A non-line-of-sight identification and mitigation method suitable for indoor positioning

Abstract

The present application relates to a kind of non-line-of-sight identification and mitigation method suitable for indoor positioning, the multipath effect caused by non-line-of-sight propagation can cause the proportion of first arrival path signal power to total received signal power to drop, according to the grouping of reference signal autocorrelation function characteristics, extract first arrival path signal and multipath signal from cross-correlation function, calculate the first arrival path signal power and total received signal power, judge the non-line-of-sight state of base station by the ratio between them;Its mitigation method, under the premise of existing redundant base station, use base station position information and signal arrival time to estimate multiple positioning coordinates and calculate its aggregation degree;When non-line-of-sight exists, reduce the signal arrival time of non-line-of-sight base station, recalculate the positioning coordinates and aggregation degree;Iterative search result that minimizes aggregation degree as non-line-of-sight mitigation corrected positioning result is found.The present application can reduce the detection error caused by non-line-of-sight, improve positioning accuracy and robustness, while having the characteristics of small amount of calculation, suitable for real-time positioning.

Description

Technical Field

[0001] This invention relates to the field of positioning, and more particularly to a non-line-of-sight identification and mitigation method suitable for indoor positioning. Background Technology

[0002] With the advancement of technology, navigation technology has developed rapidly, and location-based services are gradually entering people's daily lives. The Global Positioning System (GPS) and the BeiDou Navigation Satellite System (BDS) have brought great convenience to human life. However, satellite signals suffer severe attenuation and multipath effects due to obstruction by ground buildings, making GPS and BDS systems unusable indoors. As the demand for indoor positioning grows stronger, more and more scholars are investing in research on indoor positioning technology. Indoor positioning schemes based on various technologies such as Bluetooth, Wi-Fi, geomagnetism, infrared, RFID, and UWB have been proposed, and positioning accuracy has gradually improved. However, positioning systems still face some technical challenges in practice, one of which is non-line-of-sight (NLOS) signal propagation, especially for high-precision indoor positioning technologies such as audio and UWB, where NLOS effects are particularly pronounced.

[0003] In NLOS environments, the direct signal between the transmitter and receiver is blocked by obstacles, causing additional signal propagation delays and resulting in inaccurate distance estimations, significantly reducing the system's positioning accuracy. Therefore, the identification and mitigation of non-line-of-sight (NLS) base station signals has become a significant challenge for indoor positioning technology. Traditional methods for NLS signal identification include channel feature analysis and machine learning (ML). Channel feature analysis is dependent on the positioning environment and requires prior knowledge of that environment, resulting in low universality. Machine learning methods identify NLS signals by extracting features from the signal and training a NLS classifier with a large amount of data. The accuracy of the classifier is related to the extracted features and the training data, leading to significant data and computational demands.

[0004] There are three main methods for mitigating non-line-of-sight (NLS) signal errors: estimating NLS errors based on the propagation characteristics of the channel and scattering model, which requires prior knowledge of the positioning environment; mitigating NLS errors based on weighted methods, which is effective for mild NLS situations but not for severe NLS; and mitigating NLS errors based on sparse feature recognition using NLOS, which requires a large number of base stations and involves a large amount of computation. Summary of the Invention

[0005] The technical problem solved by this invention is to reduce the impact of non-line-of-sight (NFS) signal propagation in positioning systems. It provides a non-line-of-sight (NFS) identification and mitigation method suitable for indoor positioning. The identification method has distinct characteristics, a high recognition rate, is independent of the positioning environment, has good universality, and does not require extensive data training, thus reducing computational and data acquisition costs. The mitigation method is highly universal, providing excellent mitigation effects for both mild and severe NFS, improving positioning accuracy and robustness, and has low requirements for the number of positioning base stations.

