Wireless positioning method and device based on selective iteration refinement
By using a selective iterative refinement method, error sources in complex environments are identified and corrected, solving the accuracy and robustness problems of wireless positioning technology under non-line-of-sight and multipath effects, and achieving high-efficiency positioning accuracy and robustness.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- THE 54TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY GROUP CORPORATION
- Filing Date
- 2025-12-31
- Publication Date
- 2026-04-10
AI Technical Summary
Existing wireless positioning technologies struggle to effectively verify and correct errors in angle or distance measurements in non-line-of-sight and multipath environments, resulting in insufficient positioning accuracy and robustness, as well as low computational efficiency.
A selective iterative refinement method is adopted. Through error source identification and spatial consistency verification, only the measurement values with the most serious problems are refined. Combined with super-resolution direction finding algorithm and signal propagation model, iterative correction of angle and distance is achieved.
It significantly improves positioning accuracy and robustness in complex environments, while also enhancing computational efficiency, achieving high-precision and high-efficiency positioning results.
Smart Images
Figure CN121842610A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of wireless communication and positioning technology, in particular to a wireless positioning method and device capable of achieving high precision, high robustness and high efficiency in complex environments such as non-line-of-sight and multipath propagation in a network composed of base stations with known positions. BACKGROUND
[0002] With the rapid development of technologies such as Internet of Things, smart city and self-driving, the demand for high-precision wireless positioning technology is increasingly urgent. Existing wireless positioning technologies are mainly divided into two categories: ranging-based and direction-finding-based. Ranging-based technologies (such as Time of Arrival TOA and Time Difference of Arrival TDOA) measure the propagation distance of signals for positioning, but in non-line-of-sight environments, the signal propagation path is blocked or reflected, which can cause the distance measurement value to be significantly larger, resulting in a large positioning error. Direction-finding-based technologies (such as Angle of Arrival AOA) use antenna arrays to estimate the direction of arrival of signals, but are also susceptible to multipath effects, which can cause angle estimation errors.
[0003] Existing technologies usually process angle or distance information in isolation, or simply fuse angle and distance information (such as least squares method), but this fusion method can cause the positioning result to be severely distorted when there is a large error in either angle or distance measurement, and lacks an effective mechanism to verify and correct the contaminated signal measurement values. In addition, when multiple measurement values have errors, existing solutions often treat all data equally, resulting in large computational complexity and low efficiency, and even introducing new errors by correcting originally accurate data, which can result in insufficient reliability, accuracy and real-time performance in complex environments.
[0004] Therefore, how to overcome the influence of non-line-of-sight and multipath effects, and to achieve selective mutual verification and correction of angle-distance measurement values, so as to improve the positioning accuracy and robustness in complex environments, has become a problem to be solved in the field. SUMMARY
[0005] The present application aims to overcome the shortcomings of the prior art and provide a wireless positioning method and device based on selective iterative refinement. By introducing an error source identification mechanism, the present application intelligently selects the most problematic measurement values for targeted refinement, achieving the best balance between accuracy, robustness and computational efficiency in complex environments.
[0006] To achieve the above-mentioned purpose, the present application adopts the following technical solutions:
[0007] A wireless positioning method based on selective iterative refinement is realized in a wireless network environment composed of multiple base stations with known geographical positions, specifically comprising the following steps:
[0008] Step 1, an antenna array deployed on the target to be positioned receives wireless signals transmitted from at least three base stations whose positions are known; for each base station, an initial angle of arrival of the signal from the base station is estimated by using a super-resolution direction finding algorithm, and a signal impulse response is obtained by calculating a cross-correlation between the received signal and a locally generated positioning reference signal, and a first significant peak delay of the signal impulse response that exceeds a noise threshold is taken as a propagation delay, and an initial distance between the target to be positioned and the base station is calculated;
[0009] Step 2, the initial angle of arrival and the initial distance are subjected to spatial consistency verification; if the spatial consistency verification is passed, a positioning result of the target to be positioned is directly output; when the spatial consistency verification is not passed, Step 3 is performed.
[0010] Step 3, a global reference position is calculated according to the initial measurement values of all the base stations, and a residual error between the initial measurement value of each base station and a geometric value calculated from the global reference position is calculated; one or more base stations with the largest residual error are identified as target base stations.
