A Multipath Ghosting Removal Method for Indoor Target Localization

Abstract

This invention discloses a multipath ghosting elimination method for indoor target localization, comprising the following steps: Step 1, initial target localization; Step 2, mirror target position calculation; Step 3, theoretical position calculation of single-path first reflection + single-path LOS path; Step 4, theoretical position calculation of single-path LOS path + single-path first reflection; Step 5, multipath position matching and elimination. The method disclosed in this invention utilizes millimeter-wave radar to perform multipath elimination and subsequent localization of targets in an indoor multipath environment. Based on the propagation phenomena of electromagnetic waves, the main propagation paths of electromagnetic waves in indoor multipath scenarios are analyzed, and the theoretical positions of combined paths and two-way first reflection paths are calculated. This removes the multipath positioning points from the final target calculation results, obtaining the true target positioning trace. Experimental results demonstrate the accuracy of the method in locating true targets and the effectiveness of multipath elimination.

Description

Technical Field

[0001] This invention belongs to the field of target localization technology in indoor multipath environments, and specifically relates to a target localization method for eliminating multipath ghosting in indoor environments. Background Technology

[0002] In indoor target localization, electromagnetic waves inevitably reflect off building walls, so the true target position is often affected by multipath interference. The initial localization point includes the true target location point and multipath location points. These multipath location points can be considered as derivative targets of the true target, and they need to be eliminated to obtain the true target position coordinates. Many research institutions at home and abroad have carried out research on multipath ghosting algorithms.

[0003] In 2017, researchers from the German Research Foundation (T. Visentin, J. Hasch and T. Zwick, Analysis of multipath and DOA detection using a fully polarimetric automotive radar [C]. European Radar Conference, 2017: 45-48.) disclosed a polarization-based multipath ghosting discrimination method. This method utilizes the different polarization information of the direct-line-of-sight path and multipath reflection path of the preceding vehicle target. It performs coherent Pauli decomposition on the scattering matrix to distinguish the polarization information of different paths, estimating the direction of arrival (DOA) of the vehicle ghost, and thus distinguishing between the target and the multipath ghost. Experiments have verified that this method can distinguish between the target signal on the direct-line-of-sight path and the ghost of the primary reflection path. A drawback is that this method only achieves good discrimination results when the target is known.

[0004] In 2020, scholars from Southeast University (Liu, Chenwen, Liu Shengheng, Zhang Cheng, et al. Multipath Propagation Analysis and Ghost Target Removal for FMCW Automotive Radars, IET International Radar Conference, Chongqing, China, 2020, pp. 1-5) disclosed an electromagnetic propagation path differentiation method. This method analyzes the electromagnetic wave propagation path of a directly line-of-sight target near a radar reflector. By determining the different distances between the direct path and the reflected path, it distinguishes between them, thereby eliminating the influence of the reflected path's location point. Finally, simulation experiments verified the effectiveness of the proposed algorithm. A limitation is that this method was only numerically verified in a medium-range detection scenario. Summary of the Invention

[0005] The technical problem to be solved by this invention is to provide a multipath ghosting elimination method for indoor target localization.

[0006] The present invention adopts the following technical solution:

[0007] An improved method for multipath ghosting elimination in indoor target localization includes the following steps:

[0008] Step 1, Initial target location:

[0009] Let z be the electromagnetic wave emitted by the m-th antenna of the radar, reflected by the target, and then received by the n-th antenna of the radar. m，n (t, i), where t is the reference time and i is the number of cycles;

[0010] The radar echo signal is preprocessed using the MTI method, as shown below:

[0011] z′ m，n (t, i) = z m，n (t,i)-z m，n (t, i-1)

[0012] In the above formula, z′ m，n (t, i) represents the echo signal after preprocessing by the MTI method;

[0013] The echo signal z′ is processed using the FFT method. m，n (t, i), obtain the target distance image x m，n ;

[0014] For the target distance image x m，n Perform CFAR detection to obtain the target distance value R. For the j-th distance cell, the detection threshold TH is set. j Represented as:

[0015]

[0016] In the above formula, u m，n (h, j) represents the signal after incoherent superposition of the h-th frame, P f N represents the probability of a false alarm. r Represents u m，n Number of reference units in (h, j);

[0017] The target azimuth angle is calculated using the MVDR method, and the average output power P is then analyzed. MVDR (θ) performs a spectral peak search; the θ corresponding to the peak point is the azimuth angle of the target in the current frame. MVDR The formula for calculating (θ) is:

[0018]

[0019] In the above formula, a(θ) is the direction vector, which is related to the number of array antennas K, and its expression is as follows:

[0020]

[0021] In the above formula, λ is the phase difference between adjacent antennas, d is the receiving antenna spacing, and λ is the signal wavelength.

