphone, and n satisfies -N≤n≤N. By selecting the window support W=2x N When W=2x, only one adjustment parameter β is left; for a specific window support W, the parameter β can be adjusted to keep the beam width constant.

[0079] Then, step S40 is performed to adjust and normalize each initial beamformer weight in the microphone array to obtain a beamformer weight.

[0080] After obtaining the initial beamformer weight at each microphone position in the microphone array by sampling using the continuous Kaiser window function, the initial beamformer weight is further adjusted to a beamformer weight intermediate value, and then the beamformer weight intermediate value is normalized to obtain the beamformer weight.

[0081] In step S40, each initial beamformer weight in the microphone array is adjusted and normalized to obtain a beamformer weight, including:

[0082] Step S401: using trapezoidal integration technology, the initial beamformer weight w k (x n ) of the outermost microphone and the microphone inside the outermost microphone is multiplied and calculated with the corresponding microphone spacing Δn to obtain an adjusted beamformer weight intermediate value;

[0083] Step S402: normalizing the beamformer weight intermediate value to obtain a beamformer weight

[0084] wherein the microphone spacing Δn of the outermost microphone satisfies Δn=x N -x N-1 , |n|=N, the microphone spacing Δn of the microphone inside the outermost microphone satisfies

[0085] In this embodiment, in order to improve the adaptability to some closely spaced microphones, when adjusting the initial beamformer weight at each microphone position, the outermost microphone can be multiplied by the microphone spacing Δn (Δn satisfies Δn=x N -x N-1 , |n| = N) to perform product calculation adjustment, and the microphone inside the outermost microphone can be multiplied by the microphone spacing Δn (Δn satisfies ) to perform product calculation adjustment, so as to obtain the adjusted beamformer weight intermediate value. Wherein, the microphone spacing Δn satisfies the formula By using the trapezoidal integral technique (TIT) to adjust the weight at all frequencies of the microphone array, the influence of the central sensor with close spacing on the constant beam width can be reduced, and the required constant beam width can be maintained at low frequencies.

[0086] Then, step S50 is performed, and the estimated value corresponding to the incident direction of the incident signal source is output by the constant beam width beamformer according to the received signal model and the beamformer weight.

[0087] After obtaining the normalized beamformer weight, it is further necessary to combine the received signal model, and then output the estimated value of the expected source signal X n (f) according to the incident direction of the incident signal source through the constant beam width beamformer.

[0088] In step S50, the estimated value corresponding to the incident direction of the incident signal source is output by the constant beam width beamformer according to the received signal model and the beamformer weight, including:

[0089] Step S501: obtaining the incident direction θ s of the incident signal source x0(t);

[0090] Step S502: obtaining the output of the constant beam width beamformer as as the estimated value according to the incident direction θ s of the incident signal source x0(t);

[0091] Wherein, h(f) is the normalized beamformer weight with a length of 2N+1; * is a complex conjugate operator; H is a conjugate transpose operator; Z(f) is the estimated value of the expected source signal X0(f) in the frequency domain; d(f, θ) is a steering vector with a length of 2n+1; Y n (f) is the output noise signal in the frequency domain; X0(f) is the expected signal source in the frequency domain; y(f) is the noise signal vector in the frequency domain.

[0092] In the embodiment, in the process of outputting the estimated value of the constant beamwidth beamformer, when the incident direction of the incident signal source x0(t) is θ s , the output of the constant beamwidth beamformer can be obtained as Thus, the estimated value Z(f) of the constant beamwidth beamformer output to the frequency domain representation of the expected source signal X0(f) is obtained, that is, the output stable constant beamwidth is obtained.

[0093] Referring to Figure 2 , Figure 2 , in the embodiment, the position diagram located at the upper side is obtained by the Kaiser window constant beamwidth beamforming based on the uniform symmetric linear array, wherein the number of microphones of the uniform symmetric linear array is M=11, and the interval of the adjacent microphones is δ=3.5 cm. The position diagram located at the lower side is obtained by the Kaiser window constant beamwidth beamforming based on the non-uniform symmetric linear microphone array, wherein the number of microphones of the non-uniform symmetric linear array is M=11, and the coordinates of the microphone positions at greater than 0 (when n≥1) are [3.5, 7.5, 16.5, 33.5, 86.5] cm.

[0094] Referring to Figure 3 and Figure 4 , Figure 3 and Figure 4 , in the embodiment, it can be seen that the constant beamwidth beamforming method provided by the present application can maintain the required constant beamwidth at a lower frequency.

[0095] Referring to Figure 5 , Figure 5 , in the embodiment, it can be seen that the constant beamwidth frequency range of the constant beamwidth beamforming method provided by the present application is 1000 Hz larger than that of the method based on the uniform symmetric linear array, that is, the constant beamwidth beamforming method provided by the present application can be applied to a lower frequency, and can maintain a constant beamwidth in a larger frequency range.

