Electronic device and method for performing beamforming

By using parallel multiplication and addition operations of memristor arrays in the analog domain, the power consumption and area problems in narrowband beamforming circuits are solved, and subsequent processing needs are reduced, and multi-channel data processing of large-scale sensor arrays is adapted.

CN113824483BActive Publication Date: 2024-11-08TSINGHUA UNIVERSITY
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111106102.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-09-22
Publication Date
2024-11-08
Estimated Expiration
2041-09-22

AI Technical Summary

Technical Problem

Existing narrow-band beamforming circuits require a large number of phase shifters and variable gain amplifiers, resulting in large power consumption and circuit area overhead. Moreover, wideband beamforming depends on analog-to-digital conversion and digital signal processing, making it difficult to integrate at the device terminal. Increased calculation and transmission pressure at the sensor end.

Method used

Using electronic devices based on memristor arrays, complex weights are realized by adjusting the conductance of memristor arrays, and beamforming is directly carried out in the analog domain by using parallel multiplication and addition operations of memristor arrays to avoid integrating analog-to-digital conversion interface circuits and reduce Follow-up processing requirements.

Benefits of technology

It improves computing efficiency, reduces power consumption and circuit area, and realizes flexible beamforming signal preprocessing, adapting to the multi-channel data parallel processing requirements of large-scale sensor arrays.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113824483B_ABST
    Figure CN113824483B_ABST
Patent Text Reader

Abstract

An electronic device and method for beamforming. The electronic device includes a first beamforming calculation unit, which is configured to receive multiple input signals and perform beamforming on the multiple input signals to form a first beamforming signal, wherein the first beamforming calculation unit includes a first memristor array, a first capacitor and a first operation unit, and the first beamforming calculation unit is configured to adjust the conductance of the first memristor array to achieve a first complex weight, and multiply the first complex weight with the multiple input signals based on the first memristor array, the first capacitor and the first operation unit to perform beamforming on the multiple input signals. The electronic device and method can avoid integrating an analog-to-digital conversion interface circuit at the sensor end, improve computing efficiency, and reduce the computing power requirements and transmission bandwidth of subsequent processing modules, thereby adapting to the multi-channel data parallel processing requirements of large-scale sensor arrays.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] Embodiments of the present disclosure relate to an electronic device and method for performing beamforming. Background Art

[0002] With the continuous development of technologies such as wearable devices and autonomous driving, the increasing number of sensors in terminal devices has put forward higher requirements for data transmission and timely processing of back-end signals. Beamforming, also known as spatial filtering, is a technology that uses sensor arrays to achieve directional transmission and reception of information. Beamforming technology phase-shifts and weighted superpositions the collected signals so that signals at certain angles obtain constructive or destructive interference, thereby achieving directional selectivity of signal transmission or reception. For example, by integrating the beamforming system and microphone array, the reception of sound wave signals can be directional and the aliasing of environmental noise can be reduced. For example, applying beamforming to application scenarios such as sound source localization, sound imaging, and speech enhancement can reduce the processing pressure of subsequent modules, while blocking useless signals to reduce the requirements for transmission bandwidth. In addition, memristor arrays are considered to have great potential in the preprocessing of near-sensing signals, thanks to the programmable, non-volatile, and nearly continuous analog conductance characteristics of memristor devices, and thanks to the fact that memristor arrays can be used to achieve efficient analog multiplication and addition operations. Summary of the invention

[0003] At least some embodiments of the present disclosure provide an electronic device, including a first beamforming calculation unit, wherein the first beamforming calculation unit is configured to receive multiple input signals and beamform the multiple input signals to form a first beamforming signal, wherein the first beamforming calculation unit includes a first memristor array, a first capacitor and a first operation unit, and the first beamforming calculation unit is configured to adjust the conductance of the first memristor array to achieve a first complex weight, and multiply the first complex weight with the multiple input signals based on the first memristor array, the first capacitor and the first operation unit to beamform the multiple input signals.

[0004] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first memristor array includes a first frequency adjustment subarray, and the first beamforming calculation unit is further configured to adjust the conductance of the first frequency adjustment subarray to adjust the first beamforming signal to correspond to a first frequency band.

[0005] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first beamforming calculation unit is further configured to adaptively update the conductance of the first memristor array based on the first beamforming signal.

[0006] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first memristor array further includes a first sub-array, a second sub-array and a third sub-array, wherein the first sub-array is connected to the first end of the capacitor, the second end of the capacitor is connected to the operation unit, the first sub-array is configured to receive a voltage signal corresponding to the multiple input signals, and the first sub-array and the capacitor are configured to output a first sub-current signal to the operation unit at the second end of the capacitor; the second sub-array is connected to the operation unit, the second sub-array is configured to receive a voltage signal corresponding to the multiple input signals, and output a second sub-current signal to the operation unit; the third sub-array is connected to the operation unit, the third sub-array is configured to receive a voltage signal corresponding to the multiple input signals, and output a third sub-current signal to the operation unit; the operation unit is configured to operate on the first sub-current signal, the second sub-current signal and the third sub-current signal to form the first beamforming signal.

[0007] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first sub-array and the second sub-array are respectively configured to receive the voltage signals with the same amplitude and polarity.

[0008] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first sub-array and the second sub-array are respectively configured to receive the voltage signals with the same amplitude but opposite polarities.

[0009] For example, in the electronic device provided by at least some embodiments of the present disclosure, one or more of the first subarray, the second subarray and the third subarray each include a positive weight subarray and a negative weight subarray, and the positive weight subarray and the negative weight subarray are respectively configured to receive the voltage signals with the same amplitude but opposite polarities.

[0010] For example, in the electronic device provided by at least some embodiments of the present disclosure, the second sub-array and the third sub-array are combined into one memristor sub-array.

[0011] For example, the electronic device provided by at least some embodiments of the present disclosure also includes a second beamforming calculation unit, which is configured to receive the multi-path input signal and beamform the multi-path input signal to form a second beamforming signal, wherein the second beamforming calculation unit includes a second memristor array, a second capacitor, and a second operation unit, and the second beamforming calculation unit is configured to adjust the conductance of the second memristor array to achieve a second complex weight, and multiply the second complex weight with the multi-path input signal based on the second memristor array, the second capacitor and the second operation unit to beamform the multi-path input signal, the second memristor array includes a second frequency adjustment subarray, and the second beamforming calculation unit is also configured to adjust the conductance of the second frequency adjustment subarray to adjust the second beamforming signal to correspond to a second frequency band, wherein the first frequency band is different from the second frequency band.

