Signal processing method, storage medium, integrated circuit, device and terminal equipment

Abstract

The embodiment of the invention relates to the technical field of signal processing, and discloses a signal processing method, a storage medium, an integrated circuit, a device and terminal equipment, and the method comprises the steps: carrying out the discrete spectrum analysis of a to-be-processed signal based on an iterative compressed sensing technology, and determining a target frequency component; amplifying the target frequency component, and judging whether the current iteration is converged or not based on the amplified target frequency component; if the current iteration converges, performing frequency domain inverse transformation on the amplified target frequency component to obtain a target signal; and based on the target signal, performing signal recovery, velocity ambiguity resolution and / or direction of arrival estimation on the to-be-processed signal. According to the signal processing method provided by the embodiment of the invention, the number of iterations can be reduced on the premise that the cost is not increased, the convergence speed of the algorithm is effectively improved, the balance between the hardware cost and the calculation performance is realized, and meanwhile, the method is suitable for the signal with a very large missing sampling point proportion and has very high universality.

Description

Technical Field

[0001] The embodiments of the present application relate to the technical field of signal processing, and in particular, to a signal processing method, a storage medium, an integrated circuit, a device, and a terminal device. Background Art

[0002] Performing spectrum analysis on a tuned signal with incomplete sampling is a problem that many technologies and engineering need to face. For example, when performing direction of arrival (DOA) estimation on a sparse antenna array, a specific coefficient array signal can be regarded as an incomplete sampling of a uniformly dense array signal; for a signal sequence affected by interference, the non-affected part can be regarded as an incomplete sampling of the true signal; for non-periodic time division multiplexing (TDM) radar, velocity ambiguity resolution, etc.

[0003] The iterative method with adaptive threshold (IMAT) is a method to solve the above problems. This algorithm assumes that the signal has the characteristic of frequency domain sparsity. After transforming the time domain signal to the frequency domain, the frequency components higher than a certain threshold are retained, and other components are removed. Then, the processed frequency domain signal is inversely transformed to the time domain, and the obtained result is filled into the part with incomplete sampling. After that, the filled time domain signal is transformed to the frequency domain again, and this cycle of iteration continues until convergence.

[0004] However, the convergence speed of the IMAT algorithm is related to the proportion of missing sampling points in the signal. The larger the proportion of missing sampling points in the signal, the slower the convergence speed of the IMAT algorithm, and the more iterations are required. Limited hardware computing power is difficult to support too many iterations, which requires increasing costs to use hardware with stronger computing power. At the same time, for signals with a very large proportion of missing sampling points, the performance of the IMAT algorithm is poor. Summary of the Invention

[0005] The purpose of the embodiments of the present application is to provide a signal processing method, a storage medium, an integrated circuit, a device, and a terminal device, which can reduce the number of iterations without increasing costs, effectively improve the convergence speed of the algorithm, achieve a balance between hardware costs and computing performance, and at the same time be applicable to signals with a very large proportion of missing sampling points, and has strong universality.

[0006] To solve the above technical problems, an embodiment of the present application provides a signal processing method, which can be applied to the process of processing echo signals received by electromagnetic wave sensors such as millimeter waves and lasers. When the echo signal contains a part affected by co-frequency and / or adjacent-frequency interference, for example, after the echo signal received by the receiving channel is subjected to analog-to-digital conversion, various methods such as sorting, taking the median, and averaging can be used to determine whether the currently received echo signal (digital signal) is interfered, and the time period of the interfered part in the time dimension, etc. That is, after the position information of the time-domain echo signal that is interfered is detected and confirmed based on the interference detection, the following operations can be performed: Step S1, preprocess the interfered signal part in the echo signal as the signal to be processed; Step S2, perform spectral analysis on the signal to be processed to obtain the frequency-domain data to be processed; Step S3, obtain the amplification factor of the current amplification operation based on the missing sampling ratio of the signal to be processed and the preset target convergence speed, and use the amplification factor of the current amplification operation to amplify the frequency-domain data to be processed; Step S4, obtain the threshold of the current filtering operation based on the initial threshold and the attenuation factor, and use the threshold of the current filtering operation to filter the amplified frequency-domain data to be processed to obtain the filtered frequency-domain data; Step S5, determine whether the current iteration operation meets the preset requirements; if it does not meet the preset requirements, continue to Step S6, otherwise continue to Step S7; Step S6, inverse-transform the filtered frequency-domain data to the time domain to obtain a restored signal, and use the restored signal to replace the preprocessed part in the echo signal, and use the replaced echo signal as the signal to be processed to continue Steps S2 to S5; Step S7, inverse-transform the filtered frequency-domain data to the time domain to obtain a restored signal, and use the restored signal to replace the preprocessed part in the echo signal, and obtain the target information based on the replaced echo signal. In this embodiment, based on the technical idea of compressive sensing and sparse recovery, by transforming the interfered signal from the signal domain (i.e., the time domain) to the sparse domain (i.e., the frequency domain), amplifying it, and then filtering it, the interfered signal can be effectively and quickly denoised. At the same time, based on the continuous alternation of the signal domain (i.e., the time domain) transformation and the sparse domain (i.e., the frequency domain) amplification, combined with the convergence of each iteration and the adjustment of the threshold, it is further possible to achieve, for example, co-frequency and / or adjacent-frequency interference in the signal.

[0007] Based on the above technical idea, an embodiment of the present application further provides a signal processing method, which can be applied to signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation of a signal to be processed, including the following steps: performing discrete spectrum analysis on the signal to be processed based on iterative compressive sensing technology to determine the target frequency component; amplifying the target frequency component, and determining whether the current iteration converges based on the amplified target frequency component; if the current iteration converges, performing inverse frequency domain transformation on the amplified target frequency component to obtain the target signal; and performing signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation on the signal to be processed based on the target signal.

[0008] An embodiment of the present application also provides a computer-readable storage medium storing a computer program, which when executed by a processor can implement the signal processing method in the embodiment of the present application.

[0009] An embodiment of the present application also provides an integrated circuit, which may include: a signal transceiver channel for transmitting radio signals and receiving echo signals formed by reflection of the radio signals by a target; a signal processing module for performing signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation based on the above signal processing method. The signal processing module may include: a sampling unit for sampling the echo signal to obtain a discrete signal to be processed; a frequency domain conversion unit for performing frequency domain conversion on the signal to be processed to obtain a signal to be processed in the frequency domain; a spectrum analysis unit for performing discrete spectrum analysis on the signal to be processed in the frequency domain based on iterative compressive sensing technology to determine the target frequency component; an amplification unit for amplifying the target frequency component; a judgment unit for determining whether the current iteration converges based on the amplified target frequency component, and performing inverse frequency domain transformation on the amplified target frequency component to obtain the target signal when the current iteration converges; and an execution unit for performing signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation on the signal to be processed based on the target signal.

[0010] It should be noted that for the sampling unit, the frequency domain conversion unit, the spectrum analysis unit, the amplification unit, the judgment unit, and the execution unit, some or all of them can be shared to improve the integration degree, or the hardware structure can be set separately according to requirements to improve the overall signal processing efficiency.

[0011] An embodiment of the present application also provides a radio device, including: a carrier; the integrated circuit as described above, disposed on the carrier; and an antenna disposed on the carrier for transmitting and receiving radio signals.

[0012] An embodiment of the present application further provides a terminal device, including: a device body; and a wireless electrical device disposed on the device body as described above, where the wireless electrical device is used for target detection and / or communication.