[0096] Referring to Figure 6 , Figure 6 , in the embodiment, it can be seen that when the frequency is less than 2000 Hz, compared with the constant beamwidth beamforming method based on the uniform symmetric linear array, the directivity index (DI) of the constant beamwidth beamforming method based on the non-uniform symmetric linear microphone array provided by the present application is larger, that is, it can be realized that the directivity is higher in the low-frequency application.

[0097] Referring to FIG. 7, the application further provides a constant beamwidth beamforming system 11, comprising: a receiving module 111, configured to acquire a receiving signal of a non-uniform symmetric microphone array; a processing module 112, configured to perform time-domain and frequency-domain based signal processing on the receiving signal to obtain a receiving signal model; a sampling module 113, configured to sample each microphone position in the microphone array to obtain an initial beamformer weight; an adjusting module 114, configured to adjust and normalize each initial beamformer weight in the microphone array to obtain a beamformer weight; and an output module 115, configured to output an estimation corresponding to an incident direction of an incident signal source through a constant beamwidth beamformer according to the receiving signal model and the beamformer weight.

[0098] It should be noted that the constant beamwidth beamforming system 11 provided by the above embodiment and the constant beamwidth beamforming method provided by the above embodiment belong to the same concept, and the specific manner in which each module and unit performs operations has been described in detail in the method embodiment, which will not be described here. In actual application, the constant beamwidth beamforming system 11 provided by the above embodiment can distribute the above functions to be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the functions described above, and this is not limited here.

[0099] Referring to FIG. 7, Figure 8 The electronic device 1 can include a memory 12, a processor 13 and a bus, and can further include a computer program, such as a constant beamwidth beamforming program, stored in the memory 12 and executable on the processor 13.

[0100] The memory 12 includes at least one type of readable storage medium, including a flash memory, a mobile hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a magnetic memory, a disk, an optical disk, etc. The memory 12 can be an internal storage unit of the electronic device 1 in some embodiments, such as a mobile hard disk of the electronic device 1. The memory 12 can also be an external storage device of the electronic device 1 in other embodiments, such as a plug-in mobile hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 1. Further, the memory 12 can include both an internal storage unit and an external storage device of the electronic device 1. The memory 12 can be used not only to store application software and various data installed on the electronic device 1, such as constant beamwidth beamforming code, etc., but also to temporarily store data that has been output or will be output.

[0101] The processor 13 can be composed of integrated circuits in some embodiments, for example, can be composed of a single packaged integrated circuit, or can be composed of multiple packaged integrated circuits with the same function or different functions, including one or more central processing units (CPU), microprocessors, digital processing chips, graphics processors, combinations of various control chips, etc. The processor 13 is the control core of the electronic device 1, and is connected to various components of the electronic device 1 through various interfaces and lines, and executes programs or modules stored in the memory 12 (such as a constant beamwidth beamforming program, etc.), and calls data stored in the memory 12, to perform various functions of the electronic device 1 and process data.

[0102] The processor 13 executes an operating system of the electronic device 1 and various installed application programs. The processor 13 executes the application programs to implement the steps in the constant beamwidth beamforming method described above.

[0103] For example, the computer program can be divided into one or more modules, which are stored in the memory 12 and executed by the processor 13 to complete the present application. The one or more modules can be a series of computer program instruction segments capable of completing a specific function, which are used to describe the execution process of the computer program in the electronic device 1. For example, the computer program can be divided into a receiving module 111, a processing module 112, a sampling module 113, an adjusting module 114, and an output module 115.

[0104] The integrated units implemented in the form of software function modules described above can be stored in a computer readable storage medium, which can be non-volatile or volatile. The software function modules described above are stored in a storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a computer device, or a network device, etc.) or a processor to execute part of the constant beamwidth beamforming method described in various embodiments of the present application.

[0105] In summary, the constant beam width beam forming method, system, device and medium disclosed by the application can make the inner microphone sensors with small spacing invalid at low frequency because the weight of each beam former is multiplied by the microphone spacing under the non-uniform array, and the outer sensors with large spacing are given larger weight, and by using the symmetric property of the non-uniform linear array, the array has the same spatial response to input signals of different frequencies within the main lobe width, effectively expanding the low frequency range of maintaining the required constant beam width, thereby maintaining the constant beam width at a lower frequency, that is, maintaining the constant beam width in a larger frequency range, and improving the directivity at low frequency. Therefore, the application effectively overcomes the shortcomings of the prior art and has high industrial utilization value.

[0106] The above embodiments only exemplarily illustrate the principles and effects of the application, and are not used to limit the application. Any person skilled in the art can modify or change the above embodiments without departing from the spirit and scope of the application. Therefore, all equivalent modifications or changes completed by those skilled in the art without departing from the spirit and technical idea disclosed by the application should be covered by the claims of the application.

