A wireless positioning method based on intelligent reflective surfaces

Abstract

This invention discloses a wireless positioning method based on intelligent reflective surfaces. The method comprises the following steps: S1, acquiring the location information of a base station and at least two intelligent reflective surfaces in the positioning scenario; S2, the receiving end performs timing estimation using the received signal to obtain the arrival time difference between the reflection path and path AB passing through each intelligent reflective surface, and uses the time difference to obtain the optimal estimate of the target location using a timing estimation algorithm; S3, based on the obtained optimal estimate, the optimal solution is obtained through an optimization algorithm, which is the optimal estimate of the user's location. This method provides a method for calculating virtual base stations using multiple RIS (Reflection Sources) and a method for target positioning based on virtual base stations. The wireless positioning method based on RIS presented in this invention can effectively estimate the accurate location of the user in a positioning system under RIS scenarios.

Description

Technical Field

[0001] This invention relates to the field of wireless positioning, and in particular to a wireless positioning method based on intelligent reflective surfaces. Background Technology

[0002] Millimeter-wave (mmWave) signals and large antenna arrays are considered key technologies for future 5G networks. By utilizing the wide bandwidth of the mmWave band, data transmission rates can reach gigabits per second. However, due to significant path loss during transmission, the quality of communication is affected. To effectively improve the data transmission rate and communication service quality of future 5G networks, more flexible and competitive physical layer technologies are being explored. While the benefits of mmWave signals in achieving high data rate communication are well-known, their potential advantages in precise positioning remain largely undiscovered.

[0003] As a promising technology for future wireless communication, intelligent reflective surfaces (RIS) are low-cost, low-power, and easy to deploy. They can also intelligently control the phase shift of the incident signal through a connected controller, thereby enhancing signal transmission. Furthermore, RIS not only enhances wireless communication transmission in channels but also improves imaging and positioning applications, particularly in the positioning field.

[0004] However, in the field of positioning, when base stations are unevenly distributed, it is often difficult to solve the problem of severe signal attenuation caused by off-path, which causes serious delays in the terminal's acquisition of base station data and affects positioning accuracy.

[0005] In addition, existing wireless network positioning technologies require at least three or more base stations to achieve effective estimation of the target location. Since signal transmission is required between multiple base stations, the data exchange between them requires complex messages and synchronization instructions, which not only increases the network burden and has a high failure rate, but also cannot ensure that multiple base stations participate in positioning due to power limitations and the influence of high-frequency network coverage. Positioning accuracy and reliability cannot be guaranteed, and the deployment cost of multiple base stations is also high. Summary of the Invention

[0006] To address the shortcomings of existing technologies, a wireless positioning method based on intelligent reflective surfaces is provided. This method, which uses RIS (Radio Reflective Surface) for wireless positioning, can effectively estimate the user's accurate location in a positioning system within a RIS scenario.

[0007] To achieve the above objectives, the following technical solutions are adopted:

[0008] A wireless positioning method based on a smart reflective surface includes the following steps:

[0009] S1. Obtain the location information of the base station and at least two smart reflective surfaces in the positioning scenario, wherein the base station is located at location A, the user to be located is located at location B, and the smart reflective surfaces are located at location G. n (n = 1, 2, 3, ...), the position of RIS_1 is... The position of RIS_2 is The coordinates of the user to be located are B(x,y), where x and y are unknown;

[0010] S2. The receiver estimates the arrival time difference D between the reflection path and path AB of each smart reflective surface by performing timing estimation on the received signal. n (n = 1, 2, 3, ...), D n (n = 1, 2, 3, ...) The optimal estimate of the target to be measured is obtained through a timing estimation algorithm.

[0011] S3. Based on the best estimate obtained The optimal solution obtained through the optimization algorithm That is, the best estimate of the location of the user terminal to be tested.

[0012] S3-1. Search for samples using an optimization algorithm and calculate their fitness, then store the best fitness value for the current sample.

[0013] S3-2, Perform directional looping operation to update the position of the sample;

[0014] S3-3. Perform a replication loop, accumulating the fitness value of each sample to obtain energy. The sample position corresponding to the maximum energy value is the optimal solution. That is, the best estimate of the location of the user terminal to be tested.

[0015] Preferably, the base station has a single antenna, and the position information of the two smart reflective surfaces is obtained, with the position of RIS_1 being... The position of RIS_2 is The location of the base station is A(x) A ,y A The coordinates of the user to be located are B(x,y), where x and y are unknown.

[0016] Preferably, in step S2, the arrival time differences D1 and D2 are obtained based on the position information of the smart reflective surfaces RIS_1 and RIS_2, respectively.

