CN106950528B - direction-of-arrival estimation method based on linear frequency modulation signals - Google Patents

Abstract

The invention discloses a method for estimating the direction of arrival based on a linear frequency modulation signal. The distance between two antennas installed at the anchor node is d, and the two antennas are controlled by the same oscillator to simultaneously transmit signals with different initial frequencies and the same frequency modulation rate. LFM signal; the tag node performs A/D sampling on the received signal received by the antenna, performs signal down-conversion processing, and then performs NDFT transformation. In the line-of-sight environment, the direct signal reaching the tag node through two antennas will have a delay difference. After NDFT Estimate the period of the signal after transformation to estimate AOA; when there is multipath effect, there are different delays in the direct signals arriving at the tag node through the two antennas, and the received signal period corresponding to the delay difference of the direct signals is calculated. For the frequency of interest Carry out more NDFT at the point, find the frequency corresponding to the maximum peak value, and estimate the signal AOA. The invention has the advantages of anti-multipath effect, high dynamics, high precision, low hardware complexity and the like.

Description

A Direction of Arrival Estimation Method Based on Chirp Signal

technical field

The invention belongs to the field of wireless sensor technology and electronic communication system, and relates to a direction-of-arrival estimation method based on linear frequency modulation signals.

Background technique

Realizing the positioning of the target by measuring the angle of arrival (AOA) of the RF (radio frequency) signal is a working method often used by the RF positioning system. In recent years, with the popularization of wireless networks and the development of indoor positioning technology, Location Based Service (LBS) has received more and more attention; for example, in the fields of emergency assistance, medical care, and personalized information transmission Shows great vitality. The purpose of wireless sensor network node positioning is to measure relevant parameters of received radio waves, such as time of arrival (TOA), time difference of arrival (TDOA), angle of arrival (AOA) or signal field strength (RSS), etc. The observation formula of these parameters is a function of the target position. The observation station estimates the position of the signal source through an appropriate positioning algorithm, and then gives the absolute or relative coordinates of each node in the plane or space according to a specific algorithm.

Among the more common positioning schemes at present, the positioning scheme based on TOA/TDOA is relatively common, and the positioning scheme based on TOA/AOA has gradually attracted people's attention, and there have been a lot of researches on the positioning scheme based on AOA. Therefore, the technology of RF signal AOA estimation is also very important. With the development of RF technology and the continuous emergence of new measurement technologies, angle measurement technology has also been continuously improved and developed. Generally speaking, AOA estimation problems can be solved using traditional methods or Take a modern approach. The traditional method uses a mechanical rotating directional antenna to achieve AOA estimation, which is susceptible to interference and has a relatively low angular resolution; the modern method uses an array antenna, which can overcome the shortcomings of the traditional method, but requires a large number of antennas so the cost is relatively expensive .

To sum up, the existing AOA estimation technology has the following defects: the method based on the mechanical rotating directional antenna has relatively low resolution and the angle estimation accuracy is not high; due to the serious multipath effect in the complex environment, the acquired AOA There is a large error, and the direct delay difference of the signal cannot be accurately obtained; the method based on the array antenna requires more antennas, and has higher cost and power consumption.

Contents of the invention

In order to overcome the shortcomings of the existing AOA estimation technology, such as low estimation angle accuracy, large errors due to multipath effects, large number of antennas and high cost, the present invention discloses a method for estimating the direction of arrival based on chirp signals.

A direction-of-arrival estimation method based on chirp signals according to the present invention is composed of an anchor node with two omnidirectional antennas and a label node with one antenna;

Specifically, the distance between the two antennas installed on the anchor node is d, the anchor node has the function of transmitting linear frequency modulation LFM signals, and the two antennas are controlled by the same oscillator to simultaneously transmit signals with different initial frequencies and the same frequency modulation rate LFM signal.