[0022] Based on the target distance R and the target azimuth angle θ, the initial positioning point's coordinates (x″, y″) are calculated using the following formula:

[0023]

[0024] The position coordinates obtained from the above calculation are from the radar array's view coordinate system. Transforming them to the coordinate system of the real scene yields the initial positioning point (x, y). The coordinate transformation formula is as follows:

[0025]

[0026] In the above formula, φ is the angle between the radar array direction and the horizontal direction;

[0027] Step 2, Calculate the target position:

[0028] The coordinates (x′, y′) of the mirror target symmetrical to the reflecting surface are calculated for each positioning point (x, y) in each frame, using the following formula:

[0029]

[0030] In the above formula, L is the distance between the radar and the reflector.

[0031] Step 3, one-way single reflection + one-way LOS path 21 Theoretical position calculation:

[0032] Assuming the obtained initial positioning point is the actual direct path positioning point, calculate l 21 The path distance is:

[0033]

[0034] l 21 The azimuth of the path location point is:

[0035] θ 21 =arctan(x / y)

[0036] The theoretical l of each initial positioning point 21 The path position (x1, y1) is calculated as follows:

[0037]

[0038] Step 4, one-way LOS path + one-way reflection 12 Theoretical position calculation:

[0039] l 12 The azimuth of the path location point is:

[0040] θ 12 =arctan(x′ / y′)

[0041] The theoretical l of each initial positioning point 12 The path position (x2, y2) is calculated as follows:

[0042]

[0043] Step 5, Multipath Position Match Elimination:

[0044] l 21 l 12 The path positioning point is matched with the initial positioning point, and the distance error between the positioning points is calculated. If the calculated distance error is less than 1 meter, the initial positioning point is a multipath positioning point and is eliminated. The formula for calculating the distance error is:

[0045]

[0046]

[0047] The beneficial effects of this invention are:

[0048] The method disclosed in this invention utilizes millimeter-wave radar to locate targets in indoor multipath environments after multipath cancellation. Based on the propagation phenomena of electromagnetic waves, the main propagation paths of electromagnetic waves in indoor multipath scenarios are analyzed, and the theoretical positions of combined paths and two-way single-reflection paths are calculated. This removes multipath positioning points from the final target calculation results, yielding the true target location. Experimental results demonstrate the accuracy of the method in locating true targets and the effectiveness of multipath cancellation. Attached Figure Description

[0049] Figure 1 This is a schematic diagram of indoor detection of a target scene directly in front of the viewer;

[0050] Figure 2 This is a schematic diagram of the actual test scenario;

[0051] Figure 3 It is a single-target range image;

[0052] Figure 4 This is a schematic diagram of the initial positioning points;

[0053] Figure 5 This is a schematic diagram of the actual target localization result after multipath elimination. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0055] Example 1 discloses a multipath ghosting elimination method for indoor target localization. First, data is acquired using a two-transmitter, four-receiver millimeter-wave radar, and preprocessed using the MTI method to remove static clutter. Then, the target distance is obtained through a detection method. Next, the target azimuth is obtained using the MVDR angle measurement method, and the initial target location coordinates are calculated. The location points of each mirrored target are reconstructed based on building layout information, and then the theoretical location points of the combined path are calculated. These combined path location points are then matched and eliminated with the initial location points, while false target points outside the building layout are removed. Finally, the true location point of the target, free from multipath ghosting interference, is obtained.

[0056] Electromagnetic wave propagation path analysis: detection scenario such as Figure 1 As shown, according to Figure 1 Signal modeling shows that the two-way electromagnetic propagation of electromagnetic waves in the above scenario mainly includes a two-way LOS path, a two-way single reflection, and a one-way single reflection + one-way LOS path. The propagation paths are as follows:

[0057] Two-way LOS path (l 11 ): R→Q→R, the electromagnetic wave signal is emitted from radar R and propagates directly to the target location and returns to radar along the original path.