[0012] At least some embodiments of the present disclosure also provide a method for beamforming multiple input signals using the electronic device, comprising: adjusting the conductance of the first memristor array to achieve a first complex weight, and multiplying the first complex weight with the multiple input signals based on the first memristor array, the first capacitor and the first operating unit to beamform the multiple input signals.

[0013] For example, in the method provided by at least some embodiments of the present disclosure, the first memristor array includes a first frequency adjustment subarray, and the method further includes: adjusting the conductance of the first frequency adjustment subarray to adjust the first beamforming signal to correspond to a first frequency band.

[0014] For example, the method provided by at least some embodiments of the present disclosure further includes: adaptively updating the conductance of the first memristor array based on the first beamforming signal.

[0015] For example, in the method provided by at least some embodiments of the present disclosure, the first memristor array further includes a first sub-array, a second sub-array and a third sub-array, the first sub-array is connected to the first end of the capacitor, the second end of the capacitor is connected to the operation unit, the second sub-array is connected to the operation unit, and the third sub-array is connected to the operation unit, wherein multiplying the first complex weight with the multi-path input signal based on the first memristor array, the first capacitor and the first operation unit to perform beamforming on the multi-path input signal includes: inputting a voltage signal corresponding to the multi-path input signal to the first sub-array, and outputting a first sub-current signal at the second end of the capacitor to the operation unit; inputting a voltage signal corresponding to the multi-path input signal to the second sub-array, and outputting a second sub-current signal to the operation unit at the second sub-array; inputting a voltage signal corresponding to the multi-path input signal to the third sub-array, and outputting a third sub-current signal to the operation unit at the third sub-array; and operating the first sub-current signal, the second sub-current signal and the third sub-current signal to form the first beamforming signal.

[0016] For example, in the method provided by at least some embodiments of the present disclosure, inputting a voltage signal corresponding to the multiple input signals to the first sub-array, and inputting a voltage signal corresponding to the multiple input signals to the second sub-array, includes: inputting the voltage signals having the same amplitude and polarity to the first sub-array and the second sub-array, respectively.

[0017] For example, in the methods provided in at least some embodiments of the present disclosure, inputting a voltage signal corresponding to the multiple input signals to the first subarray and inputting a voltage signal corresponding to the multiple input signals to the second subarray include: inputting the voltage signals having the same amplitude but opposite polarities to the first subarray and the second subarray, respectively.

[0018] For example, in the method provided by at least some embodiments of the present disclosure, one or more of the first subarray, the second subarray and the third subarray each include a positive weight subarray and a negative weight subarray, a voltage signal corresponding to the multiple input signals is input to the first subarray, a voltage signal corresponding to the multiple input signals is input to the second subarray, and a voltage signal corresponding to the multiple input signals is input to the third subarray, and the method further includes: inputting the voltage signals with the same amplitude but opposite polarities to the positive weight subarray and the negative weight subarray, respectively.

[0019] For example, in the methods provided by at least some embodiments of the present disclosure, the second sub-array and the third sub-array are combined into one memristor sub-array.

[0020] For example, in the method provided by at least some embodiments of the present disclosure, the electronic device also includes a second beamforming calculation unit, which is configured to receive multiple input signals and beamform the multiple input signals to form a second beamforming signal, and the second beamforming calculation unit includes a second memristor array, a second capacitor, a second operation unit and a second frequency adjustment subarray. The method also includes: adjusting the conductance of the second memristor array to achieve a second complex weight; multiplying the second complex weight with the multiple input signals based on the second memristor array, the second capacitor and the second operation unit to beamform the multiple input signals; and adjusting the conductance of the second frequency adjustment subarray to adjust the second beamforming signal to correspond to a second frequency band, wherein the first frequency band is different from the second frequency band. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments will be briefly introduced below. Obviously, the drawings in the following description only relate to some embodiments of the present disclosure, but are not intended to limit the present disclosure.

[0022] Figure 1 A schematic diagram of a narrowband beamforming system processing microphone array signals is illustrated;

[0023] Figure 2 A block diagram illustrating an example of a beamforming calculation unit of an electronic device provided by at least some embodiments of the present disclosure is shown;

[0024] Figure 3A A schematic diagram illustrating an example of a beamforming calculation unit of an electronic device provided by at least some embodiments of the present disclosure is illustrated;

[0025] Figures 3B-1 to 3B-3 Pictured Figure 3A An equivalent principle schematic diagram of a beamforming calculation unit is shown;

[0026] Figure 4A A flowchart illustrating an example of beamforming using an electronic device provided by at least some embodiments of the present disclosure;

[0027] Figure 4B A flowchart illustrating another example of beamforming using an electronic device provided by at least some embodiments of the present disclosure;

[0028] Figures 5A-5D A schematic diagram illustrating an example of a sub-array of complex weights implementing bipolarity provided by at least some embodiments of the present disclosure;

[0029] Figure 6A schematic diagram illustrating an example of an electronic device for implementing broadband beamforming provided by at least some embodiments of the present disclosure is shown;

[0030] Figure 7 The diagram shows the amplitude response of the desired beamforming system to signals at different incident directions;

[0031] Figures 8A-8C The figure shows the time domain response of the electronic device provided by the present disclosure receiving a pulse modulation signal with noise;

[0032] Fig. 9 A beamforming method provided by at least some embodiments of the present disclosure is illustrated. DETAILED DESCRIPTION

[0033] In order to make the purpose, technical solution and advantages of the embodiments of the present disclosure clearer, the technical solution of the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings of the embodiments of the present disclosure. Obviously, the described embodiments are part of the embodiments of the present disclosure, not all of the embodiments. Based on the described embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present disclosure.

[0034] Unless otherwise defined, the technical terms or scientific terms used in the present disclosure should be understood by people with ordinary skills in the field to which the present disclosure belongs. The "first", "second" and similar words used in the present disclosure do not indicate any order, quantity or importance, but are only used to distinguish different components. Similarly, similar words such as "one", "one" or "the" do not indicate quantity restrictions, but indicate that there is at least one. Similar words such as "include" or "comprise" mean that the elements or objects appearing before the word cover the elements or objects listed after the word and their equivalents, without excluding other elements or objects. Similar words such as "connect" or "connected" are not limited to physical or mechanical connections, but can include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0035] The present disclosure is described below by several specific embodiments. In order to keep the following description of the embodiments of the present invention clear and concise, the detailed description of known functions and known components may be omitted. When any component of the embodiments of the present invention appears in more than one accompanying drawings, the component is represented by the same reference numeral in each of the accompanying drawings.