[0006] The principle of this invention is as follows: The multipath effect caused by non-line-of-sight (NLS) signal propagation leads to a decrease in the power of the first arrival path signal and an increase in the power of the multipath signal, thereby reducing the proportion of the first arrival path signal power to the total received signal power. The cross-correlation function between the positioning tag's received signal and the reference signal can be regarded as the superposition of the reference signal's autocorrelation function after amplitude modulation and time delay. Therefore, the first arrival path signal and multipath signal can be extracted in groups based on the autocorrelation characteristics of the reference signal, and the ratio of the first arrival path signal power to the total received signal power can be calculated. This ratio is compared with a set threshold to determine the non-line-of-sight status of the base station. For non-line-of-sight base stations, the time of arrival (TOA) is adjusted, and the aggregation degree of multiple positioning coordinates is used as a judgment condition. The signal arrival time and positioning result with the highest aggregation degree are searched as the corrected result, thereby mitigating the positioning error caused by non-line-of-sight signal propagation.

[0007] The technical solution of this invention is a non-line-of-sight identification and mitigation method suitable for indoor positioning. It utilizes the multipath effect caused by non-line-of-sight signal propagation, which leads to a decrease in the power of the first arrival path signal and an increase in the power of the multipath signal, thereby causing a decrease in the proportion of the first arrival path signal power to the total received signal power. This ratio is compared with a set threshold to determine whether the base station is in a non-line-of-sight state, thus obtaining the non-line-of-sight base station.

[0008] The base station transmits a reference signal. The cross-correlation function between the received signal and the reference signal is considered as a superposition of the autocorrelation function of the reference signal after amplitude modulation and time delay. Based on the waveform characteristics of the autocorrelation function of the reference signal, the cross-correlation functions of the received signal and the reference signal are grouped. The group that first appears and matches the autocorrelation waveform characteristics is searched, and this group is taken as the group containing the first arrival path. Thus, the signal power of the first arrival path and the multipath signal power are calculated. Finally, the ratio of the signal power of the first arrival path to the total received signal power is obtained.

[0009] The method specifically includes the following steps:

[0010] Step 1: Extract the effective data segment r(t) containing the base station's transmitted reference signal s(t) from the received signal.

[0011] The signal interception method is as follows:

[0012] The received signal is divided into frames of data. Assuming the frame length is FL and the frame sliding step size is SL, the cosine value of the i-th frame of data is defined as follows:

[0013]

[0014] Where “·” is the dot product operator, “||||” is the modulo operator, and F 1×M [i] represents the vector of the i-th frame, constructed as follows: divide the data within the frame into M sub-segments, and then use the sum of the absolute values ​​of the sub-segments as the elements of the vector. It is a vector of all 1s with the same dimension as the frame vector. When the amplitude change of the data in the i-th frame is very small, that is, it does not contain a valid data segment, the cosine value is close to 1. When the frame slides to a point where the amplitude of the data changes significantly, that is, when it contains a valid data segment, the cosine value suddenly drops, indicating that a valid signal has been detected. From this point, a valid data segment r(t) can be appropriately truncated according to the length of the original signal. r(t) can be regarded as a linear superposition of the original signal after different path delays, that is:

[0015]

[0016] Where l is the total number of paths, ρ i t represents the degree of decay along different paths. i Let n(t) represent the delay of different paths, and n(t) represent the noise.

[0017] Step 2: Calculate the autocorrelation function of the reference signal The cross-correlation function between the valid data segment and the original signal in step 1.

[0018] Step 3: Treat the cross-correlation function between the received signal and the reference signal calculated in Step 2 as a superposition of the autocorrelation function of the reference signal after amplitude modulation and time delay, and extract the waveform features and waveform width T from the reference signal autocorrelation function. s Based on the characteristics of the autocorrelation function, the position t of the signal arrival time (TOA) in the cross-correlation function is detected. TOA And find the maximum value MAX of the cross-correlation function, and set T0 = [t TOA -T s / 2,t TOA +T s / 2] is the group containing the first arrival path, with T s To move the step size backward, that is When T i Maximum value within a group i Stop moving when ωmax is less than ωMAX.

