A near-field positioning method and system for multi-user equipment based on antenna array

Abstract

The present invention belongs to the field of wireless communication technology and device positioning, and in particular, to a multi-user device near-field positioning method based on an antenna array. By deploying a very large-scale array antenna at a base station, the present invention constructs a very large-scale array-assisted multi-user near-field positioning system. Based on the antenna radiation pattern and the free space path loss, an actual near-field channel model is established. We use a subarray partitioning strategy to divide the near-field channel into multiple subarray channels, and perform probabilistic modeling on the multi-user positioning problem. We use a message passing algorithm to iteratively update the messages between the nodes of the factor graph, make full use of the mapping relationship between the near-field channel and the user position, and achieve high-precision user positioning.

Description

Technical Field

[0001] The present invention belongs to but is not limited to the field of wireless communication technology and device positioning, and in particular relates to a multi-user device near-field positioning method and system based on an antenna array. Background Art

[0002] The sixth generation of mobile communication systems (6G) is expected to achieve significant performance improvements in access capacity, communication rate and reliability. The key to fully realizing the potential of 6G lies in the precise positioning of user equipment. High-precision user equipment location information is crucial for applications such as autonomous driving, augmented reality and the Internet of Things. Very large-scale antenna arrays are an advanced evolution of large-scale multiple-input multiple-output array technology, which provides a promising solution for user equipment positioning. Compared with the existing large-scale multiple-input multiple-output arrays, very large-scale antenna arrays further increase the number of antennas, resulting in a significant increase in the array aperture size. At this time, the wavefront curvature of the electromagnetic wave becomes significant, and near-field modeling of the wireless channel is required. Under the near-field spherical wave model, the base station can estimate the distance between the user equipment and the antenna array based on the wavefront curvature of the electromagnetic wave reaching the antenna array. In addition, as the array size increases, the angular resolution and angle estimation accuracy of the very large-scale antenna array are significantly improved. Combined with high-precision angle information and distance information, the base station can achieve high-precision user equipment positioning.

[0003] Existing near-field positioning methods are mainly divided into two categories, namely, subspace-based positioning methods and maximum likelihood-based estimation methods, which face many problems in actual deployment. For subspace-based methods, multiple signal measurements are usually required to construct a reliable sample covariance matrix. However, in mobile communication scenarios, due to the fast fading of the wireless channel, it is often difficult to obtain enough observations within a coherence time. In addition, when the user equipment is in the near-field area, the array response no longer has a Vandermonde structure, making the spatial smoothing method unsuitable. For the maximum likelihood-based method, the high angular resolution characteristics of the ultra-large-scale antenna array are used to achieve accurate angle estimation. However, the high angular resolution also makes the corresponding maximum likelihood problem highly non-convex. This problem is particularly prominent in near-field multi-user positioning, where the position of each user is parameterized by two angles and a distance, which leads to a large feasible domain of the optimization problem and it is difficult to obtain the optimal estimate of the user's position. In addition, none of the existing near-field positioning schemes consider the radiation pattern of the array antenna. In the near-field channel model, for the same user device, different antenna units have different wave directions. When the radiation pattern of the antenna unit is directional, the array response vector will be distorted. Existing literature has not yet explained how the antenna radiation pattern affects the near-field user positioning performance. To this end, by considering the radiation pattern of the array antenna and the near-field channel modeling, the present invention proposes a reliable and efficient method for multi-user near-field positioning using a large-scale receiving array. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention provides a near-field positioning method and system for multi-user equipment based on an antenna array.

[0005] The present invention is implemented as follows: a near-field positioning method for multi-user equipment based on an antenna array, the method comprising:

[0006] One with A base station with antenna units and A single-antenna terminal device; the base station is equipped with The fully connected analog beamforming architecture is used between the RF chain and the array antenna; the partially connected analog beamforming architecture can be used as a special case of the system currently considered; in the three-dimensional Cartesian coordinate system, The user terminal position coordinates are marked as , No. The base station antenna unit position coordinates are marked as At this time, The user terminal and The channel coefficients between base station antennas are modeled as