[0013] For the signal processing method, storage medium, integrated circuit, device, and terminal device provided by the embodiments of the present application, first, discrete spectrum analysis is performed on the signal to be processed based on the iterative compressive sensing technology to obtain target frequency components, and then the target frequency components are amplified. Subsequently, it is determined whether the current iteration converges based on the amplified target frequency components. When the current iteration converges, inverse Fourier transform is performed on the amplified target frequency components to obtain a target signal. Finally, based on the target signal, signal recovery, velocity ambiguity resolution, and / or direction-of-arrival estimation are performed on the signal to be processed. Considering that the conventional iterative compressive sensing technology has a relatively slow convergence speed and requires a large number of iterations, and it is difficult for hardware with limited computing power to support too many iterations, therefore, in the process of iterative compression in the embodiments of the present application, the target frequency components are reasonably amplified, and the number of iterations is reduced without increasing the cost. With fewer iterations, the same effect as that of the conventional iterative compressive sensing technology with multiple iterations can be achieved, effectively improving the convergence speed of the algorithm, achieving a balance between hardware cost and computing performance. At the same time, there are no special requirements for the signal to be processed itself, and it is also applicable to signals with a very large proportion of missing sampling points, and has strong robustness and universality.

[0014] In some alternative embodiments, the amplifying the target frequency components includes: determining an amplification factor corresponding to the current iteration according to the number of iterations of the current iteration, the missing sampling ratio of the signal to be processed, and a preset target convergence speed; and amplifying the target frequency components according to the amplification factor. According to the signal energy theory, the larger the proportion of missing sampling points in the signal to be processed, that is, the lower the missing sampling ratio, the slower the convergence speed of the algorithm and the more iterations are required. That is, the convergence speed of the algorithm is limited by the missing sampling ratio of the signal to be processed. Therefore, in the present application, a scientific and reasonable amplification factor is determined based on the number of iterations of the current iteration, the missing sampling ratio of the signal to be processed, and in combination with the target convergence speed, which can well accelerate the algorithm convergence process.

[0015] In some alternative embodiments, the determining the amplification factor corresponding to the current iteration according to the number of iterations of the current iteration, the missing sampling ratio of the signal to be processed, and a preset target convergence speed is implemented by the following formula: where n is the number of iterations of the current iteration, k is the missing sampling ratio of the signal to be processed, β is the target convergence speed, and α[n] is the amplification factor corresponding to the current iteration.

[0016] In some alternative embodiments, the value range of the target convergence rate is from 0.2 to 0.8, such as 0.2, 0.3, 0.4, 0.45, 0.5, 0.6, 0.65, 0.7 or 0.8, etc.

[0017] In some alternative embodiments, the iterative compressive sensing technology includes an adaptive threshold iterative algorithm. The discrete spectrum analysis of the frequency-domain signal to be processed based on the iterative compressive sensing technology to determine the target frequency component includes: sequentially determining whether each frequency component in the frequency-domain signal to be processed is greater than the adaptive iterative threshold, and determining the frequency component greater than the adaptive iterative threshold as the target frequency component; wherein, the adaptive iterative threshold is obtained based on the adaptive threshold iterative algorithm.

[0018] In some alternative embodiments, after determining whether the current iteration converges based on the amplified target frequency component, the method further includes: if the current iteration does not converge, updating the adaptive iterative threshold according to the iteration number of the current iteration, the initial adaptive iterative threshold, and a preset attenuation factor; sequentially determining whether each frequency component in the frequency-domain signal to be processed is greater than the updated adaptive iterative threshold, and determining the frequency component greater than the updated adaptive iterative threshold as the target frequency component. In the case of non-convergence, iteration needs to continue and the adaptive iterative threshold is updated.

[0019] In some alternative embodiments, updating the adaptive iterative threshold according to the iteration number of the current iteration, the initial adaptive iterative threshold, and a preset attenuation factor: where ε is the initial adaptive iterative threshold, is the preset attenuation factor, n is the iteration number of the current iteration, Thr n+1 is the updated adaptive iterative threshold.

[0020] In some alternative embodiments, determining whether the current iteration converges based on the amplified target frequency component includes: calculating the total spectral energy of the amplified target frequency component corresponding to the current iteration; determining whether the current iteration converges according to the total spectral energy of the amplified target frequency component corresponding to the current iteration, the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, and a preset change threshold.

[0021] In some alternative embodiments, the following formula is used to determine whether the current iteration converges according to the total spectral energy of the amplified target frequency component corresponding to the current iteration, the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, and a preset change threshold: where P nis the total spectral energy of the amplified target frequency component corresponding to the current iteration, P n-1 is the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, δ is the preset change threshold, f = 1 indicates that the current iteration converges, and f = 0 indicates that the current iteration does not converge.

[0022] In some alternative embodiments, the preset change threshold is from 0.5% to 1.5%, such as 0.5%, 0.6%, 0.7%, 0.8%, 0.9%, 1.0%, 1.1%, 1.15%, 1.2%, 1.3%, 1.4%, 1.45% or 1.5%, etc.

[0023] In some alternative embodiments, determining whether the current iteration converges based on the amplified target frequency component includes: sorting the amplified target frequency component corresponding to the current iteration according to the frequency value to obtain a target sequence; wherein, the sorting is in ascending order or descending order; using the frequency value at a preset position in the target sequence as the noise floor estimate value corresponding to the current iteration; calculating the difference between the noise floor estimate value corresponding to the current iteration and the noise floor estimate value corresponding to the previous iteration of the current iteration; if the absolute value of the difference is within a preset threshold range, determining that the current iteration converges; if the absolute value of the difference is outside the preset threshold range, determining that the current iteration does not converge.

[0024] In some alternative embodiments, when the signal processing method is applied to perform signal recovery on a signal to be processed, before performing discrete spectrum analysis on the frequency-domain signal to be processed based on the iterative compressive sensing technology, the method further includes: performing interference detection on the signal to be processed, and removing the interfered part in the signal to be processed through filtering; performing signal recovery on the signal to be processed based on the target signal includes: replacing the signal amplitude of the interfered part in the signal to be processed with the signal amplitude corresponding to the interfered part in the target signal to obtain a signal with recovery completed.

[0025] In some alternative embodiments, removing the interfered part in the signal to be processed through filtering includes: setting the signal amplitude of the interfered part in the signal to be processed to zero to obtain the filtered signal to be processed.

[0026] In some alternative embodiments, after performing signal recovery, velocity ambiguity resolution, and / or direction-of-arrival estimation on the signal to be processed, the method further includes: processing the signal after signal recovery, velocity ambiguity resolution, and / or direction-of-arrival estimation to obtain the distance, velocity, and / or angle, etc. of the target object.

[0027] In some alternative embodiments, the signal to be processed is a unit signal such as a signal frame or a chirp signal.

[0028] In some alternative embodiments, the waveform of the signal to be processed is a continuous wave whose frequency linearly changes with time.

[0029] In some alternative embodiments, the continuous wave includes at least one of frequency modulated continuous wave (FMCW), stepped frequency continuous wave (SFCW), MBC, etc.

[0030] In some alternative embodiments, the signal to be processed is obtained through the following steps: acquiring an echo signal; mixing the echo signal to obtain an intermediate frequency signal; performing analog-to-digital conversion on the intermediate frequency signal to obtain a first discrete signal, and using the first discrete signal as the signal to be processed.

[0031] In some alternative embodiments, using the first discrete signal as the signal to be processed includes: performing digital signal processing on the first discrete signal at least once to obtain a second discrete signal; using the second discrete signal as the signal to be processed. Description of the Drawings

[0032] One or more embodiments are exemplarily illustrated by pictures in the corresponding drawings, and these exemplary illustrations do not limit the embodiments.