[0019] T sThe determination method is as follows:

[0020] A. Extract the position t of the maximum value of the autocorrelation function. max And calculate the absolute value of the autocorrelation function |R r SUM is the sum of (τ)|.

[0021] B. Calculate {t} max -τ,t max The absolute value of the autocorrelation function within the group +τ} |R r The sum of (τ)| is SUM0, and τ gradually increases from 0, let T s =τ min , where represents the minimum τ value that makes SUM0 / SUM greater than 0.9;

[0022] Step 4: Take the average value of group T0 as the first arrival path power W origin , sequentially grouping T i (i = 1, 2, ..., N) Take the average and sum them to get the total power W of the multipath sharding caused by non-line-of-sight. multi Define γ = W origin / (W origin +W multi γ represents the ratio of the first arriving path signal power to the total received signal power, serving as the basis for determining the non-line-of-sight (NOS) level of the base station. When γ < α, the base station is determined to be NOS; otherwise, it is line-of-sight. α is the threshold for determining whether it is NOS, and γ is the subsequent weighted calculation target (x). t ,y t When providing weighting coefficients, it can reduce positioning errors caused by non-line-of-sight to some extent.

[0023] The non-line-of-sight mitigation method of the present invention uses redundant base stations to calculate multiple positioning coordinates and their clustering degree, adjusts the signal arrival time or signal arrival time difference including the non-line-of-sight base stations, and searches for the result that minimizes the clustering degree of the positioning coordinates as the corrected result, thereby mitigating the positioning error caused by non-line-of-sight signal propagation.

[0024] The method specifically includes the following steps:

[0025] Step 1: Utilize base station location information (x i ,y i (i = 1, 2, 3...K) and signal arrival time TOA i (i = 1, 2, 3...K) or signal arrival time difference TDOA ij (i≤ji,j=1,2,3……K), where K is the number of base stations, TOA i TDOA represents the signal arrival time from the i-th base station to the location tag. ijThis represents the difference in signal arrival time between the i-th base station and the j-th base station to the positioning tag. This method requires redundant base stations, therefore K≥4. Based on the three-base station positioning method, multiple positioning coordinates (x, y, j) are calculated. ti ,y ti (i = 1, 2, 3...S) and calculate its degree of polymerization δ, where The formula for calculating the degree of polymerization δ is as follows:

[0026]

[0027]

[0028] Among them, D ij Represents the positioning coordinates (x) ti ,y ti ) and (x tj ,y tj The distance between them;

[0029] Step 2: When the i-th base station is in a non-line-of-sight state, the signal arrival time (TOA) of the non-line-of-sight base station is taken into account. i The value is too large compared to the actual value, so it is reduced to TOA'. i =TOA i -θ i ε (0≤θ i ≤1), where θ i The ratio γ of the first arrival path power to the total received signal power of the i-th base station i Relatedly, for line-of-sight base stations θ i =0, non-line-of-sight base station γ i The smaller θ i The larger the value, the more likely a new signal arrival time set TOA' is obtained. i (i = 1, 2, 3...K) or the set of signal arrival time differences TDOA' ij After (i≤ji,j=1,2,3……K), the positioning result (x') is recalculated. ti ,y' ti i = 1, 2, 3...S and aggregation degree δ';

[0030] Step 3: Adjust ε iteratively (from ε) min The INC value gradually increases to ε using a fixed increment. max The search finds the set of signal arrival time differences that make argmin{δ} and the positioning result as the corrected positioning result, thereby mitigating the positioning error caused by non-line-of-sight signal propagation.

[0031] The advantages of this invention compared with the prior art are as follows: the identification method proposed in this invention has obvious features, high recognition rate, does not depend on the positioning environment, has good universality and does not require a large amount of data for training, thus reducing the cost of calculation and data collection; at the same time, the non-line-of-sight mitigation method proposed in this invention for indoor positioning has strong universality, has a good mitigation effect on both mild and severe non-line-of-sight situations, improves positioning accuracy and robustness, and does not require a large number of positioning base stations. Attached Figure Description

[0032] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0033] Figure 1 This is a flowchart of a non-line-of-sight identification and mitigation method for indoor positioning, as described in this embodiment of the invention.

