A radar channel high-resolution parameter estimation method based on original received signal

Abstract

The application relates to a radar channel high-resolution parameter estimation method based on an original receiving signal, which comprises the following steps: S1, acquiring an original receiving signal of a vehicle-mounted radar; S2, determining a frequency domain model of the original radar receiving signal by combining a time domain model of a radar transmitting signal, a frequency domain model of the radar transmitting signal and a channel frequency response model; S3, estimating a multipath component parameter in a frequency domain by taking the frequency domain signal of the original radar receiving signal as input and taking the frequency domain model of the original radar receiving signal as an a priori model; and S4, analyzing multipath information according to the estimation result of the multipath component parameter, wherein the multipath information comprises time delay, Doppler frequency shift, wave direction and wave separation direction, and the original receiving signal is reconstructed. Compared with the prior art, the application utilizes rich channel information in the original receiving signal, and improves the resolution and accuracy of target estimation.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a method for high-resolution parameter estimation of radar channels based on raw received signals. Background Technology

[0002] With the rapid development of autonomous driving and the Internet of Vehicles, automotive millimeter-wave radar plays a crucial role in advanced driver assistance systems and vehicle environmental perception. Automotive radar channels possess certain unique characteristics, such as the high carrier frequency of the millimeter-wave band, the co-location of the transmitter and receiver, and time-varying conditions in specific and typical environments. These unique characteristics present both challenges and necessities for radar channel research.

[0003] Current radar signal processing utilizes beat frequency signals for analysis and research. This involves transmitting the signal through a channel, receiving it with a receiving antenna, and then processing the echo signal through mixing and anti-aliasing filtering to obtain the beat frequency signal. While this method effectively converts high frequencies to low frequencies, reducing the sampling frequency at the receiver, it overlooks the frequency-selective fading inherent in the channel. This leads to the loss of some channel information and also reduces resolution and accuracy. Summary of the Invention

[0004] The purpose of this invention is to provide a high-resolution parameter estimation method for radar channels based on the original received signal.

[0005] The objective of this invention can be achieved through the following technical solutions:

[0006] A high-resolution parameter estimation method for radar channels based on raw received signals includes:

[0007] Step S1: Acquire the raw received signal from the vehicle radar;

[0008] Step S2: Determine the frequency domain model of the original radar received signal by combining the time domain model and frequency domain model of the radar transmitted signal with the channel frequency response model.

[0009] Step S3: Using the frequency domain signal of the original radar received signal as input and the frequency domain model of the original radar received signal as a priori model, estimate the multipath component parameters in the frequency domain.

[0010] Step S4: Based on the estimation results of the multipath component parameters, obtain the multipath information, which includes time delay, Doppler frequency shift, direction of arrival, and direction of departure, and reconstruct the original received signal.

[0011] The vehicle-mounted radar is a millimeter-wave linear frequency modulated radar.

[0012] The mathematical expression for the time-domain model of the radar transmitted signal is:

[0013]

[0014] Where: s t (t s ,t f ) represents the time-domain model of the radar transmitted signal, t s t represents the slow time of signal transmission. f Let T be the fast time of the transmitted signal, l be the slow time index, L be the number of signal sequences, δ(·) be the Dirac function, and T be the slow time index. cp T is the pulse repetition time, Rect(·) is the window function, and T chirp For the duration of the signal, A t f is the complex amplitude of the transmitted signal. c Here, k is the center frequency, and k1 is the ratio of signal bandwidth to signal duration.

[0015] The mathematical expression for the fast time of the transmitted signal is:

[0016] t f =n s / F s

[0017] Wherein: F s Where n is the sampling frequency s These are the sampling points.

[0018] The mathematical expression for the frequency domain model of the radar transmitted signal is:

[0019]

[0020] Wherein: S t (f) is the frequency domain model of the radar transmitted signal, where f is the frequency at each time point. It is a Fresnel function.

[0021] The mathematical expression for the channel frequency response model is:

[0022]

[0023] Where: H(f) is the channel frequency response model, α n Let τ be the complex amplitude of the nth path, N be the number of paths, and τ be the time delay of the nth path.