A method for estimating the direction of arrival based on a chirp signal proposed by the present invention specifically includes the following steps:

Step 1: The anchor node is two antennas installed with a distance of d, and transmits LFM signals with different initial frequencies and the same frequency modulation rate:

Among them: the time-domain waveform of the signal transmitted by antenna A is:

The instantaneous frequency of this signal is:

f _A (t) = f _a t + βt, (2)

The time-domain waveform of the signal transmitted by antenna B is:

The instantaneous frequency of this signal is:

f _B (t) = f _b +βt, (4)

Where: f _a > f _b , and

f _a is the initial frequency of antenna A, f _b is the initial frequency of antenna B, t represents time, β represents the frequency modulation rate, Denotes the initial phase of antenna A, Indicates the initial phase of antenna B;

Step 2: Due to the small indoor range, it is considered that the transmitted signal is propagated in the near field. In the LOS (line-of-sight) environment, the tag node performs A/D sampling on the signal received by the antenna, and at the same time performs signal down-conversion processing , the sampling period is T _s , and the sampling frequency The received signal is:

s _RSS =||r(t)|| ² +w(t), (6)

Among them: w(t) represents the received white noise, and r(t) is the signal after the down-conversion of the received signal of the antenna:

Where: P and Q represent the multipath numbers of antennas A and B, τ _Ai and τ _Bi represent the attenuation and delay of antenna A and B on the i-th path, respectively, where τ _Ai and τ _Bi are the LOS direct path delay of antenna A and B (under LOS environment), and the indoor environment has obvious When there is multipath effect, go to step 5, otherwise go to step 3;

Step 3: In the LOS environment, there are different time delays in the direct signals arriving at the tag node through the two antennas, which will cause a time delay difference. At this time, P=1 and Q=1 in the formula (7), at this time, the formula (7) is substituted into the formula (6), and the received signal is sorted out to obtain:

in:

θ is the direction of arrival AOA of the signal, d is the distance between antennas A and B, and c is the speed of light;

Under the LOS environment, the received signal is a sine wave having a frequency of (f _a -f _b -β(τ _A1 -τ _B1 )), and then proceeds to step eight, and performs NDFT transformation on the received signal; then proceeds to step four;

Step 4: Find the period of the signal after performing NDFT changes on the signal Obtain the delay difference τ _A1 -τ _B1 , and then calculate the direction of arrival AOA from the formula (9) through the geometric relationship, as shown in the formula (10), turn to step 9;

Where: c represents the speed of light, and d is the distance between antennas A and B.

Step 5: When there is multipath effect, there will be different delays in the direct signal reaching the label node through two antennas. At this time, in formula (7), P and Q are not 1 at the same time. Substitute formula (7) into formula In (6), the received signal is arranged to obtain the following expression:

Where: τ _k =τ _Ai -τ _Bj represents the time delay difference between the i-th path of antenna A and the j-th path of antenna B, and Π represents a constant.

In a multipath environment, the received signal is the superposition of signals of different periods (f _a -f _b -βτ _k ) related to the multipath delay difference τ _k . Since the amplitude of the direct signal is strong, two antennas are useful for AOA estimation The direct path delay difference of the transmitted signal τ _k =τ _A1 -τ _B1 , the corresponding signal frequency is:

At this time, the frequency range of the key area is: (f _a -f _b ±βd/c), assuming that f _a -f _b =K f _s , f _s >2βd/c, the frequency range of the key area at this time is:

(K-1)f _s ～(K+1)f _s , (13)

Set the maximum number of points within these ranges, turn to step 8 for NDFT transformation, and then turn to step 6;

Step 6: Perform NDFT changes on the signal to determine whether the transformed signal peak value exceeds the specified threshold. If it does not exceed the specified threshold, go to step 7. Otherwise, search for the frequency corresponding to the position where the transformed signal peak appears. As the estimated value of f _RSS , AOA can be obtained by substituting into formula (14), and then turn to step 9;

Wherein: the distance between the antennas d<c/2βT _s .

Step 7: When the searched peak value does not exceed the predetermined threshold, it is considered that there is no direct path signal in the current environment, which is an NLOS (non-line-of-sight) environment, and the frequency range of key areas is increased, and then enter step 5 to estimate the suboptimal AOA;

Step 8: Perform NDFT transformation on the signal. The FFT transformation sampling points are evenly distributed on the unit circle, so the method based on FFT transformation cannot achieve sampling with smaller frequency intervals at the frequency of key areas; the NDFT transformation sampling points are on the unit circle. It is randomly distributed, and the frequency range of the key area is set at as many sampling points as possible, and the NDFT change of the frequency interval is as small as possible, so as to estimate the delay difference of the first path in the multipath environment. The NDFT transformation is shown in the following formula :

Among them: z ₀ , z ₁ ,…, z _N-1 are any different N points selected on the unit circle; formula (15) is written in the form of matrix (16), and the NDFT transformation can be quickly performed in the matrix form:

S=Z ^T s, (16)

Where: S＝[S _NDFT (z ₀ ),S _NDFT (z ₁ ),…,S _NDFT (z _N-1 )] ^T , (17)

s＝[ _sRSS [0], _sRSS [1],... _sRSS [N-1]] ^T , (18)

The number of sampling points in the range of (K-1)f _s ~ (K+1)f _s is greater than the number of sampling points in other places;

Step 9: Carry out Kalman filter processing on the AOA to eliminate the impact of noise on the system, further improve the estimation accuracy, and output the AOA in the direction of arrival of the signal.

