Transmit link frequency conversion method for 92-94GHz millimeter-wave radar systems

Abstract

This application relates to a transmit link frequency conversion method for a 92-94 GHz millimeter-wave radar system, comprising: acquiring a 79-81 GHz input radio frequency signal and a 13 GHz local oscillator signal; performing mixing processing through a mixer to generate a mixed output signal containing a 92-94 GHz target frequency component and a non-target frequency component; determining the frequency parameters, amplitude parameters, and phase parameters of the non-target frequency component based on the sampled data sequence; generating a compensation signal according to the frequency parameters, amplitude parameters, and phase parameters of the non-target frequency component; and synthesizing the compensation signal with the mixed output signal to obtain the 92-94 GHz output signal. This application improves the spectral purity of the 92-94 GHz output signal.

Description

Technical Field

[0001] This application relates to the field of radio frequency processing technology, and in particular to a transmit link frequency conversion method for a 92-94 GHz millimeter-wave radar system. Background Technology

[0002] Currently, traffic millimeter-wave radar mainly operates in the 76-81GHz frequency band and relies on mature automotive-grade SoC chips. However, to meet higher performance requirements, emerging millimeter-wave frequency bands such as 92-94GHz are being explored, but they face a bottleneck due to the lack of existing SoC solutions, preventing a smooth and direct migration of the system. This expansion of frequency bands makes achieving efficient frequency up-conversion through mixers a key technical challenge.

[0003] During frequency conversion, due to the physical characteristics of the components, mixers inevitably generate non-target frequency components when converting the input signal to the target frequency. To suppress these non-target frequency components, existing technologies typically incorporate hardware bandpass filters at the mixer's output. These filters, through their designed frequency selectivity, physically block or attenuate signals outside the target frequency band, thereby allowing only the desired 92-94 GHz target frequency components to pass through.

[0004] However, in the 92-94 GHz millimeter-wave high-frequency band, even high-performance filters struggle to completely suppress non-target frequency components with high power or frequencies very close to the target signal, resulting in residual stray signals in the output signal. Therefore, current filter solutions are significantly inadequate in ensuring high-purity millimeter-wave signals. Summary of the Invention

[0005] Based on this, it is necessary to provide a transmit link frequency conversion method for 92-94GHz millimeter-wave radar systems that improves the spectral purity of the 92-94GHz output signal by accurately identifying and actively canceling non-target frequency components in the mixing output, in order to address the aforementioned technical problems.

[0006] This application provides a transmit link frequency conversion method for a 92-94GHz millimeter-wave radar system. The method is executed by a signal processing unit and includes:

[0007] Acquire 79-81GHz input RF signal and 13GHz local oscillator signal;

[0008] A 79-81GHz input RF signal is input to the first input port of the mixer, and a 13GHz local oscillator signal is input to the second input port of the mixer. The mixer performs mixing processing to generate a mixed output signal containing a 92-94GHz target frequency component and non-target frequency components. The non-target frequency components are frequency components generated during the mixing process other than the 92-94GHz target frequency component.

[0009] The mixed output signal is sampled to obtain a sampled data sequence;

[0010] Based on the sampled data sequence, determine the frequency parameters, amplitude parameters, and phase parameters of the non-target frequency components;

[0011] A compensation signal is generated based on the frequency, amplitude, and phase parameters of the non-target frequency component; wherein, the compensation signal has the characteristic of having the same amplitude and opposite phase as the non-target frequency component at the frequency of the non-target frequency component.

[0012] The compensation signal and the mixing output signal are combined to obtain the 92-94GHz output signal.

[0013] In one embodiment, after the mixing process and before generating the compensation signal, the method further includes:

[0014] Continuously monitor the spectral characteristics of the mixer output signal;

[0015] Real-time analysis of spectral characteristics and mixer environmental operating parameters dynamically identifies the instantaneous nonlinear transfer function of the mixer nonlinear unit that causes non-target frequency components, and obtains the identification results;

[0016] Based on the identification results, an optimization strategy is determined to adjust multiple controllable physical parameters of the mixer nonlinear unit; among which, multiple controllable physical parameters include, but are not limited to, multi-point bias voltage, adjustable element parameters of the active matching network, internal feedback loop gain, and local oscillator injection lock-in parameters.

[0017] Based on the optimization strategy, the controllable physical parameters of the mixer's nonlinear unit are dynamically adjusted to reconstruct the mixer's instantaneous nonlinear transfer function.

[0018] Among them, the reconstructed instantaneous nonlinear transfer function has the characteristic of suppressing the generation of non-target frequency components during the mixing process, or has the characteristic of guiding the energy of non-target frequency components to non-sensitive frequency bands outside the 92-94GHz target frequency components.

[0019] In one embodiment, the compensation signal is combined with the mixing output signal, including:

[0020] The signal transmission path after the compensation signal is input to the output port of the mixer;

[0021] Based on the frequency, amplitude, and phase parameters of the non-target frequency components, the coupling coefficient and phase delay parameters in the signal transmission path are determined so that the compensation signal and the non-target frequency components achieve amplitude matching and phase inversion, thus completing signal synthesis.

[0022] In one embodiment, the coupling coefficient and phase delay parameters in the signal transmission path are determined based on the frequency parameters, amplitude parameters, and phase parameters of the non-target frequency components, including:

[0023] Calculate the coupling coefficient in the signal transmission path based on the amplitude parameters of the non-target frequency components;

[0024] Calculate the phase delay parameter in the signal transmission path based on the phase parameter of the non-target frequency component.

[0025] In one embodiment, the coupling coefficient in the signal transmission path is calculated based on the amplitude parameter of the non-target frequency component, including:

[0026] Set a small constant to prevent division by zero, and calculate the sum of the amplitude parameters of non-target frequency components and the small constant;

[0027] Determine the first proportional constant and divide it by the sum of the amplitude parameter and the small constant to obtain the coupling coefficient in the signal transmission path.

[0028] In one embodiment, the phase delay parameter in the signal transmission path is calculated based on the phase parameter of the non-target frequency component, including:

[0029] Calculate the difference between the phase parameters of π and the non-target frequency components;

[0030] Determine the second proportional constant and multiply it by the difference to obtain the phase delay parameter in the signal transmission path.

[0031] In one embodiment, determining the frequency parameters of non-target frequency components based on the sampled data sequence includes:

[0032] Perform a discrete Fourier transform on the sampled data sequence to obtain the frequency domain signal data;

[0033] Discrete frequency points with power values ​​greater than a preset noise threshold are identified from the frequency domain signal data and used as frequency parameters of non-target frequency components; wherein, the preset noise threshold is dynamically determined based on the system noise floor.

[0034] In one embodiment, the amplitude and phase parameters of the non-target frequency components are determined based on the sampled data sequence, including:

[0035] Based on the complex values ​​of discrete frequency points in the frequency domain signal data, the amplitude and phase parameters of non-target frequency components are calculated.

[0036] In one embodiment, before inputting the 13GHz local oscillator signal to the second input port of the mixer, the method further includes:

[0037] Acquire the spectral characteristics data of the 13GHz local oscillator signal;

[0038] Based on the difference between the spectral characteristic data and the preset ideal spectrum, calculate the operating bias voltage adjustment parameters of the mixer's nonlinear unit;

[0039] The operating bias voltage adjustment parameter is applied to the mixer nonlinear unit.