[0033] Figure 1 are schematic diagrams of several incomplete sampling signals;

[0034] Figure 2 is a schematic diagram of the principle of the IMAT algorithm;

[0035] Figure 3 is the flow of the signal processing method provided in an embodiment of the present application Figure 1 ;

[0036] Figure 4 is the module schematic diagram of the signal processing method provided in an embodiment of the present application;

[0037] Figure 5 is Figure 4 the module schematic diagram of the amplification module and the adaptive iteration loop in

[0038] Figure 6 is the flow of a signal processing method provided in an embodiment of the present application Figure 2 ;

[0039] Figure 7 is the flowchart of amplifying the target frequency component and determining whether the current iteration converges based on the amplified target frequency component in an embodiment of the present application;

[0040] Figure 8 In an embodiment of the present application, it is an iterative schematic diagram of signal processing based on the conventional iterative compressive sensing technology;

[0041] Figure 9 In an embodiment of the present application, it is an iterative schematic diagram of signal processing based on the iterative compressive sensing technology with an amplification process set;

[0042] Figure 10 In an embodiment of the present application, it is a flowchart for judging whether the current iteration converges according to the spectrum energy convergence condition;

[0043] Figure 11 In an embodiment of the present application, it is a flowchart for judging whether the current iteration converges according to the noise floor convergence condition;

[0044] Figure 12 In another embodiment of the present application, it is a schematic diagram of an integrated circuit provided;

[0045] Figure 13 In an embodiment of the present application, it is a schematic diagram of a signal processing module provided. Specific embodiments

[0046] To make the objectives, technical solutions, and advantages of the embodiments of the present application clearer, the embodiments of the present application will be elaborated in detail below with reference to the accompanying drawings. However, those of ordinary skill in the art can understand that in the embodiments of the present application, many technical details are provided to help readers better understand the present application. However, even without these technical details and various changes and modifications based on the following embodiments, the technical solutions claimed in the present application can still be implemented. The following division of each embodiment is for convenience of description and should not constitute any limitation on the specific implementation manner of the present application. Each embodiment can be combined and cross-referenced with each other on the premise of no contradiction.

[0047] Performing spectrum analysis on signals with incomplete sampling is a technical problem commonly faced in many actual scenarios. For example, when performing direction of arrival (DOA) estimation on a coefficient antenna array, a specific sparse array signal can be regarded as an incomplete sampling of a uniform dense array signal. Another example is when performing signal recovery on an interfered signal, the non-interfered part of the signal can be regarded as an incomplete sampling of the true signal. Figure 1 Several signals with incomplete sampling are shown, and it can be seen that Figure 1 the signals in all have missing parts.

[0048] The IMAT algorithm is a technology that can solve the above problems. This algorithm assumes that the signal has the characteristic of sparsity in the frequency domain. After transforming the time-domain signal to the frequency domain, a small number of frequency components higher than a certain threshold are retained (such as Figure 2 as shown), other components are removed, and then it is inverse-transformed to the time domain. The obtained result is used to fill the incomplete part of the sampling. Then, the frequency-domain transformation is performed on the time-domain signal after filling. Such iteration is carried out until the loop iteration converges. Among them, the frequency-domain threshold gradually decreases as the number of iterations increases.

[0049] However, the convergence speed of the IMAT algorithm is related to the proportion of missing sampling points in the signal. The larger the proportion of missing sampling points in the signal, the slower the convergence speed of the IMAT algorithm, and the more iterations are required. Limited hardware computing power is difficult to support too many iterations, which requires increasing the cost to use more powerful hardware. If the number of iterations required by the IMAT algorithm can be reduced and the convergence process of the IMAT algorithm can be accelerated, the applicability of the IMAT algorithm can be greatly expanded.

[0050] To solve the above technical problems such as the slow convergence speed of the IMAT algorithm and the large number of iterations required, an embodiment of the present application proposes a signal processing method for signal restoration, velocity ambiguity resolution, and / or direction-of-arrival estimation of a signal to be processed. Specifically, it can be applied to a terminal or a processor. This embodiment and each of the following embodiments are described by taking the processor as an example. The implementation details of the signal processing method of this embodiment are specifically described below. The following content is only the implementation details provided for convenient understanding and is not necessary for implementing this solution.

[0051] The embodiment of the present application provides a signal processing method, which can be applied to the process of processing echo signals received by electromagnetic wave sensors such as millimeter waves and lasers. When there are two or more identical or similar transmitting signal sensors in the same application scenario, there is likely to be co-frequency and / or adjacent-frequency interference between the electromagnetic wave signals transmitted by the sensors, which seriously affects the performance of sensor target detection. At this time, for the echo signal received by the current electromagnetic wave sensor, interference detection can be first performed on the echo signal. It can be applied to the process of processing echo signals received by electromagnetic wave sensors such as millimeter waves and lasers. When the echo signal contains a part affected by co-frequency and / or adjacent-frequency interference, for example, after analog-to-digital conversion of the echo signal received by the receiving channel, various methods such as sorting, taking the median, and averaging can be used to determine whether the currently received echo signal (digital signal) is interfered, and the time period of the interfered part in the time dimension, etc. That is, based on the position information of the time-domain echo signal that is detected to be interfered after interference detection, the following operations can be performed as Figure 3 shown:

[0052] Step S1, after preprocessing the power value of the interference signal part in the echo signal, such as zeroing or setting it to a preset value, the preprocessed echo signal is used as the signal to be processed, that is, the interference signal is preprocessed in the time domain first. Among them, the above preprocessing operation is an operation on a part of the echo signal, that is, in a processing unit of the received echo signal, only a part of the signal in this processing unit is interfered. For example, when the sensor is an FMCW sensor and the processing unit of the received echo signal is a chirp, the interference signal part is a part of the frequency points within a chirp, that is, preprocessing operations are performed on all the interfered frequency points within the current chirp (taking the frequency point as the smallest preprocessing unit), such as the 1D-FFT (such as distance dimension FFT) interference recovery implemented for the burst interference within a pulse; similarly, if the processing unit of the received echo signal is a frame signal, the interference signal part is a part of the chirps in this frame signal, that is, preprocessing operations are performed on all the interfered chirps within the current frame (taking the chirp as the smallest preprocessing unit), such as the 2D-FFT (such as Doppler dimension FFT) interference recovery implemented for the burst interference within a frame; and when the processing unit of the received echo signal contains multiple frame signals, for example, a group contains 12 frame signals, the interference signal part is a part of the frame signals in this group, that is, preprocessing operations are performed on all the interfered frame signals within the current group (taking the frame signal as the smallest preprocessing unit).

[0053] Step S2, perform spectrum analysis operations such as Fourier transform on the signal to be processed to convert the signal to be processed in the time domain to the frequency domain and obtain the data of the signal to be processed in the frequency domain.

[0054] Step S3, based on the missing sampling ratio of the signal to be processed (that is, the proportion of the missing part in the entire signal to be processed) and a preset target convergence speed (which can be specifically set according to big data analysis or empirical values combined with the scenario), obtain the amplification factor of the current amplification operation, and use the amplification factor of the current amplification operation to amplify the data of the signal to be processed in the frequency domain in Step S2 to obtain the amplified data of the signal to be processed in the frequency domain. The amplification operation can effectively improve the convergence speed.

[0055] Step S4, after obtaining the threshold of the current filtering operation based on the initial threshold (which can be specifically set according to big data analysis or empirical values combined with the scenario) and the attenuation factor, etc., use the threshold of the current filtering operation to filter the amplified data of the signal to be processed in the frequency domain in Step S3 to obtain the filtered data in the frequency domain, that is, complete the filtering operation of the interference signal and the preprocessed signal in the frequency domain.