Compared with the prior art, the present invention has the following beneficial effects: adopting the LFM signal system, high time resolution can be obtained by setting a large bandwidth, thereby improving the accuracy of AOA estimation; Uniform discrete Fourier transform, sampling more points in the frequency range of key areas, and sampling less points in other places, improves the frequency resolution performance by observing the fine structure of the spectrum, and more accurately estimates AOA under multipath conditions; at the same time The invention has the advantages of anti-multipath effect, high dynamics, high precision, low hardware complexity and the like.

Description of drawings

The drawings described here are used to provide further explanation of the present invention, and constitute a major part of the present invention, but do not constitute a limitation to the present invention. In the attached picture:

Fig. 1 is the block diagram of the hardware composition of the direction of arrival estimation method based on the chirp signal proposed by the present invention

Fig. 2 is the instantaneous frequency of the transmitted signal (solid line) and the received signal (dashed line) of the direction of arrival estimation method based on chirp signal of the present invention

Fig. 3 is the distribution of NDFT (figure b) sampling points on the z plane unit circle in the direction of arrival estimation method based on chirp signal of the present invention

Fig. 4 is the implementation flow chart of the direction of arrival estimation method based on chirp signal of the present invention

Detailed ways

The present invention will be described in more detail and completeness below in conjunction with the accompanying drawings and specific implementation cases. It is stated here that the described specific implementation cases are only used for an explanation of the present invention, and are not intended to limit the present invention. In addition, it should be noted that for the convenience of description, only the parts related to the present invention are shown in the drawings but not the whole content.

The present invention designs a method for estimating the direction of arrival based on a linear frequency modulation signal. Two antennas transmit LFM signals at the same time. Focus on sampling within the range to further eliminate the impact of multipath effects on the system and improve the accuracy of AOA estimation. The purpose of the present invention is to overcome the shortcomings of the existing AOA estimation technology in the multipath environment, such as low precision and large error, and design an AOA with anti-signal multipath effect, high dynamics, high precision and low hardware complexity. Estimation method.

In order to solve the above-mentioned technical problems, the present invention adopts the following technical solutions:

As shown in Figure 1, a block diagram of the hardware composition of a direction-of-arrival estimation method based on chirp signals, it can be seen from the figure that the present invention consists of an anchor node with two omnidirectional antennas and a label with one antenna Node composition; the two antennas installed by the anchor node have a fixed interval and have the function of transmitting a linear frequency modulation signal LFM, and the two antennas are controlled by an oscillator to simultaneously transmit LFM signals with different initial frequencies and the same frequency modulation rate; the The tag node performs A/D sampling on the received signal received by the antenna, and at the same time performs signal down-conversion processing, and performs NDFT transformation on the pre-processed data; due to the small indoor range, it can be considered that the transmitted signal is propagated in the near field , as shown in Figure 2; in the LOS environment, the direct signal arriving at the label node through two antennas will have different delays, resulting in a delay difference. After NDFT transformation, the period of the signal can be estimated, and the AOA can be further estimated according to the geometric relationship As shown in formula (10); when there is multipath effect, there will be different delays in the direct signals arriving at the tag node through the two antennas, and multipath signals will also produce complex delays, so the received signal is a variety of For the superposition of signals, since the amplitude of the direct signal is strong, the received signal period corresponding to the delay difference of the direct signal can be calculated according to the distance of the transmitting antenna and the sampling rate, and more NDFTs are performed on the frequency points in key areas, as shown in the figure As shown in 3, find the frequency corresponding to the maximum peak value, that is, the delay difference of the direct signal, and then accurately estimate the signal AOA, and filter the AOA to further improve the estimation accuracy; when the maximum peak value does not exceed the predetermined threshold At that time, it is considered that the current environment has no direct signal and is a non-line-of-sight (NLOS) environment. It is necessary to expand the frequency range of key areas and continue to use NDFT transformation to find the suboptimal AOA in the NLOS environment.