[0040] In one embodiment, the operating bias voltage adjustment parameters of the mixer's nonlinear unit are calculated based on the difference between the spectral characteristic data and a preset ideal spectrum, including:

[0041] Obtain the phase noise spectrum of the 13GHz local oscillator signal;

[0042] Based on the phase noise spectrum, the Wiener filtering algorithm is used to calculate the filter coefficients of the local oscillator signal purification filter;

[0043] The 13GHz local oscillator signal is passed through a local oscillator signal purification filter to obtain a local oscillator signal with phase noise suppression;

[0044] Based on the signal characteristics of the local oscillator signal after phase noise suppression, the operating bias voltage adjustment parameters of the mixer's nonlinear unit are calculated.

[0045] The aforementioned frequency conversion method for the transmit link of a 92-94 GHz millimeter-wave radar system, as described in this application, achieves precise acquisition of the frequency, amplitude, and phase parameters of non-target frequency components by digitally sampling the mixed output signal. Based on these precise parameters, a compensation signal with the same frequency and amplitude as the non-target frequency components but with opposite phase is generated. Subsequently, this compensation signal is precisely synthesized with the mixed output signal, achieving effective and active cancellation of the non-target frequency components. This active cancellation mechanism based on digital signal processing overcomes the inherent limitations of traditional filters in the millimeter-wave band. Especially for spurious signals with frequencies close to or high power than the target signal, it can achieve more thorough and flexible suppression, thereby significantly improving the spectral purity of the 92-94 GHz output signal and enhancing the overall system performance and reliability. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is an application environment diagram of a transmit link frequency conversion method for a 92-94GHz millimeter-wave radar system in one embodiment.

[0048] Figure 2 This is a flowchart illustrating a transmit link frequency conversion method for a 92-94GHz millimeter-wave radar system in one embodiment.

[0049] Figure 3 This is a flowchart illustrating the dynamic optimization steps for the nonlinear characteristics of a mixer in one embodiment.

[0050] Figure 4 This is a flowchart illustrating the steps of combining the compensation signal and the mixing output signal in one embodiment.

[0051] Figure 5 This is a flowchart illustrating the steps for determining the frequency parameters, amplitude parameters, and phase parameters of a non-target frequency component in one embodiment.

[0052] Figure 6 This is a flowchart illustrating the step of applying operating bias voltage adjustment parameters to the nonlinear unit of the mixer in one embodiment. Detailed Implementation

[0053] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0054] In one exemplary embodiment, such as Figure 1 As shown, a system for transmit link frequency conversion in a 92-94 GHz millimeter-wave radar system is provided. The system includes a signal processing unit and a mixer. The method is executed by the signal processing unit, such as... Figure 2 As shown, the method includes:

[0055] S100 acquires 79-81GHz input RF signals and 13GHz local oscillator signals.

[0056] The 79-81GHz input radio frequency signal is the radio frequency excitation signal of the radar system's original operating frequency band, representing the standard operating frequency band of current automotive millimeter-wave radar systems. The 13GHz local oscillator (LO) signal is a reference signal used for mixing with the input radio frequency signal.

[0057] When an 80GHz radio frequency signal is mixed with a 13GHz local oscillator signal, two main frequency components are generated: 93GHz (80+13) and 67GHz (80-13), with the 93GHz component falling within the target frequency band of 92-94GHz. This design avoids the technical challenge of directly generating a 94GHz signal in traditional millimeter-wave radar systems, reducing system implementation complexity. In practical applications, this design can effectively reduce system power consumption, improve signal stability, and provide a solid foundation for subsequent mixing processing.

[0058] S200: Input the 79-81GHz input RF signal to the first input port of the mixer, and input the 13GHz local oscillator signal to the second input port of the mixer. The mixer performs mixing processing to generate a mixed output signal containing the 92-94GHz target frequency component and non-target frequency components.

[0059] In this embodiment, the mixer is a nonlinear device, typically employing a Schottky diode or transistor mixing structure. The non-target frequency components are frequency components generated during the mixing process other than the 92-94 GHz target frequency component.

[0060] Assume the input radio frequency signal and local oscillator signal ;

[0061] in, : Represents the instantaneous voltage of the input radio frequency signal that varies with time;

[0062] ): Represents the instantaneous voltage of the local oscillator signal that varies with time;

[0063] : Indicates the amplitude of the input radio frequency signal;

[0064] : Indicates the amplitude of the local oscillator signal;

[0065] : Represents the angular frequency of the input radio frequency signal, where =2π ;

[0066] : Represents the angular frequency of the local oscillator signal, where =2π ;

[0067] : Indicates the initial phase of the input radio frequency signal;

[0068] : Represents the initial phase of the local oscillator signal; t: Represents the time variable.

[0069] At this point, the mixer output is:

[0070]

[0071] in, : represents the instantaneous voltage of the mixer's output signal; k is the mixer gain constant;

[0072] Then, using trigonometric identities Expanding, we get:

[0073]

[0074]

[0075] S300: Sample the mixing output signal to obtain a sampled data sequence.

[0076] In this embodiment, sampling refers to digitizing the analog signal output from the mixer using a high-speed analog-to-digital converter (ADC) to obtain a sampled data sequence. Specifically, bandpass sampling technology is employed, setting the sampling clock frequency to 95GHz, thus "folding" the 92-94GHz frequency band signal into the baseband 0-3GHz range, thereby reducing the performance requirements of the ADC. The sampled data sequence is a discrete-time signal generated during the sampling process, denoted by x[n], where n is the sampling sequence number. The sampling process can be represented as:

[0077]

[0078] in, The output signal is a continuous-time mixing signal. Let x[n] be the sampling period, and x[n] be the discrete sampling sequence.

[0079] This embodiment employs bandpass sampling technology, by using the sampling frequency f s Set to 95GHz, so that the center frequency is f c A mixer output signal with a frequency of 93 GHz and a bandwidth of B = 2 GHz (i.e., 92-94 GHz) can be folded down to a lower frequency range (e.g., 1-3 GHz) for processing without aliasing.

[0080] S400. Based on the sampled data sequence, determine the frequency parameters, amplitude parameters, and phase parameters of the non-target frequency components.

[0081] In this embodiment, the specific frequency value of the non-target frequency component is, for example, in actual implementation, a significant non-target frequency component is detected at 67.2 GHz, with a frequency parameter of 67.205 GHz, which is determined by spectral interpolation technology with an accuracy of ±1 MHz.

[0082] Amplitude parameters refer to the signal strength of non-target frequency components, usually expressed in dBm. For example, the amplitude parameter of the 67.2GHz non-target frequency component mentioned above is -12.5dBm, indicating that its power level is 12.5dB lower than 1mW.

[0083] Phase parameters refer to the phase information of non-target frequency components, expressed in angles or radians. For example, the phase parameter of the aforementioned 67.2 GHz non-target frequency component is 45°, indicating its phase offset relative to the reference signal.