[0056] Step S5: Determine whether the current iterative operation meets the preset requirements. If it does not meet the preset requirements, proceed to step S6; otherwise, proceed to step S7. Specifically, it can be determined whether the current iterative operation meets the preset requirements based on whether at least one of the energy change and noise floor change between two adjacent iterative operations meets the preset convergence condition, and / or whether the number of times of the current iterative operation reaches the maximum number of iterative times. As long as one of the above conditions of meeting the preset convergence condition and meeting the maximum number of iterative times is satisfied, the iterative operation can be stopped. For example, it can first be determined whether the maximum number of iterative times has been reached, and then whether the preset convergence condition is met. That is, if the maximum number of iterative times has been reached, step S7 can be directly performed; if the maximum number of iterative times has not been reached but the preset convergence condition is met, step S7 can also be directly performed; if the maximum number of iterative times has not been reached and the preset convergence condition is not met, step S6 is performed.

[0057] Step S6: Inverse-transform the filtered frequency-domain data to the time domain to obtain a restored signal, and use the restored signal to replace the preprocessed part in the echo signal. The replaced echo signal is used as the signal to be processed, and steps S2 to S5 are continued, that is, cyclic iteration between the signal domain (i.e., the time domain) and the sparse domain (i.e., the frequency domain) is realized. While effectively removing the interference signal, the restoration operation of the interfered part of the signal is realized.

[0058] Step S7: Inverse-transform the filtered frequency-domain data to the time domain to obtain a restored signal, and use the restored signal to replace the preprocessed part in the echo signal. Based on the replaced echo signal, information such as the distance, speed, azimuth angle, elevation angle, and / or micro-Doppler motion characteristics of the target relative to the sensor is obtained.

[0059] In this embodiment, based on the technical idea of compressive sensing and sparse recovery, by transforming the interfered signal from the signal domain (i.e., the time domain) to the sparse domain (i.e., the frequency domain), amplifying and filtering it, the interfered signal can be effectively and quickly denoised. That is, in a scenario where the requirement for interference recovery accuracy is relatively low, steps S5 and S6 can be omitted, and step S7 can be directly performed after step S4 to realize the suppression of the interference signal and the recovery operation of the interfered signal (that is, the signal processing method at this time includes steps S1 to S4 and step S7). At the same time, based on the alternating iteration between the signal domain (i.e., the time domain) transformation and the sparse domain (i.e., the frequency domain) in steps S5 and S6, combined with the convergence of each iteration and the adjustment of the threshold, it is further possible to highly suppress interference signals such as co-frequency and / or adjacent-frequency in the signal and achieve high-quality recovery of the interfered signal.

[0060] Figure 4 It is a schematic diagram of the module of the signal processing method provided in an embodiment of the present application.Figure 5 Yes Figure 4 It is a schematic diagram of the amplification module and the adaptive iterative loop in the middle.

[0061] Based on the above basic idea, see Figure 2 and Figure 3 As shown, in the signal processing of the FMCW millimeter-wave radar, after the echo signal is first converted into a digital signal through ADC, and after the interference detection of the digital signal, preprocessing operations such as setting the interference signal to zero are performed to suppress and weaken (truncate) the interference signal. Then, operations such as FFT are continued to perform spectral analysis on the preprocessed digital signal. Subsequently, the analyzed frequency components are amplified to facilitate subsequent threshold filtering (threshold judgment) of the amplified frequency components in the frequency domain; after convergence analysis (convergence judgment) based on the filtered frequency-domain signal, inverse transformation operations such as IFFT are performed to convert the filtered frequency-domain signal into a time-domain signal, and at the same time, the time-domain signal is used to replace the signal after the above suppression and weakening. At this time, if the result of the convergence analysis is convergence, operations such as range-dimension FFT, velocity-dimension FFT, constant false alarm detection, velocity ambiguity resolution, and direction-of-arrival estimation can be performed based on the replaced time-domain signal to obtain parameters such as the distance, velocity, azimuth angle, elevation angle, and even micro-Doppler motion characteristics of the target; if the result of the convergence analysis is non-convergence, spectral analysis operations such as FFT can be continued on the replaced time-domain signal, that is, continue the next time-domain-frequency-domain-time-domain iterative loop, but the threshold of the threshold filtering in the frequency domain can be adaptively adjusted each time the iterative loop is performed, so that the finally output time-domain signal meets the interference suppression requirement and returns to the quality requirement without interference, while the number of iterative loops or the computational resource consumption is also within the expected range.

[0062] Based on the above basic idea, this embodiment further elaborates on the signal processing method, that is, the above signal processing method can also be applied to the signal processing of communication or sensing such as sparse arrays and transmit antenna permutations (Tx permutation). For the specific process, it can be as Figure 6 shown, including:

[0063] Step 101, perform discrete spectral analysis on the signal to be processed based on iterative compressive sensing technology to determine the target frequency components.

[0064] In a specific implementation, after the processor obtains the signal to be processed, it can perform discrete spectral analysis on the signal to be processed based on iterative compressive sensing technology to determine the target frequency components. It can be understood that the signal to be processed is a time-domain signal. When the processor performs discrete spectral analysis on the signal to be processed, it first performs a frequency-domain transformation on the signal to be processed to obtain the signal to be processed in the frequency domain, and then performs discrete spectral analysis on the transformed signal to be processed in the frequency domain based on iterative compressive sensing technology to determine the target frequency components.

[0065] In some examples, the frequency-domain transformation of the signal to be processed may be a Fast Fourier Transform (FFT), a Z-transform, a Laplace transform, etc.

[0066] In some examples, the signal to be processed by the processor may be a signal frame or a chirp signal, and the processor may also select multiple consecutive chirp signals as the signal to be processed.

[0067] In some examples, the waveform of the signal to be processed by the processor is a continuous wave whose frequency changes linearly with time, and the continuous wave includes at least one of FMCW and SFCW.

[0068] In some examples, the iterative compressive sensing technology may select the IMAT algorithm, or the Iterative Hard Thresholding (IHT) algorithm, etc.

[0069] In some examples, the processor can obtain the signal to be processed through the following steps: obtain the echo signal; mix the obtained echo signal to obtain an intermediate-frequency signal; perform analog-to-digital conversion, sampling, etc. on the mixed intermediate-frequency signal to obtain a first discrete signal, and use this first discrete signal as the signal to be processed. That is, in this embodiment, the first discrete signal obtained after analog-to-digital conversion can be directly used as the signal to be processed, that is, signal processing is performed at the ADC threshold, which has less impact on subsequent applications of the radar.

[0070] In some other examples, the processor can obtain the signal to be processed through the following steps: obtain the echo signal; mix the obtained echo signal to obtain an intermediate-frequency signal; perform analog-to-digital conversion on the mixed intermediate-frequency signal to obtain a first discrete signal, perform at least one digital signal processing on this first discrete signal to obtain a second discrete signal; use this second discrete signal as the signal to be processed. That is, in this embodiment, the second discrete signal obtained after digital signal processing can be used as the signal to be processed, that is, signal processing is performed at the DSP threshold.

[0071] In some examples, the processor performs discrete spectrum analysis on the signal to be processed (the frequency-domain transformation is the signal to be processed in the frequency domain) based on the IMAT algorithm. The processor sequentially determines whether each frequency component in the signal to be processed in the frequency domain is greater than the adaptive iteration threshold of the IMAT algorithm, and determines the frequency components greater than the adaptive iteration threshold as the target frequency components. Among them, the adaptive iteration threshold is obtained based on the IMAT algorithm itself.