[0007] ,

[0008] in For the The user terminal and The Euclidean distance between base station antennas, Corresponding free space path loss, Corresponding to The base station antenna unit is The antenna radiation gain in the direction of the user terminal is a function of the user terminal position, The complex gain coefficient of the channel that is independent of the terminal position is included; the planar antenna array of the base station is divided into subarrays, each with antenna units, satisfying , . No. The user terminal and The channels of the subarrays can be written as ,in For the The channel coefficients of the subarray reference antennas, is a vector-valued function of the device position; the reference antenna channel coefficients for different subarrays are written as vectors , which can be further written as ,in is a vector value function of the device position; the received signal model of the base station receiving array is

[0009] ,

[0010] in For the Device reference array gain, To send a signal, is the transmit power, It is The beamforming matrix of the subarrays is is the Gaussian white noise at the receiving end; by defining the auxiliary variable , , , , we get an equivalent signal model

[0011] ,

[0012] Perform probability modeling on the above signal model.

[0013] Will and The prior distribution of is modeled as a complex Gaussian distribution and a Gaussian distribution

[0014] ,

[0015] because Is the device location Function, get the conditional probability density function

[0016] ,

[0017] in , , is the Dirac delta function. The observation likelihood function is

[0018] Then the joint probability density function can be written as

[0019] ,

[0020] in , ; Based on the joint probability density function, draw its corresponding factor graph representation, where the hollow circular box and the black rectangle represent the variable node and the check node respectively; the variable node includes , and ; Check nodes include , , (denoted as )and (denoted as ); the message is calculated and updated on the factor graph, and then a multi-user device near-field positioning algorithm is proposed. The multi-user near-field positioning method includes the following steps:

[0021] S1. Order .

[0022] S2. Get device location Initialization of Device Locations During initialization, the previous Initialization of user locations , ; Construct objective function

[0023] ,

[0024] in For device location Function. Using device location Polar coordinate representation of ,in is the distance from the device to the center of the array, (or ) is the device orientation and array Axis (or axis); construct the grid

[0025] .

[0026] S3. Discrete grid Perform an exhaustive search to find the maximum Grid points .

[0027] S4, As a starting point, perform the gradient ascent algorithm and obtain The local maximum of the gradient ascent algorithm is updated as follows:

[0028] .

[0029] Where n is the iteration round, step length , and Determined by the Armijo criterion; after the gradient descent algorithm converges, it is obtained from the convergence point Initialization , the convergence condition of the gradient descent algorithm is .

[0030] S5, ,if , return to S2.

[0031] S6. Initialize message mean , , ; Initialize message covariance , in is a large positive value such as .

[0032] S7, Update and for

[0033] .

[0034] in , .

[0035] S8, Update , ,in , is through Substitution Got it.

[0036] S9, Update and .

[0037] S10. Construct objective function

[0038] .

[0039] renew To obtain Maximum value, expressed in polar coordinates ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is

[0040] .

[0041] Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is .

[0042] S11, Update for

[0043] .

[0044] S12, Update and ,in , is through Substitution Got it.

[0045] S13, Update and .

[0046] S14. Construct objective function

[0047] .

[0048] renew To obtain Maximum value, expressed in polar coordinates ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is .

[0049] Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is .

[0050] S15, Update for

[0051] .

[0052] S16, Update and for

[0053] .

[0054] S17, Calculation variable node News from

[0055] .

[0056] Determine variable node Whether the message at the location has converged, if it has converged, the device position estimate is obtained ; If not converged, return to step S7.

[0057] Another object of the present invention is to provide a multi-user device near-field positioning system based on an antenna array based on the multi-user device near-field positioning method based on an antenna array, the system specifically comprising:

[0058] Parameter setting module, the system operating frequency is 6 GHz, the wavelength is 0.05m, and the base station is equipped with Antennas are arranged in a uniform array with a half-wavelength spacing. The number of base station RF chains is 160; the coordinates of the central antenna unit of the base station antenna array are (0 m, 0 m, 0 m), and the two direction vectors of the base station antenna array are (1, 0, 0) and (0, 1, 0) respectively; the area where the user equipment is located is a cone area with a cone area vertex angle of 120 degrees. The distance between the equipment and the base station array ranges from 5 meters to 10 meters. The base station locates 3 user equipment at the same time; the user terminal equipment is equipped with a single uniform antenna, and the user equipment position is uniformly sampled in the target area; the beamforming matrix The elements of have unit modulus and random phase.