[0034] Figure 2 This is a diagram illustrating non-line-of-sight signal propagation in an embodiment of the present invention;

[0035] Figure 3 This is a time-domain waveform diagram of the autocorrelation (left) and cross-correlation of the reference signal and the received signal in an embodiment of the present invention;

[0036] Figure 4 This is a schematic diagram of four base station positioning in an embodiment of the present invention; Detailed Implementation

[0037] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of the present invention.

[0038] The steps of this invention are as follows: N positioning base stations of the positioning system are deployed in the positioning area according to the specific environment. In this embodiment, N=4. The positioning tag can be a mobile phone or other terminal device. The specific non-line-of-sight base station is identified by the signal received by the positioning tag. The signal arrival time of the non-line-of-sight base station or the signal arrival time difference containing the non-line-of-sight base station is adjusted. The reliability of the adjusted signal arrival time or arrival time difference is judged according to the clustering degree of multiple target positioning coordinates. The most reliable signal arrival time or arrival time difference is searched to correct the positioning result and reduce the positioning error caused by non-line-of-sight.

[0039] like Figure 1As shown, a non-line-of-sight identification and mitigation method suitable for indoor positioning uses the multipath effect caused by non-line-of-sight signal propagation, which leads to a significant decrease in the proportion of the first arrival path signal power to the total received signal power, as an identification feature. Then, a threshold comparison is used to determine whether the base station is in a non-line-of-sight state.

[0040] like Figure 2 As shown, assuming the base station transmits a linear frequency modulated signal (Chirp signal) in the 18-15kHz band, the cross-correlation function between the received signal and the reference Chirp signal can be considered as the superposition of the autocorrelation function of the reference Chirp signal after amplitude modulation and time delay. Figure 3 As shown, the autocorrelation function R of the reference Chirp signal s (τ) clusters within a relatively short timeframe (approximately 2 ms), due to the influence of non-line-of-sight (NLOS), R r (τ) exhibits the characteristic of grouping and clustering, that is, clustering in multiple shorter time intervals, which can be utilized using R. r (τ) The characteristic of intragroup clustering and sufficiently large intragroup intervals affects R. r (τ) Grouping is performed, and the first waveform segment that meets the autocorrelation waveform characteristics is searched. This segment is taken as the group containing the first arrival path, thereby quantifying the power of the first arrival path and the overall signal reception function power. The method specifically includes the following steps:

[0041] Step 1: Extract the effective data segment r(t) containing the original signal s(t) transmitted by the base station from the received signal. The signal extraction method is as follows:

[0042] The received signal is divided into data segments, with frame length FL and frame sliding step size SL. The cosine value of the i-th frame of data is defined as follows:

[0043]

[0044] Where “·” is the dot product operator, “||||” is the modulo operator, and F 1×M [i] represents the vector of the i-th frame, constructed as follows: divide the data within the frame into M sub-segments, and then use the sum of the absolute values ​​of the sub-segments as the elements of the vector. It is a vector of all 1s with the same dimension as the frame vector. When the amplitude change of the data in the i-th frame is small, the cosine value is close to 1. When the frame slides to a point where the amplitude of the data changes significantly, the cosine value suddenly drops, indicating that a valid signal has been detected. From this point, a valid data segment r(t) can be appropriately truncated according to the length of the original signal. r(t) can be regarded as a linear superposition of the original signal after different path delays, that is:

[0045]

[0046] Where l is the total number of paths, ρ i t represents the degree of decay along different paths. i Let n(t) represent the delay of different paths, and n(t) represent the noise.