[0058] Two-way single reflection (l 22 ): R→P→Q→P→R, the electromagnetic wave signal is emitted from radar R, reflected by the wall, propagates to the target location, and returns to radar along the original path.

[0059] One-way LOS path + one-way single reflection (l 12 ): R→Q→P→R, or R→P→Q→R, the electromagnetic wave signal reaches the target through the LOS path and is reflected by the wall upon return to the radar.

[0060] One-way single reflection + one-way LOS path (l 21 ): R→P→Q→R, the electromagnetic wave signal is emitted from radar R, propagated through the reflection of the wall to the target location, and returns to the radar directly through the LOS path when returning to the radar.

[0061] The method disclosed in this embodiment specifically includes the following steps:

[0062] Step 1, Initial target location:

[0063] Let z be the electromagnetic wave emitted by the m-th antenna of the radar, reflected by the target, and then received by the n-th antenna of the radar. m，n (t, i), where t is the reference time and i is the number of cycles;

[0064] To suppress the effects of static background echoes and antenna coupling in radar echo signals, the radar echo signal is preprocessed using the MTI (Moving Target Indicator) method, as shown below:

[0065] z′ m，n (t, i) = z m，n (t,i)-z m，n (t, i-1)

[0066] In the above formula, z′ m，n (t, i) represents the echo signal after preprocessing by the MTI method;

[0067] The echo signal z′ is processed using the FFT method. m，n (t, i), obtain the target distance image x m，n ;

[0068] For the target distance image x m，n Perform CFAR detection to obtain the target distance value R. For the j-th distance cell, the detection threshold TH is set. j Represented as:

[0069]

[0070] In the above formula, u m，n (h, j) represents the signal after incoherent superposition of the h-th frame, P f N represents the probability of a false alarm. r Represents u m，n Number of reference units in (h, j);

[0071] The target azimuth angle is calculated using the MVDR method, and the average output power P is then analyzed. MVDR (θ) performs a spectral peak search; the θ corresponding to the peak point is the azimuth angle of the target in the current frame. MVDR The formula for calculating (θ) is:

[0072]

[0073] In the above formula, a(θ) is the direction vector, which is related to the number of array antennas K, and its expression is as follows:

[0074]

[0075] In the above formula, λ is the phase difference between adjacent antennas, d is the receiving antenna spacing, and λ is the signal wavelength.

[0076] Based on the target distance R and the target azimuth angle θ, the initial positioning point's coordinates (x″, y″) are calculated using the following formula:

[0077]

[0078] The position coordinates obtained from the above calculation are from the radar array's view coordinate system. Transforming them to the coordinate system of the real scene yields the initial positioning point (x, y). The coordinate transformation formula is as follows:

[0079]

[0080] In the above formula, φ is the angle between the radar array direction and the horizontal direction;

[0081] Step 2, Calculate the target position:

[0082] According to the laws of electromagnetic wave reflection, the real target and its mirror image will be symmetrical with respect to the reflecting surface.

[0083] The coordinates (x′, y′) of the mirror target symmetrical to the reflecting surface are calculated for each positioning point (x, y) in each frame, using the following formula:

[0084]

[0085] In the above formula, L is the distance between the radar and the reflector.

[0086] Step 3, one-way single reflection + one-way LOS path 21 Theoretical position calculation:

[0087] After obtaining the initial positioning point coordinates, we can use l 21 The path location point and the direct path have the same azimuth angle, and the distance is half the sum of the direct path distance and the reflection path distance. Assuming that the obtained initial location point is the true direct path location point, calculate l. 21 The path distance is:

[0088]

[0089] l 21 The azimuth of the path location point is:

[0090] θ 21 =arctan(x / y)

[0091] The theoretical l of each initial positioning point 21 The path position (x1, y1) is calculated as follows:

[0092]

[0093] Step 4, one-way LOS path + one-way reflection 12 Theoretical position calculation:

[0094] l 12 The azimuth of the path location point is:

[0095] θ 12 =arctan(x′ / y′)

[0096] The theoretical l of each initial positioning point 12 The path position (x2, y2) is calculated as follows:

[0097]

[0098] Step 5, Multipath Position Match Elimination:

[0099] l 21 l 12 The path positioning point is matched with the initial positioning point, and the distance error between the positioning points is calculated. If the calculated distance error is less than 1 meter, the initial positioning point is a multipath positioning point and is eliminated. The formula for calculating the distance error is:

[0100]

[0101]

[0102] In other words, if error1 or error2 is less than 1, it means that the initial positioning point and the multipath positioning point are successfully matched. At this time, the initial positioning point (x, y) will be eliminated, and only the real target positioning point will be retained.