[0011] Step 4, for the target base station, an initial reference position of the target to be measured is determined according to the measurement values of the non-target base stations, the distance estimation value of the reference position and the base station is used to constrain the angle search space of the current round, the angle search range of the super-resolution direction finding algorithm is limited within the angle search space, a spectral peak search is performed to obtain a refined angle of arrival estimation value; the current refined angle of arrival estimation value is used to construct a signal propagation model to filter out abnormal measurement values caused by multipath, and distance refinement is performed to obtain new angle of arrival and distance estimation values.
[0012] Step 5, the angle of arrival and distance estimation values of all the base stations are used to calculate a current target position estimation value in combination with the known positions of the base stations; it is judged whether the difference between the current target position estimation value and the target position estimation value of the last round is less than a preset convergence threshold; if yes, the current target position estimation value is output as a final positioning result; otherwise, Step 2 is returned to start the next iteration.
[0013] Further, the spatial consistency verification process of Step 2 is specifically as follows: according to the initial angle of arrival and the initial distance from different base stations, in combination with the known positions of the base stations, a plurality of estimated positions of the target to be positioned are calculated, and a dispersion degree among the estimated positions is calculated; if the dispersion degree is greater than a preset threshold, it is determined that the initial measurement values are inconsistent, the spatial consistency verification is not passed, and Step 3 is executed.
[0014] Further, the angle search space is specifically: taking the known position of the base station as the center and the distance estimation value as the radius, and setting a preset distance tolerance error, an annular search region is formed, and the angle search range of the super-resolution direction finding algorithm is limited within the angle interval corresponding to the annular region.
[0015] A wireless positioning device for implementing the above-mentioned wireless positioning method based on selective iterative refinement, characterized by comprising:
[0016] An antenna array for receiving wireless signals from a plurality of base stations;
[0017] A radio frequency front-end module connected to the antenna array for down-conversion, filtering and analog-to-digital conversion of the received signals;
[0018] A memory for storing known position coordinates of the plurality of base stations and sequence information of positioning reference signals.
[0019] A signal processing unit connected to the radio frequency front-end module and the memory, configured to perform the following operations: initial angle of arrival and distance estimation, spatial consistency check, error source identification, the selective iterative refinement process and final positioning result calculation.
[0020] A communication interface for outputting the positioning result.
[0021] The beneficial effects of the present application are:
[0022] (1) High precision and high robustness: through the iterative correction of angle and distance, effectively against non-line-of-sight and multipath effects, significantly improving the positioning accuracy and reliability in complex environments.
[0023] (2) High computational efficiency: through error source identification, only the most prominent measurement values are refined, avoiding full base station iteration, greatly improving the real-time performance of the algorithm.
[0024] (3) Easy to implement in engineering: based on standard positioning reference signals and mature signal processing technology, the technical path is clear and easy to implement on existing hardware platforms. BRIEF DESCRIPTION OF DRAWINGS
[0025] Figure 1 is the overall flowchart of the wireless positioning method provided by the embodiment of the present application;
[0026] Figure 2 is the detailed flowchart of the error source identification and selective iterative refinement process in the embodiment of the present application;
[0027] Figure 3 is the structural block diagram of the wireless positioning device provided by the embodiment of the present application. DETAILED DESCRIPTION
[0028] The present application will be described in detail below in conjunction with the drawings and embodiments. It should be noted that all the drawings of the present patent are schematic diagrams, which are intended to clearly show the structural composition and working process of the technical scheme of the present application.
[0029] In a first aspect, the present application provides a wireless positioning method based on selective iterative refinement, which is implemented in a wireless network environment consisting of a plurality of base stations with known geographical positions, and comprises the following steps:
[0030] S1: Signal reception and initial parameter estimation.
[0031] An antenna array deployed on the target to be positioned receives wireless signals transmitted from at least three base stations with known positions; for each base station, an initial angle of arrival of the signal from the base station is estimated using a super-resolution direction finding algorithm (such as the MUSIC algorithm), and at the same time, the signal impulse response is obtained by calculating the cross-correlation of the received signal and the locally generated positioning reference signal, and the first significant peak delay exceeding the noise threshold is taken as the propagation delay, and the initial distance between the target to be positioned and the base station is calculated.
[0032] S2: Spatial consistency check.
[0033] The initial angle of arrival and the initial distance are subjected to spatial consistency check. Specifically, in a positioning scenario, a plurality of estimated positions of the target to be positioned are calculated from the initial angle of arrival and the initial distance from different base stations, combined with the known positions of the base stations, and the dispersion between these estimated positions is calculated; if the dispersion is greater than a preset threshold, it is determined that there is inconsistency in the initial measurement values, the spatial consistency check fails, and the error source identification step is performed.