[0024] The mathematical expression for the frequency domain model of the original radar received signal is:

[0025]

[0026] Wherein: S r (f) represents the frequency domain model of the original radar received signal, where f is the frequency at each time point, k1 is the ratio of signal bandwidth to signal duration, and f c For the center frequency, For Fresnel functions, α n Let τ be the complex amplitude of the nth path. n Let be the delay of the nth path.

[0027] The process of estimating the multipath component parameters is implemented using a spatial iterative generalized expectation-maximization algorithm.

[0028] The multipath component parameters include time delay and corresponding distance and latitude information.

[0029] The process of obtaining multipath information in step S4 is based on channel feature analysis using power delay spectrum.

[0030] Compared with the prior art, the present invention has the following beneficial effects:

[0031] 1. Traditional vehicle radar signal processing involves receiving signals through mixing and anti-aliasing filtering to obtain beat frequency signals. In this invention, the original radar received signal is not subjected to mixing calculations. Instead, the original received signal is directly processed and analyzed, avoiding the impact of frequency-selective fading caused by the channel and improving the resolution and accuracy of parameter estimation results.

[0032] 2. It better reflects the actual working environment of vehicle-mounted radar and derives the frequency domain model of the original radar received signal without mixing operation. This frequency domain model is used as the prior model for the high-precision channel parameter estimation algorithm to perform channel parameter estimation. It utilizes the rich channel information in the original received signal to improve the resolution and accuracy of the parameter estimation results. Attached Figure Description

[0033] Figure 1 This is a flowchart of the present invention;

[0034] Figure 2 This is a schematic diagram of a single-target multipath simulation scenario in an embodiment of the present invention;

[0035] Figure 3 This is the power delay spectrum of the original received signal in a single-target multipath scenario according to an embodiment of the present invention;

[0036] Figure 4This is a comparison chart showing the results before and after reconstruction of the power delay spectrum after high-precision channel parameter estimation using the original received signal in a single-target multipath scenario according to an embodiment of the present invention.

[0037] Figure 5 This is the power delay spectrum of a traditional beat frequency signal in a single-target multipath scenario according to an embodiment of the present invention;

[0038] Figure 6 This is a comparison of the power delay spectrum reconstruction results before and after high-precision channel parameter estimation using traditional beat frequency signals in a single-target multipath scenario according to an embodiment of the present invention.

[0039] Figure 7 This is a comparison diagram of the power delay spectrum of the original received signal and the traditional beat frequency signal in a single-target multipath scenario according to an embodiment of the present invention. Detailed Implementation

[0040] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. These embodiments are based on the technical solution of the present invention and provide detailed implementation methods and specific operating procedures. However, the scope of protection of the present invention is not limited to the following embodiments.

[0041] A high-resolution parameter estimation method for radar channels based on raw received signals, such as... Figure 1 As shown, it includes:

[0042] Step S1: Obtain the original received signal from the vehicle-mounted radar. The vehicle-mounted radar is a millimeter-wave linear frequency modulated radar. The transmitted signal is received by the receiving antenna after being transmitted through the channel, and the received echo signal is obtained.

[0043] In this embodiment, the vehicle-mounted channel environment and simulation scenario in step S1 are as follows: Figure 2 As shown, in a radar scenario, the vehicle-mounted radar channel has certain unique characteristics, with the transmitter (Tx) and receiver (Rx) located very close to each other. Therefore, in the simulation, it is assumed that Tx and Rx are in the same position, such as... Figure 2 The pentagram in the middle represents the transmitter / receiver (Tx / Rx). Figure 2 The "Zhongxing" symbol represents the set target point, located 50m from the transmitter / receiver. The circles represent random scatterers in the environment, set to 5.

[0044] Step S2: Determine the frequency domain model of the original radar received signal by combining the time domain model and frequency domain model of the radar transmitted signal with the channel frequency response model.