[0036] Beamforming systems are often used for directional transmission and reception positioning of signals that propagate in the form of waves. For the receiving end, the principle is to perform phase shifting and linear superposition on the signals received by the sensor array so that the signals transmitted in a specific direction can be constructively superimposed after being received by the sensor, while the signals in other directions can be destructively superimposed, thereby generating a specific spatial filtering pattern.

[0037] For example, Figure 1 The figure shows a schematic diagram of a narrowband beamforming system processing microphone array signals. Figure 1 In the figure, the uniformly distributed linear microphone array MA includes microphones 1 to 2m-1, the propagation direction of the narrowband sound wave signal received by the linear microphone array MA and the angle between the linear microphone array MA is "θ", and due to the propagation delay, there is a phase difference Δφ between the sound wave signals received by adjacent microphones in the linear microphone array MA. If it is assumed that the phase shift of the sound wave signal received by microphone 1 is 0, the phase shift of the sound wave signal received by the linear microphone array MA can be expressed as a vector Φ = [0, Δφ, 2Δφ, ..., (2m-2)Δφ] T , where Δφ=2π·f ac ·d·cosθ / v ac Or Δφ=ω ac ·d·cosθ / v ac , f ac is the frequency of the sound wave signal, ω ac is the angular frequency of the sound wave signal, d is the distance between adjacent microphones, v ac is the speed of sound.

[0038] like Figure 1 As shown, the narrowband beamforming system performs beamforming on the narrowband sound wave signal received by the linear microphone array MA by performing the following beamforming process: phase shifting is performed on the sound wave signals received by microphones 1 to 2m-1, for example, Figure 1 As shown, the sound wave signals received by microphones 1 to 2m-1 are delayed respectively; the phase-shifted (or delayed) sound wave signals are multiplied by corresponding real number weights (for example, Figure 1 The real number weights in A1 to A 2m-1 ) and superimposed. It can be noted that Figure 1 As shown, the above beamforming process can be equivalent to multiplying the sound wave signals received by microphones 1 to 2m-1 by the corresponding complex weights and superimposing them. The complex weight vector is w = [w1, w2, w3, ..., w 2m-1 ] T , then the beamformed acoustic wave signal can be expressed as s = w T By adjusting the complex weight vector Φ, a beamformed acoustic wave signal s with arbitrary spatial response can be achieved.

[0039] However, the narrowband beamforming circuits currently used in radar systems require a large number of circuit units such as phase shifters and variable gain amplifiers, and the implementation of wideband beamforming relies on analog-to-digital conversion interfaces and digital signal processing (DSP) systems. Therefore, integrating the beamforming system and large-scale sensor arrays at the device terminal will undoubtedly bring great power consumption and circuit area overhead.

[0040] The present disclosure utilizes the characteristic of a memristor array that can realize multiplication and addition operations in parallel, and proposes an electronic device and method for beamforming based on a memristor array. The electronic device and method can directly beamform the sensor signal in the analog domain, thereby avoiding the integration of an analog-to-digital conversion interface circuit on the sensor end, and utilizes the parallelism of the multiplication and addition operations of the memristor array to further improve the operation efficiency; in addition, the electronic device and method can also be integrated on the sensor end to perform flexible beamforming signal preprocessing, thereby reducing the computing power requirements and transmission bandwidth of subsequent processing modules, thereby adapting to the multi-channel data parallel processing requirements of large-scale sensor arrays.

[0041] The electronic device and method for performing beamforming proposed in the present disclosure are described in detail below with reference to the accompanying drawings.

[0042] For example, some embodiments of the present disclosure provide an electronic device including a beamforming calculation unit. Figure 2 FIG. 1 is a block diagram of an example of a beamforming calculation unit of an electronic device provided by at least some embodiments of the present disclosure. Figure 2 In the example shown, the electronic device includes a first beamforming computing unit 200, which includes a first memristor array 210, a first capacitor 220 and a first operation unit 230. The first beamforming computing unit 200 is configured to receive multiple input signals and beamform the multiple input signals to form a first beamforming signal, wherein the first beamforming computing unit 200 is configured to adjust the conductance of the first memristor array 210 to achieve a first complex weight, and multiply the first complex weight with the multiple input signals based on the first memristor array 210, the first capacitor 220 and the first operation unit 230 to beamform the multiple input signals.

[0043] For example, the multi-channel input signal is a signal to be beamformed. The embodiment of the present disclosure does not limit the type of the multi-channel input signal. For example, the multi-channel input signal is a sensor signal. For example, the multi-channel input signal is Figure 1 The acoustic wave signal in the example described. In addition, the embodiments of the present disclosure do not limit the form of the multi-channel input signal. For example, the multi-channel input signal can be in the form of a voltage signal.

[0044] For example, the beamforming signal is a signal generated by beamforming multiple input signals. The embodiments of the present disclosure do not limit the form of the beamforming signal. For example, the beamforming signal may be in the form of a current signal or a voltage signal.

[0045] For example, the complex weights are complex weights required for beamforming for multiple input signals (eg, Figure 1 The complex weight vector w) in the example described, the embodiment of the present disclosure does not limit the type of complex weight. For example, the complex weight is a fixed weight calculated based on the spatial frequency response. For example, the complex weight is an optimal weight designed in the time domain based on methods such as the minimum mean square error (LMS). For another example, the complex weight is an adaptive weight that changes continuously with multiple input signals.

[0046] It should be noted that Figure 2 This is only exemplary and not a limitation of the present disclosure, and the first beamforming calculation unit 200 may include other components according to actual conditions.

[0047] It should also be noted that, in the embodiment of the present disclosure, the electronic device may include other beamforming calculation units in addition to the first beamforming calculation unit 200, and the embodiment of the present disclosure is not limited to this.

[0048] In an embodiment of the present disclosure, the electronic device implements complex weights based on the conductance value of the memristor array, and utilizes the multiplication and addition operations of the memristor array to implement the multiplication and addition operations in beamforming, and can directly perform beamforming on multiple input signals in the analog domain, thereby improving the computing efficiency and avoiding the integration of analog-to-digital conversion interface circuits on the sensor end. In addition, the electronic device can be integrated on the sensor end to perform flexible beamforming signal preprocessing, thereby reducing the computing power requirements and transmission bandwidth of subsequent processing modules.