[0072] Step 102, amplify the target frequency components and determine whether the current iteration converges based on the amplified target frequency components.

[0073] In a specific implementation, the processor performs discrete spectrum analysis on the signal to be processed based on the iterative compressive sensing technology. After determining the target frequency component, the target frequency component can be multiplied by the obtained amplification factor to amplify the target frequency component, and it is determined whether the current iteration converges based on the amplified target frequency component. The design of amplifying the target frequency component can reduce the number of iterations without increasing the cost. With fewer iterations, the same effect as that of the conventional iterative compressive sensing technology with multiple iterations can be achieved, effectively improving the convergence speed of the algorithm.

[0074] In some examples, the processor can multiply the target frequency component by a preset amplification factor to amplify the target frequency component. Among them, the preset amplification factor can be set by those skilled in the art according to actual needs, and the embodiments of the present application do not make specific limitations on this.

[0075] In some examples, the processor amplifies the target frequency component and determines whether the current iteration converges based on the amplified target frequency component, which can be implemented through the following Figure 7 sub-steps as shown, specifically including:

[0076] Sub-step 1021, determining the amplification factor corresponding to the current iteration according to the number of iterations of the current iteration, the missing sampling ratio of the signal to be processed, and the preset target convergence speed.

[0077] In a specific implementation, the adaptive iteration threshold of the IMAT algorithm changes with the increase in the number of iterations. If the same preset amplification factor is always used to amplify the target frequency component after each iteration, it is difficult to well control the amplification scale of the target frequency component, which may affect the convergence of the algorithm. Therefore, the amplification factor after each iteration by the processor should also be different. The processor can determine the amplification factor corresponding to the current iteration according to the number of iterations of the current iteration, the missing sampling ratio of the signal to be processed, and the preset target convergence speed. The preset target convergence speed can be set by those skilled in the art according to actual needs. Generally, the value range is set to 0.2 to 0.8. For example, the target convergence speed can be set to 0.5. According to the signal energy theory, the larger the proportion of missing sampling points in the signal to be processed, that is, the larger the missing sampling ratio, the slower the convergence speed of the algorithm, and the more iterations are required. That is, the convergence speed of the algorithm is limited by the missing sampling ratio of the signal to be processed. Therefore, the processor determines a scientific and reasonable amplification factor based on the number of iterations of the current iteration, the missing sampling ratio of the signal to be processed, and in combination with the target convergence speed, which can well accelerate the algorithm convergence process.

[0078] It should be noted that the missing sampling ratio is the proportion of the incomplete part in the signal to be processed in the entire signal. The corresponding one is the effective sampling ratio (i.e., the proportion of the complete part in the entire signal). The missing sampling ratio can reflect the completeness of the signal to be processed. The preset target convergence speed is actually the convergence speed estimated by the technical personnel, and the amplification process is based on this.

[0079] In a specific implementation, let the missing sampling ratio of the signal to be processed be k, and the energy of the complete signal to be processed be S. Before starting the iteration, the signal energy can be regarded as: E = (1 - k)S. After n iterations of filling by the IMAT algorithm, the signal energy of the signal to be processed can be regarded as: When n approaches infinity, E[n] approaches S, that is, all the energy is restored. It can be seen that the convergence speed in the formula is limited by the missing sampling ratio k. Amplifying the target frequency component can improve the convergence performance of the IMAT algorithm and does not destroy its convergence focus. Let the preset target convergence speed be β, that is, it is required that E[n] = S(1 - β n+1 ). Based on this, the processor determines the amplification coefficient corresponding to the current iteration according to the iteration number of the current iteration, the missing sampling ratio of the signal to be processed, and the preset target convergence speed, which can be achieved through the following formula: In the formula, n is the iteration number of the current iteration, k is the missing sampling ratio of the signal to be processed, β is the target convergence speed, and α[n] is the amplification coefficient corresponding to the current iteration.

[0080] Sub-step 1022, amplify the target frequency component according to the amplification coefficient, and judge whether the current iteration converges based on the amplified target frequency component.

[0081] In a specific implementation, the processor determines the amplification coefficient corresponding to the current iteration, can amplify the target frequency component according to the amplification coefficient, and judge whether the current iteration converges based on the amplified target frequency component, that is, judge whether the amplified target frequency component meets the preset convergence condition. If the amplified target frequency component meets the preset convergence condition, the processor can determine that the current iteration converges.

[0082] Step 103, if the current iteration converges, perform an inverse Fourier transform on the amplified target frequency component to obtain the target signal.

[0083] Step 104, based on the target signal, perform signal restoration, velocity ambiguity resolution, and / or direction-of-arrival estimation on the signal to be processed.

[0084] In a specific implementation, if the processor determines that the current iteration converges based on the amplified target frequency component, it can perform an inverse frequency domain transform on the amplified target frequency component to obtain a target signal. The target signal is a time domain signal, and the processor can perform signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation on the signal to be processed based on the target signal.

[0085] In some examples, the processor can perform an Inverse Fast Fourier Transform (IFFT) on the amplified target frequency component to obtain the target signal.

[0086] In some examples, the processor performs signal processing based on the IMAT algorithm. If the processor determines that the current iteration does not converge based on the amplified target frequency component, it can update the adaptive iteration threshold according to the iteration number of the current iteration, the initial adaptive iteration threshold, and a preset attenuation factor, that is, enter the next iteration. After entering the new iteration, the processor sequentially determines whether each frequency component in the frequency domain signal to be processed in the new iteration is greater than the updated adaptive iteration threshold, determines the frequency components greater than the updated adaptive iteration threshold as the new target frequency components, and then amplifies the new target frequency components according to the amplification factor corresponding to the new iteration. The processor then determines whether the new iteration converges based on the amplified new target frequency components.

[0087] In some examples, the processor updates the adaptive iteration threshold according to the following formula based on the iteration number of the current iteration, the initial adaptive iteration threshold, and a preset attenuation factor: In this formula, ε is the initial adaptive iteration threshold, is the preset attenuation factor, n is the iteration number of the current iteration, Thr n+1 is the updated adaptive iteration threshold. The value range of is 0.5 - 10, and ε can be selected according to the frequency value in the frequency domain signal to be processed, for example, taking the spectral value of the first FFT.

[0088] In some examples, after the processor performs signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation on the signal to be processed, it can also process the signal after signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation to obtain the distance and / or velocity of the target object.

[0089] In this embodiment, the processor first performs discrete spectrum analysis on the signal to be processed based on the iterative compressive sensing technology to obtain the target frequency components. Subsequently, the target frequency components are amplified, and based on the amplified target frequency components, it is determined whether the current iteration converges. In the case where the current iteration converges, the inverse Fourier transform is performed on the amplified target frequency components to obtain the target signal. Finally, based on the target signal, signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation are performed on the signal to be processed. Considering that the conventional iterative compressive sensing technology has a relatively slow convergence speed and requires a large number of iterations, and it is difficult for hardware with limited computing power to support a large number of iterations, the embodiments of the present application reasonably amplify the target frequency components during the iterative compression process, reduce the number of iterations without increasing costs, and can achieve the same effect as multiple iterations of the conventional iterative compressive sensing technology with fewer iterations, effectively improving the convergence speed of the algorithm, achieving a balance between hardware costs and computing performance, and having no special requirements for the signal to be processed itself, and is also applicable to signals with a very large proportion of missing sampling points, with strong robustness and universality.