[0059] Initialization module, connected with parameter setting module, ; Get device location Initialization of Device Locations During initialization, the previous Initialization of user locations , ; Construct objective function

[0060] .

[0061] in For device location function.

[0062] Leverage device location Polar coordinate representation of ,in is the distance from the device to the center of the array, (or ) is the device orientation and array Axis (or axis); construct the grid

[0063] .

[0064] Discrete grid Perform an exhaustive search to find the maximum Grid points ;by As a starting point, perform the gradient ascent algorithm and obtain The local maximum of the gradient ascent algorithm is updated as follows:

[0065] .

[0066] Where n is the iteration round, step length , and Determined by the Armijo criterion; after the gradient descent algorithm converges, it is obtained from the convergence point Initialization , the convergence condition of the gradient descent algorithm is .

[0067] ,if , execute the above grid exhaustive search and gradient ascent algorithm again.

[0068] Initialize message mean , , ; Initialize message covariance , in is a large positive value such as .

[0069] The message update module is connected to the initialization module to update and for

[0070] .

[0071] in , .

[0072] renew , ,in , is through Substitution Got it.

[0073] renew and ;

[0074] Constructing the objective function

[0075] .

[0076] renew To obtain Maximum value, expressed in polar coordinates ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is

[0077] .

[0078] Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is .

[0079] renew for

[0080] .

[0081] renew and ,in , is through Substitution Got it.

[0082] renew and ;

[0083] Constructing the objective function

[0084] .

[0085] renew To obtain Maximum value, expressed in polar coordinates ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is

[0086] .

[0087] Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is .

[0088] renew for

[0089] ,

[0090] renew and for .

[0091] The edge message calculation module is connected to the target function construction module to calculate the variable node News from

[0092] .

[0093] Determine variable node Whether the message at the location has converged, if it has converged, the device position estimate is obtained ; If it does not converge, execute the above grid exhaustive search and gradient ascent algorithm again.

[0094] Another object of the present invention is to provide a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the multi-user device near-field positioning method based on an antenna array.

[0095] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, enables the processor to perform the steps of the multi-user device near-field positioning method based on an antenna array.

[0096] Another object of the present invention is to provide an information data processing terminal, which is used to implement the multi-user device near-field positioning system based on the antenna array.

[0097] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0098] First, the present invention realizes the simultaneous positioning of multiple terminal devices by deploying ultra-large-scale array antennas at the base station end. Multiple terminal devices use the same time domain and frequency domain resources to reduce the overhead of wireless air interface resources. The signal modeling under the limited number of RF chains of the base station is considered, and the antenna unit radiation pattern is considered when modeling the near-field channel. The array grouping strategy and message passing algorithm are used to reliably and efficiently solve the highly non-convex problem of multi-user positioning. The performance of the positioning algorithm proposed by the present invention can be close to the theoretical lower bound, and the algorithm is easy to deploy on the wireless communication infrastructure without incurring additional hardware overhead.

[0099] Second, the expected benefits and commercial value after the transformation of the technical solution of the present invention are: the positioning algorithm proposed in the present invention is easily deployed on the wireless communication infrastructure, enabling the wireless communication system to provide positioning services, and supporting the implementation of new applications such as Internet of Vehicles, virtual reality, and digital twins. When executing the positioning algorithm of the present invention, multiple devices can transmit signals in the same time domain and frequency domain resources, and do not rely on orthogonal resource allocation schemes, which can effectively reduce the air interface resources required for positioning. In addition, the algorithm of the present invention is suitable for scenarios with a limited number of base station RF chains, so the requirements for base station hardware are low and the algorithm is easier to deploy.

[0100] The technical solution of the present invention fills the technical gap in the industry at home and abroad:

[0101] By considering the radiation pattern of antenna units, a more accurate near-field channel model is established, and the near-field positioning problem of devices under this accurate model is studied. Under the near-field channel model, multiple devices use the same time-frequency resources to communicate with the base station, and there is interference between different devices. This algorithm effectively solves the mutual interference between devices and realizes the accurate positioning of multiple devices.