[0049] For example, in the electronic device provided by at least some embodiments of the present disclosure, Figure 2As shown, the first memristor array 210 further includes a first sub-array 211, a second sub-array 212, and a third sub-array 213. For example, the first sub-array 211 is connected to the first end of the capacitor 220, the second end of the capacitor 220 is connected to the operation unit 230, the first sub-array 210 is configured to receive a voltage signal corresponding to the multi-path input signal, the first sub-array 211 and the capacitor 220 are configured to output a first sub-current signal to the operation unit 230 at the second end of the capacitor 220; the second sub-array 212 is connected to the operation unit 230, the second sub-array 212 is configured to receive a voltage signal corresponding to the multi-path input signal, and output a second sub-current signal to the operation unit 230; the third sub-array 213 is connected to the operation unit 230, the third sub-array 213 is configured to receive a voltage signal corresponding to the multi-path input signal, and output a third sub-current signal to the operation unit 230; the operation unit 230 is configured to operate the first sub-current signal, the second sub-current signal, and the third sub-current signal to form a first beamforming signal.

[0050] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first memristor array 210 includes a first frequency adjustment sub-array 214, such as Figure 2 The first beamforming calculation unit 200 is further configured to adjust the conductance of the first frequency adjustment subarray 214 to adjust the first beamforming signal to correspond to the first frequency band.

[0051] It should be noted that in the embodiments of the present disclosure, there is no limitation on the types of memristors in the first memristor array 210 and each sub-array included therein. For example, the memristor may adopt a 1R structure, that is, only one resistive switching element. For example, the memristor may adopt a 1T1R structure, that is, one transistor and one resistive switching element. For another example, the memristor may also adopt other structures such as a 1S1R structure.

[0052] It should also be noted that in the embodiments of the present disclosure, there is no limitation on the type of the operation unit. For example, the operation unit may be a current operation circuit and a transimpedance amplifier circuit based on an operational amplifier.

[0053] For example, Figure 3A A schematic diagram of an example of a beamforming calculation unit of an electronic device provided by at least one embodiment of the present disclosure is illustrated. Figure 3A In the example shown, the conductance of each memristor in the memristor array is represented by G i,j , each memristor in the memristor array is connected to a source line (e.g., SL[0], SLa[0], SL[1]), a bit line (e.g., BL[0] to BL[k]), and a word line (not shown), where i, j, k are positive integers, and where the bit lines BL[0] to BL[k-1] receive the same power as channels CH0 to CHk-1 Corresponding voltage signals, the bit line BL[k] is grounded (zero potential), the source line SL[0] is grounded (zero potential) through the capacitor C0, and the source lines SLa[0] and SL[1] are directly grounded (zero potential). It should be noted that in this article, for the sake of convenience, the memristor G i,j The conductivity is represented by G i,j It should be noted that, in the embodiment of the present disclosure, channels CH0 to CH k-1 Corresponding to k sensor input signals, the following is exemplarily and not restrictively described with channels CH0 to CH k-1 Corresponds to Figure 1 The sound wave signals received by the k microphones shown are taken as an example for explanation.

[0054] For example, in Figure 3A In the example shown, the memristor G 0,0 To G k-1,0 To implement the first sub-array 211, the memristor G a0,0 To G ak-1,0 The second sub-array 212 is implemented with memristors G 0,1 To G k-1,1 To implement the third sub-array 213, the memristor G k,0 、Memristor G ak,0 and memristor G k,1 The first frequency adjustment subarray 214 can be implemented, and the first capacitor 220 can be implemented by the capacitor C0 (for ease of explanation, C0 can also refer to the capacitance value of the capacitor C0), and the first operation unit 230 can be implemented by the operation unit T0. In this case, I f0 Corresponding to the first sub-current signal, I n0 Corresponding to the second sub-current signal, I n1 Corresponding to the third sub-current signal. It should be noted that, Figure 3A This is only exemplary and not a limitation of the present disclosure. For example, the memristor subarrays used to implement the first subarray 211, the second subarray 212, and the third subarray 213 may be arranged separately from each other to receive the channels CH0 to CH1 respectively. k-1 For another example, the memristor subarray used to implement the first frequency adjustment subarray 214 can be arranged separately from the memristor subarray used to implement the first subarray 211, the second subarray 212, and the third subarray 213.

[0055] For example, Figures 3B-1 to 3B-3 Picture shows Figure 3A The equivalent principle diagram of the beamforming calculation unit is shown below. Figures 3B-1 to 3B-3 right Figure 3A The working principle of the beamforming calculation unit shown is explained.

[0056] According to the circuit superposition theorem, the voltage signal V CH0 to I f0 The current contribution is BL[1] to BL[k] grounded (e.g. Figure 3B-1 As shown, and for Figure 3B-1 The situation shown, Figure 3B-2 Lieutenant General G 1,0 To G k,0 Equivalent to a conductivity G psum ) after the current I flowing through capacitor C0 f0-CH0 , according to the Laplace transform, the current I f0-CH0 It can be expressed as:

[0057]

[0058] As shown in formula (1), the above frequency response characteristics are determined by the capacitance of capacitor C0 at the end of source line SL[0] and the sum of all the conductances connected to source line SL[0].

[0059] Similarly, the voltage signal V CH0 to I n0 The contribution of current is I n0-CH0 =V CH0 G a0,0 (For example, in this case, Figure 3B-3 Lieutenant General G a1,0 To G ak,0 Equivalent to a conductivity G apsum ),Pick According to the Laplace transform and Fourier transform, we can calculate:

[0060]

[0061] Replace s in formula (2) with jω and convert it into Fourier transform spectrum representation:

[0062]

[0063] in,

[0064]

[0065] As shown in formula (3), by adjusting the conductance but (The adjustment range of β is 0 to π), where, for example, the capacitance of capacitor C0 can be preset.

[0066] Similarly, the voltage signal of CH0 channel is n1 The contribution of I n1-CH0 =V CH0 G0,1 .

[0067] Combined with the above formula, the contribution of the voltage signal of the CH0 channel to the output of T0 is:

[0068]

[0069] For k input signals, the output signal of all superimposed signals is:

[0070]

[0071] That is, the k-path input signals are multiplied by the corresponding complex weights and then superimposed, thereby realizing beamforming for the k-path input signals.

[0072] In the embodiment of the present disclosure, the combination of the first sub-array 211 and the second sub-array 212 (for example, Figure 3A The memristor G shown in 0,0 To G k-1,0 and memristor G a0,0 To G ak-1,0 The combination of) can realize the imaginary part of the complex weight, the third subarray 213 (for example, Figure 3A The memristor G shown in 0,1 To G k-1,1 ) can realize the real part of the complex weight, and the conductance of the first sub-array 211, the second sub-array 212 and the third sub-array 213 can be determined according to the complex weight to be realized, and the first frequency adjustment sub-array 214 (for example, Figure 3A The memristor G shown in k,0 , G ak,0 and G k,1 )'s conductivity.