[0084] Optionally, this parameter determination process is implemented using digital signal processing technology. Specifically, firstly, the sampled data sequence is windowed to obtain a windowed sampled data sequence (a Hanning window is used in this embodiment). The window length is dynamically adjusted according to the expected bandwidth of the non-target frequency component (N=1024 in this embodiment). Then, a Discrete Fourier Transform (DFT) is performed on the windowed data to obtain frequency domain signal data. Discrete frequency points with power values ​​greater than a preset noise threshold are identified from the frequency domain data as frequency parameters of the non-target frequency component. Finally, the amplitude and phase parameters of the non-target frequency component are calculated based on the complex values ​​in the frequency domain data.

[0085] Specifically, the key formula for the parameter extraction process is:

[0086]

[0087] Where x[n] is the sampled data sequence, w[n] is the window function, and X[k] is the frequency domain signal data.

[0088] The preset noise threshold is dynamically determined based on the system noise floor, specifically the average noise floor value plus three times the standard deviation. For identified non-target frequency components, their precise frequency positions are determined using spectral interpolation techniques, and their amplitude and phase are calculated based on the complex value of X[k].

[0089] S500 generates a compensation signal based on the frequency, amplitude, and phase parameters of the non-target frequency components.

[0090] The compensation signal has the characteristic that it has the same amplitude and opposite phase as the non-target frequency component at the frequency of the non-target frequency component.

[0091] In this embodiment, the compensation signal is generated based on the parameters extracted by S400, and it has the characteristic of having the same amplitude but opposite phase as the non-target frequency component at the same frequency. Specifically, the compensation signal generator generates the signal based on the frequency parameters of the non-target frequency component. Amplitude parameters and phase parameters Construct a compensation signal of the following form. :

[0092]

[0093] The phase offset π ensures that the compensation signal is out of phase with the non-target frequency component.

[0094] For example, for a non-target frequency component at 67.2 GHz with an amplitude of -12.5 dBm and a phase of 45°, the generated compensation signal has an amplitude of -12.5 dBm and a phase of 225° (45° + 180°) at 67.2 GHz. For example, when the phase of the non-target frequency component is 45°, the phase of the compensation signal is set to 225°, and the phase difference between the two is 180°.

[0095] Specifically, the compensation signal generation employs Direct Digital Frequency Synthesis (DDS) technology, implemented using a Field-Programmable Gate Array (FPGA). The FPGA internally includes a phase accumulator, a phase-to-amplitude conversion table, and a digital-to-analog converter (DAC). The initial phase of the phase accumulator is set... n +π ensures that the phase of the signal is opposite to that of the non-target frequency component. The amplitude An of the compensation signal is strictly equal to the amplitude of the non-target frequency component, which is achieved by adjusting the amplitude control word of the DDS.

[0096] S600 combines the compensation signal with the mixing output signal to obtain a 92-94GHz output signal.

[0097] Signal synthesis is the process of combining two or more signals into a single signal. In this embodiment, signal synthesis is used to cancel out the non-target frequency components in the compensation signal and the mixer output signal. For example, a directional coupler is used as the signal synthesizer, with the mixer output signal input to the main input port and the compensation signal input to the coupling port.

[0098] Amplitude matching means that the amplitude of the compensation signal at a non-target frequency is equal to the amplitude of the component at that frequency in the mixer output signal. For example, when the amplitude of the 67.2GHz component in the mixer output signal is -12.5dBm, the amplitude of the compensation signal at 67.2GHz is also set to -12.5dBm.

[0099] In this embodiment, the signal synthesis principle is based on linear superposition, specifically:

[0100] Assuming the mixer output signal Includes a target frequency component and a non-target frequency component to be suppressed The compensation signal generated in this embodiment Intended to offset ,use This represents the synthesized output signal.

[0101] : Represents the output signal of the mixer, which includes:

[0102] , representing the target frequency component;

[0103] , representing a specific non-target frequency component;

[0104] : Represents the compensation signal generated by the system, and its form is:

[0105] The signal and Having the same amplitude and angular frequency But the phases are opposite. ;

[0106] in, The amplitude of the target frequency component; Angular frequency of the target frequency component (e.g., its center frequency is 93 GHz); : The initial phase of the target frequency component; : Amplitude of non-target frequency components; Angular frequency of non-target frequency components (e.g., angular frequency corresponding to 67.2 GHz); : Initial phase of the non-target frequency component; t: Time variable.

[0107] Based on the principle of linear superposition, the synthesized output signal for:

[0108]

[0109]

[0110]

[0111] Using trigonometric identities The above formula can be further simplified to:

[0112]

[0113] This demonstrates that, through signal synthesis, specific non-target frequency components are effectively suppressed, while only the target frequency components are retained.

[0114] In one exemplary embodiment, such as Figure 3 As shown, after the mixing process and before generating the compensation signal, the method further includes:

[0115] S201. Continuously monitor the spectral characteristics of the mixing output signal.

[0116] Spectral characteristics refer to the signal's behavior in the frequency domain, including key parameters such as amplitude spectrum, phase spectrum, phase noise, and spurious distribution. In this embodiment, the focus is on the signal characteristics near the target frequency band of 92-94 GHz and the main non-target frequency band of 66-68 GHz. For example, in one specific implementation, it was observed that when the mixer operating temperature increased from 25°C to 65°C, the power of the non-target frequency components in the 66-68 GHz range increased by approximately 8 dB, and the phase characteristics also changed significantly.

[0117] Continuous monitoring refers to an uninterrupted, periodic signal measurement process. In this embodiment, the monitoring system employs a sliding window technique, analyzing the signal within a 1ms time window each time and acquiring spectral data through a Fast Fourier Transform (FFT). For example, complete spectral data is collected every 10ms to form a time series for trend analysis. The spectral data for each time window can be obtained by averaging multiple FFT results.

[0118] S202. Real-time analysis of spectral characteristics and mixer environmental operating parameters, dynamic identification of the instantaneous nonlinear transfer function of the mixer nonlinear unit that causes non-target frequency components, and obtaining the identification result.

[0119] Environmental operating parameters refer to external condition parameters that affect the mixer's operating state. In this embodiment, these mainly include the mixer chip's package temperature, its supply voltage, the output power and frequency stability of the local oscillator (LO) signal, and the strength of the input radio frequency (RF) signal. For example, in a specific test, when the ambient temperature rises from 25°C to 65°C, the characteristics of the mixer diodes change, causing the coefficient characterizing the third-order nonlinear characteristic to increase by approximately 60%.

[0120] The nonlinear transfer function is a mathematical model describing the input-output relationship of a mixer, reflecting the nonlinear characteristics of the mixer. In this embodiment, the nonlinear transfer function of the mixer can be expressed as:

[0121]

[0122] x(t): represents the input signal of the mixer; y(t): represents the output signal of the mixer; : Represents the nonlinear coefficient of the mixer; N: Represents the nonlinear order; t: Represents the time variable.

[0123] For example, when the system detects a cubic nonlinear coefficient through real-time parameter estimation. From the initial state of -0.05 (dBV / V) 3 Increased to -0.08 (dBV / V) 3 This explains why the power of non-target frequency components increases.

[0124] In practical implementation, dynamic identification of real-time nonlinear transfer functions refers to parameter estimation based on the spectral characteristics monitored by S201 (especially the amplitude and phase changes of non-target frequency components in the mixer output) and real-time acquired environmental operating parameters, using a pre-established nonlinear model or lookup table.