[0102] Third, the technical solution of the present invention solves many problems in the prior art in industrial applications and has achieved significant technical progress, as shown in the following:

[0103] 1. The problem of insufficient accuracy of traditional positioning

[0104] Existing user equipment positioning methods, especially in multi-user and near-field scenarios, often rely on traditional far-field models, resulting in poor positioning accuracy. In particular, in dense user environments or complex indoor environments, the positioning error is large and it is difficult to meet high-precision application requirements.

[0105] The present invention realizes a high-precision positioning method based on near field by using antenna array and multi-subarray structure. This method effectively improves the positioning accuracy in multi-user scenarios by combining probability density function and factor graph calculation, and is particularly suitable for complex indoor scenarios, dense user scenarios and short-range near-field communication environments, ensuring high-precision estimation of user device locations and significantly improving positioning accuracy.

[0106] 2. Low positioning efficiency in multi-user scenarios

[0107] In multi-user scenarios, traditional positioning algorithms often have high computational complexity, and are prone to problems such as low computational efficiency and large latency. They are unable to respond to the positioning needs of multiple user devices in a timely manner, affecting the quality of communication and application services.

[0108] The multi-user near-field positioning method based on antenna array proposed in the present invention realizes efficient update and calculation of message transmission by distributed modeling and factor graph optimization of joint probability density function. The method can quickly iteratively optimize the location information of multiple user devices, and combines grid search and gradient ascent algorithm to reduce the computational complexity and significantly improve the positioning efficiency, so that it can meet the positioning needs of multi-user devices in real time.

[0109] 3. Existing systems have poor adaptability to complex scenarios

[0110] Traditional positioning technology performs poorly in complex scenarios, especially when faced with complex buildings, obstacle obstructions, reflections and multipath interference. Positioning accuracy and system stability are often seriously affected and cannot adapt to complex application scenarios.

[0111] The present invention can effectively adapt to multipath and channel fading problems in complex environments by introducing antenna radiation gain model, RF link simulation beamforming architecture and dynamic modeling of channel coefficients. This solution can not only accurately model the near-field channel characteristics of user equipment, but also achieve high robustness and stable positioning performance in different complex scenarios through the gradient ascent algorithm and factor graph message update mechanism. It is suitable for applications in complex indoor environments, dense urban buildings and complex traffic scenarios.

[0112] 4. Insufficient system scalability and flexibility

[0113] Many existing positioning systems lack flexibility in hardware architecture and algorithm design, are difficult to adapt to different scenario requirements, and have poor scalability, making them difficult to apply in diverse industrial environments.

[0114] The positioning method of the present invention uses two types of analog beamforming architectures, partial connection and full connection, which can not only adapt to the current system requirements, but also have good scalability. Through the flexible division and subarray structure of the antenna array, the system can be adjusted according to the actual application scenario, suitable for application scenarios of different scales and requirements, and has high flexibility and adaptability. This makes the technology applicable not only to existing communication systems, but also to a wide range of fields such as autonomous driving, drone positioning, and the Internet of Things.

[0115] The present invention solves the problems of low positioning accuracy, low efficiency, poor adaptability to complex environments, and insufficient system scalability in the prior art by adopting a multi-user device near-field positioning method based on an antenna array. Its significant technological progress is reflected in improved positioning accuracy, optimized computing efficiency in multi-user scenarios, enhanced adaptability in complex scenarios, and improved system flexibility and scalability. It has broad industrial application prospects and is particularly suitable for 5G communications, the Internet of Things, autonomous driving, drone positioning, and other fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0116] Figure 1 It is a system model provided by an embodiment of the present invention.

[0117] Figure 2 This is an example of a factor graph representation of a near-field multi-user positioning probability model provided by an embodiment of the present invention.

[0118] Figure 3 This is the near-field device positioning performance under different array sizes and grouping strategies provided by the embodiments of the present invention.

[0119] Figure 4 This is the near-field device positioning performance under different distances between the device and the base station provided by the embodiment of the present invention.

[0120] Figure 5It is a comparison between the solution of the present invention provided by the embodiment of the present invention and the existing baseline solution. DETAILED DESCRIPTION

[0121] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0122] The following are two specific industrial application embodiments based on the multi-user device near-field positioning method:

[0123] Example 1: User equipment positioning in a 5G communication system.