[0073] For example, Figure 4A A flowchart of an example of beamforming using an electronic device provided by at least some embodiments of the present disclosure is illustrated. Figure 4A As shown, using an electronic device (e.g., including Figure 2 or Figure 3AThe electronic device of the beamforming calculation unit shown in the figure) performs beamforming including steps S410a to S450a. At step S410a, the complex weights are determined, for example, the complex weights to be implemented are determined according to the required spatial response (frequency domain) or the target output signal (time domain); at step S420a, the conductance of the memristor used to implement the complex weights is determined according to the complex weights, for example, the conductance of the memristor (for example, the first subarray 211, the second subarray 212, and the third subarray 213) used to implement the complex weights is determined according to the above formulas (1) to (6); at step S430a, the conductance of the memristor used to adjust the frequency is determined, for example, the conductance of the memristor (for example, the first subarray 211, the second subarray 212, and the third subarray 213) used to implement the complex weights is determined according to the above formulas (1) to (6). The method comprises the following steps: determining the conductance of a memristor (e.g., the first frequency adjustment subarray 214) for adjusting the frequency according to a conductance, a capacitance of a preset capacitor (e.g., the capacitance of capacitor C0), and the frequency of a beamforming signal to be implemented; at step S440a, mapping the determined conductance of the memristor for implementing the complex weights and the conductance of the memristor for adjusting the frequency to corresponding memristors, for example, changing the conductance of the corresponding memristor through a write operation to implement the mapping; at step S450a, beamforming is performed on the multiple input signals to form a beamforming signal, for example, beamforming is performed on the multiple input signals through the first beamforming calculation unit 200 to form a beamforming signal.

[0074] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first beamforming calculation unit 200 is further configured to adaptively update the conductance of the first memristor array 210 based on the first beamforming signal.

[0075] For example, Figure 4B FIG. 2 is a flowchart of another example of using an electronic device provided by at least one embodiment of the present disclosure to perform beamforming. Figure 4B As shown, using an electronic device (e.g., including Figure 2 or Figure 3AThe electronic device of the beamforming calculation unit shown in the figure) performs beamforming including steps S410b to S470b. At step S410b, the initial complex weights are determined, for example, the complex weights to be implemented are determined according to the required spatial response (frequency domain) or the target output signal (time domain); at step S420b, the conductance of the memristor used to implement the complex weights is determined according to the complex weights, for example, the conductance of the memristor (for example, the first subarray 211, the second subarray 212, and the third subarray 213) used to implement the complex weights is determined according to the above formulas (1) to (6); at step S430b, the conductance of the memristor used to adjust the frequency is determined, for example, according to the conductance of the memristor (for example, the first subarray 211, the second subarray 212, and the third subarray 213) used to implement the complex weights, the capacitance of the pre-set capacitor (for example, the capacitance of the capacitor C0), and the beamforming signal to be implemented. The conductance of the memristor (e.g., the first frequency adjustment subarray 214) for adjusting the frequency is determined by the frequency; at step S440b, the conductance of the memristor determined for implementing the complex weight and the conductance of the memristor for adjusting the frequency are respectively mapped to the corresponding memristors, for example, the conductance of the corresponding memristor is changed by a write operation to achieve the mapping; at step S450b, a plurality of input signals are beamformed to form a beamformed signal, for example, the multi-path input signal is beamformed by the first beamforming calculation unit 200 to form a beamformed signal; at step S460b, an error is obtained, for example, by comparing the formed beamforming signal with the ideal beamforming signal to obtain the error; at step S470b, the complex weight is updated, and the process returns to step S420b. In this example, the electronic device can adaptively adjust by obtaining the error and updating the complex weight to perform beamforming more flexibly and accurately.

[0076] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first sub-array and the second sub-array are respectively configured to receive voltage signals with the same amplitude and polarity. Figure 3A In the example shown, the memristor G used to implement the first sub-array 0,0 To G k-1,0 and a memristor G for realizing the second sub-array a0,0 To G ak-1,0 The bit lines BL[0] to BL[k-1] receive voltage signals with the same amplitude and polarity. In this case, the operation unit T0 is configured to convert the first sub-current signal I f0 and the second sub-current signal I n0 Subtract, for example Figure 3A As shown in formula (3).

[0077] For example, in the electronic device provided by at least some embodiments of the present disclosure, the first sub-array and the second sub-array are respectively configured to receive voltage signals with the same amplitude but opposite polarities. Figure 3A In a variation of the example shown, the memristor G used to implement the first subarray is 0,0 To G k-1,0 and a memristor G for realizing the second sub-array a0,0 To G ak-1,0 The bit lines BL[0] to BL[k-1] receive voltage signals with the same amplitude but opposite polarities (for example, using the same method as will be described later). Figure 5A In a similar manner to the example of , one of the first sub-array and the second sub-array directly receives the input voltage signal through the bit line, and the other of the first sub-array and the second sub-array receives the voltage signal with the same amplitude but opposite polarity as the input voltage signal through the inverter). In this case, the operation unit T0 is configured to convert the first sub-current signal I f0 and the second sub-current signal I n0 Add (with Figure 3A This is the opposite of the example shown in formula (3).

[0078] It should be noted that Figure 3A The example shown can only realize the real part and imaginary part of the complex weight of a single polarity, but the present disclosure is not limited thereto. Figure 3A The examples shown are appropriately modified to implement the real and imaginary parts of the bipolar complex weights.

[0079] For example, in the electronic device provided by at least some embodiments of the present disclosure, one or more of the first subarray, the second subarray, and the third subarray each include a positive weight subarray and a negative weight subarray.

[0080] For example, the positive weight sub-array and the negative weight sub-array are respectively configured to receive voltage signals with the same amplitude and polarity. Figure 5A FIG. 2 is a schematic diagram showing an example of a subarray (eg, the third subarray) implementing a real weight. Figure 5A In the example shown, the positive weight subarray and the negative weight subarray are implemented as two separate columns of memristors ( Figure 5A In this case, the operation unit is configured to subtract the current signals output by the positive weight subarray and the negative weight subarray.

