Radio frequency computing method and system based on time encoding metasurface

Abstract

The application discloses a radio frequency computing method and system based on a time coding metasurface. The metasurface is composed of a metal patch layer, an F4B dielectric layer and a metal backplate layer. By applying a specific periodic control signal, the metasurface can modulate the electromagnetic information of the incident wave. The baseband signal of the incident electromagnetic wave represents a time function that needs to be operated, and the control signal of the metasurface corresponds to the coding sequence of the required operation. After the electromagnetic wave interacts with the metasurface, the reflected wave will carry the information of the operation result, and the Fourier transform and convolution operation in the radio frequency domain can be realized. The application eliminates the conversion of the radio frequency signal, and directly completes the calculation in the radio frequency domain, which can effectively reduce the calculation delay and energy consumption. At the same time, since the traditional radio frequency link is omitted, the radio frequency calculator constructed by the metasurface has low hardware complexity and cost, a simple structure and is suitable for application fields of communication, radar, sensing and the like under the premise of ensuring the calculation accuracy.

Description

Technical Field

[0001] This invention relates to a radio frequency calculator based on a time-coded metasurface, which can perform Fourier transform and convolution operations in the radio frequency domain, and belongs to the field of novel artificial electromagnetic metasurface technology. Background Technology

[0002] The field of information science has long been dominated by digital computing, using digital means to analyze and process diverse information. However, traditional digital computing requires frequent analog-to-digital and digital-to-analog conversions, resulting in high energy consumption and low information retention. This approach not only has strict requirements on energy consumption but also suffers from problems such as information distortion and computational latency. With the rapid increase in the amount of information, the existing computing model urgently needs to be transformed.

[0003] Radio frequency (RF) computing, as an emerging computing paradigm, uses RF signals as both information carriers and computational objects, completing information processing directly in the RF domain, thus avoiding frequent conversions between the RF, analog, and digital domains. This method not only effectively reduces computational latency but also decreases computational energy consumption, showing broad application prospects. However, due to imperfect theoretical foundations and methodological guidance, current RF computing lacks suitable hardware support and device design methods. Summary of the Invention

[0004] Technical issues:

[0005] The problem this invention aims to solve is to achieve direct processing of radio frequency (RF) signals carrying information using a time-coded metasurface without requiring frequency conversion. Specifically, when the RF signal to be calculated is incident on the time-coded metasurface, a control signal corresponding to the required operation is applied to the metasurface. The reflected echo from the metasurface carries the calculation result information. By designing the transformation form of the control signal, two different processing methods can be performed on the RF information: Fourier transform and convolution.

[0006] Technical solution:

[0007] To address the aforementioned technical problems, this invention proposes a time-coded metasurface capable of radio frequency (RF) calculations. The core structure of this invention is a time-coded metasurface, a two-dimensional ultrathin artificial surface formed by arranging M×L subwavelength-scale units in a specific manner. Each unit integrates an active tunable device (varactor diode). By connecting to an FPGA, the bias voltage of the active device can be controlled, thereby modulating the reflected phase of the electromagnetic wave while ensuring a strong reflection amplitude.

[0008] Each unit is made of double-sided copper-clad F4B dielectric substrate (dielectric constant 2.65, loss tangent 0.001), and each unit integrates 4 varactor diodes (SMV-2019) and 4 surface mount capacitors (0.1pF).

[0009] The control module of the metasurface consists of a field-programmable gate array (FPGA) and a digital-to-analog converter (DAC). This control module is capable of generating arbitrary periodic voltage signals that vary with time within the range of 0V to 14V.

[0010] When the control voltage is switched between any value in the range of 0V to 14V, the metasurface can achieve continuous variation of the reflection phase while maintaining a high reflection amplitude in the frequency band of 4.0GHz to 4.3GHz, and its adjustable range exceeds 360°.

[0011] Among them, the metasurface can complete the information processing of the incident signal in the radio frequency domain by applying the control voltage waveform of the corresponding operation, and directly obtain the result of the corresponding operation.