[0124] In 5G networks, accurate user device positioning is an important technology for achieving efficient communication and network optimization. The multi-user device near-field positioning method of the invention can be used for user terminal positioning of 5G base stations. When multiple user devices (such as mobile phones or IoT devices) access the 5G network, the base station uses an antenna array to locate the near-field position of each user device in real time. Through this method, the base station can allocate resources to the user device based on its precise location information, optimize the network transmission path, and improve communication efficiency. In addition, this method can also be used in indoor positioning scenarios, such as smart factories and large indoor event venues, to achieve high-precision user positioning and navigation services.

[0125] Example 2: V2X (Vehicle to Everything) positioning of an autonomous vehicle.

[0126] In the field of autonomous driving, communication and collaborative operation between vehicles and the surrounding environment (including other vehicles, traffic facilities, and pedestrians) are crucial. Using this multi-user device near-field positioning method, a high-precision location information sharing mechanism can be established between vehicles and base stations (or roadside units RSU). Through the antenna array equipped with the base station, the system can accurately determine the near-field position of each autonomous driving vehicle, providing real-time traffic situation awareness and path optimization. This method can effectively support the collaborative control of vehicles and surrounding equipment in the vehicle-to-everything (V2X) network, ensure traffic safety, optimize road use efficiency, and realize refined management of intelligent transportation systems.

[0127] like Figure 1 As shown, the embodiment of the present invention provides that the present invention is implemented in this way: a near-field positioning method for a multi-user device based on an antenna array, the method comprising:

[0128] One with A base station with antenna units and A single-antenna terminal device; the base station is equipped with The fully connected analog beamforming architecture is used between the RF chain and the array antenna; the partially connected analog beamforming architecture can be used as a special case of the system currently considered; in the three-dimensional Cartesian coordinate system, The user terminal position coordinates are marked as , No. The base station antenna unit position coordinates are marked as At this time, The user terminal and The channel coefficients between base station antennas are modeled as

[0129] .

[0130] in For the The user terminal and The Euclidean distance between base station antennas, Corresponding free space path loss, Corresponding to The base station antenna unit is The antenna radiation gain in the direction of the user terminal is a function of the user terminal position, The complex gain coefficient of the channel that is independent of the terminal position is included; the planar antenna array of the base station is divided into subarrays, each with antenna units, satisfying , . No. The user terminal and The channels of the subarrays can be written as ,in For the The channel coefficients of the subarray reference antennas, is a vector-valued function of the device position; the reference antenna channel coefficients for different subarrays are written as vectors , which can be further written as ,in is a vector value function of the device position; the received signal model of the base station receiving array is

[0131] .

[0132] in For the Device reference array gain, To send a signal, is the transmit power, It is The beamforming matrix of the subarrays is is the Gaussian white noise at the receiving end; by defining the auxiliary variable , , , , we get an equivalent signal model

[0133] ,

[0134] Perform probability modeling on the above signal model.

[0135] Will and The prior distribution of is modeled as a complex Gaussian distribution and a Gaussian distribution

[0136] .

[0137] because Is the device location Function, get the conditional probability density function

[0138] .

[0139] in , , is the Dirac delta function. The observation likelihood function is

[0140] .

[0141] Then the joint probability density function can be written as

[0142] .

[0143] in , ; Based on the joint probability density function, draw its corresponding factor graph representation, where the hollow circular box and the black rectangle represent the variable node and the check node respectively; the variable node includes , and ; Check nodes include , , (denoted as )and (denoted as ); the message is calculated and updated on the factor graph, and then a multi-user device near-field positioning algorithm is proposed. The multi-user near-field positioning method includes the following steps:

[0144] S1. Order .

[0145] S2. Get device location Initialization of Device Locations During initialization, the previous Initialization of user locations , ; Construct objective function

[0146] .

[0147] in For device location Function. Using device location Polar coordinate representation of ,in is the distance from the device to the center of the array, (or ) is the device orientation and array Axis (or axis); construct the grid

[0148] .

[0149] S3. Discrete grid Perform an exhaustive search to find the maximum Grid points .

[0150] S4, As a starting point, perform the gradient ascent algorithm and obtain The local maximum of the gradient ascent algorithm is updated as follows:

[0151] .

[0152] Where n is the iteration round, step length , and Determined by the Armijo criterion; after the gradient descent algorithm converges, it is obtained from the convergence point Initialization , the convergence condition of the gradient descent algorithm is .