[0090] In one embodiment, the signal processing based on the conventional iterative compressive sensing technology and the signal processing based on the iterative compressive sensing technology with an amplification process can be respectively performed on the same signal to be processed. The target convergence speed during the amplification process is 0.5, and the missing sampling ratio of the signal to be processed is 0.75. The iterative situation of the conventional signal processing based on the iterative compressive sensing technology is as Figure 8 shown, and the iterative situation of the signal processing based on the iterative compressive sensing technology with an amplification process is as Figure 9 shown. According to Figure 8 、 Figure 9 it can be seen that according to the signal processing method shown in Figure 9 , the effect of 12 iterations of the conventional signal processing can be achieved after 6 iterations.

[0091] In one embodiment, the preset convergence condition includes the spectrum energy convergence condition. The processor determines whether the current iteration converges according to the spectrum energy convergence condition, which can be implemented through the steps shown in Figure 10 , specifically including:

[0092] Step 201, calculate the total spectrum energy of the amplified target frequency components corresponding to the current iteration.

[0093] In a specific implementation, after the processor completes each iteration, it can calculate the total spectrum energy of the amplified target frequency components corresponding to the current iteration based on Parseval's theorem, and save the total spectrum energy of the amplified target frequency components corresponding to each iteration.

[0094] Step 202: Determine whether the current iteration converges based on the total spectral energy of the amplified target frequency component corresponding to the current iteration, the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, and a preset change threshold.

[0095] In a specific implementation, after the processor calculates the total spectral energy of the amplified target frequency component corresponding to the current iteration, it can determine whether the current iteration converges based on the total spectral energy of the amplified target frequency component corresponding to the current iteration, the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, and a preset change threshold. The change in spectral energy can well measure the difference between two adjacent iterations, and based on this, it can accurately determine whether the current iteration converges.

[0096] In some examples, the processor can use the following formula to determine whether the current iteration converges based on the total spectral energy of the amplified target frequency component corresponding to the current iteration, the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, and a preset change threshold:

[0097]

[0098] where P n is the total spectral energy of the amplified target frequency component corresponding to the current iteration, P n-1 is the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, δ is the preset change threshold, f = 1 indicates that the current iteration converges, and f = 0 indicates that the current iteration does not converge. The value range of the preset change threshold is 0.5% to 1.5%. For example, the preset change threshold can be set to 1%. That is, when the change amplitude of the total spectral energy of the amplified target frequency component corresponding to two adjacent iterations is less than 1%, it can be determined that the current iteration converges.

[0099] In one embodiment, the preset convergence condition includes a noise floor convergence condition. The processor determines whether the current iteration converges according to the noise floor convergence condition, which can be implemented through the following steps as shown Figure 11 and specifically includes:

[0100] Step 301: Sort the amplified target frequency components corresponding to the current iteration according to the frequency values to obtain a target sequence.

[0101] In a specific implementation, the processor sorts the amplified target frequency components corresponding to the current iteration according to the frequency values to obtain a target sequence. This sorting can be in ascending order or descending order.

[0102] Step 302: Use the frequency value at the preset position in the target sequence as the noise floor estimate value corresponding to the current iteration.

[0103] In a specific implementation, the processor uses the frequency value at a preset position in the target sequence as the noise floor estimation value corresponding to the current iteration, and the preset position can be set to one of 50%-60% in the target sequence.

[0104] Step 303: Calculate the difference between the noise floor estimation value corresponding to the current iteration and the noise floor estimation value corresponding to the previous iteration of the current iteration.

[0105] Step 304: Determine whether the absolute value of the difference is within a preset threshold range. If so, execute Step 305; otherwise, execute Step 306.

[0106] Step 305: Determine that the current iteration converges.

[0107] Step 306: Determine that the current iteration does not converge.

[0108] In a specific implementation, after the processor obtains the noise floor estimation value corresponding to the current iteration, it can calculate the difference between the noise floor estimation value corresponding to the current iteration and the noise floor estimation value corresponding to the previous iteration of the current iteration, and determine whether the absolute value of the difference is within a preset threshold range. If the absolute value of the difference is within the preset threshold range, it can be determined that the current iteration converges; if the absolute value of the difference is outside the preset threshold range, it can be determined that the current iteration does not converge. Among them, the preset threshold range can be set by those skilled in the art according to actual needs. The calculation of the noise floor estimation value is simple, and based on the change of the noise floor, it can quickly determine whether the current iteration converges.

[0109] In one embodiment, when the signal processing method proposed in this application is applied to signal recovery of a signal to be processed, before the processor performs discrete spectrum analysis on the frequency-domain signal to be processed based on the iterative compressive sensing technology, it is also necessary to perform interference detection on the signal to be processed, and filter out the interfered part of the signal to be processed. For example, set the signal amplitude of the interfered part in the signal to be processed to zero to obtain the filtered signal to be processed. When the processor performs signal recovery on the signal to be processed based on the target signal, it can replace the signal amplitude of the interfered part in the signal to be processed with the signal amplitude corresponding to the interfered part in the target signal to obtain the signal with the recovery completed.

[0110] In one embodiment, when the signal processing method proposed in this application is applied to perform velocity ambiguity resolution and / or direction-of-arrival estimation on a signal to be processed, it is applicable to a radar with a receiving array composed of a plurality of receiving elements. The receiving array is an equally spaced linear array composed of N element positions, and among the N positions, only M positions are equipped with real elements. The signal to be processed is obtained by this radar based on this receiving array. By using a receiving array that is an equally spaced linear array composed of N element positions but only M positions are equipped with real elements to obtain the signal to be processed, the velocity ambiguity resolution and / or direction-of-arrival estimation problem can be transformed into a frequency estimation problem of a missing sampled signal. Then, based on the fast-converging iterative compressive sensing technology provided in the embodiments of this application, the velocity ambiguity resolution and / or direction-of-arrival estimation can be quickly performed.

[0111] The step division of the above various methods is only for clear description. When implemented, they can be combined into one step or some steps can be split into multiple steps. As long as they include the same logical relationship, they are all within the protection scope of this patent; adding insignificant modifications to the algorithm or process or introducing insignificant designs, but not changing the core design of its algorithm and process are all within the protection scope of this patent.

[0112] Another embodiment of this application relates to a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the above method embodiments are implemented.

[0113] That is, those skilled in the art can understand that all or part of the steps in implementing the above method embodiments can be completed by instructing relevant hardware through a program. This program is stored in a storage medium, including several instructions to enable a device (which can be a single-chip microcomputer, a chip, etc.) or a processor to execute all or part of the steps of the methods described in various embodiments of this application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memories (ROM), random access memories (RAM), magnetic disks, or optical discs that can store program codes.

[0114] Another embodiment of this application relates to an integrated circuit. The details of the integrated circuit in this embodiment are specifically described below. The following content is only implementation details provided for convenient understanding and is not necessary for implementing this solution. The schematic diagram of the integrated circuit in this embodiment can be as Figure 12 shown, including:

[0115] A signal transceiver channel 401 for transmitting radio signals and receiving echo signals formed by the reflection of the radio signals by a target.

[0116] A signal processing module 402 for performing signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation based on the signal processing method described above.

[0117] In some examples, the signal processing module 402 may be as Figure 13 shown and includes:

[0118] A sampling unit 4021 for sampling the echo signal to obtain a discrete signal to be processed.

[0119] A frequency domain conversion unit 4022 for performing frequency domain conversion on the signal to be processed to obtain a frequency domain signal to be processed.

[0120] A spectrum analysis unit 4023 for performing discrete spectrum analysis on the frequency domain signal to be processed based on iterative compressive sensing technology to determine the target frequency components.

[0121] An amplification unit 4024 for amplifying the target frequency components.