[0034] S3: Error source identification.
[0035] When the spatial consistency check fails, a global reference position is calculated from the initial measurement values of all base stations, and the residual error between the initial measurement value of each base station and the geometric value calculated from the global reference position is calculated; one or more base stations with the largest residual error are identified as target base stations.
[0036] S4: Selective iterative refinement process. This process is performed for the target base stations identified in S3, and includes the following sub-steps:
[0037] • S41: Angle refinement under distance constraint. For a target base station, an initial reference position of the target to be measured is determined from the measurement values of non-target base stations, and the distance estimate value of the reference position and the base station is used to constrain the angle search space of the current round. Specifically, the known position of the base station is taken as the center of a circle, the distance estimate value is taken as the radius, and a preset distance tolerance error is set to form a ring-shaped search region; the angle search range of the super-resolution direction finding algorithm is limited to the angle interval corresponding to the ring-shaped region, and spectral peak search is performed to obtain the refined angle of arrival estimate value.
[0038] • S42: Distance refinement under angle constraint. The current refined angle of arrival estimate is used to construct a signal propagation model to filter out abnormal measurements caused by multipath.
[0039] S5: Convergence judgment
[0040] The current target position estimate is calculated using the angle of arrival and distance estimates of all base stations combined with their known positions. The difference between the current target position estimate and the last target position estimate is judged. If the difference is smaller than a preset convergence threshold, the current target position estimate is output as the final positioning result. Otherwise, the next iteration is started from step S2.
[0041] In a second aspect, the present application provides a wireless positioning device for implementing the above method, comprising:
[0042] (1) Antenna array: for receiving wireless signals from multiple base stations.
[0043] (2) RF front-end module: connected to the antenna array, for down-conversion, filtering and analog-to-digital conversion of received signals.
[0044] (3) Memory: for storing known position coordinates of the multiple base stations and sequence information of positioning reference signals.
[0045] (4) Signal processing unit: connected to the RF front-end module and the memory, configured to perform the following operations: initial angle of arrival and distance estimation, spatial consistency check, error source identification, selective iterative refinement process and final positioning result calculation.
[0046] (5) Communication interface: for outputting the positioning result.
[0047] In the embodiment part, specific data is selected as an exemplary value to better illustrate the implementation process and effect of the present application.
[0048] Embodiment 1:
[0049] Referring to Figure 1 and Figure 2 , a wireless positioning method based on selective iterative refinement of the present embodiment is run in a network composed of three 5G base stations (BS1, BS2, BS3) with known position information, including the following steps:
[0050] Step 1: The target device receives downlink positioning reference signals (5G NR-PRS) from the three base stations through the uniform linear antenna array on it.
[0051] Step two: For each base station signal, use the MUSIC algorithm to search for spectral peaks in the global space, calculate the initial angle of arrival θ0. At the same time, correlate the received PRS with the locally generated PRS to get the global channel impulse response (CIR), and take the time delay of the first significant peak as the initial propagation time delay, and calculate the initial distance d0.
[0052] Step three: Spatial consistency check: According to the three groups of (θ0, d0) data, combined with the known position coordinates of the base station, calculate three estimated positions by the method of triangulation. Calculate the maximum distance of the centroid of these three positions and each position. The maximum distance is 25 meters, which exceeds the threshold of 15 meters, so the check fails.
[0053] Step four: Use the least squares method to calculate a total reference position P initial from the three sets of measurement values; calculate the residual distances of each base station, where the distance residual of BS1 is 18 meters, the distance residual of BS1 is 5 meters, and the distance residual of BS1 is 4 meters; identify BS1 as the most problematic target base station.
[0054] Step five: Selective iterative refinement (for BS1).
[0055] (1) Angle refinement: Take the known coordinates (0, 0) of BS1 as the center and the distance of 120 meters between the non-target base station and the measured target position as the radius, set the error tolerance to ±20 meters, and form a ring-shaped area. Restrict the search range of the MUSIC algorithm to the corresponding angle interval in this area for fine search, and find the accurate angle θ1(BS1).
[0056] (2) Distance refinement: Use the current refined angle estimate θ1(BS1) to construct a signal propagation model to calculate the refined distance d1(BS1) = 105 meters.
[0057] Step six: Use the refined (θ1, d1) and the initial measurement values of BS2 and BS3 to recalculate the current position P1.