[0153] S5, ,if , return to S2.

[0154] S6. Initialize message mean , , ; Initialize message covariance , in is a large positive value such as .

[0155] S7, Update and for

[0156] .

[0157] in , .

[0158] S8, Update , ,in , is through Substitution Got it.

[0159] S9, Update and .

[0160] S10. Construct objective function

[0161] .

[0162] renew To obtain Maximum value, using polar coordinates of the position ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is

[0163] .

[0164] Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is .

[0165] S11, Update for

[0166] .

[0167] S12, Update and ,in , is through Substitution Got it.

[0168] S13, Update and .

[0169] S14. Construct objective function

[0170] .

[0171] renew To obtain Maximum value, using polar coordinates of the position ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is

[0172] .

[0173] Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is .

[0174] S15, Update for

[0175] .

[0176] S16, Update and for .

[0177] S17, Calculation variable node News from

[0178] .

[0179] Determine variable node Whether the message at the location has converged, if it has converged, the device position estimate is obtained ; If not converged, return to step S7.

[0180] 1. Specific application fields or related products of the present invention.

[0181] An embodiment of the present invention provides a multi-user device near-field positioning system based on an antenna array based on the multi-user device near-field positioning method based on an antenna array, and the system specifically includes:

[0182] Parameter setting module, the system operating frequency is 6 GHz, the wavelength is 0.05m, and the base station is equipped with Antennas are arranged in a uniform array with a half-wavelength spacing. The number of base station RF chains is 160; the coordinates of the central antenna unit of the base station antenna array are (0 m, 0 m, 0 m), and the two direction vectors of the base station antenna array are (1, 0, 0) and (0, 1, 0) respectively; the area where the user equipment is located is a cone area with a cone area vertex angle of 120 degrees. The distance between the equipment and the base station array ranges from 5 meters to 10 meters. The base station locates 3 user equipment at the same time; the user terminal equipment is equipped with a single uniform antenna, and the user equipment position is uniformly sampled in the target area; the beamforming matrix The elements of have unit modulus and random phase.

[0183] Initialization module, connected with parameter setting module, ; Get device location Initialization of Device Locations During initialization, the previous Initialization of user locations , ; Construct objective function

[0184] .

[0185] in For device location Function. Using device location Polar coordinate representation of ,in is the distance from the device to the center of the array, (or ) is the device orientation and array Axis (or axis); construct the grid

[0186] .

[0187] For discrete grid Perform an exhaustive search to find the maximum Grid points ;by As a starting point, perform the gradient ascent algorithm and obtain The local maximum of the gradient ascent algorithm is updated as follows:

[0188] .

[0189] Where n is the iteration round, step length , and Determined by the Armijo criterion; after the gradient descent algorithm converges, it is obtained from the convergence point Initialization , the convergence condition of the gradient descent algorithm is .

[0190] ,if , execute the above grid exhaustive search and gradient ascent algorithm again.

[0191] Initialize message mean , , ; Initialize message covariance , in is a large positive value such as .

[0192] The message update module is connected to the initialization module to update and for

[0193] .

[0194] in , .

[0195] renew , ,in , is through Substitution Got it.

[0196] renew and .

[0197] Constructing the objective function

[0198] .

[0199] renew To obtain Maximum value, expressed in polar coordinates ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is

[0200] .

[0201] Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is .

[0202] renew for

[0203] .

[0204] renew and ,in , is through Substitution Got it.

[0205] renew and .

[0206] Constructing the objective function

[0207] .

[0208] renew To obtain Maximum value, using polar coordinates of the position ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is

[0209] ,

[0210] Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is .

[0211] renew for

[0212] .

[0213] renew and for

[0214] .

[0215] The edge message calculation module is connected to the target function construction module to calculate the variable node News from

[0216]

[0217] Determine variable node Whether the message at the location has converged, if it has converged, the device position estimate is obtained ; If it does not converge, execute the above grid exhaustive search and gradient ascent algorithm again.

[0218] An embodiment of the present invention provides a computer device, which includes a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the multi-user device near-field positioning method based on an antenna array.

[0219] An embodiment of the present invention provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the multi-user device near-field positioning method based on an antenna array.

[0220] An embodiment of the present invention provides an information data processing terminal, which is used to implement the multi-user device near-field positioning system based on the antenna array.