[0047] Step 2: Calculate the autocorrelation function of the original signal Cross-correlation function between valid data segment and original signal

[0048] Step 3: Treat the cross-correlation function of the received signal and the reference signal as a superposition of the autocorrelation function of the reference signal after amplitude modulation and time delay, and extract the waveform features and waveform width T from the autocorrelation function of the reference signal. s T s The determination method is as follows:

[0049] A. Extract the position t of the maximum value of the autocorrelation function. max And calculate the absolute value of the autocorrelation function |R r SUM is the sum of (τ)|.

[0050] B. Calculate {t} max -τ,t max The absolute value of the autocorrelation function within the group +τ} |R r The sum of (τ)| is SUM0, and τ gradually increases from 0, let T s =τ min , where represents the minimum τ value that makes SUM0 / SUM greater than 0.9;

[0051] Detecting the position t of the signal arrival time (TOA) in the cross-correlation function based on the characteristics of the autocorrelation function. TOA And find the maximum value (MAX) of the cross-correlation function. The TOA detection method is as follows:

[0052] A. Based on cluster analysis, |R r Grouping the time series of (τ)|

[0053]

[0054] diffTS r (q)=TS r (q+1)-TS r (q) q=1,2,…,N-1 (2)

[0055] The time series can be constructed using equation (1) above, where MAX R(τ) The maximum peak value of the received signal. Using a scaling factor, gradually increasing from 0.02 to 1, with a step size of 0.01, the time series TS can be obtained. r(n). Then, using equation (2) above, the time series is differenced to calculate diffTS. r (q) reflects TS r The degree of aggregation of (n) is compared with diffTS. r (q) and clustering parameter TH doa Time difference series can be grouped, TH doa The value should be slightly larger than all values ​​in the time difference sequence of the autocorrelation signal. The grouping method is as follows: Let τ = TS r (1) , so that q gradually increases from 1 until TS. r (q) is greater than TH doa TS r (1) to TS r (q) forms a group, and the starting point of the next group is TS. r (q+1), and so on, until P groups are obtained.

[0056] B. Determine the group to which TOA belongs. Estimate the group to which TOA belongs using the ratio of maximum signal value to noise, as shown in the following formula:

[0057] MNR r (i)=|R r (τ' i ))| / mean(|R r (0:τ i )|) i=1,2,…,P

[0058] Among them, MNR r (i) represents the ratio of the maximum value to the noise in the i-th group, |R r (τ' i ))| represents the maximum value of the received signal in the i-th group, mean(|R) r (0:τ i )|) represents the range from 0 to τ i The average value of the signal at time TH. Then, the first value greater than the threshold parameter TH. mnr The group containing the value is considered to be the group containing TOA.

[0059] C. Determining TOA within a group

[0060]

[0061] τ TOA The estimation formula is shown in the above equation, where τ a τ' is the starting time of the grouping. a This is the end time. TH span The value is determined by the actual experimental conditions.

[0062] Let T0 = [t TOA -Ts / 2,t TOA +T s / 2] is the group containing the first arrival path, with T s To move the step size backward, that is When T i Maximum value within a group i Stop moving when ωmax is less than ωMAX.

[0063] Step 4: Take the average value of group T0 as the first arrival path power W origin , sequentially grouping T i (i = 1, 2, ..., N) Take the average and sum them to get the total power W of the multipath sharding caused by non-line-of-sight. multi Define γ = W origin / (W origin +W multi γ represents the ratio of the first arrival path power to the total received signal power, used as a basis for determining the non-line-of-sight (NOS) level of the base station. When γ < α, the base station is determined to be NOS; otherwise, it is line-of-sight. α is the threshold for determining whether it is NOS, and γ can be used to calculate the target (x) in subsequent weighted calculations. t ,y t When providing weighting coefficients, it can reduce positioning errors caused by non-line-of-sight to some extent.