[0122] A judgment unit 4025 for judging whether the current iteration converges based on the amplified target frequency components, and performing inverse frequency domain transformation on the amplified target frequency components to obtain a target signal when the current iteration converges.

[0123] An execution unit 4026 for performing signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation on the signal to be processed based on the target signal.

[0124] It is worth mentioning that each module involved in this embodiment is a logic module. In practical applications, a logic unit can be a physical unit, a part of a physical unit, or a combination of multiple physical units. In addition, to highlight the innovative part of this application, units not closely related to solving the technical problems proposed in this application are not introduced in this embodiment, but this does not mean that there are no other units in this embodiment.

[0125] In some alternative embodiments, the above integrated circuit may be a millimeter-wave radar chip or a lidar chip (such as an FMCW lidar chip), etc., for obtaining information such as the distance, angle, velocity, shape, size, surface roughness, and dielectric properties of a target. Optionally, the integrated circuit may be an Antenna-In-Package (AiP) chip structure, an Antenna-On-Package (AoP) chip structure, or an Antenna-On-Chip (AoC) chip structure, etc.

[0126] In an optional embodiment, different integrated circuits (such as chips) can be combined with each other to form a cascade structure. For the sake of simplicity, it will not be elaborated here, but it should be understood that all technologies that those skilled in the art should know based on the content recorded in this application should be included within the scope recorded in this application.

[0127] Another embodiment of this application relates to a radio device, which includes a carrier, the integrated circuit as described above disposed on the carrier, and an antenna disposed on the carrier for receiving and transmitting radio signals. The antenna can be integrated with the integrated circuit into an integrated device and disposed on the carrier (that is, at this time, the antenna can be the antenna provided in the AiP or AoC structure), and the integrated circuit can also be two separate components from the antenna, and a system-on-chip (SoC) structure is formed through connection. Among them, the carrier can be a printed circuit board (PCB), such as a development board, a data acquisition board, or the main board of a device, etc., and the first transmission line can be a PCB trace.

[0128] In some optional embodiments, this application also provides a terminal device, which may include a device body and the radio device described in any of the above embodiments disposed on the device body; wherein, the radio device can be used to implement functions such as target detection and / or wireless communication.

[0129] Specifically, on the basis of the above embodiments, in some optional embodiments of this application, the radio device can be disposed outside the device body or inside the device body, and in other optional embodiments of this application, a part of the radio device can also be disposed inside the device body and a part can be disposed outside the device body. The embodiments of this application do not limit this, and it can be determined according to the specific situation.

[0130] In some alternative embodiments, the above device body can be components and products applied to fields such as smart cities, smart homes, transportation, smart home appliances, consumer electronics, security monitoring, industrial automation, in-cabin detection (such as smart cockpits), medical devices, and healthcare. For example, the device body can be a smart transportation device (such as a car, bicycle, motorcycle, ship, subway, train, etc.), a security device (such as a camera), a liquid level / flow rate detection device, a smart wearable device (such as a bracelet, glasses, etc.), a smart home appliance (such as a floor cleaning robot, door lock, TV, air conditioner, smart light, etc.), various communication devices (such as a mobile phone, tablet computer, etc.), as well as a barrier gate, smart traffic lights, smart signs, traffic cameras, and various industrial robotic arms (or robots), and can also be various instruments for detecting vital sign parameters and various devices equipped with such instruments, such as in-car vital sign detection in a car, indoor personnel monitoring, smart medical devices, consumer electronic devices, etc.

[0131] The radio device can be the radio device described in any embodiment of the present application. The structure and working principle of the radio device have been described in detail in the above embodiments and will not be elaborated here one by one.

[0132] It should be noted that the radio device can achieve functions such as target detection and / or communication by transmitting and receiving radio signals, so as to provide detection target information and / or communication information to the device body, thereby assisting or even controlling the operation of the device body.

[0133] For example, when the above device body is applied to an Advanced Driving Assistance System (ADAS), the radio device as a vehicle-mounted sensor (such as a millimeter-wave radar, lidar, etc.) can assist the ADAS system to achieve application scenarios such as adaptive cruise control, Autonomous Emergency Braking (AEB), Blind Spot Detection (BSD), Lane Change Assist (LCA), Rear Cross Traffic Alert (RCTA), parking assistance, warning of rear vehicles, anti-collision, pedestrian detection, etc. At the same time, it can also be applied to application scenarios such as anti-collision when opening the door of a car.

[0134] The technical features of the above-described embodiments can be combined arbitrarily. For the sake of brevity of description, 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, it should be considered to be within the scope described in this specification.

[0135] The above-described embodiments merely represent the preferred embodiments of the present application and the technical principles applied. Their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the invention patent. Those skilled in the art can make various obvious changes, readjustments, and substitutions without departing from the protection scope of the present application. Therefore, although the present application has been described in relatively detail through the above embodiments, the present application is not limited to the above embodiments. Without departing from the concept of the present application, it can also include more other equivalent embodiments, and the protection scope of the present application is determined by the scope of the appended claims.

Claims

1. A signal processing method, characterized in that, Applied in the process of processing the echo signal received by the sensor, the echo signal includes a part affected by co-frequency and / or adjacent-frequency interference, and the method includes: Step S1, preprocessing the interference signal part in the echo signal as the signal to be processed; Step S2, performing spectral analysis on the signal to be processed to obtain the frequency-domain data to be processed; Step S3, obtaining the amplification factor of the current amplification operation based on the missing sampling ratio of the signal to be processed and the preset target convergence speed, and amplifying the frequency-domain data to be processed by using the amplification factor of the current amplification operation; Step S4, obtaining the threshold of the current filtering operation based on the initial threshold and the attenuation factor, and filtering the amplified frequency-domain data to be processed by using the threshold of the current filtering operation to obtain the filtered frequency-domain data; Step S5, determining whether the current iterative operation meets the preset requirements; if it does not meet the preset requirements, continue to step S6, otherwise continue to step S7; Step S6, inverse-transforming the filtered frequency-domain data to the time domain to obtain the restored signal, and using the restored signal to replace the preprocessed part in the echo signal, and using the replaced echo signal as the signal to be processed to continue steps S2 to S5; Step S7, inverse-transforming the filtered frequency-domain data to the time domain to obtain the restored signal, and using the restored signal to replace the preprocessed part in the echo signal, and obtaining the target information based on the replaced echo signal.

2. The signal processing method according to claim 1, wherein The target information includes the distance, speed, azimuth angle, elevation angle and / or micro-Doppler motion characteristics of the target object.

3. The signal processing method according to claim 1, wherein The sensor is an FMCW sensor, and the interference signal part is a part of the frequency points within a chirp, a part of the chirp signals within a frame, or a part of the frame signals within a preset number of frames.

4. The signal processing method according to any one of claims 1 to 3, characterized in that The preprocessing includes performing a zeroing operation on the interference signal part in the echo signal in the time domain.

5. The signal processing method according to any one of claims 1 to 4, characterized in that The determination of whether the current iterative operation meets the preset requirements includes: Determining whether at least one of the energy change and the noise floor change between two adjacent iterative operations meets the preset convergence condition, and / or, Determining whether the number of times of the current iterative operation reaches the maximum number of iterative times.

6. A signal processing method, which is applied to perform signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation on a signal to be processed, and is characterized in that, The method includes: Performing discrete spectral analysis on the signal to be processed based on the iterative compressive sensing technology to determine the target frequency component; Amplifying the target frequency component, and determining whether the current iteration converges based on the amplified target frequency component; If the current iteration converges, performing inverse frequency-domain transformation on the amplified target frequency component to obtain the target signal; Based on the target signal, performing signal restoration, velocity ambiguity resolution and / or direction of arrival estimation on the signal to be processed.