[0012] This invention also provides a radio frequency calculation method based on a time-coded metasurface, comprising the following steps:

[0013] The mapping relationship between the echo signal R and the reflection coefficient encoding Γ of the time-coded metasurface can be represented in the following matrix form:

[0014]

[0015] Among them, I=[I[n]] N×1 Let R be the incident signal matrix. K×1 ,Γ K×N These are the echo signal matrix and the reflection coefficient encoding matrix, respectively, n = 0, ..., N-1, representing the index of the matrix in the time dimension, and N and K representing the lengths of the input vector and the output vector, respectively;

[0016] Design the reflection coefficient encoding matrix Γ according to radio frequency calculation requirements. K×N This allows the echo signal to carry information about the calculation result;

[0017] By demodulating the echo signal from the time-coded metasurface, waveform information containing the desired calculation results can be obtained.

[0018] Preferably, the radio frequency calculation is a Fourier transform, then the reflection coefficient coding matrix is ​​designed as follows:

[0019]

[0020] The reflected echo from the metasurface will carry information about the Fourier transform of the incident wave.

[0021] Preferably, the radio frequency calculation is a convolution operation, then the designed reflection coefficient coding matrix is ​​Γ. K×N =[Γ0,…,Γ k ,…,Γ (N-M+1) ]T , where Γ k =[[0] 1×k H 1×M [0] 1×(K-k-1) ], where H 1×M This is the impulse response matrix.

[0022] The present invention also provides a radio frequency computing system based on a time-coded metasurface, including a time-coded metasurface, a vector signal transceiver, a waveform generator, and a radio frequency calculator; the vector signal transceiver is used to generate the input signal to be calculated and to receive the echo signal to demodulate the calculation result; the waveform generator is used to generate a periodic voltage control signal to realize the periodic switching of the metasurface reflection phase; the radio frequency calculator is used to perform the above method for radio frequency computing.

[0023] Beneficial effects:

[0024] 1. Reduced computation latency and energy consumption: Compared to traditional digital computing, this invention eliminates the frequent conversions between the radio frequency (RF) and analog-to-digital domains, directly using RF signals as the information carrier and computational object. Completing information processing directly in the RF space effectively reduces computation latency and energy consumption.

[0025] 2. Providing a high-precision RF computing tool: This invention uses a time-coded metasurface to modulate RF signals, enabling Fourier transform and convolution operations in the RF domain. This method provides a high-precision and efficient new tool for RF computing, filling the gap in existing technologies that lack suitable hardware support.

[0026] 3. Low hardware complexity, simple structure and low cost: The radio frequency calculator based on time-coded metasurface involved in this invention completes the electromagnetic wave processing directly on the metasurface, eliminating the complex radio frequency links in traditional electronic systems, and has the advantages of low cost, simple structure and low hardware complexity. Attached Figure Description

[0027] Figure 1 This is a schematic diagram illustrating the working principle of the radio frequency calculator based on a time-coded metasurface in this invention.

[0028] Figure 2 This is an exploded view of the unit structure of the time-coded metasurface in this invention;

[0029] Figure 3 The present invention describes the relationship between the reflection amplitude and phase of the metasurface unit and the bias voltage at 4.1 GHz, 4.2 GHz and 4.3 GHz when the control voltage varies from 0 to 14 V.

[0030] Figure 4 This is a scenario diagram illustrating the implementation of radio frequency computing in this invention;

[0031] Figure 5 This shows the Fourier transform results based on radio frequency (RF) calculations in this invention. ac) RF calculation spectrum display for three incident waveforms: a single sine wave, a mixed sine wave, and a frequency-modulated continuous wave. df) Time-domain waveform (real part and phase) reconstructed from the RF calculation results using IFFT.

[0032] Figure 6 This is the result of the convolution operation based on radio frequency computing in this invention (comparison of theoretical and experimental results). The target positions added to the simulator are a: 1Km, b: 3Km, c: 10Km, d: 14Km. Detailed Implementation

[0033] This invention provides a radio frequency calculator based on a time-coded metasurface for implementing two fundamental operations crucial in signal processing: Fourier transform and convolution. Its conceptual diagram is shown below. Figure 1 As shown. In this embodiment, a 16×8 all-phase modulation reflective metasurface sample was designed, with the unit structure as shown. Figure 2 As shown, each unit integrates four varactor diodes and four surface-mount capacitors. By adjusting the voltage across the varactor diodes on the metasurface unit, the metasurface can achieve continuous adjustment of the reflection coefficient within the range of 0° to 360°. Figure 3 Figure ab shows the relationship between the reflection amplitude and reflection phase of the metasurface and the control voltage at operating frequencies of 4.1 GHz, 4.2 GHz, and 4.3 GHz. The results show that the metasurface can maintain a high reflection amplitude in the above operating frequency bands, and the adjustable range of the reflection phase exceeds 360°.