Furthermore, in this case, in order to let those skilled in the related art understand a method for estimating the direction of arrival based on chirp signals of the present invention, the chirp pulse signal is used to realize the function of AOA estimation, and the The overall flowchart of scheme implementation, as shown in accompanying drawing 4, the flow process of this method comprises the following steps:

SS1. Initialize various parameters: the initial frequency f _a , f _b and frequency modulation rate β of the LFM signal transmitted by antenna A and B, the repetition period T of the LFM signal and the distance d between the two antennas, the sampling period T _s of the received signal, sampling frequency The number of sampling points is N, so that f _a -f _b = K·f _s , f _s >2βd/c, and the threshold Δ of the transformed peak value is judged;

SS2. The anchor node periodically transmits LFM signals, and the label node samples the received signal of the antenna through the A/D module, and performs down-conversion operation, and saves the original data obtained by sampling;

SS3. Perform NDFT changes on the saved data, and focus on sampling the frequency range of key areas, for example: (K-1)f _s ~ (K+1)f _s , by sampling more points at these frequency points, Improve frequency resolution performance by observing the fine mechanism of the spectrum;

SS4. Search for the peak value of the transformed data and record the frequency point corresponding to the peak value Judging whether the peak value exceeds a given threshold: If it exceeds a given threshold Δ, it is considered that there is a direct signal under this path, and it is transferred to SS5; otherwise, it is considered that there is no direct signal on the current path, and it is a non-line-of-sight path (NLOS). After eliminating non-line-of-sight errors through related mechanisms, and expanding the frequency range of key areas, transfer to SS3 to find the suboptimal AOA in the current environment;

SS5. According to the formula (12) f _RSS =f _a -f _b -β(τ _A1 -τ _B1 ), estimate the direct path delay difference (τ _A1 -τ _B1 ) after the signal passes through the antennas A and B, according to the formula ( 14) Estimate the direction of arrival AOA of the signal;

SS6. Estimated Carry out filtering processing such as Kalman and particle filtering to further eliminate the impact of noise on the system and improve the accuracy of estimation.

After the above steps, the angle between the label node and the anchor node can be obtained in real time at the label node, so as to realize high-precision estimation of AOA.

The invention is low in cost and can reduce errors caused by multipath effects. After many times of actual experimental measurements, the angle measurement accuracy of the invention is high, and is especially suitable for the following scenarios: positioning of cars driving in indoor parking lots, pedestrians shopping in shopping malls, etc. In the application, the positioning system based on AOA can realize effective direction finding and positioning.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit them, and those of ordinary skill in the art should understand; it is still possible to modify the solutions of the present invention or replace some technologies. It is used to limit the scope of patent protection of the present invention, and the scope of protection of the present invention is still subject to the claims; all similar structural changes made by using the contents of the drawings or instructions of the present invention should be included in the scope of protection of the patent in the same way Inside.

Claims (1)