[0081] For example, the positive weight sub-array and the negative weight sub-array are respectively configured to receive voltage signals with the same amplitude but opposite polarities. Figure 5B FIG. 4 is a schematic diagram showing another example of a subarray (eg, the third subarray) implementing a real weight. Figure 5B In the example shown, the positive weight subarray and the negative weight subarray are alternately arranged as the same column of memristors ( Figure 5B , and one of the positive weight sub-array and the negative weight sub-array (e.g., Figure 5B The subarray for implementing the positive weight in the positive weight subarray) directly receives the input voltage signal through the bit line, while the other of the positive weight subarray and the negative weight subarray (for example, Figure 5B The subarray used to implement negative weights in the embodiment of the present invention receives a voltage signal having the same amplitude as the input voltage signal but opposite polarity through an inverter.

[0082] Similarly, the positive weight subarray and the negative weight subarray are respectively configured to receive voltage signals with the same amplitude and polarity or configured to receive voltage signals with the same amplitude but opposite polarity, and bipolar imaginary weights can also be achieved. Figure 5C A schematic diagram illustrating an example of a subarray (eg, a first subarray and a second subarray) implementing an imaginary weight, Figure 5D FIG. 2 is a schematic diagram showing another example of a subarray (eg, a first subarray and a second subarray) implementing an imaginary weight. Figure 5C In the example shown, the positive weight subarray and the negative weight subarray of the first subarray and the second subarray are each implemented as two separate columns of memristors; Figure 5D In the example shown, the positive weight subarray and the negative weight subarray of the first subarray are implemented as two separate columns of memristors. The positive weight subarray and the negative weight subarray of the second subarray are alternately arranged as the same column of memristors. The principle and circuit structure of implementing the bipolar imaginary weight are similar to those of implementing the bipolar real weight, which will not be repeated here.

[0083] It should be noted that the above Figure 5A-5B Any one of Figure 5C-5D The combination of any one of the above can realize the bipolar imaginary weight and real weight at the same time. And it should also be noted that the above Figures 5A-5D This is only exemplary, and different circuit structures may be used to implement bipolar imaginary weights and real weights according to specific circumstances, and the present disclosure does not limit this.

[0084] For example, in the electronic device provided by at least some embodiments of the present disclosure, the second sub-array and the third sub-array are combined into a memristor sub-array. Figure 3AThe memristor connected to the source line Sla[0] and the memristor connected to the source line SL[1] shown in have no phase shift (i.e., they are directly connected to the operation unit, unlike the memristor connected to SL[0] which is connected to the operation unit through a capacitor and thus has a phase shift), and can be combined into a column of memristor arrays. In this case, it can still be assumed that there are a second sub-array and a third sub-array, and the conductance of the assumed second sub-array and the third sub-array can be determined according to the above formulas (1) to (6), and then the determined conductance of the assumed second sub-array and the third sub-array is combined to determine the actual conductance of the combined column of memristor arrays. In addition, in the case of realizing bipolar imaginary weights and real weights, the conductance columns without phase shift can also be combined.

[0085] It should be noted that although the above Figure 3A The example shown is applicable to narrowband beamforming, but the present disclosure is not limited thereto, and the electronic device provided by the present disclosure may also be applicable to wideband beamforming.

[0086] For example, the electronic device provided by at least one embodiment of the present disclosure also includes a second beamforming calculation unit, which is configured to receive multiple input signals and perform beamforming on the multiple input signals to form a second beamforming signal, wherein the second beamforming calculation unit includes a second memristor array, a second capacitor, and a second operation unit, and the second beamforming calculation unit is configured to adjust the conductance of the second memristor array to achieve a second complex weight, and multiply the second complex weight with the multiple input signals based on the second memristor array, the second capacitor, and the second operation unit to perform beamforming on the multiple input signals, the second memristor array includes a second frequency adjustment subarray, and the second beamforming calculation unit is further configured to adjust the conductance of the second frequency adjustment subarray to adjust the second beamforming signal to correspond to a second frequency band, wherein the first frequency band is different from the second frequency band. It should be noted that the second beamforming calculation unit is similar in structure to the first beamforming calculation unit described above, and will not be repeated here.

[0087] For example, Figure 6 FIG. 1 is a schematic diagram of an example of an electronic device for implementing broadband beamforming provided by at least one embodiment of the present disclosure. Figure 6 In the example shown, multiple beamforming calculation units are included, and the multiple beamforming calculation units are configured to receive multiple input signals, perform beamforming on the multiple input signals to form multiple beamforming signals corresponding to different narrowbands, and the multiple beamforming models can form multiple narrowband bandpass filter signals through corresponding bandpass filtering, and the multiple narrowband bandpass filter signals can realize wideband beamforming signals by superposition. In the embodiment of the present disclosure, the electronic device can realize wideband beamforming that satisfies joint filtering in the spatial domain and the frequency domain.

[0088] It should be noted that in Figure 6 In the example shown, the conductance columns without phase shift have been merged for clarity, but this is exemplary only and not a limitation of the present disclosure.

[0089] It should also be noted that the application of the electronic device provided in the present disclosure to the above-mentioned narrowband beamforming and broadband beamforming is merely exemplary and not a limitation of the present disclosure, and it may also be applied to other scenarios, such as generalized sidelobe cancellation (GSC).

[0090] Combine the following Figure 7 and Figures 8A-8C Authentication of electronic devices provided by the present disclosure is described.

[0091] For example, Figure 7 The figure shows the amplitude response of the desired beamforming system to signals at different incident directions. Figure 7 In the simulation example shown, 30 linearly evenly arranged sensors receive an input signal with a frequency of 100 kHz, and the beamforming system beamforms the input signal so that the amplitude of the spatial response to the input signal incident at 45° is 1. According to the embodiments disclosed herein, an electronic device (for example, including a memristor array) for implementing beamforming is constructed. Figure 2 or Figure 3A The electronic device of the beamforming calculation unit shown in the figure) has a simulation circuit, wherein the conductance range of the memristor array is "2μS~20μS", the conductance of the memristor is obtained by linear mapping of the complex weight according to the above formulas (1) to (6), and the phase shift signal received by the sensor is simulated by a sinusoidal voltage source with different phases. Table 1 presents the amplitude response of the simulation circuit under different selected incident angles, wherein serial numbers 1-11 correspond to the test incident angles selected in the simulation circuit. It can be seen that the electronic device provided by the present disclosure can effectively realize the spatial filtering function.