[0045] The mathematical expression for the time-domain model of radar transmitted signals is:

[0046]

[0047] Where: s t (t s , t f ) represents the time-domain model of the radar transmitted signal, t s t represents the slow time of signal transmission. f Let T be the fast time of the transmitted signal, l be the slow time index, L be the number of signal sequences, δ(·) be the Dirac function, and T be the slow time index. cp T is the pulse repetition time, Rect(·) is the window function, and T chirp For the duration of the signal, A t f is the complex amplitude of the transmitted signal. c Here, k is the center frequency, and k1 is the ratio of signal bandwidth to signal duration.

[0048] The mathematical expression for the fast time of signal transmission is:

[0049] t f =n s / F s

[0050] Wherein: F s Where n is the sampling frequency s For sampling points, N s n is the number of sampling points, 0 ≤ n s ≤N s -1.

[0051] The frequency domain model of the radar transmitted signal is obtained by Fourier transforming the time domain expression of the radar transmitted signal. Its mathematical expression is:

[0052]

[0053] Wherein: S t (f) is the frequency domain model of the radar transmitted signal, where f is the frequency at each time point. It is a Fresnel function.

[0054] The derivation process is as follows:

[0055] S t (f)=∫s t (t s , t f )·exp(-j2πft f )dt f

[0056]

[0057] make

[0058]

[0059] make

[0060]

[0061] Therefore, the frequency domain formula for the transmitted signal is:

[0062]

[0063] in, It is a Fresnel function.

[0064]

[0065] The mathematical expression for the channel frequency response model is:

[0066]

[0067] Where: H(f) is the channel frequency response model, α n Let τ be the complex amplitude of the nth path, N be the number of paths, and τ be the complex amplitude of the path. n Let be the delay of the nth path.

[0068] The original radar received signal is obtained from the frequency domain. The frequency domain model of the radar transmitted signal is obtained by performing a Fourier transform, and then multiplied by the channel frequency response H(f) to obtain the frequency domain model of the original radar received signal. In this embodiment, this frequency domain approach is used to obtain the original radar received signal.

[0069] like Figure 3 The image shows the power delay profile (PDP) of the original radar received signal. The channel impulse response (CIR) is first obtained by deconvolution of the frequency domain signal of the original radar received signal, and then the PDP is calculated. Figure 3The asterisk line represents the power delay curve calculated by deconvolution of the original received signal. The solid line represents the power delay curve of the real channel generated using a graph theory model. The vertical dashed line represents the location of the target. As can be seen from the figure, the power delay curve obtained using the original received signal largely coincides with the power delay curve of the real channel, indicating that it can represent the information of the real channel relatively completely. Furthermore, the target location delay corresponding to the real channel is 0.3333 μs, while the delay of the power delay curve obtained using the original received signal at the corresponding target location is 0.3286 μs, maintaining high accuracy and precisely representing the target's location. It also relatively completely reflects other scatterers in the environment; several scatterers in close proximity are also clearly shown, indicating high resolution. It can identify targets or scatterers at relatively close distances, accurately reflecting the surrounding environment of the vehicle and aiding in subsequent radar interference detection and suppression technologies.

[0070] The mathematical expression for the frequency domain model of the original radar received signal is:

[0071]

[0072] Wherein: S r (f) represents the frequency domain model of the original radar received signal, where f is the frequency at each time point, k1 is the ratio of signal bandwidth to signal duration, and f c For the center frequency, For Fresnel functions, α n Let τ be the complex amplitude of the nth path. n Let be the delay of the nth path.

[0073] Traditional vehicle radar signal processing involves receiving signals through mixing and anti-aliasing filtering to obtain beat frequency signals. In this application, the original radar received signal is not subjected to mixing calculations; instead, the original received signal is directly processed and analyzed.

[0074] Step S3: Using the frequency domain signal of the original radar received signal as input and the frequency domain model of the original radar received signal as the prior model, estimate the multipath component parameters in the frequency domain.

[0075] In this embodiment, the process of estimating multipath component parameters is implemented using a spatial iterative generalized expectation-maximization algorithm, and the multipath component parameters include time delay and corresponding distance-dimensional information. In this embodiment, the information of time delay and distance dimension is mainly utilized, the number of estimated paths is 10, and the number of iterations is 2.