[0125] Specifically, the recursive least squares (RLS) algorithm is used for online parameter estimation, and the nonlinear coefficients are updated in real time. The estimated value is obtained to acquire the real-time transfer function model of the mixer's nonlinear unit. The time constant of the RLS algorithm is set to 5ms to ensure that the algorithm can quickly track changes in mixer characteristics within the millisecond range. To improve the robustness and accuracy of identification, a correlation model between spectral characteristics and environmental parameters can be established: ΔF=J(θ)·ΔP. Where ΔF can represent the nonlinear transfer function parameters (such as...). The deviation between the coefficient vector and the ideal state, ΔP is the change vector of environmental parameters, and J(θ) is the Jacobian matrix at a specific operating point θ.

[0126] S203. Based on the identification results, determine an optimization strategy for adjusting multiple controllable physical parameters of the mixer's nonlinear unit.

[0127] Among them, controllable physical parameters refer to hardware parameters that can be adjusted to change the nonlinear characteristics of the mixer. In this embodiment, they mainly include:

[0128] Multi-point bias voltage: bias voltage of the mixer diode, adjustable from -0.5V to 0V, with a resolution of 1mV;

[0129] Adjustable component parameters of the active matching network: such as adjustable capacitance value, ranging from 0.1pF to 1pF, with a resolution of 0.01pF;

[0130] Internal feedback loop gain: The gain coefficient of the feedback network, ranging from 0 to 20 dB, with a resolution of 0.1 dB;

[0131] The power / frequency / phase control parameters of the local oscillator injection signal are parameters used to improve the quality of the local oscillator signal (such as phase noise and frequency stability). For example, the amplitude of the injection signal or the impedance matching parameters of the injection circuit, the adjustment range and resolution need to be determined according to the specific injection-locked oscillator characteristics.

[0132] For example, in one specific implementation, when a third nonlinear coefficient is identified... When increasing the voltage, it is determined that the negative bias voltage of the mixer diode needs to be increased by about 15mV, and the capacitance value of the active matching network needs to be fine-tuned by about 0.05pF.

[0133] An optimization strategy refers to a method for determining how to adjust controllable physical parameters to achieve the best results. In this embodiment, the core of the optimization strategy is to establish a relationship model between the nonlinear transfer function and the controllable physical parameters:

[0134]

[0135] in, : Represents the vector of changes in the controllable physical parameters that need to be adjusted; : Represents the deviation vector between the nonlinear transfer function (or its key coefficients) identified by S202 and the ideal state; : Represents the deviation of the nonlinear transfer function at the current operating point θ. With controllable physical parameter changes Jacobian matrix of the relationship between them The inverse matrix of the mixer. This matrix describes how the nonlinear characteristics of the mixer can be changed by adjusting controllable physical parameters.

[0136] The optimization strategy takes into account the interplay of parameter adjustments and system stability. In practice, the impact of each parameter adjustment on overall performance is evaluated to avoid over-adjusting a single parameter and causing deterioration in other performance metrics. For example, when a cubic nonlinear coefficient is identified... When increasing the voltage, not only is the negative bias voltage of the mixer diode considered, but the impact of this adjustment on conversion loss and noise figure is also evaluated to ensure optimal overall performance. In practical applications, the gradient descent method is used to find the optimal adjustment strategy, ensuring that the maximum non-target frequency component suppression effect is obtained with the minimum adjustment.

[0137] S204. Based on the optimization strategy, dynamically adjust the controllable physical parameters of the mixer's nonlinear unit to reconstruct the instantaneous nonlinear transfer function of the mixer.

[0138] The reconstructed instantaneous nonlinear transfer function exhibits the characteristic of suppressing the generation of non-target frequency components during the mixing process, or the characteristic of guiding the energy of non-target frequency components to a non-sensitive frequency band outside the 92-94GHz target frequency component. This dynamic adjustment refers to the real-time adjustment of the controllable physical parameters inside the mixer through a precise analog control circuit based on the optimization strategy determined in S203. Specifically, the controllable physical parameters of the mixer are precisely adjusted by outputting control voltage or current through a digital-to-analog converter (DAC).

[0139] The instantaneous nonlinear transfer function refers to the nonlinear characteristic model of the mixer at a specific moment, reflecting the input-output relationship under the current operating state. In this embodiment, the reconstructed instantaneous nonlinear transfer function can be expressed as:

[0140]

[0141] Reconfiguration refers to altering the nonlinear characteristics of a mixer by adjusting controllable physical parameters. In this embodiment, the reconfigured instantaneous nonlinear transfer function has two possible optimization directions:

[0142] Directly suppress the generation of non-target frequency components, thereby reducing the corresponding nonlinear coefficients (such as cubic nonlinear coefficients). Minimize the absolute value of ).

[0143] Directing the energy of non-target frequency components to a non-sensitive frequency band outside the 92-94 GHz target frequency component, for example by adjusting... If the value is positive, energy will be transferred to a higher frequency band.

[0144] In one exemplary embodiment, such as Figure 4 As shown, S600 synthesizes the compensation signal and the mixing output signal, including:

[0145] S60a, the signal transmission path after the compensation signal is input to the output port of the mixer.

[0146] The purpose of this step is to provide the necessary physical basis for the effective injection of the compensation signal and its interaction with the mixer output signal. The signal transmission path refers to the physical channel from the mixer output port to the final output. In this embodiment, it mainly consists of a low-loss 50Ω characteristic impedance microstrip line, for example, a 3mm long, 0.15mm wide microstrip line laid on an RF substrate with a dielectric constant of 3.55. The compensation signal is injected into this main signal path through a coupling node. The coupling node is a specific location on the compensation signal path used for physical connection with the main signal path and for signal injection.

[0147] In terms of circuit layout, the mixer output port is connected to the main input arm of a signal synthesizer (e.g., a broadband directional coupler or a Wilkinson power synthesizer) via a microstrip line, while the compensation signal is injected into the compensation input arm of the synthesizer through its dedicated compensation path. This design avoids system stability problems that may be caused by directly injecting the compensation signal into the mixer, and significantly reduces the possibility of phase mismatch because the signals propagate in the same transmission medium.

[0148] S60b: Based on the frequency, amplitude, and phase parameters of the non-target frequency components, adjust the coupling coefficient and phase delay parameters in the signal transmission path.

[0149] This step is crucial for achieving precise amplitude matching and phase inversion between the compensation signal and the non-target frequency components, resulting in optimal suppression. The compensation signal gain / attenuation (or coupling coefficient) refers to the gain or attenuation of the compensation signal path at the signal synthesis point, designed to ensure its amplitude is precisely proportional to the amplitude of the non-target frequency components in the mixer output signal. The phase delay coefficient refers to the phase offset of the compensation signal relative to the mixer output signal; its precise adjustment is necessary for phase inversion cancellation. The system calculates and implements these adjustments based on the specific parameters (frequency, amplitude, phase) of the non-target frequency components identified in step S400.