[0064] like Figure 4 As shown, taking a 4-base station positioning system as an example, the non-line-of-sight identification and mitigation method in the embodiments of the present invention will be further described in detail. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0065] like Figure 1 and Figure 4 As shown, a non-line-of-sight (NOS) mitigation method for indoor positioning involves deploying four positioning base stations in the positioning area, calculating multiple positioning coordinates and their clustering using redundant base stations, adjusting the signal arrival time or signal arrival time difference (SATD) of base stations containing NOS, and searching for the result that minimizes the clustering of positioning coordinates as the corrected result. This mitigates positioning errors caused by NOS propagation. The method specifically includes the following steps:

[0066] Step 1: Utilize base station location information (x i ,y i (i = 1, 2, 3, 4) and signal arrival time TOA i (i = 1, 2, 3, 4) or the time difference of arrival (TDOA) ij (i≤ji,j=1,2,3,4), where K is the number of base stations, TOA i TDOA represents the signal arrival time from the i-th base station to the location tag.ij This represents the difference in signal arrival time between the i-th base station and the j-th base station to the positioning tag. Based on the three-base station positioning method, the coordinates (x, y) of multiple target positioning points are calculated. ti ,y ti (i = 1, 2, 3) and calculate its degree of polymerization δ, where The formula for calculating the degree of polymerization δ is as follows:

[0067]

[0068]

[0069] Among them, D ij Represents the positioning coordinates (x) ti ,y ti ) and (x tj ,y tj The distance between them;

[0070] Step 2: When non-line-of-sight exists, it will cause the signal arrival time (TOA) of the non-line-of-sight base station to be affected. i The actual TOA is larger, reducing the signal arrival time (TOA') of non-line-of-sight base stations. i =TOA i -θ i ε (0≤θ i ≤1), where θ i The ratio γ of the first arrival path power to the total received signal power at the i-th base station i Inversely proportional to θ i =μ / γ i Where μ is a constant coefficient, a new set of signal arrival times TOA' is obtained. i (i = 1, 2, 3, 4) or the set of signal arrival time differences TDOA' ij (i≤ji,j=1,2,3,4), and recalculate the positioning result (x') ti ,y' ti i = 1, 2, 3 and aggregation degree δ';

[0071] Step 3: Adjust ε iteratively, ε from ε min The INC value gradually increases to ε using a fixed increment. max Let ε min =0, INC=0.1, ε max =min{TOA m The search is performed on m = 1, 2, 3, ..., K. The set of signal arrival time differences that make argmin{δ} equal to the positioning result is used as the corrected result, thereby mitigating the positioning error caused by non-line-of-sight.

[0072] In summary, this invention provides a non-line-of-sight (NOS) identification and mitigation method for indoor positioning. It treats the cross-correlation function between the received signal of the positioning target and the reference signal transmitted by the base station as a superposition of multiple autocorrelation functions of the reference signals after amplitude modulation and time delay. Based on the characteristics of the autocorrelation function of the reference signals, it extracts the first arrival path signal and multipath signals in groups. Considering that the multipath effect caused by NOS propagation leads to a significant decrease in the proportion of the first arrival path signal power to the total received signal power, the power of the first arrival path signal and the total received signal power are quantitatively calculated. The NOS state of the base station can be determined by the ratio between the two. The mitigation method includes: Step 1: Under the premise of redundant base stations, multiple positioning coordinates are estimated using base station location information and signal arrival time, and their clustering degree is calculated; Step 2: When NOS exists, the signal arrival time of the NOS base station is reduced and adjusted, and the positioning coordinates and clustering degree are recalculated; Step 3: The result that minimizes the clustering degree is iteratively searched and used as the positioning result after NOS mitigation correction. The method proposed in this invention can reduce the detection error caused by NOS, improve positioning accuracy and robustness, and has the characteristics of low computational load, making it suitable for real-time positioning.

[0073] The above description is only one specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the protection scope of the present invention.