[0017] Preferably, the timing estimation algorithm in step S2 includes the following steps:

[0018] definition:

[0019] Where R1 represents the distance from the BS to the user, R2 represents the distance from the BS to RIS_1, R3 represents the distance from RIS_1 to the user to be located, R4 represents the distance from the BS to RIS_2, and R5 represents the distance from RIS_2 to the user to be located. c is the speed of light.

[0020] In ΔABG1 formed by base station A, RIS_1, and user B, the coordinates (x, y) of point A' are calculated using the law of cosines. A' ,y A' ),

[0021] Similarly, for ΔABG2 formed by base station A, RIS_2, and user B, the coordinates (x, y) of A” can be calculated using the law of cosines. A” ,y A” );

[0022] An arc is drawn with the location of RIS_1 as the center and R2 as the radius, intersecting the extension line G1B of the reflection path from RIS_1 to the user. Let the intersection point be A', then the distance from point A' to RIS_1 is R2. An arc is drawn with the location of RIS_2 as the center and R4 as the radius, intersecting the extension line G2B of the reflection path from RIS_2 to the user. Let the intersection point be A”, then the distance from point A” to RIS_2 is R4.

[0023] definition:

[0024]

[0025]

[0026]

[0027]

[0028] Define a vector γ, which can be represented as:

[0029] γ=[a(x,y),b(x,y),c(x,y),d(x,y)] T

[0030] The vector γ is a function containing unknown parameters x and y, denoted as X = [x, y];

[0031] Let the objective function be F(x,y)=||γ|| 2 The following optimization algorithm is used to search for X that minimizes F(x,y), namely:

[0032]

[0033] when When the objective function F(x,y) reaches its minimum value, then... This is the best estimate of the target to be measured.

[0034] Preferably, the specific method of step S3-1 is as follows:

[0035] Initialize the dimension of the search space by parameters, and obtain the total number of samples S = 50 to 20,000 through optimization algorithm.

[0036] Next, for each sample i, calculate its fitness function:

[0037] J i (j,k)=F(P i (j,k))

[0038] Among them, P i (j,k) represents the position of sample i after the j-th approach operation and the k-th copy operation. Assume the initial position of each sample is P. i (1,1)=(0,0), let This represents the best fitness value currently stored for sample i.

[0039] Preferably, the specific method for step S3-2 is as follows:

[0040] Perform a directional loop operation to update the position function of the sample:

[0041] P i (j+1,k)=P i (j,k)+δ*Δ(i)

[0042] Where Δ(i) (i = 1, 2, ..., S) is a random number with a value in the range [-1, 1], and the step size is δ = 0.001.

[0043] Calculate the corresponding fitness value:

[0044] βJ i (j+1,k)=F(P i (j+1,k))

[0045] If the current fitness value J i (j+1,k) is greater than Then save to the current value

[0046]

[0047] Calculate the new direction of movement for each sample:

[0048]

[0049]

[0050] in

[0051] Where C1 and C2 are random numbers between 0 and 2, and Q1 and Q2 are two independent random numbers in the range [0,1].

[0052] The effect of inter-sample movement on the fitness function, J cc (j+1,k) is defined as:

[0053]

[0054] Therefore, the mathematical expression after adding the sample shift operation to the approach loop is:

[0055] J i (j+1,k)=J i (j,k)+J cc (j+1,k)

[0056] Among them, V i (j,k) represents the velocity of sample i after the j-th approach operation and the k-th replication operation. The initial velocity V of each sample i is set. i (1,1)=0.15, This indicates the location of a local extremum in a directional operation. F represents the location of the global extremum in a trending operation. x ' represents taking the partial derivative of the objective function F(x,y) with respect to x, F y ' indicates that the objective function F(x,y) is taken as a partial derivative with respect to y.

[0057] If j≤N c If the trend operation is successful, continue with the trend operation; otherwise, exit the trend loop and proceed to step S3-3.

[0058] Preferably, the specific method of step S3-3 is as follows:

[0059] Perform a replication loop: accumulate the fitness value of each sample to obtain energy, expressed as:

[0060]

[0061] Where, the trend operand N c =100~1000,

[0062] Then the energy value of each sample Arrange the samples in ascending order of energy value, discard the first S / 2 samples with smaller energy values, and select the last S / 2 samples with larger energy values ​​for replication. Each sample is replicated into two identical samples. That is, for the first S / 2 samples with smaller energy values, their position function P... i (j,k) is changed to:

[0063] P i (j,k)=P i+S / 2 (j,k)(i=0,1,2,...,S / 2-1)

[0064] After each copy operation, k = k + 1, if k ≤ N re If the copy operation succeeds, continue; otherwise, exit the copy loop.