[0103] Below is a real-world test example, focusing on a single target in an indoor scene. The test scenario is as follows: Figure 2 As shown, a single target is swaying in place at position Q1 (-2.52, 4.6). The radar is 5.33 meters away from the wall, and the radar and the horizontal direction are facing the wall at an angle of 54.3 degrees.

[0104] First, the single-target range image is obtained using the MTI method and pulse compression method, such as... Figure 3 As shown. The initial positioning point of the target is calculated as follows. Figure 4 As shown. The theoretical location of the combined path is calculated using the method described above, and then matching and elimination are performed to obtain the location points after multipath elimination, as shown. Figure 5 As shown in the figure, the experimental results only retain the actual target location coordinates, indicating that the method of the present invention can accurately locate indoor targets, thus verifying the accuracy and effectiveness of the method of the present invention.

Claims

1. A multipath ghosting elimination method for indoor target localization, characterized in that, Includes the following steps: Step 1, Initial target location: Let z be the electromagnetic wave emitted by the m-th antenna of the radar, reflected by the target, and then received by the n-th antenna of the radar. m，n (t, i), where t is the reference time and i is the number of cycles; The radar echo signal is preprocessed using the MTI method, as shown below: z′ m，n (t,i)=z m，n (t,i)-z m，n (t,i-1) In the above formula, z′ m，n (t, i) represents the echo signal after preprocessing by the MTI method; The echo signal z′ is processed using the FFT method. m，n (t, i), obtain the target distance image x m，n ; For the target distance image x m，n Perform CFAR detection to obtain the target distance value R. For the j-th distance cell, the detection threshold TH is set. j Represented as: In the above formula, u m，n (h, j) represents the signal after incoherent superposition of the h-th frame, P f N represents the probability of a false alarm. r Represents u m，n Number of reference units in (h, j); The target azimuth angle is calculated using the MVDR method, and the average output power P is then analyzed. MVDR (θ) performs a spectral peak search; the θ corresponding to the peak point is the azimuth angle of the target in the current frame. MVDR The formula for calculating (θ) is: In the above formula, a(θ) is the direction vector, which is related to the number of array antennas K, and its expression is as follows: In the above formula, λ is the phase difference between adjacent antennas, d is the spacing between receiving antennas, and λ is the signal wavelength. Based on the target distance R and the target azimuth angle θ, the initial positioning point's coordinates (x″, y″) are calculated using the following formula: The position coordinates obtained from the above calculation are from the radar array's view coordinate system. Transforming them to the coordinate system of the real scene yields the initial positioning point (x, y). The coordinate transformation formula is as follows: In the above formula, φ is the angle between the radar array direction and the horizontal direction; Step 2, Calculate the target position: The coordinates (x′, y′) of the mirror target symmetrical to the reflecting surface are calculated for each positioning point (x, y) in each frame, using the following formula: In the above formula, L is the distance between the radar and the reflector. Step 3, one-way single reflection + one-way LOS path 21 Theoretical position calculation: Assuming the initial location point is the actual direct path location point, calculate l. 21 The path distance is: l 21 The azimuth of the path location point is: θ 21 =arctan(x / y) The theoretical l of each initial positioning point 21 The path position (x1, y1) is calculated as follows: Step 4, one-way LOS path + one-way reflection 12 Theoretical position calculation: l 12 The azimuth of the path location point is: θ 12 =arctan(x′ / y′) The theoretical l of each initial positioning point 12 The path position (x2, y2) is calculated as follows: Step 5, Multipath Position Match Elimination: l 21 l 12 The path positioning point is matched with the initial positioning point, and the distance error between the positioning points is calculated. If the calculated distance error is less than 1 meter, the initial positioning point is a multipath positioning point and is eliminated. The formula for calculating the distance error is:

Images