[0221] 2. Relevant evidence of the technical effects obtained by the embodiments of the present invention.

[0222] For specific implementation results, see Figure 3 , Figure 4 and Figure 5 . Figure 3 The figure shows the impact of the number of antenna elements in the base station antenna array and the array grouping strategy on the positioning performance. Equipping the antenna array with more antenna elements can improve the spatial resolution of the array, thereby improving the positioning accuracy of the terminal device. In addition, Figure 3 As shown in Figure 2, with the same number of array antennas, using different subarray division strategies has similar positioning performance. When , the 3×20 and 4×15 subarray partitioning strategies have similar performance; similarly, for the number of base station antennas ,The 3×25 and 5×15 subarray partitioning strategies have similar performance. Figure 4The positioning performance of the proposed multi-user positioning algorithm for users in different areas is evaluated, where the distance ranges between the base station and the user equipment are set to [5,10], [10,15], and [15,20], respectively. As the distance between the base station and the user equipment increases, the positioning accuracy gradually decreases. Figure 5 In the paper, we compare with two existing baseline algorithms, where the baseline algorithm ES-GA uses grid search and gradient descent to obtain user positions; the baseline algorithm P-SIGW is given by the document "M.Cui and L. Dai, "Channel estimation for extremely large-scale MIMO: far-fieldor near-field?" IEEE Transactions on Communications, vol. 70, no. 4, pp.2663–2677, 2022.". The multi-user positioning accuracy of the present invention is higher than that of the baseline scheme and can approach the lower bound of the error of the positioning performance. In an embodiment, the positioning accuracy is generally at the decimeter level, and the positioning accuracy can reach the centimeter level under a high signal-to-noise ratio.

[0223] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. It can be understood by a person of ordinary skill in the art that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0224] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. A near-field positioning method for multi-user devices based on antenna arrays, characterized in that: The method comprises the following steps: 1) Using a base station with multiple antenna elements and multiple single-antenna terminal devices. The base station is equipped with multiple RF chains and adopts a fully connected analog beamforming architecture or a partially connected analog beamforming architecture; 2) Model the channel coefficients between the user terminal and the base station antenna, and describe the channel coefficients through Euclidean distance, free space path loss, antenna radiation gain, and complex gain coefficient; 3) Divide the antenna array of the base station into multiple sub-arrays and construct a received signal model based on the channel coefficients of the sub-arrays; 4) A joint probability density function is constructed based on the channel model, and multi-user near-field positioning of user equipment is performed through a factor graph. The multi-user near-field positioning method includes the following steps: S1. Order ; S2. Get device location Initialization of Device Locations During initialization, the previous Initialization of user locations , ; Construct objective function ， in For device location Function that uses the device location Polar coordinate representation of ,in is the distance from the device to the center of the array, or is the device orientation and array Axis or Cosine of the angle between the axes; constructing the grid ； S3. Discrete grid Perform an exhaustive search to find the maximum Grid points ; S4, As a starting point, perform the gradient ascent algorithm and obtain The local maximum of the gradient ascent algorithm is updated as follows: Where n is the iteration round, step length , and Determined by the Armijo criterion; after the gradient descent algorithm converges, it is obtained from the convergence point Initialization , the convergence condition of the gradient descent algorithm is ; S5, ,if , return to S2; S6. Initialize message mean , , ; Initialize message covariance , in is a large positive value such as ; S7, Update and for in , ; S8, Update , ,in , is through Substitution obtained; S9, Update and ; S10. Construct objective function renew To obtain Maximum value, expressed in polar coordinates ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is ， Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is ; S11, Update for ； S12, Update and ,in , is through Substitution obtained; S13, Update and ; S14. Construct objective function ， renew To obtain Maximum value, expressed in polar coordinates ;make , ,by The polar coordinates corresponding to the old value of are used as the starting point. and Execute the gradient ascent algorithm. The update formula of the gradient ascent algorithm is ， Where n is the iteration round, step length , Determined by the Armijo criterion; after the gradient descent algorithm converges, the and renew The convergence condition of the gradient descent algorithm is ; S15, Update for ； S16, Update and for ； S17, Calculation variable node News from ， Determine variable node Whether the message at the location has converged, if it has converged, the device position estimate is obtained ; If not converged, return to step S7; The positioning method further comprises the following steps: 1) Initialize the user equipment position and represent the user equipment in polar coordinates; 2) Find the grid point that maximizes the objective function by exhaustively searching the discrete grid; 3) Based on the exhaustive search results, a gradient ascent algorithm is executed to obtain the local maximum of the user device location; 4) Repeat the above process until all user device locations are initialized.