[0065] The final energy value J health (i,N re The sample position corresponding to the maximum value of ) The position coordinates at which the objective function F(x,y) reaches its minimum value. Right now:

[0066]

[0067] Among them, the maximum step size N in the search space s =3~6, copy operands N re =4, i max The number of digits in the total number of samples containing the sample with the highest fitness value after homing and replication operations can be represented as:

[0068]

[0069] Therefore, the optimal solution obtained by the above optimization algorithm It is the best estimate of the location of the user terminal to be tested in the positioning system.

[0070] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0071] Based on the existence of a line-of-sight (LOS) path, the RIS (Reflection Path Array) can add multiple auxiliary non-line-of-sight (NLOS) reflection paths. The location of the base station is known. We construct a mirror-symmetric point of the base station based on the plane where the RIS is located. Then, using this mirror-symmetric point and the formation mechanism of the virtual base station, we construct a virtual base station and calculate the equivalent virtual base station location based on the angle of signal reflection from the RIS to the user and the distance between the base station and the RIS. This allows for target location estimation using only a single positioning base station. This invention presents a method for calculating virtual base stations using multiple RIS and a method for target localization based on virtual base stations.

[0072] Compared to conventional positioning scenarios, this invention requires only one positioning base station. By constructing a "virtual base station" through RIS_1 and RIS_2, the number of positioning base stations required in the system is reduced, the complexity of the positioning system is lowered, and the positioning calculation is performed on the terminal to be positioned, thus meeting the user's privacy and confidentiality requirements.

[0073] By solving the objective function using the optimization algorithm employed in this invention, an effective estimation of the user's location coordinates is achieved, resulting in a more accurate location estimate. Attached Figure Description

[0074] Figure 1 This is a system model diagram of a RIS-based two-dimensional positioning scenario for implementing the positioning of this invention; Detailed implementation method:

[0075] To more clearly illustrate the embodiments of the present invention, the embodiments will be further described below with reference to the accompanying drawings. Obviously, the embodiments described below are only some embodiments of the present invention. For those skilled in the art, other embodiments can be obtained based on these embodiments without creative effort and also fall within the protection scope of the present invention.

[0076] like Figure 1 As shown, the positioning system used in this invention consists of a base station. This embodiment uses two RIS (Radio Router Arrays) as an example. Here, it is assumed that the base station is a single antenna, and RIS_1, RIS_2, and the receiving user are a multi-antenna uniform linear array (ULA), with L and N array elements respectively, and an element spacing of d. Assume the base station location is A(x... A ,y A The position of RIS_1 is... The position of RIS_2 is The coordinates of the user to be located are B(x,y), where x and y are unknown. R1 represents the distance from BS to the user, R2 represents the distance from BS to RIS_1, R3 represents the distance from RIS_1 to the user to be located, R4 represents the distance from BS to RIS_2, and R5 represents the distance from RIS_2 to the user to be located.

[0077] With the location of RIS_1 as the center and R2 as the radius, draw an arc that intersects the extension line G1B of the reflection path from RIS_1 to the user. Let the intersection point be A', then the distance from point A' to RIS_1 is R2. With the location of RIS_2 as the center and R4 as the radius, draw an arc that intersects the extension line G2B of the reflection path from RIS_2 to the user. Let the intersection point be A”, then the distance from point A” to RIS_2 is R4.

[0078] Step 1: The receiver performs timing estimation using the received signal to obtain the arrival time difference D1 between the reflection path AG1B via RIS_1 and path AB, and the arrival time difference D2 between the reflection path AG2B via RIS_2 and path AB. D1 and D2 can be obtained using common timing estimation algorithms. For example, a pseudo-random sequence can be sent by the transmitter, and the receiver can perform timing estimation through cross-correlation.

[0079] Step 2: Construct a system of equations:

[0080]

[0081]

[0082]

[0083]

[0084] in,

[0085] The x-coordinate of A' is... A' Represented as y-axis A' Represented as The x-coordinate of A” A” Represented as y-axis A” Represented as c is the speed of light.

[0086] in,

[0087]

[0088]

[0089]

[0090]

[0091] Step 3: Define the optimization vector γ for the optimization algorithm:

[0092] γ=[a(x,y),b(x,y),c(x,y),d(x,y)] T

[0093] The vector γ is a function containing unknown parameters x and y, denoted as X = [x, y].

[0094] Step 4: Take the objective function as F(x,y)=||γ|| 2 The optimal solution is obtained by using an optimization algorithm.