[0092] Table 1 Gain results of simulations at different incident angles

[0093]

[0094]

[0095] Figures 8A-8C The figure shows the time domain response of the electronic device provided by the present disclosure receiving a pulse modulated signal with noise. For example, Fig. 8A The figure shows the original pulse signal. Figure 8B The figure shows the noise-contaminated signal received by the first sensor channel (the received noise-contaminated signal is scaled to between 0.1V and 0.3V). Figure 8C The output filtering result of the electronic device provided by the present disclosure is shown in FIG. Fig. 8AAs shown in the figure, the source sends out a pulse with an amplitude of 1 and an incident direction of 45° at 2ms (i.e. the amplitude of the wave changes from 0 to 1). Figure 8B As shown in the figure, due to noise pollution, the electrical signals received and converted by each sensor channel have a lot of noise, which makes it impossible to distinguish the original waveform, resulting in loss of effective information. Figure 8C As shown, the pulse amplitude is obviously restored, and the original pulse amplitude change is well restored. Therefore, the electronic device provided by the present disclosure can effectively realize beamforming.

[0096] According to the above-mentioned embodiments of the present disclosure, the electronic device provided by the present disclosure realizes complex weights based on the conductance value of the memristor array, and utilizes the multiplication and addition operation of the memristor array to realize the multiplication and addition operation in beamforming, and can directly perform beamforming on multiple input signals in the analog domain, thereby improving the computing efficiency and avoiding the integration of analog-to-digital conversion interface circuits on the sensor end. Moreover, the electronic device can be integrated on the sensor end to perform flexible beamforming signal preprocessing, thereby reducing the computing power requirements and transmission bandwidth of subsequent processing modules.

[0097] In addition, at least some embodiments of the present disclosure also provide a method for performing beamforming, which can be performed using an electronic device (for example, including) provided by an embodiment of the present disclosure. Figure 2 or Figure 3A The electronic device of the beamforming calculation unit shown in the figure) performs beamforming on the multiple input signals, Fig. 9 The flowchart of the method is shown in FIG. Fig. 9 As shown, the method includes steps S910 and S920.

[0098] S910: adjusting the conductance of the first memristor array to achieve a first complex weight, and,

[0099] S920: Multiply the first complex weight and the multiple input signals based on the first memristor array, the first capacitor, and the first operation unit to perform beamforming on the multiple input signals.

[0100] For example, in the methods provided by at least some embodiments of the present disclosure, the first memristor array includes a first frequency adjustment subarray, and the method further includes: adjusting the conductance of the first frequency adjustment subarray to adjust the first beamforming signal to correspond to the first frequency band.

[0101] For example, the method provided by at least some embodiments of the present disclosure further includes: adaptively updating the conductance of the first memristor array and the first frequency adjustment subarray based on the first beamforming signal.

[0102] For example, in the method provided by at least some embodiments of the present disclosure, the first memristor array further includes a first sub-array, a second sub-array, and a third sub-array, the first sub-array is connected to a first end of the capacitor, the second end of the capacitor is connected to a computing unit, the second sub-array is connected to the computing unit, and the third sub-array is connected to the computing unit, wherein multiplying the first complex weight with the multi-path input signal based on the first memristor array, the first capacitor, and the first computing unit to perform beamforming on the multi-path input signal includes: inputting a voltage signal corresponding to the multi-path input signal to the first sub-array, and outputting a first sub-current signal to the computing unit at the second end of the capacitor; inputting a voltage signal corresponding to the multi-path input signal to the second sub-array, and outputting a second sub-current signal to the computing unit at the second sub-array; inputting a voltage signal corresponding to the multi-path input signal to the third sub-array, and outputting a third sub-current signal to the computing unit at the third sub-array; and operating the first sub-current signal, the second sub-current signal, and the third sub-current signal to form a first beamforming signal.

[0103] For example, in the method provided by at least some embodiments of the present disclosure, inputting a voltage signal corresponding to the multiple input signals to the first sub-array and inputting a voltage signal corresponding to the multiple input signals to the second sub-array include: inputting voltage signals with the same amplitude and polarity to the first sub-array and the second sub-array, respectively.

[0104] For example, in the methods provided in at least some embodiments of the present disclosure, inputting a voltage signal corresponding to the multiple input signals to the first subarray and inputting a voltage signal corresponding to the multiple input signals to the second subarray include: inputting voltage signals with the same amplitude but opposite polarities to the first subarray and the second subarray, respectively.

[0105] For example, in the method provided by at least some embodiments of the present disclosure, one or more of the first subarray, the second subarray, and the third subarray each include a positive weight subarray and a negative weight subarray, a voltage signal corresponding to the multiple input signals is input to the first subarray, a voltage signal corresponding to the multiple input signals is input to the second subarray, and a voltage signal corresponding to the multiple input signals is input to the third subarray, and the method further includes: inputting voltage signals with the same amplitude but opposite polarities to the positive weight subarray and the negative weight subarray, respectively.

[0106] For example, in the methods provided by at least some embodiments of the present disclosure, the second sub-array and the third sub-array are combined into one memristor sub-array.

[0107] For example, in the method provided by at least some embodiments of the present disclosure, the electronic device also includes a second beamforming calculation unit, which is configured to receive multiple input signals and beamform the multiple input signals to form a second beamforming signal. The second beamforming calculation unit includes a second memristor array, a second capacitor, a second operation unit, and a second frequency adjustment subarray. The method also includes: adjusting the conductance of the second memristor array to achieve a second complex weight; multiplying the second complex weight with the multiple input signals based on the second memristor array, the second capacitor, and the second operation unit to beamform the multiple input signals; and adjusting the conductance of the second frequency adjustment subarray to adjust the second beamforming signal to correspond to a second frequency band, wherein the first frequency band is different from the second frequency band.

[0108] Embodiments of the present disclosure provide a method for using an electronic device (eg, including Figure 2 or Figure 3A The electronic device or the beamforming calculation unit shown Figure 6 The technical effect of the method for performing beamforming on an electronic device (for example, including Figure 2 or Figure 3A The electronic device or the beamforming calculation unit shown Figure 6 The corresponding description of the electronic device shown in the figure) will not be repeated here.

[0109] There are a few points to note:

[0110] (1) The drawings of the embodiments of the present disclosure only relate to the structures related to the embodiments of the present disclosure, and other structures may refer to the general design.

[0111] (2) In the absence of conflict, the embodiments of the present disclosure and the features therein may be combined with each other to obtain new embodiments.

[0112] The above description is merely an exemplary embodiment of the present disclosure and is not intended to limit the protection scope of the present disclosure. The protection scope of the present disclosure is determined by the appended claims.

Claims

1. An electronic device, comprising a first beamforming calculation unit, wherein the first beamforming calculation unit is configured to receive a plurality of input signals and perform beamforming on the plurality of input signals in an analog domain to form a first beamforming signal, wherein: The first beamforming calculation unit includes a first memristor array, a first capacitor and a first operation unit. The first beamforming calculation unit is configured to adjust the conductance of the first memristor array to achieve a first complex weight, and multiply the first complex weight with the multi-path input signal based on the first memristor array, the first capacitor and the first operation unit to beamform the multi-path input signal, where the multi-path input signal is an analog signal to be beamformed.