[0034] To demonstrate that the time-coded metasurface in this invention can realize radio frequency computing, a structure as follows was constructed... Figure 4 The illustrated radio frequency computing system uses a vector signal transceiver to generate the input signal to be calculated, which is incident on the metasurface in the normal direction. A waveform generator generates a periodic voltage control signal, which can be converted into a continuous control voltage waveform to achieve periodic switching of the metasurface reflection phase. According to... Figure 3 The voltage-phase relationship curve of the metasurface at a working frequency of 4.2 GHz, shown in Figure b, provides the required control voltage encoding. By demodulating the reflected echo signal from the metasurface, waveform information containing the desired calculation results can be obtained.

[0035] The working principle of the radio frequency calculator of this invention is as follows: Let the reflection coefficient of the time-coded metasurface be Γ(t), when subjected to a carrier frequency of f... c Furthermore, when the baseband signal I(t) is incident with the signal, the reflected echo signal R(t) can be expressed as:

[0036] R(t)=I(t)×Γ(t) (1)

[0037] Clearly, the reflected signal is the product of the incident signal and the reflection coefficient. The core idea of ​​time-coded metasurfaces is to achieve electromagnetic wave manipulation by discretely encoding the electromagnetic properties of the metasurface units. For ease of analysis, the above expression is converted from continuous signal form to discrete vector form. Let the discrete sample sequence of the incident signal be I = [I[n]]. N×1 The reflection coefficient sequence Γ at time k k =[Γ k [n] N×1 Where n = 0, ..., N-1, the reflected signal is represented as:

[0038] R k =Γ k ·I (2)

[0039] In practical electronic information systems, different types of system responses exhibit diverse processing methods for input signals. The output of any linear system can be considered as the result of the interaction between the system input f(t) and the system response h(t,m), expressed as:

[0040]

[0041] exist Figure 1 In the demonstrated RF calculator based on a time-coded metasurface, different types of system responses can be constructed by flexibly controlling the metasurface using an FPGA to achieve the desired target output g(m). In this architecture, the input signal f(t) is replaced by the incident wave signal I(t), and the system function h(t,m) corresponds to the time-varying reflection coefficient Γ(t) of the time-coded metasurface. The discrete summation form of the output expression for the linear system is given below:

[0042]

[0043] Encoding method of time-encoded metasurfaces Γ k The design of [n] will directly affect the system response of the constructed linear system. To intuitively construct the mapping relationship between the system response and the encoded atoms, it can be converted into matrix form:

[0044]

[0045] The system output at this time is represented as follows:

[0046] R K×1 =[Γ0 T ·I,Γ1 T ·I,…,Γ K-1T ·I] (6)

[0047] In the constructed RF computing system, it is only necessary to design a reflection coefficient code Γ that meets the requirements. i If the system response (i = 0, ..., K-1) meets the requirements of the RF computing system, then the construction scheme of the RF computing system based on the time-coded metasurface is feasible. To enhance the rigor of the theoretical derivation, test cases of RF computing based on the time-coded metasurface are provided for Fourier transform and convolution operations, respectively.

[0048] Example 1: Experimental verification of Fourier transform using time-coded metasurface.

[0049] Figure 5 The example in section 'ac' presents the RF calculation results for three incident waveforms: a single sine wave, a mixed sine wave, and a frequency-modulated continuous wave. To verify the accuracy of the RF calculations, this example also provides the theoretical incident signal spectrum calculated using numerical methods. The comparison results show that the RF calculation results are highly consistent with the numerical calculation results, accurately reflecting the spectral characteristics of the incident signal. Figure 5 The diagram below (df) shows the time-domain waveform reconstruction after inverse Fourier transform of the spectrum obtained from RF calculations. The results show that the original waveform information is recovered relatively completely, indicating the correctness of the RF calculation results and verifying the effectiveness of the time-coded metasurface in RF Fourier transform. Therefore, Example 1 demonstrates that the time-coded metasurface can realize the Fourier transform of RF signals.

[0050] Example 2: Experimental verification of convolution operation using time-coded metasurface.