[0058] Step seven: Determine if the position converges. If |P1-P initial | < 0.1 meters, output the final result (step six), otherwise return to step three for the next round of error source identification and iteration.
[0059] Example 2:
[0060] Referring to Figure 3 , this embodiment provides a wireless positioning device, comprising:
[0061] Antenna array 401, using a 4-element uniform linear array.
[0062] RF front end 402, responsible for signal amplification, mixing and sampling.
[0063] Memory 403, stores the known position coordinates and PRS sequence information of all base stations in the network.
[0064] Signal processing unit 404, which can be implemented by FPGA or DSP, integrates:
[0065] (1) Initial estimation module 4041: running MUSIC algorithm and cross-correlation processing.
[0066] (2) Verification and identification module 4042: based on the known coordinates of the base station to conduct spatial consistency test and error source identification.
[0067] (3) Selective iterative refinement module 4043: core processing unit.
[0068] (4) Positioning solution module 4044: performs the final position calculation
[0069] Communication interface 405, for outputting the positioning coordinates.
[0070] The above is only a specific embodiment of the present application, but the protection scope of the present application is not limited to this. Any skilled person in the art can easily think of changes and replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application.
Claims
1. A wireless positioning method based on selective iterative refinement, characterized in that, This is implemented in a wireless network environment consisting of multiple geographically known base stations, and specifically includes the following steps: Step 1: The antenna array deployed on the target to be located receives wireless signals transmitted from at least three base stations with known locations. For each base station, the initial angle of arrival of the signal from that base station is estimated using a super-resolution direction finding algorithm. At the same time, the signal impulse response is obtained by calculating the cross-correlation between the received signal and the locally generated positioning reference signal. The delay of the first significant peak exceeding the noise threshold is taken as the propagation delay, and the initial distance between the target to be located and the base station is calculated. Step 2: Perform a spatial consistency check on the initial angle of arrival and initial distance; if the spatial consistency check passes, directly output the positioning result of the target to be located; if the spatial consistency check fails, proceed to step 3. Step 3: Calculate an overall reference position based on the initial measurements of all base stations, and calculate the residual between the initial measurements of each base station and the geometric value calculated from the overall reference position. Identify one or more base stations with the largest residuals as target base stations; Step 4: For the target base station, determine an initial reference position of the target to be measured based on the measurement values of the non-target base station. Use the distance estimate between this reference position and the base station to constrain the angle search space of the current round, limiting the angle search range of the super-resolution direction finding algorithm to the angle search space. Perform spectral peak search to obtain a refined angle of arrival estimate. Use the current refined angle of arrival estimate to construct a signal propagation model, filter out abnormal measurement values caused by multipath, and refine the distance to obtain new angle of arrival and distance estimates. Step 5: Using the angle of arrival and distance estimates of all base stations, combined with their known locations, calculate the current target location estimate; determine whether the difference between the current target location estimate and the target location estimate from the previous round is less than a preset convergence threshold; if it is less, output the current target location estimate as the final positioning result. Otherwise, return to step 2 and begin the next iteration.
2. The wireless positioning method based on selective iterative refinement according to claim 1, characterized in that, The spatial consistency verification process in step 2 is as follows: Based on the initial angle of arrival and initial distance from different base stations, and combined with the known positions of the base stations, multiple estimated positions of the target to be located are calculated, and the dispersion between these estimated positions is calculated. If the dispersion is greater than the preset threshold, it is determined that the initial measurement value is inconsistent, the spatial consistency check fails, and step 3 is executed.
3. The wireless positioning method based on selective iterative refinement according to claim 1, characterized in that, The angle search space is specifically defined as follows: a circular search area is formed with the known location of the base station as the center, the estimated distance value as the radius, and a preset distance tolerance error is set. The angle search range of the super-resolution direction finding algorithm is limited to the angle interval corresponding to this circular area.
4. A wireless positioning device for implementing a wireless positioning method based on selective iterative refinement as described in any one of claims 1 to 3, characterized in that, include: Antenna array, used to receive wireless signals from multiple base stations; The radio frequency front-end module, connected to the antenna array, is used for down-conversion, filtering, and analog-to-digital conversion of the received signal; The memory is used to store the known location coordinates and sequence information of the positioning reference signals of the plurality of base stations. The signal processing unit, connected to the radio frequency front-end module and the memory, is configured to perform the following operations: initial angle of arrival and distance estimation, spatial consistency verification, error source identification, the selective iterative refinement process, and the calculation of the final positioning result. The communication interface is used to output the positioning results.