[0150] For example, when the S400 detects a non-target frequency component at 67.2 GHz with an amplitude of -12.5 dBm and a phase of 45°, it will automatically calculate and adjust the gain / attenuation of the compensation signal to -12.5 dB (i.e., ensure that the amplitude of the compensation signal at this synthesis point is also -12.5 dBm). Simultaneously, the phase delay of the compensation signal will be adjusted to be precisely out of phase with the non-target component. The formula for calculating the required phase delay θ is: ,in This is the initial phase of the non-target frequency component. According to this formula, when When the angle is 45°, the required phase delay is 45° + 180° = 225°.

[0151] Specifically, based on the frequency, amplitude, and phase parameters of the non-target frequency components, the coupling coefficient and phase delay parameters in the signal transmission path are adjusted, including:

[0152] The first step is to calculate the coupling coefficient in the signal transmission path based on the amplitude parameters of the non-target frequency components.

[0153] This step aims to accurately calculate the required compensation signal gain / attenuation based on the real-time monitored amplitude of the non-target frequency components, ensuring that the compensation signal reaches an amplitude that matches the non-target components at the synthesis point. A small constant ε is introduced during the calculation, for example, set to 0.001. This value was determined through statistical analysis of a large amount of experimental data, and its main function is to prevent errors due to the amplitude of the non-target frequency components during calculation. To prevent division by zero errors or numerical instability caused by values ​​that are too small (close to zero), thus ensuring the robustness of the calculation process. For example, when the amplitude of non-target frequency components... When the value is 0.0001 (corresponding to -40dBm), the calculation will use... + =0.0011, which effectively avoids numerical instability problems, and has minimal impact on the calculation results under normal working conditions (less than 0.1dB).

[0154] The calculation also introduced a proportionality constant. Its function is to convert theoretically calculated values ​​into gain factors that convert control parameters that can be executed by actual hardware. The determination is based on offline calibration and online calibration, and dynamically considers the influence of ambient temperature. For example, The calculation formula is:

[0155]

[0156] in, The base value is α, where α is the temperature coefficient and T is the current temperature. This is a reference temperature.

[0157] Based on these parameters, the coupling coefficient The calculation formula is:

[0158]

[0159] in, It is a small constant (usually set to 0.001) to prevent division by zero errors and to ensure the numerical stability of the calculation process when the amplitude of non-target frequency components is too small.

[0160] The second step is to calculate the phase delay parameter in the signal transmission path based on the phase parameter of the non-target frequency component.

[0161] This step aims to provide a precise method for calculating phase delay, ensuring that the compensation signal and the non-target frequency component are perfectly out of phase, thus achieving optimal cancellation. The calculation is based on phase parameters. This refers to the phase information of non-target frequency components, typically expressed in radians, with a value range of [0, 2π]. In this embodiment, the phase parameter... The phase is obtained through frequency domain analysis in step S400, taking into account phase dewinding and fixed phase offset correction. For example, when the original phase measurement is 0.785 radians (corresponding to 45°), after system-preset fixed phase offset correction (such as...), =0.1 radians), corrected phase parameters It is 0.685 radians.

[0162] To compensate for the non-ideal characteristics in the system, a second proportionality constant is introduced in the calculation. . K2 is a scaling factor, determined based on offline calibration and compensation for environmental changes. The formula for calculating K2 is:

[0163]

[0164] in, The base value is α, the temperature coefficient is T, the current temperature is , and the reference temperature is .

[0165] The introduction of this feature provides the system with additional degrees of adjustment freedom, allowing for adjustments in specific application scenarios. Set to a value slightly greater than or less than 1 to compensate for a fixed phase shift present in the signal transmission path. For example, when there is a fixed 2° phase shift in the signal transmission path, it can be... Setting it to 0.98 allows the phase delay parameter to automatically compensate for this deviation. Tests show that this design improves the phase matching accuracy by about 3°, significantly improving the interference suppression effect.

[0166] Based on the above parameters, the phase delay parameter The calculation formula is:

[0167]

[0168] This calculation process is performed by a digital signal processor, taking into account several engineering details to ensure accuracy. For example, the phase parameter of the non-target frequency component... When the value is 0.785 radians (corresponding to 45°) and the ambient temperature T is 65°C, Therefore, the phase delay parameter .

[0169] Furthermore, the compensation signal and the non-target frequency components in the mixing output signal are amplitude matched and phase inverted, thereby completing signal synthesis.

[0170] This step achieves active cancellation by precisely matching amplitude and inverting phase, superimposing the compensation signal with the non-target frequency component in the mixer output signal. Amplitude matching ensures that the amplitude of the compensation signal at the synthesis point is exactly equal to the amplitude of the non-target component; phase inversion means that the phase of the compensation signal is precisely 180° out of phase with the phase of the non-target component. For example, if the S400 identifies a 67.2GHz non-target component with an amplitude of -12.5dBm and a phase of 45°, the compensation signal amplitude is precisely set to -12.5dBm, and the phase is set to 225° (45° + 180°).

[0171] Signal synthesis is performed using a signal synthesizer, for example, a directional coupler with a directivity better than 25dB, ensuring the accuracy and amplitude-phase stability of the synthesis process. In a typical test scenario, the mixer output signal contains a 93GHz target component and a -12.5dBm, 45° non-target component at 67.2GHz. The system sets the amplitude of the compensation signal to -12.5dBm and the phase to 225°. After synthesis, the power of the 67.2GHz component drops from -12.5dBm to -42.5dBm, achieving a suppression effect of 30dB. Simultaneously, the 93GHz target signal power is reduced by only 0.2dB, remaining almost intact. This active cancellation mechanism significantly eliminates non-target frequency components, resulting in a clean 92-94GHz target output signal.

[0172] In one exemplary embodiment, such as Figure 5 As shown, S400 determines the frequency parameters, amplitude parameters, and phase parameters of the non-target frequency components, including:

[0173] S40a. Perform a discrete Fourier transform on the sampled data sequence to obtain the frequency domain signal data.

[0174] The Discrete Fourier Transform (DFT) is a mathematical transformation that converts a discrete-time signal into a discrete-frequency signal. In this embodiment, the mathematical definition of DFT is:

[0175]

[0176] Where x[n] is the sampled data sequence, N is the number of transformation points (N=1024 in this embodiment), and X[k] is the frequency domain signal data.

[0177] Frequency domain signal data is the result of DFT transformation, representing the distribution of the signal in the frequency domain. For example, in a specific scenario, the sampled data sequence x[n] contains a 93GHz target signal and a 67GHz non-target signal. After DFT processing, the frequency domain signal data X[k] shows obvious peaks near k=731 (corresponding to 67GHz) and near k=995 (corresponding to 93GHz).

[0178] S40b: Identify discrete frequency points in the frequency domain signal data whose power values ​​are greater than a preset noise threshold.

[0179] Specifically, discrete frequency points that exceed a preset noise threshold are used as frequency parameters for non-target frequency components.

[0180] The preset noise threshold is a power threshold used to distinguish between signal and noise. In this embodiment, the noise threshold... The calculation formula is:

[0181]

[0182] Where μ is the mean of the spectrum basis, σ is the standard deviation, and β is the threshold coefficient (β=3 in this embodiment).

[0183] Discrete frequency points are frequency points in frequency domain signal data where the power value exceeds the noise threshold, representing potential signal components. For example, in a test scenario, if the average power of the noise floor is -95dBm and the standard deviation is 2dBm, then the noise threshold is set to -89dBm (-95dBm + 3 × 2dBm). Using this threshold, a discrete frequency point at 67.2GHz (power of -85dBm) was successfully identified, while noise fluctuations below -90dBm were ignored.