[0076] Step S4: Based on the estimation results of the multipath component parameters, obtain the multipath information, which includes time delay, Doppler frequency shift, direction of arrival, and direction of departure. Reconstruct the original received signal. The process of obtaining the multipath information is based on channel feature analysis using the power delay spectrum.

[0077] like Figure 4 The figure shows a comparison between the PDP of the reconstructed signal obtained by the high-precision channel parameter estimation algorithm and the original radar received signal, i.e., the PDP before estimation. The asterisk lines represent the power delay curve of the original received signal, i.e., before estimation, and... Figure 3 The asterisks correspond to the solid lines; the solid lines represent the power delay curve of the signal reconstructed using a high-precision channel parameter estimation algorithm; the vertical dashed lines indicate the location of the target. From Figure 4 As can be seen, the power delay curve of the reconstructed signal basically coincides with the power delay curve before estimation, which also proves the correctness of the derived frequency domain model of the radar transmitted and received signals. Meanwhile, the target position delay corresponding to the power delay curve of the reconstructed signal using the high-precision channel parameter estimation algorithm is also 0.3286μs, consistent with the delay of the original received signal at the corresponding target position. This indicates that the accuracy of the high-precision channel parameter estimation algorithm remains stable before and after estimation, which helps in accurately identifying the target. Furthermore, it can be observed that even the reconstructed signal maintains high resolution, comprehensively reflecting the scattering information in the vehicle environment, which is helpful for subsequent detection and discrimination of radar interference signals and improves the capability of radar interference suppression technology.

[0078] like Figure 5 The image shows the power delay spectrum of a traditional beat frequency signal. The channel impulse response can be obtained by directly performing a Fourier transform on the beat frequency signal obtained by the mixing operation, and then the PDP can be calculated. Figure 5The central star line represents the power delay curve obtained by mixing the beat frequency signal. The solid line represents the power delay curve of the real channel generated using the graph theory model, and the vertical dashed line represents the location of the target. As can be seen from the figure, the power delay curve obtained using the traditional beat frequency signal does not coincide with the curve of the real channel, only reflecting a portion of the real channel. Due to the mixing calculation, the signal bandwidth is narrowed, which is reflected in the fact that the main lobe width of the beat frequency signal is wider than that of the real channel. Furthermore, the mixing calculation method also affects the signal power. Simultaneously, the target location delay corresponding to the real channel is 0.3333 μs, while the power delay curve obtained by the traditional beat frequency signal has a delay of 0.35 μs at the corresponding target location. Its accuracy is lower than that of the original radar received signal, leading to errors in target distance judgment and potentially even false detections. More importantly, the figure shows that the traditional beat frequency signal cannot fully represent the real environment; nearby scattering objects are not effectively identified, and its resolution is lower than that of processing the original received signal, thus affecting the subsequent radar interference detection and suppression results.

[0079] like Figure 6 The figure shows a comparison between the PDP of the beat frequency reconstructed signal obtained by the high-precision channel parameter estimation algorithm and the PDP of the traditional beat frequency signal before estimation. The asterisk lines represent the power delay curve of the traditional beat frequency signal before estimation. Figure 5 The asterisks correspond to the solid lines; the solid lines represent the power delay curve of the signal reconstructed using a high-precision channel parameter estimation algorithm; the vertical dashed lines indicate the location of the target. From Figure 6 As can be seen, although the power delay curve of the reconstructed signal basically coincides with the power delay curve before estimation, the reconstructed power delay curve is not consistent with the real channel. Furthermore, the corresponding target position delay is 0.35μs, which is lower than the accuracy of the original radar received signal. Because the reference signal for high-precision channel parameter estimation using traditional beat frequency signals is the beat frequency signal, the resolution of the estimation result is also consistent with the beat frequency signal. Scattering objects at close range are not estimated or reconstructed. Therefore, the estimation result can only reflect a portion of the real channel and cannot accurately distinguish scattering objects at close range, potentially leading to missed detections or false detections.