[0095] We use the following optimization algorithm to search for X that minimizes F(x,y), namely:

[0096]

[0097] when When the objective function F(x,y) reaches its minimum value, then... This is the best estimate of the target to be measured.

[0098] Step 5: Initialize the dimension of the parameter search space N = 14, the total number of samples searched in the optimization algorithm S = 50~200, and the number of convergence operations N. c =100~1000, the maximum step size N in the search space s =3~6, copy operands N re =4, step size unit δ = 0.001, Δ(i) (i = 1, 2, ..., S) is a random number with a value in [-1, 1], C1 and C2 are random numbers with a value in 0 to 2, and Q1 and Q2 are two independent random numbers in [0, 1].

[0099] Step 6: For each sample i, calculate its fitness function:

[0100] J i (j,k)=F(P i (j,k))

[0101] Among them, P i (j,k) represents the position of sample i after the j-th approach operation and the k-th copy operation. Assume the initial position of each sample is P. i (1,1) = (0,0). Let This represents the best fitness value currently stored for sample i.

[0102] Step 7: Perform a directional loop operation to update the sample's position function:

[0103] P i (j+1,k)=P i (j,k)+δ*Δ(i)

[0104] Calculate the fitness value of the corresponding sample:

[0105] βJ i (j+1,k)=F(P i (j+1,k))

[0106] If the current fitness value J i (j+1,k) is greater than Then save to the current value

[0107]

[0108] Calculate the new direction of movement for each sample:

[0109]

[0110]

[0111] in

[0112] The effect of inter-sample movement on the fitness function, J cc (j+1,k) is defined as:

[0113]

[0114] Therefore, the mathematical expression after adding the sample shift operation to the approach loop is:

[0115] J i (j+1,k)=J i (j,k)+J cc (j+1,k)

[0116] Among them, V i (j,k) represents the velocity of sample i after the j-th approach operation and the k-th replication operation. The initial velocity V of each sample i is set. i (1,1)=0.15, This indicates the location of a local extremum in a directional operation. F represents the location of the global extremum in a trending operation. x ' represents taking the partial derivative of the objective function F(x,y) with respect to x, F y ' indicates that the objective function F(x,y) is partially derived with respect to y.

[0117] If j≤N c If the trend continues, the trend operation will continue. Otherwise, the trend loop will be terminated.

[0118] Step 8: Perform a replication loop: Accumulate the fitness value of each sample to obtain energy, expressed as:

[0119]

[0120] Then the energy value of each sample Arrange the samples in ascending order of energy value, discard the first S / 2 samples with smaller energy values, and select the last S / 2 samples with larger energy values ​​for replication. Each sample is replicated into two identical samples.

[0121] That is, for the first S / 2 samples with smaller energy values, their position function P i (j,k) is changed to:

[0122] P i (j,k)=P i+S / 2 (j,k)(i=0,1,2,...,S / 2-1)

[0123] After each copy operation, k = k + 1. If k ≤ N re If the copy operation succeeds, continue. Otherwise, exit the copy loop.

[0124] Step 9: Calculate position coordinates The final energy value J health (i,N re The sample position corresponding to the maximum value of ) The position coordinates at which the objective function F(x,y) reaches its minimum value. Right now:

[0125]

[0126] Among them, i max The number of digits in the total number of samples containing the sample with the highest fitness value after homing and replication operations can be represented as:

[0127]

[0128] The optimal solution obtained by the above optimization algorithm It is the best estimate of the location of the user terminal to be tested in the positioning system.

[0129] The examples above are merely illustrative to aid in understanding the method and core ideas of this invention. It should be noted that those skilled in the art can make various improvements and modifications to this invention without departing from its principles, and these improvements and modifications also fall within the scope of protection of the claims. Furthermore, the target position estimated using the optimization algorithm still deviates somewhat from the user's actual location coordinates; further research will focus on developing algorithms with higher positioning accuracy.