[0184] Optionally, the dynamic determination process of the noise floor is as follows: divide the spectrum into multiple sub-bands and exclude known signal regions; calculate the average power and standard deviation of each sub-band; estimate the system noise floor using robust statistical methods (such as median absolute deviation); and dynamically set the noise threshold based on the estimation results.

[0185] For example, in a test scenario, by analyzing regions far from the signal components, such as 50-60GHz and 95-100GHz, the average power of the system noise floor was determined to be -95dBm, with a standard deviation of 2dBm. The noise threshold was then set to -89dBm. Using this threshold, a discrete frequency point at 67.2GHz (with a power of -85dBm) was successfully identified, while noise fluctuations below -90dBm were ignored.

[0186] S40c calculates the amplitude and phase parameters of non-target frequency components based on the complex values ​​of discrete frequency points in the frequency domain signal data.

[0187] The amplitude parameter refers to the signal strength of the non-target frequency component, typically expressed in dBm. In this embodiment, the amplitude parameter... The calculation formula is:

[0188]

[0189] in, Discrete frequency points Complex values ​​at, |X[ ] | is its modulus.

[0190] Phase parameters refer to the phase information of non-target frequency components, expressed in angles or radians. In this embodiment, the formula for calculating the phase parameters is:

[0191]

[0192] in, Represents the phase angle of a complex number.

[0193] In practical applications, due to the discrete nature of the DFT, the precise frequency of a non-target frequency component may lie between two DFT points. This embodiment uses spectral interpolation to determine the precise frequency location:

[0194]

[0195] Precise frequency The calculation formula is:

[0196]

[0197] Where δ is the frequency offset (|δ|<0.5). For precise frequency, Where N is the sampling rate and N is the number of DFT points.

[0198] For example, when the true frequency of the non-target frequency component is 67.205 GHz, the DFT can only estimate it to 67.2 GHz (resolution 92.97 MHz). The frequency estimation accuracy can be improved to 67.205 ± 0.1 MHz by using spectral interpolation techniques.

[0199] Calculate the amplitude and phase parameters based on the precise frequency location:

[0200]

[0201] in, Obtained through linear interpolation or higher-order interpolation.

[0202] In another implementation, after obtaining the aforementioned discrete frequency points, in order to obtain more accurate amplitude and phase parameters of the non-target frequency components, this embodiment also provides another implementation, including:

[0203] (a) The sampled data sequence is windowed with the time position corresponding to the discrete frequency point as the center, and the window length is dynamically determined according to the expected bandwidth of the non-target frequency component.

[0204] (b) Perform adaptive weighted discrete Fourier transform on the windowed sampled data sequence to obtain high-resolution transform results, wherein the weight coefficients are dynamically adjusted based on the signal-to-noise ratio determined by the amplitude parameters of the non-target frequency components calculated based on the complex values ​​of discrete frequency points in the frequency domain signal data.

[0205] (c) Construct a signal model based on the high-resolution transformation results in the neighborhood of discrete frequency points:

[0206]

[0207] in: M is the frequency domain index corresponding to the discrete frequency point, M is the number of points in the adaptive weighted discrete Fourier transform, δ is the subbin frequency offset and |δ|<0.5, and A is the amplitude parameter. For phase parameters;

[0208] (d) Solve the signal model using the least squares method to obtain the optimal estimates of the model parameters. ,in:

[0209] The preliminary amplitude parameter obtained by calculating the amplitude and phase parameters of non-target frequency components based on the complex values ​​of discrete frequency points in frequency domain signal data is equal to |X[ The initial phase parameter is equal to ∠X[ ];

[0210] The precise amplitude parameter obtained by solving the signal model using the least squares method is equal to Â, and the precise phase parameter is equal to Â. The precise frequency parameter is equal to ( + )·f_s / M; where f_s is the sampling frequency.

[0211] In one exemplary embodiment, before the 79-81 GHz input RF signal and the 13 GHz local oscillator signal are input to the mixer for mixing processing, as follows: Figure 6 As shown, the method also includes:

[0212] Sa, acquire the spectral characteristics data of the 13GHz local oscillator signal.

[0213] The spectral characteristics data refer to the signal's performance in the frequency domain, including key parameters such as amplitude spectrum, phase noise, and spurious distribution. In this embodiment, the focus is on the phase noise characteristics of the 13GHz local oscillator signal, particularly the phase noise within the 10Hz to 1MHz offset range. For example, in one specific implementation, the phase noise of the 13.000GHz local oscillator signal was detected to be -105dBc / Hz at a 1MHz offset, and -85dBc / Hz at a 10kHz offset.

[0214] Real-time monitoring refers to the continuous and periodic measurement and recording of the local oscillator signal. In this embodiment, the spectral data of the local oscillator signal is acquired at an update rate of 100kHz, the spectral analysis range covers 12-14GHz, and the frequency resolution reaches 100kHz. For example, spectral characteristic data is acquired every 10ms to form a time series for trend analysis.

[0215] Optionally, the spectral characteristics of the local oscillator signal will vary with environmental conditions (such as temperature and power supply voltage) and operating time. In practical applications, superheterodyne spectrum analysis is used to downconvert the 13GHz signal to an intermediate frequency through two frequency conversions for analysis.

[0216] The first local oscillator is set to 12.5GHz, and the 13GHz signal is downconverted to 500MHz;

[0217] The second local oscillator is set to 450MHz, and the 500MHz signal is further down-converted to a 50MHz intermediate frequency;

[0218] Digital processing is performed using a 100MSPS ADC.

[0219] Sb, obtain the phase noise spectrum of the 13GHz local oscillator signal.

[0220] Among them, the phase noise spectrum is a key indicator describing the phase stability of the local oscillator signal, and is defined as:

[0221]

[0222] in, Where is the carrier frequency (13 GHz), f is the offset frequency, and Δf is the measurement bandwidth. To be at carrier frequency At f, where f is the offset frequency, The single-sideband noise power measured within the measurement bandwidth. This refers to the carrier power.

[0223] Phase noise measurement: Specific technical means for obtaining the phase noise spectrum. In this embodiment, a combination of the direct spectrum method and the phase detector method is used: the direct spectrum method is suitable for phase noise measurement with a large offset (100kHz); the phase detector method is suitable for phase noise measurement with a small offset (<100kHz). For example, in a test scenario, the phase noise spectrum of a 13GHz local oscillator signal is measured as follows:

[0224] 10Hz offset: -65dBc / Hz;

[0225] 100Hz offset: -85dBc / Hz;

[0226] 1kHz offset: -95dBc / Hz;

[0227] 10kHz offset: -100dBc / Hz;

[0228] 100kHz offset: -105dBc / Hz;

[0229] 1MHz offset: -110dBc / HzSc. Based on the phase noise spectrum, the filter coefficients of the local oscillator signal purification filter are calculated using the Wiener filtering algorithm.

[0230] Sc. Based on the phase noise spectrum, the Wiener filtering algorithm is used to calculate the filter coefficients of the local oscillator signal purification filter.