7. The signal processing method according to claim 1, characterized in that The amplification of the target frequency component includes: Determining the amplification factor corresponding to the current iteration according to the number of iterations of the current iteration, the missing sampling ratio of the signal to be processed and the preset target convergence speed; Amplifying the target frequency component according to the amplification factor.

8. The signal processing method according to claim 1, wherein Determine the amplification factor corresponding to the current iteration according to the number of iterations of the current iteration, the missing sampling ratio of the signal to be processed, and a preset target convergence rate, which is achieved by the following formula: Where n is the number of iterations of the current iteration, k is the missing sampling ratio of the signal to be processed, β is the target convergence rate, and α[n] is the amplification factor corresponding to the current iteration.

9. The signal processing method according to claim 8, wherein The value range of the target convergence rate is from 0.2 to 0.

8.

10. The signal processing method according to any one of claims 6 to 9, characterized in that The iterative compressive sensing technology includes an adaptive threshold iterative algorithm. The discrete spectrum analysis of the signal to be processed in the frequency domain based on the iterative compressive sensing technology to determine the target frequency component includes: Successively determine whether each frequency component in the signal to be processed in the frequency domain is greater than the adaptive iteration threshold, and determine the frequency components greater than the adaptive iteration threshold as the target frequency components; wherein, the adaptive iteration threshold is obtained based on the adaptive threshold iterative algorithm.

11. The signal processing method according to claim 10, wherein, After determining whether the current iteration converges based on the amplified target frequency component, the method further includes: If the current iteration does not converge, update the adaptive iteration threshold according to the number of iterations of the current iteration, the initial adaptive iteration threshold, and a preset attenuation factor; Successively determine whether each frequency component in the signal to be processed in the frequency domain is greater than the updated adaptive iteration threshold, and determine the frequency components greater than the updated adaptive iteration threshold as the target frequency components.

12. The signal processing method according to claim 11, wherein Update the adaptive iteration threshold according to the number of iterations of the current iteration, the initial adaptive iteration threshold, and a preset attenuation factor through the following formula: where ε is the initial adaptive iteration threshold, is the preset attenuation factor, n is the number of iterations of the current iteration, Thr n+1 is the updated adaptive iteration threshold.

13. The signal processing method according to any one of claims 6 to 9, characterized in that Determining whether the current iteration converges based on the amplified target frequency component includes: Calculate the total spectral energy of the amplified target frequency component corresponding to the current iteration; Determine whether the current iteration converges according to the total spectral energy of the amplified target frequency component corresponding to the current iteration, the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, and a preset change threshold.

14. The signal processing method according to claim 13, wherein Determine whether the current iteration converges through the following formula according to the total spectral energy of the amplified target frequency component corresponding to the current iteration, the total spectral energy of the amplified target frequency component corresponding to the previous iteration of the current iteration, and a preset change threshold: where P n is the total spectral energy of the magnified target frequency component corresponding to the current iteration, and P n-1 is the total spectral energy of the magnified target frequency component corresponding to the previous iteration of the current iteration. δ is the preset change threshold. f = 1 indicates that the current iteration converges, and f = 0 indicates that the current iteration does not converge.

15. The signal processing method according to claim 14, characterized in that The value range of the preset change threshold is from 0.5% to 1.5%.

16. The signal processing method according to any one of claims 6 to 9, characterized in that Determining whether the current iteration converges based on the amplified target frequency component includes: Sort the amplified target frequency components corresponding to the current iteration according to the frequency value to obtain a target sequence; wherein, the sorting is in ascending order or descending order; Use the frequency value at a preset position in the target sequence as the noise floor estimation value corresponding to the current iteration; Calculate the difference between the noise floor estimation value corresponding to the current iteration and the noise floor estimation value corresponding to the previous iteration of the current iteration; If the absolute value of the difference is within a preset threshold range, determine that the current iteration converges; If the absolute value of the difference is outside the preset threshold range, it is determined that the current iteration has not converged.

17. The signal processing method according to any one of claims 6 to 9, characterized in that, When the signal processing method is applied to perform signal recovery on the signal to be processed, before performing discrete spectrum analysis on the signal to be processed in the frequency domain based on the iterative compressive sensing technology, the method further includes: Performing interference detection on the signal to be processed, and removing the interfered part in the signal to be processed through filtering; Performing signal recovery on the signal to be processed based on the target signal, including: Replacing the signal amplitude of the interfered part in the signal to be processed with the signal amplitude corresponding to the interfered part in the target signal to obtain a signal with completed recovery.

18. The signal processing method according to claim 17, characterized in that, The removing the interfered part in the signal to be processed through filtering includes: Setting the signal amplitude of the interfered part in the signal to be processed to zero to obtain the filtered signal to be processed.

19. The signal processing method according to any one of claims 6 to 9, characterized in that, After performing signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation on the signal to be processed, the method further includes: Processing the signal after signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation to obtain the distance and / or velocity of the target object.

20. The signal processing method according to any one of claims 6 to 9, characterized in that The signal to be processed is a signal frame or a chirp signal.

21. The signal processing method according to any one of claims 6 to 9, characterized in that, The waveform of the signal to be processed is a continuous wave with a frequency linearly changing over time.

22. According to the signal processing method of claim 21, the continuous wave includes at least one of a frequency-modulated continuous wave (FMCW) and a stepped-frequency continuous wave (SFCW).

23. The signal processing method according to any one of claims 6 to 9, characterized in that The signal to be processed is obtained through the following steps: Obtaining an echo signal; Performing frequency mixing on the echo signal to obtain an intermediate-frequency signal; Performing analog-to-digital conversion on the intermediate-frequency signal to obtain a first discrete signal, and using the first discrete signal as the signal to be processed.

24. The signal processing method according to claim 23, wherein The using the first discrete signal as the signal to be processed includes: Performing digital signal processing on the first discrete signal at least once to obtain a second discrete signal; Using the second discrete signal as the signal to be processed.

25. A computer-readable storage medium stores a computer program, characterized in that, When the computer program is executed by a processor, it implements the signal processing method according to any one of claims 1-5, 6-24.

26. An integrated circuit, characterized in that, Including: A signal transceiver channel for transmitting a radio signal and receiving an echo signal formed by reflection of the radio signal by a target; A signal processing module for performing signal recovery, velocity ambiguity resolution, and / or direction of arrival estimation based on the method according to any one of claims 1-5, 6-24.

27. The integrated circuit according to claim 26, wherein, The signal processing module includes: A sampling unit for sampling the echo signal to obtain a discrete signal to be processed; A frequency-domain conversion unit for performing frequency-domain conversion on the signal to be processed to obtain a signal to be processed in the frequency domain; A spectrum analysis unit for performing discrete spectrum analysis on the signal to be processed in the frequency domain based on the iterative compressive sensing technology to determine the target frequency component; An amplification unit for amplifying the target frequency component; A judgment unit for judging whether the current iteration converges based on the amplified target frequency component, and performing inverse frequency-domain transformation on the amplified target frequency component to obtain a target signal when the current iteration converges. An execution unit for performing signal restoration, velocity ambiguity resolution, and / or direction-of-arrival estimation on the signal to be processed based on the target signal.

28. A radio device, characterized in that, Comprising: A carrier; The integrated circuit according to any one of claims 26 to 27, disposed on the carrier; An antenna disposed on the carrier for transmitting and receiving radio signals.

29. A terminal device, characterized in that, Comprising: A device body; The radio device according to claim 28 disposed on the device body, the radio device being used for target detection and / or communication.

Images