Claims

1. A non-line-of-sight recognition and mitigation method suitable for indoor positioning, characterized in that: Includes the following steps: For each of the four base stations of the positioning system , The base station transmits a reference signal. The positioning tag obtains the corresponding received signal. Calculate the autocorrelation function of the reference signal. and the received signal With the reference signal cross-correlation function According to the autocorrelation function The waveform characteristics from the cross-correlation function The arrival time of the signal for determining the first arrival path is determined in the middle. and the maximum value of the cross-correlation function. ; Determine the main peak width of the autocorrelation function of the reference signal. And construct the first arrival path group. by Construct subsequent groups sequentially based on step size. And in subsequent groups Maximum value within Stop constructing subsequent packets; retrieve the first arrival path packet. The mean of the absolute values ​​of the cross-correlation functions within the range is used as the first arrival path power. Take each subsequent group The mean of the absolute values ​​of the cross-correlation functions within the multipath is taken and summed to obtain the total power of the multipath shunt. ,calculate And in Time to determine the first The first base station is a non-line-of-sight base station; otherwise, the second base station is determined to be a non-line One of the base stations is a line-of-sight base station, among which... This is the grouping stopping threshold. This is the threshold for non-line-of-sight determination; Using the location information of the four base stations and their corresponding signal arrival times or signal arrival time difference , and Calculate the location coordinates of the location tags for each of the three base station combinations. ,in and The clustering degree of the positioning coordinates is calculated according to the following formula: For the first [unit / item] identified as a line-of-sight base station One base station, making For the first base station identified as a non-line-of-sight base station One base station, making in These are constant coefficients; for each candidate correction value Using the same candidate correction amount The signal arrival time of all non-line-of-sight base stations is determined according to their respective... Simultaneous reduction yields... And based on the corrected signal arrival time Obtain the corrected signal arrival time difference Recalculate the positioning coordinates and clustering degree corresponding to all three base station combinations. ; Make the candidate correction amount from With a fixed increment Gradually increase to Determine the degree of aggregation. Minimum correction amount and will The corresponding corrected signal arrival time or signal arrival time difference and the positioning result are used as the result after non-line-of-sight mitigation.

2. The non-line-of-sight recognition and mitigation method for indoor positioning according to claim 1, characterized in that, Extract the reference signal from the received signal. valid data segment include: The received signal is divided into data frames, with a frame length of [missing information]. The frame sliding step size is , No. The cosine value of the frame data is: in, Indicates the first The frame vector divides the intra-frame data into... The vector is composed of sub-segments, and the sum of the absolute values ​​of each sub-segment is used as an element of the vector. It is a vector of all 1s with the same dimension as the vector; when the... When the frame data does not contain a valid data segment, the data amplitude changes little and the cosine value is close to 1. When the frame slides to a position containing a valid data segment, the data amplitude changes, causing the cosine value to decrease, and from that position, the valid data segment is truncated according to the reference signal length. The valid data segment satisfy: in, The total number of paths, For the first The attenuation coefficient of the path, For the first The delay of the path, It is noise.

3. The non-line-of-sight recognition and mitigation method for indoor positioning according to claim 1, characterized in that, The arrival time of the signal is determined from the cross-correlation function based on the waveform characteristics of the autocorrelation function of the reference signal. include: Let the scaling factor The value gradually increases from 0.02 to 1 in increments of 0.01, following the... Constructing time series And calculate the difference sequence of adjacent time points. The difference sequence is compared with a clustering threshold determined based on the autocorrelation function of the reference signal. The comparisons are made to divide the time series into multiple candidate groups. ; Regarding the first Calculate candidate groups The first one will be satisfied. The candidate group is determined as the candidate group containing the first arrival path, and according to Determine the arrival time of the signal .

4. The non-line-of-sight recognition and mitigation method for indoor positioning according to claim 1, characterized in that, The main peak width of the autocorrelation function of the reference signal The determination includes: extracting the location of the maximum value of the autocorrelation function of the reference signal. Calculate the sum of the absolute values ​​of the autocorrelation functions of the reference signal. ;from Gradually increase the interval parameter to both sides Calculate the interval The sum of the absolute values ​​of the autocorrelation functions of the reference signal mentioned above , will make The smallest The value is determined as .

5. The non-line-of-sight recognition and mitigation method for indoor positioning according to claim 1, characterized in that, ,and .

Images