[0231] In this embodiment, the Wiener filtering algorithm is one of the core innovations, used to design the optimal local oscillator signal purification filter. The goal of the Wiener filter is to minimize the mean square error between the output signal and the ideal local oscillator signal, and its transfer function H(f) is:

[0232]

[0233] in, The power spectrum of the desired signal (ideal local oscillator) This is the noise power spectrum.

[0234] A Wiener filter is a minimum mean square error (MMSE) filter used to extract signals from noise. In this embodiment, the Wiener filter is used to purify the local oscillator signal and suppress phase noise. For example, when the phase noise spectrum is -110 dBc / Hz at a 1 MHz offset, the calculated gain of the Wiener filter at that frequency is:

[0235]

[0236] This indicates that the filter passes through almost the entire frequency component because its noise power is much lower than the signal power.

[0237] The local oscillator signal purification filter is a digital filter designed based on the Wiener filtering algorithm, used to improve the quality of the local oscillator signal. In this embodiment, the filter order is set to 64.

[0238] The specific calculation process is as follows:

[0239] 1. From the phase noise spectrum Calculate the noise power spectrum :

[0240]

[0241] in, ;

[0242] 2. Assume the power spectrum of the ideal local oscillator signal is as follows:

[0243] Since the ideal delta function cannot be physically realized, this embodiment uses a Gaussian function approximation. To simulate an ideal local oscillator signal with limited bandwidth:

[0244]

[0245] in, Here, σ represents the peak power density of the Gaussian spectrum, and σ is the standard deviation of the Gaussian function in the time domain (σ=10 in this embodiment). -7 (seconds), corresponding to an equivalent bandwidth of approximately 1.59 MHz.

[0246] 3. Calculate the frequency response of the Wiener filter. :

[0247]

[0248] 4. Obtain the time-domain filter coefficients through inverse Fourier transform. :

[0249]

[0250] in, This represents the inverse Fourier transform.

[0251] Sd. The 13GHz local oscillator signal is passed through a local oscillator signal purification filter to obtain a local oscillator signal with phase noise suppression.

[0252] The local oscillator signal purification filter is a digital filter implemented based on Wiener filter coefficients, used to improve the quality of the local oscillator signal. In this embodiment, the filter adopts a 64th-order FIR structure, and the coefficients are calculated using the Wiener filtering algorithm in step Sc.

[0253] The local oscillator signal after phase noise suppression is the local oscillator signal processed by a purification filter, resulting in lower phase noise. For example, when the phase noise of the original local oscillator signal at a 10kHz offset is -85dBc / Hz, the phase noise of the purified local oscillator signal at that offset is improved to -93dBc / Hz. The mathematical description of the signal purification process is as follows:

[0254]

[0255] in, This is the original local oscillator signal. The impulse response of the Wiener filter is given. This represents the convolution operation. This is the purified local oscillator signal.

[0256] In actual circuit implementation, the process is divided into three stages: downconversion stage: using a 12.5GHz local oscillator to downconvert the 13GHz signal to a 500MHz intermediate frequency; digital filtering stage: digitizing the 500MHz intermediate frequency signal at a 1GSPS sampling rate and applying a 64th-order FIR filter; upconversion stage: upconverting the filtered intermediate frequency signal back to 13GHz.

[0257] Se, based on the signal characteristics of the local oscillator signal after phase noise suppression, calculate the operating bias voltage adjustment parameters of the mixer's nonlinear unit.

[0258] The operating bias voltage adjustment parameter refers to the amount of bias voltage adjustment that needs to be applied to the nonlinear unit of the mixer. In this embodiment, the adjustment parameter... The calculation formula is:

[0259]

[0260] in, This is the optimal bias voltage. This is the current bias voltage.

[0261] The optimal bias voltage is the bias voltage value that optimizes the nonlinear characteristics of the mixer. In this embodiment, the optimal bias voltage... The calculation formula is:

[0262]

[0263] in, As a reference bias voltage, The adjustment factor is ΔL, which represents the improvement in phase noise of the local oscillator signal (unit: dB).

[0264] Specifically, the calculation of the operating bias voltage adjustment parameters is based on the mixer's nonlinear model:

[0265]

[0266] Where x(t) is the input signal and y(t) is the output signal. These are nonlinear coefficients. This embodiment focuses on the cubic nonlinear coefficients. This is because it primarily affects the energy transfer between the sum and difference frequency components.

[0267] Operating bias voltage V b With the third nonlinear coefficient The relationship can be approximated as:

[0268]

[0269] in, This is the optimal bias voltage. and These are constants related to the device characteristics.

[0270] In this embodiment, the calculation of the operating bias voltage adjustment parameters takes into account several factors:

[0271] Factor 1, Phase Noise Spectrum Weighting: Different phase noise offsets have different effects on mixer performance. This embodiment uses an empirical weighting function:

[0272]

[0273] in, =100kHz, used to calculate the weighted average phase noise improvement, γ is the steepness coefficient (γ=2 in this embodiment), used to calculate the weighted average phase noise improvement.

[0274] Factor 2, Temperature Compensation: Considering the impact of temperature on mixer characteristics, a secondary compensation is applied to the calculation results:

[0275]

[0276] Where β is the temperature coefficient (in this embodiment) ).

[0277] Factor 3, Safety Boundary: To prevent over-adjustment from causing system instability, an upper limit is set for the adjustment amount. .

[0278] Sf applies the operating bias voltage adjustment parameter to the mixer nonlinear unit.

[0279] The mixer nonlinearity unit refers to the key component in the mixer that generates nonlinear effects; in this embodiment, it is a Schottky diode. The nonlinear characteristics of the mixer are mainly determined by this unit, and the nonlinear characteristics of the mixer can be changed by adjusting its operating point.

[0280] Operating point adjustment refers to adjusting the bias voltage to ensure the mixer's nonlinear unit operates at its optimal state. In this embodiment, the bias voltage adjustment range is -0.5V to 0V, with a resolution of 1mV.

[0281] For example, when calculated When the bias voltage is 0.08V, it will not immediately adjust the bias voltage from -0.25V to -0.17V, but will do so in 5 steps, adjusting 0.016V in each step, with an interval of 20ms.

[0282] Specifically, the key technical details of the application process include:

[0283] First, voltage conversion: converting the digitally calculated adjustment parameters... The conversion formula for converting to an analog voltage signal is as follows:

[0284]

[0285] in, The reference voltage is -0.25V in this embodiment.

[0286] Secondly, gradual adjustment: To avoid system instability caused by sudden changes, a gradual adjustment strategy is adopted.

[0287]

[0288] in, This is the step factor (η=0.2 in this embodiment). The target is the optimal bias voltage.

[0289] Then, feedback correction: the system monitors the quality of the mixer output signal and fine-tunes the bias voltage based on the actual effect.

[0290]

[0291] Wherein, γ is the feedback gain (in this embodiment, γ = 0.001V / dB). The measured power of the non-target frequency component. The target power.

[0292] For example, when calculated When the bias voltage is 0.08V, it will not immediately adjust from -0.25V to -0.17V, but will be done in 5 steps, adjusting 0.016V each time, with an interval of 20ms. At the same time, the power of the 67GHz non-target frequency component in the mixer output is monitored. If the suppression effect is found to be unsatisfactory (e.g., only reducing by 6dB instead of the expected 8dB), an additional adjustment of 0.005V is made through feedback correction.