[0051] To verify the feasibility of using time-coded metasurfaces to perform convolution operations, this embodiment uses matched filtering of linear frequency modulated signals as an example for experimentation. Matched filtering is widely used in various fields such as communication, radar, and sensing for signal detection and recognition, improving anti-interference performance, and signal parameter estimation. The experimental setup is as follows: a radar simulator is connected between the signal transmission module and the transmitting antenna to simulate a real radar ranging environment. After the RF signal passes through the simulator, the retransmitted signal carries complex white noise, range delay, and Doppler frequency. The transmitted signal parameters under the simulated environment are: signal bandwidth 1MHz, pulse repetition period (PRT) 100μs, corresponding to a maximum unambiguous distance of 15km and a range resolution of 50m. Within the range measurement interval of 0-15km, the results of RF calculations and their deviations from the theoretical positions are recorded in 1km increments. Specific data are shown in Table 1. The peak positions of each test group all fall within the allowable range of the resolution, verifying the application effect of time-coded metasurfaces in matched filtering. Figure 6The comparison between the RF calculation waveforms and theoretical numerical calculation results is shown when the target distances are 1 km, 3 km, 10 km, and 14 km. The results show that the RF calculation waveforms exhibit a Sinc function shape, which is highly consistent with the numerical calculation results. This demonstrates that the time-coded metasurface can still achieve accurate RF calculations in complex, multi-interference, and multi-noise environments. This embodiment proves the effectiveness and accuracy of convolution operations performed by the time-coded metasurface, providing theoretical and experimental basis for its application in practical signal processing.

[0052] In summary, this invention utilizes a time-coded metasurface to implement Fourier transform and convolution operations in signal processing by designing a simple periodic control voltage signal. This radio frequency calculator is low-cost, simple in structure, easy to integrate, and has broad application prospects and significant research value, applicable to multiple fields such as communication, radar, and sensing.

[0053] The basic principles, main features, and advantages of this invention have been shown and described in detail in the foregoing. Those skilled in the art should understand that this invention is not limited to these embodiments, and the descriptions in the embodiments and specification are only for explaining the principles of the invention. Various modifications and improvements can be made to this invention without departing from its spirit and scope, and all such modifications and improvements are within the protection scope of this invention. The protection scope of this invention is defined by the appended claims and their equivalents.

Claims

1. A radio frequency calculation method based on time-coded metasurfaces, characterized in that, Includes the following steps: The mapping relationship between the echo signal R and the reflection coefficient encoding Γ of the time-coded metasurface can be represented in the following matrix form: Among them, I=[I[n]] N×1 Let R be the incident signal matrix. K×1 ,Γ K×N These are the echo signal matrix and the reflection coefficient encoding matrix, respectively, n = 0, ..., N-1, representing the index of the matrix in the time dimension, and N and K representing the lengths of the input vector and the output vector, respectively; Design the reflection coefficient encoding matrix Γ according to radio frequency calculation requirements. K×N This allows the echo signal to carry information about the calculation result; By demodulating the echo signal from the time-coded metasurface, waveform information containing the required calculation results can be obtained; The time-coded metasurface is a reflective structure composed of M×L basic units, in which units in the same column share the same control signal, enabling dynamic control of the radio frequency signal. The basic unit consists of a metal patch, a dielectric substrate, and a metal ground plane from top to bottom. A varactor diode is integrated on the metal patch, and the reflection phase of the metasurface can be controlled within the range of 0 to 360° by adjusting the bias voltage across its terminals. The radio frequency calculation is a Fourier transform, so the design of the reflection coefficient coding matrix is ​​as follows: The reflected echo from the metasurface will carry information about the Fourier transform of the incident wave. The radio frequency calculation is a convolution operation, so the designed reflection coefficient coding matrix is ​​Γ. K×N =[Γ0,…,Γ k ,…,Γ (N-M+1) ] T , where Γ k =[[0] 1×k H 1×M [0] 1×(K-k-1) ], where H 1×M This is the impulse response matrix.

2. A radio frequency computing system based on a time-coded metasurface, characterized in that, It includes a time-coded metasurface, a vector signal transceiver, a waveform generator, and an RF calculator; the vector signal transceiver is used to generate the input signal to be calculated and to receive the echo signal to demodulate the calculation result; The waveform generator is used to generate a periodic voltage control signal to achieve periodic switching of the metasurface reflection phase; the RF calculator is used to perform RF calculations according to the method of claim 1.

Images