Claims

1. A wireless positioning method based on intelligent reflective surfaces, characterized in that, Includes the following steps: S1. Obtain the location information of the base station and at least two smart reflective surfaces in the positioning scenario, wherein the location of the base station is... Users to be located The location of the intelligent reflective surface is The position of RIS_1 is The position of RIS_2 is The coordinates of the user to be located are ,in , It is unknown; S2. The receiver estimates the arrival time difference between the reflection path and path AB of each smart reflective surface by performing timing estimation on the received signal. , The optimal estimate of the target is obtained through a timing estimation algorithm. ; S3. Based on the best estimate obtained The optimal solution obtained through the optimization algorithm That is, the best estimate of the location of the user terminal to be tested. S3-1. Search for samples using an optimization algorithm and calculate their fitness, then store the best fitness value for the current sample. The specific method for step S3-1 is as follows: Initialize the dimension of the search space by parameters, and obtain the total number of samples by performing an optimization algorithm. , Then for each sample Calculate its fitness function: ； in, Indicates sample In the Second trend operation, the first The position after the copy operation, assuming the initial position of each sample. ,make , indicating sample The best fitness value currently stored; S3-2, Perform directional looping operation to update the position of the sample; The specific method for step S3-2 is as follows: Perform a directional loop operation to update the position function of the sample: ； in, A random number with a value in the range [-1, 1], with a step size of [unit]. , Calculate the corresponding fitness value: ； If the current fitness value Greater than Then save to the current value. : ； Calculate the new direction of movement for each sample: ； ； ； in, and For the numerical value in random numbers, and They are two independent random numbers in the range [0,1]. The impact of inter-sample movement on the fitness function Defined as: ； Therefore, the mathematical expression after adding the sample shift operation to the approach loop is: ； in, Indicates sample In the Second trend operation, the first The speed after each copy operation, setting the speed for each sample. initial velocity , This indicates the location of a local extremum in a directional operation. This indicates the location of the global extremum in a trending operation. Describe the objective function right Find the partial derivative. Describe the objective function right Find the partial derivative. if If the trend operation continues, then proceed with the trend operation; otherwise, exit the trend loop and proceed to step S3-3. S3-3. Perform a replication loop, accumulating the fitness value of each sample to obtain energy. The sample position corresponding to the maximum energy value is the optimal solution. That is, the best estimate of the location of the user terminal to be tested.

2. The wireless positioning method based on intelligent reflective surfaces according to claim 1, characterized in that, The base station has a single antenna, and the position information of the two smart reflective surfaces is obtained, with the position of RIS_1 being... The position of RIS_2 is The location of the base station is The coordinates of the user to be located are ,in , It is unknown.

3. The wireless positioning method based on intelligent reflective surfaces according to claim 2, characterized in that, In step S2, the arrival time difference is obtained based on the position information of the intelligent reflective surfaces RIS_1 and RIS_2. and .

4. The wireless positioning method based on intelligent reflective surfaces according to claim 3, characterized in that, In step S2, the specific steps of the timing estimation algorithm are as follows: definition: ; in, Indicates the distance from the B / S (Browser / Server) to the user. This represents the distance from BS to RIS_1. This represents the distance from RIS_1 to the user to be located. This represents the distance from BS to RIS_2. This represents the distance from RIS_2 to the user to be located. , , , , , It's the speed of light. In the system consisting of base station A, RIS_1, and user B In the middle, the point is calculated using the Law of Cosines. coordinates , Similarly, for the combination of base station A, RIS_2, and user B... In the middle, it is calculated using the Law of Cosines. coordinates ; With the location of RIS_1 as the center, Draw an arc with radius and the extension of the reflection path from RIS_1 to the user. They intersect, let the intersection point be... Then point The distance to RIS_1 is With the location of RIS_2 as the center, Draw an arc with radius and the extension of the reflection path from RIS_2 to the user. They intersect, let the intersection point be... Then point The distance to RIS_2 is ; definition: ； Define vector , can be represented as: ； vector It contains unknown parameters and The function, with unknown parameters denoted as ; Take the objective function as The following optimization algorithm is used to search. Make Minimize, that is: ； when When, the objective function The minimum value will be reached at this time. This is the best estimate of the target to be measured.

5. The wireless positioning method based on intelligent reflective surfaces according to claim 4, characterized in that, The specific method for step S3-3 is as follows: Perform a replication loop: accumulate the fitness value of each sample to obtain energy, expressed as: ； Among them, the trend operand , Then the energy value of each sample Arrange them in ascending order, eliminating the top ones. Select the samples with smaller energy values. The samples with higher energy values ​​are replicated, and each sample is replicated into two identical samples. That is, for the first... For samples with smaller energy values, their position functions Change to: ； After each copy operation is completed, ,if If the copy operation succeeds, continue; otherwise, exit the copy loop. The final energy value The sample position corresponding to the maximum value To make the objective function Position coordinates when the minimum value is obtained ,Right now: ； Among them, the maximum step size of the search space Copy operands , The number of digits in the total number of samples containing the sample with the highest fitness value after homing and replication operations can be represented as: ； Therefore, the optimal solution obtained by the above optimization algorithm It is the best estimate of the location of the user terminal to be tested in the positioning system.

Images