2. The electronic device according to claim 1, wherein: The first memristor array includes a first frequency adjustment subarray, and the first beamforming calculation unit is further configured to adjust the conductance of the first frequency adjustment subarray to adjust the first beamforming signal to correspond to a first frequency band.

3. The electronic device according to claim 2, wherein: The first beamforming calculation unit is further configured to adaptively update the conductance of the first memristor array based on the first beamforming signal.

4. The electronic device according to claim 3, wherein: The first memristor array further includes a first sub-array, a second sub-array and a third sub-array, wherein: The first sub-array is connected to a first end of the capacitor, a second end of the capacitor is connected to the operation unit, the first sub-array is configured to receive a voltage signal corresponding to the multiple input signals, and the first sub-array and the capacitor are configured to output a first sub-current signal to the operation unit at the second end of the capacitor; The second sub-array is connected to the operation unit, and the second sub-array is configured to receive a voltage signal corresponding to the multiple input signals and output a second sub-current signal to the operation unit; The third sub-array is connected to the operation unit, and the third sub-array is configured to receive a voltage signal corresponding to the multiple input signals and output a third sub-current signal to the operation unit; The operation unit is configured to operate the first sub-current signal, the second sub-current signal, and the third sub-current signal to form the first beamforming signal.

5. The electronic device according to claim 4, wherein: The first sub-array and the second sub-array are respectively configured to receive the voltage signals with the same amplitude and polarity.

6. The electronic device according to claim 4, wherein: The first sub-array and the second sub-array are respectively configured to receive the voltage signals with the same amplitude but opposite polarities.

7. The electronic device according to claim 4, wherein: One or more of the first sub-array, the second sub-array and the third sub-array each include a positive weight sub-array and a negative weight sub-array, and the positive weight sub-array and the negative weight sub-array are respectively configured to receive the voltage signals with the same amplitude but opposite polarities.

8. The electronic device according to claim 4, wherein: The second sub-array and the third sub-array are combined into a memristor sub-array.

9. The electronic device according to claim 2, further comprising a second beamforming calculation unit, the second beamforming calculation unit being configured to receive the multi-path input signal and perform beamforming on the multi-path input signal to form a second beamforming signal, wherein: The second beamforming calculation unit includes a second memristor array, a second capacitor, and a second operation unit. The second beamforming calculation unit is configured to adjust the conductance of the second memristor array to achieve a second complex weight, and multiply the second complex weight with the multi-path input signal based on the second memristor array, the second capacitor, and the second operation unit to perform beamforming on the multi-path input signal. The second memristor array includes a second frequency adjustment subarray, and the second beamforming calculation unit is further configured to adjust the conductance of the second frequency adjustment subarray to adjust the second beamforming signal to correspond to a second frequency band. The first frequency band is different from the second frequency band.

10. A method for beamforming multiple input signals in an analog domain using the electronic device of claim 1, comprising: adjusting the conductance of the first memristor array to achieve a first complex weight, and, The first complex weight is multiplied by the multi-path input signal based on the first memristor array, the first capacitor, and the first operation unit to perform beamforming on the multi-path input signal, where the multi-path input signal is an analog signal to be beamformed.

11. The method according to claim 10, wherein: The first memristor array includes a first frequency adjustment subarray, and the method further includes: The conductance of the first frequency adjustment subarray is adjusted to adjust the first beamforming signal to correspond to a first frequency band.

12. The method according to claim 11, further comprising: The conductance of the first memristor array is adaptively updated based on the first beamforming signal.

13. The method according to claim 10, wherein: The first memristor array further includes a first sub-array, a second sub-array, and a third sub-array, the first sub-array is connected to the first end of the capacitor, the second end of the capacitor is connected to the operation unit, the second sub-array is connected to the operation unit, and the third sub-array is connected to the operation unit, wherein multiplying the first complex weight by the multiple input signals based on the first memristor array, the first capacitor, and the first operation unit to beamform the multiple input signals includes: Inputting a voltage signal corresponding to the multiple input signals to the first sub-array, and outputting a first sub-current signal to the operation unit at the second end of the capacitor; Inputting a voltage signal corresponding to the multiple input signals to the second sub-array, and outputting a second sub-current signal to the operation unit at the second sub-array; Inputting a voltage signal corresponding to the multi-path input signal to the third sub-array, and outputting a third sub-current signal to the operation unit in the third sub-array; The first sub-current signal, the second sub-current signal and the third sub-current signal are operated to form the first beamforming signal.

14. The method according to claim 13, wherein: Inputting a voltage signal corresponding to the multiple input signals to the first sub-array, and inputting a voltage signal corresponding to the multiple input signals to the second sub-array, comprises: The voltage signals with the same amplitude and polarity are input to the first sub-array and the second sub-array respectively.

15. The method according to claim 13, wherein: Inputting a voltage signal corresponding to the multiple input signals to the first sub-array, and inputting a voltage signal corresponding to the multiple input signals to the second sub-array, comprises: The voltage signals having the same amplitude but opposite polarities are input to the first sub-array and the second sub-array respectively.

16. The method according to claim 13, wherein: One or more of the first subarray, the second subarray, and the third subarray each include a positive weight subarray and a negative weight subarray, a voltage signal corresponding to the multiple input signals is input to the first subarray, a voltage signal corresponding to the multiple input signals is input to the second subarray, and a voltage signal corresponding to the multiple input signals is input to the third subarray, further comprising: The voltage signals having the same amplitude but opposite polarities are input to the positive weight sub-array and the negative weight sub-array respectively.

17. The method according to claim 13, wherein: The second sub-array and the third sub-array are combined into a memristor sub-array.

18. The method according to claim 11, wherein: The electronic device further includes a second beamforming calculation unit, the second beamforming calculation unit is configured to receive a plurality of input signals and perform beamforming on the plurality of input signals to form a second beamforming signal, the second beamforming calculation unit includes a second memristor array, a second capacitor, a second operation unit and a second frequency adjustment subarray, and the method further includes: adjusting the conductance of the second memristor array to achieve a second complex weight; multiplying the second complex weights by the multiple input signals based on the second memristor array, the second capacitor, and the second operation unit to beamform the multiple input signals; and adjusting the conductance of the second frequency adjustment subarray to adjust the second beamforming signal to correspond to a second frequency band, The first frequency band is different from the second frequency band.