[0293] It should be understood that although the steps in the flowcharts of the above embodiments are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the above embodiments may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages of other steps.

[0294] Based on the same inventive concept, this application also provides a transmit link frequency conversion device for a 92-94GHz millimeter-wave radar system for implementing the aforementioned transmit link frequency conversion method for a 92-94GHz millimeter-wave radar system. The solution provided by this device is similar to the implementation described in the above method. Therefore, the specific limitations of one or more embodiments of the transmit link frequency conversion device for a 92-94GHz millimeter-wave radar system provided below can be found in the above-described limitations of the transmit link frequency conversion method for a 92-94GHz millimeter-wave radar system, and will not be repeated here.

[0295] The modules in the aforementioned transmit link frequency conversion device for 92-94GHz millimeter-wave radar systems can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in hardware within or independently of the processor in a computer device, or stored in software within the memory of a computer device, so that the processor can call and execute the corresponding operations of each module.

[0296] In one exemplary embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described transmit link frequency conversion method for a 92-94 GHz millimeter-wave radar system.

[0297] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a processor, implements the above-described transmit link frequency conversion method for a 92-94 GHz millimeter-wave radar system.

[0298] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the above-described transmit link frequency conversion method for a 92-94 GHz millimeter-wave radar system.

[0299] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0300] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0301] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A transmit link frequency conversion method for 92-94 GHz millimeter wave radar systems, characterized by, The method is executed by a signal processing unit, and the method comprises: acquiring a 79-81 GHz input radio frequency signal and a 13 GHz local oscillator signal; inputting the 79-81 GHz input radio frequency signal into a first input port of a mixer and inputting the 13 GHz local oscillator signal into a second input port of the mixer, performing mixing processing through the mixer to generate a mixing output signal containing a 92-94 GHz target frequency component and a non-target frequency component; wherein the non-target frequency component is a frequency component generated in the mixing process other than the 92-94 GHz target frequency component; sampling the mixing output signal to obtain a sample data sequence; based on the sample data sequence, determining frequency parameters, amplitude parameters and phase parameters of the non-target frequency component; generating a compensation signal according to the frequency parameters, amplitude parameters and phase parameters of the non-target frequency component; wherein the compensation signal has a characteristic of equal amplitude and opposite phase to the non-target frequency component at the frequency of the non-target frequency component; synthesizing the compensation signal and the mixing output signal to obtain a 92-94 GHz output signal.

2. The transmit chain frequency conversion method of claim 1, wherein, After the mixing processing and before the generation of the compensation signal, the method further comprises: continuously monitoring the spectral characteristics of the mixing output signal; real-time analyzing the spectral characteristics and the environmental working parameters of the mixer to dynamically identify the instantaneous nonlinear transfer function of the nonlinear unit of the mixer that causes the non-target frequency component, to obtain an identification result; based on the identification result, determining an optimization strategy for adjusting a plurality of controllable physical parameters of the nonlinear unit of the mixer; wherein the plurality of controllable physical parameters include but are not limited to multi-point bias voltage, adjustable element parameters of the active matching network, internal feedback loop gain and local oscillator injection locking parameters; according to the optimization strategy, dynamically adjusting the controllable physical parameters of the nonlinear unit of the mixer to reconstruct the instantaneous nonlinear transfer function of the mixer; wherein the reconstructed instantaneous nonlinear transfer function has the characteristics of suppressing the generation of the non-target frequency component or directing the energy of the non-target frequency component to a non-sensitive frequency band other than the 92-94 GHz target frequency component in the mixing process.

3. The transmit chain frequency conversion method of claim 1, wherein, Synthesizing the compensation signal and the mixing output signal comprises: inputting the compensation signal into a signal transmission path after an output port of the mixer; based on the frequency parameters, amplitude parameters and phase parameters of the non-target frequency component, determining the coupling coefficient and phase delay parameter in the signal transmission path to achieve amplitude matching and phase inversion between the compensation signal and the non-target frequency component, and completing signal synthesis.

4. The transmit chain frequency conversion method of claim 3, wherein, Based on the frequency parameters, amplitude parameters and phase parameters of the non-target frequency component, determining the coupling coefficient and phase delay parameter in the signal transmission path comprises: calculating the coupling coefficient in the signal transmission path according to the amplitude parameter of the non-target frequency component; According to the phase parameter of the non-target frequency component, a phase delay parameter in the signal transmission path is calculated.

5. The transmit chain frequency conversion method of claim 4, wherein, According to the amplitude parameter of the non-target frequency component, a coupling coefficient in the signal transmission path is calculated, including: A small constant preventing zero operation is set, and a sum of the amplitude parameter of the non-target frequency component and the small constant is calculated; A first proportional constant is determined, and the first proportional constant is divided by the sum of the amplitude parameter and the small constant to obtain the coupling coefficient in the signal transmission path.

6. The transmit chain frequency conversion method of claim 4, wherein, According to the phase parameter of the non-target frequency component, a phase delay parameter in the signal transmission path is calculated, including: The difference between π and the phase parameter of the non-target frequency component is calculated; A second proportional constant is determined, and the second proportional constant is multiplied by the difference to obtain the phase delay parameter in the signal transmission path.

7. The transmit chain frequency conversion method of claim 1, wherein, Based on the sampling data sequence, a frequency parameter of the non-target frequency component is determined, including: The sampling data sequence is subjected to a discrete Fourier transform to obtain frequency domain signal data; From the frequency domain signal data, a discrete frequency point with a power value greater than a preset noise threshold is identified as the frequency parameter of the non-target frequency component; wherein the preset noise threshold is dynamically determined according to a system noise floor.

8. The transmit chain frequency conversion method of claim 7, wherein, Based on the sampling data sequence, an amplitude parameter and a phase parameter of the non-target frequency component are calculated, including: Based on a complex value of the discrete frequency point in the frequency domain signal data, the amplitude parameter and the phase parameter of the non-target frequency component are calculated.

9. The transmit chain frequency conversion method of claim 2, wherein, Before the 13GHz local oscillator signal is input to the second input port of the mixer, the method further includes: Obtaining spectral characteristic data of the 13GHz local oscillator signal; According to the difference between the spectral characteristic data and a preset ideal spectrum, a working bias voltage adjustment parameter of a mixer non-linear unit in the mixer is calculated; The working bias voltage adjustment parameter is applied to the mixer non-linear unit.

10. The transmit chain frequency conversion method of claim 9, wherein, According to the difference between the spectral characteristic data and a preset ideal spectrum, a working bias voltage adjustment parameter of a mixer non-linear unit in the mixer is calculated, including: Obtaining a phase noise spectrum of the 13GHz local oscillator signal; Based on the phase noise spectrum, a Wiener filtering algorithm is used to calculate a filter coefficient of a local oscillator signal purification filter; The 13GHz local oscillator signal passes through the local oscillator signal purification filter to obtain a local oscillator signal after phase noise suppression; According to the signal characteristics of the local oscillator signal after phase noise suppression, the working bias voltage adjustment parameter of the mixer non-linear unit is calculated.

Images