[0080] like Figure 7The figure shows a comparison of the power delay spectra of the original received signal and the traditional beat frequency signal. The solid line represents the power delay curve obtained by processing the original received signal and performing deconvolution in this invention. The asterisk line represents the power delay curve after mixing and calculation of the traditional beat frequency signal. The vertical dashed line represents the location of the target. As can be seen from the figure and the enlarged view, the target information obtained using the original received signal is closer to the real target and has higher accuracy. Furthermore, processing the original received signal accurately reflects nearby scattering objects in the environment, resulting in higher resolution, while these scattering objects cannot be identified using the traditional beat frequency signal.

[0081] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

Claims

1. A high-resolution parameter estimation method for radar channels based on raw received signals, characterized in that, include: Step S1: Acquire the raw received signal from the vehicle radar; Step S2: Determine the frequency domain model of the original radar received signal by combining the time domain model and frequency domain model of the radar transmitted signal with the channel frequency response model. Step S3: Using the frequency domain signal of the original radar received signal as input and the frequency domain model of the original radar received signal as a priori model, estimate the multipath component parameters in the frequency domain. Step S4: Based on the estimation results of the multipath component parameters, obtain the multipath information, wherein the multipath information includes time delay, Doppler frequency shift, direction of arrival, and direction of departure, and reconstruct the original received signal; The raw received signal from the vehicle-mounted radar is not mixed and processed directly, thus avoiding the impact of frequency-selective fading caused by the channel and improving the resolution and accuracy of the parameter estimation results.

2. The method for high-resolution parameter estimation of radar channels based on raw received signals according to claim 1, characterized in that, The vehicle-mounted radar is a millimeter-wave linear frequency modulated radar.

3. The method for high-resolution parameter estimation of radar channels based on the original received signal according to claim 2, characterized in that, The mathematical expression for the time-domain model of the radar transmitted signal is: in: This is a time-domain model of the radar transmitted signal. For the slow time of signal transmission, For the fast transmission time of the signal, For slow time indexing, Let be the number of signal sequences, and δ(·) be the Dirac function. Let Rect(·) be the pulse repetition time, and Rect(·) be the window function. For signal duration, The complex amplitude of the transmitted signal. For the center frequency, It is the ratio of signal bandwidth to signal duration.

4. The method for high-resolution parameter estimation of radar channels based on the original received signal according to claim 3, characterized in that, The mathematical expression for the fast time of the transmitted signal is: = / in: Sampling frequency, n s These are the sampling points.

5. The method for high-resolution parameter estimation of radar channels based on the original received signal according to claim 3, characterized in that, The mathematical expression for the frequency domain model of the radar transmitted signal is: in: This is a frequency domain model of the radar transmitted signal. The frequencies corresponding to each time point, , is a Fresnel function. , .

6. The high-resolution parameter estimation method for radar channels based on the original received signal according to claim 5, characterized in that, The mathematical expression for the channel frequency response model is: in: For the channel frequency response model, For the first The complex amplitude of the path, For the number of paths, For the first The delay of each path.

7. The method for high-resolution parameter estimation of radar channels based on the original received signal according to claim 6, characterized in that, The mathematical expression for the frequency domain model of the original radar received signal is: in: This is the frequency domain model of the original radar received signal. The frequencies corresponding to each time point, It is the ratio of signal bandwidth to signal duration. For the center frequency, For the first The complex amplitude of the path, For the first The delay of each path.

8. The method for high-resolution parameter estimation of radar channels based on the original received signal according to claim 1, characterized in that, The process of estimating the multipath component parameters is implemented using a spatial iterative generalized expectation-maximization algorithm.

9. The method for high-resolution parameter estimation of radar channels based on the original received signal according to claim 1, characterized in that, The multipath component parameters include time delay and corresponding distance and latitude information.

10. The high-resolution parameter estimation method for radar channels based on the original received signal according to claim 1, characterized in that, The process of obtaining multipath information in step S4 is based on channel feature analysis using power delay spectrum.

Images