1. A direction-of-arrival estimation method based on a chirp signal, the device based on the method is composed of an anchor node installing two omnidirectional antennas and a tag node installing an antenna; Specifically, the distance between the two antennas installed on the anchor node is d, the anchor node has the function of transmitting linear frequency modulation LFM signals, and the two antennas are controlled by the same oscillator to simultaneously transmit signals with different initial frequencies and the same frequency modulation rate LFM signal; It is characterized in that the method specifically includes the following steps: Step 1: The anchor node is two antennas installed with a distance of d, and transmits LFM signals with different initial frequencies and the same frequency modulation rate: Among them: the time-domain waveform of the signal transmitted by antenna A is: The instantaneous frequency of this signal is: f _A (t) = f _a +βt, (2) The time-domain waveform of the signal transmitted by antenna B is: The instantaneous frequency of this signal is: f _B (t) = f _b +βt, (4) Where: f _a > f _b , and f _a is the initial frequency of antenna A, f _b is the initial frequency of antenna B, t represents time, β represents the frequency modulation rate, Denotes the initial phase of antenna A, Indicates the initial phase of antenna B; Step 2: Due to the small indoor range, it is considered that the transmitted signal is propagated in the near field. In the LOS environment, the tag node performs A/D sampling on the signal received by the antenna, and at the same time performs signal down-conversion processing. The sampling period is T _s , sampling frequency The received signal is: s _RSS =||r(t)|| ² +w(t), (6) Among them: w(t) represents the received white noise, and r(t) is the signal after the down-conversion of the received signal of the antenna: Among them: P and Q represent the number of multipaths of antennas A and B, α _Ai , α _Bi , τ _Ai , τ _Bi represent the attenuation and time delay of antennas A and B on the i-th path respectively, where τ _A1 , τ _B1 is the LOS direct path delay of antennas A and B. When there is obvious multipath effect in the indoor environment, go to step 5, otherwise go to step 3; Step 3: In the LOS environment, there are different delays in the direct signals arriving at the label node through the two antennas, and a delay difference will occur; at this time, P=1 and Q=1 in the formula (7), and the formula ( 7) Substituting into formula (6), the received signal is sorted to get: in: θ is the direction of arrival AOA of the signal, d is the distance between antennas A and B, c is the speed of light, α _A1 and α _B1 are the attenuation on the LOS direct path of antenna A and B, τ _A1 and τ _B1 are the attenuation of antennas A and B In the LOS environment, the received signal is a sine wave with a frequency of (f _a -f _b -β(τ _A1 -τ _B1 )), and the step 8 is performed to perform NDFT transformation on the received signal; Go to step four; Step 4: Calculate the frequency of the signal after performing NDFT transformation on the signal Obtain the delay difference τ _A1 -τ _B1 , and then calculate the direction of arrival AOA from the formula (9) through the geometric relationship, as shown in the formula (10), turn to step 9; Where: c represents the speed of light, d is the distance between antennas A and B; Step 5: When there is multipath effect, there will be different delays in the direct signal reaching the label node through two antennas. At this time, in formula (7), P and Q are not 1 at the same time. Substitute formula (7) into formula In (6), the received signal is arranged to obtain the following expression: Where: τ _k =τ _Ai -τ _Bj represents the time delay difference between the i-th path of antenna A and the j-th path of antenna B, and Π represents a constant; In a multipath environment, the received signal is the superposition of signals of different frequencies (f _a -f _b -βτ _k ) related to the multipath delay difference τ _k . Since the amplitude of the direct signal is strong, two antennas are useful for AOA estimation The direct path delay difference of the transmitted signal is τ ₁ =τ _A1 -τ _B1 , and the corresponding signal frequency is: At this time, the frequency range of the key area is: (f _a -f _b ±βd/c), assuming that f _a -f _b =K f _s , f _s >2βd/c, the frequency range of the key area at this time is: (K-1)f _s ～(K+1)f _s , (13) Set the maximum number of points within these ranges, turn to step 8 for NDFT transformation, and then turn to step 6; Step 6: Perform NDFT transformation on the signal to determine whether the transformed signal peak value exceeds the specified threshold. If it does not exceed the specified threshold, go to step 7. Otherwise, search for the frequency corresponding to the position where the transformed signal peak appears. As the estimated value of f _RSS , AOA can be obtained by substituting into formula (14), and then turn to step 9; Among them: the distance between antennas d<c/2βT _s ; Step 7: When the searched peak value does not exceed the predetermined threshold, it is considered that there is no direct path signal in the current environment, which is an NLOS environment, and the frequency range of key areas is increased, and then enter step 5 to estimate the suboptimal AOA; Step 8: Perform NDFT transformation on the signal. FFT transformation sampling points are evenly distributed on the unit circle, so the method based on FFT transformation cannot achieve sampling with smaller frequency intervals at the frequency of key areas; NDFT transformation sampling points are on the unit circle It is randomly distributed, and the frequency range of key areas is set at as many sampling points as possible, and the NDFT transformation with the smallest possible frequency interval is used to estimate the delay difference of the first path in a multipath environment. The NDFT transformation is shown in the following formula : Among them: z ₀ ,z ₁ ,…,z _N-1 are any different N points selected on the unit circle, formula (15) is written in the form of matrix (16), and the NDFT transformation can be performed quickly through the matrix form: S=Z ^T s, (16) Where: S＝[S _NDFT (z ₀ ),S _NDFT (z ₁ ),…,S _NDFT (z _N-1 )] ^T , (17) s＝[ _sRSS [0], _sRSS [1],... _sRSS [N-1]] ^T , (18) The number of sampling points in the range of (K-1)f _s ~ (K+1)f _s is greater than the number of sampling points in other places; Step 9: Carry out Kalman filter processing on the AOA to eliminate the impact of noise on the system, further improve the estimation accuracy, and output the AOA in the direction of arrival of the signal.