Claims

1. A constant beamwidth beamforming method, characterized in that, include: Acquire the received signal from a non-uniformly symmetrical microphone array; The received signal is processed based on time and frequency domains to obtain a received signal model; Samples are taken at the positions of each microphone in the microphone array to obtain the initial beamformer weights; The initial beamformer weights in the microphone array are adjusted and normalized to obtain the beamformer weights. Based on the received signal model and the beamformer weights, an estimated value corresponding to the incident direction of the incident signal source is obtained through the output of the constant beamwidth beamformer.

2. The constant beamwidth beamforming method according to claim 1, characterized in that: The received signal is processed based on time and frequency domains to obtain a received signal model, including: The received signal is processed within a preset time domain to obtain a time-domain received signal; The received signal in the time domain is subjected to a short-time Fourier transform to obtain the received signal in the frequency domain; Based on the received signal in the frequency domain, generate a received signal model in the frequency domain; The received signal model is rearranged to obtain the vector form of the received signal model.

3. The constant beamwidth beamforming method according to claim 1, characterized in that: Based on the received signal in the frequency domain, a received signal model in the frequency domain is generated, including: The positions of the microphones are obtained from a non-uniform symmetric linear array with an odd number of microphones. ; when At that time, depending on the position of the microphone Get the microphone position Corresponding incident signal source of interest Frequency domain representation ,in, ; In a non-uniform linear array, the first In the frequency domain received signal at each microphone, the output noisy signal is based on the frequency domain representation of the frequency domain received signal. Expected source signal Additive noise signal and incident signal source The corresponding time frequency in the frequency domain is generated as Received signal model: ; in, For the microphone's location index. satisfy , The number of microphones in a non-uniform symmetric linear array; For time frequency and ; It is an imaginary unit and has ; Speed ​​of sound; The incident direction of the source signal is denoted as . For time domain time.

4. The constant beamwidth beamforming method according to claim 3, characterized in that: The received signal model is rearranged to obtain its vector form, including: The received signal model By arranging the vectors, the vector form of the received signal model is obtained as follows: ; According to the vector form of the received signal model, To obtain the noisy signal vector , expected source signal vector Additive noise signal vector and length is The guide vector ; in, It is a permutation operator.

5. The constant beamwidth beamforming method according to claim 1, characterized in that: Sampling is performed at each microphone position in the microphone array to obtain the initial beamformer weights, including: The initial beamformer weights are obtained by sampling at different microphone positions in the microphone array using a continuous Kaiser window. ; in, Indicates the position of the microphone in the array, and Satisfy the formula , For the microphone's location index, and satisfy ; It is a zeroth-order modified Bessel function of the first kind; Let be the window shape factor, and Indicates an adjustable parameter; Provides support for the window.

6. The constant beamwidth beamforming method according to claim 5, characterized in that: The initial beamformer weights in the microphone array are adjusted and normalized to obtain beamformer weights, including: Using trapezoidal integration, the initial beamformer weights of the outermost and innermost microphones are combined. Distance from the corresponding microphone Perform product calculations and adjustments to obtain the adjusted intermediate values ​​of the beamformer weights; The intermediate values ​​of the beamformer weights are normalized to obtain the beamformer weights. ; Among them, the microphone spacing of the outermost microphone satisfy The microphone spacing between the outermost and innermost microphones satisfy .

7. The constant beamwidth beamforming method according to claim 1, characterized in that: Based on the received signal model and the beamformer weights, an estimated value corresponding to the incident direction of the incident signal source is obtained through the output of a constant beamwidth beamformer, including: Acquire the incident signal source incident direction ; According to the incident signal source incident direction The output of the constant beamwidth beamformer is obtained as follows: , to be used as the estimated value; in, For length is Normalized beamformer weights; * represents the complex conjugate operator; It is the conjugate transpose operator; The desired source signal in the frequency domain The estimated value; For length is The guide vector; The output noisy signal is represented in the frequency domain. The desired signal source is represented in the frequency domain; It is a noisy signal vector represented in the frequency domain.

8. A constant beamwidth beamforming system, characterized in that, include: The receiving module is used to acquire the received signals from a non-uniformly symmetrical microphone array; The processing module is used to perform time-domain and frequency-domain signal processing on the received signal to obtain a received signal model; The sampling module is used to sample the positions of each microphone in the microphone array to obtain the initial beamformer weights; An adjustment module is used to adjust and normalize the weights of each initial beamformer in the microphone array to obtain beamformer weights. as well as The output module is used to obtain an estimated value corresponding to the incident direction of the incident signal source by means of a constant beamwidth beamformer, based on the received signal model and the beamformer weights.

9. An electronic device, characterized in that: The electronic device includes: One or more processors; A storage device for storing one or more programs that, when executed by one or more processors, cause the electronic device to implement the constant beamwidth beamforming method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, It stores a computer program that, when executed by the computer's processor, causes the computer to perform the constant beamwidth beamforming method according to any one of claims 1 to 7.

Images