2. The multi-user device near-field positioning method according to claim 1, characterized in that: The method includes: One with A base station with antenna units and A single-antenna terminal device; the base station is equipped with The fully connected analog beamforming architecture is used between the RF chain and the array antenna; the partially connected analog beamforming architecture is used as a special case of the system under consideration; in the three-dimensional Cartesian coordinate system, The user terminal position coordinates are marked as , No. The base station antenna unit position coordinates are marked as At this time, The user terminal and The channel coefficients between base station antennas are modeled as ， in For the The user terminal and The Euclidean distance between base station antennas, Corresponding free space path loss, Corresponding to The base station antenna unit is The antenna radiation gain in the direction of the user terminal is a function of the user terminal position, The complex gain coefficient of the channel that is independent of the terminal position is included; the planar antenna array of the base station is divided into subarrays, each with antenna units, satisfying , ;No. The user terminal and The channels of the subarrays are written as ,in For the The channel coefficients of the subarray reference antennas, is a vector-valued function of the device position; the reference antenna channel coefficients for different subarrays are written as vectors , which can be further written as ,in is a vector value function of the device position; the received signal model of the base station receiving array is , in For the Device reference array gain, To send a signal, is the transmit power, It is The beamforming matrix of the subarrays is is the Gaussian white noise at the receiving end; by defining the auxiliary variable , , , , we get an equivalent signal model ； Carry out probability modeling on the above signal model; and The prior distribution of is modeled as a complex Gaussian distribution and a Gaussian distribution: ， because Is the device location Function, get the conditional probability density function: ， in , , is the Dirac delta function, and the observation likelihood function is: ， Then the joint probability density function is written as: ， in , ; Based on the joint probability density function, draw its corresponding factor graph representation, where the hollow circular box and the black rectangle represent the variable node and the check node respectively; the variable node includes , and ; Check nodes include , , (denoted as )and (denoted as ); the message calculation and update will be performed on the factor graph, and then a near-field positioning algorithm for multi-user devices is proposed.

3. A multi-user device near-field positioning system based on an antenna array based on the multi-user device near-field positioning method based on an antenna array as claimed in claim 1, characterized in that: The system specifically includes: Parameter setting module, the system operating frequency is 6 GHz, the wavelength is 0.05m, and the base station is equipped with antennas are arranged in a uniform array with a half-wavelength spacing; the number of base station RF chains is 160; the coordinates of the central antenna unit of the base station antenna array are (0 m, 0 m, 0 m), and the two direction vectors of the base station antenna array are (1, 0, 0) and (0, 1, 0) respectively; the area where the user equipment is located is a cone area with a vertex angle of 120 degrees. The distance between the equipment and the base station array ranges from 5 meters to 10 meters. The base station locates 3 user equipment at the same time; the user terminal equipment is equipped with a single uniform antenna, and the user equipment position is uniformly sampled in the target area; the beamforming matrix The elements of have unit modulus and random phase; Initialization module, connected with parameter setting module, ; Get device location Initialization of Initialize message mean , , ; Initialize message covariance , in is a large positive value such as ; The message update module is connected to the initialization module to update and ;renew , ;renew and ;renew ; renew for renew and ,renew and ;renew ; renew for renew and for The edge message calculation module is connected to the target function construction module to calculate the variable node The message at the variable node is judged Whether the message at the location has converged, if it has converged, the device position estimate is obtained ; If it does not converge, execute the above grid exhaustive search and gradient ascent algorithm again.

4. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the multi-user device near-field positioning method based on antenna array as claimed in claim 1.

5. A computer-readable storage medium storing a computer program, wherein when the computer program is executed by a processor, the processor executes the steps of the near-field positioning method for a multi-user device based on an antenna array as claimed in claim 1.

6. An information data processing terminal, characterized in that: The information data processing terminal is used to implement the multi-user device near-field positioning system based on antenna array as described in claim 3.

Images