Multi-target positioning method and system based on intelligent reflection surface collaborative cellular network

Abstract

The invention discloses a multi-target positioning method and system based on an intelligent reflection surface collaborative cellular network in the field of target sensing, and aims to solve the problems of vehicle association complexity and weak echo signals of a traditional cellular base station in a complex urban environment. The method comprises the following steps: acquiring priori two-dimensional position information of a vehicle, and optimizing a matching association relationship between a cellular network base station and an intelligent reflecting surface deployed at the top end of the vehicle; inputting the optimized matching incidence relation into a pre-constructed radar sensing model to obtain an echo signal received by the cellular network base station; estimating an estimated distance between the target vehicle and the cellular network base station by using the echo signal; obtaining a Cramer-Rao bound of the position estimation error according to the estimation distance; and taking the maximum value of the Cramer-Rao bound of the position estimation error as a target function to carry out minimization optimization, and outputting the optimized target vehicle position information. According to the invention, positioning precision and reliability are substantially improved, and signal interference of the base station is reduced.

Description

Technical Field

[0001] This invention relates to a multi-target localization method and system based on a smart reflective surface cooperative cellular network, belonging to the field of target perception technology. Background Technology

[0002] The applications of autonomous driving and intelligent transportation have significantly increased the demand for high-precision, low-latency positioning services. Currently, positioning technologies can be broadly categorized into device-based sensing and deviceless sensing. Device-based sensing requires the target to transmit and receive signals, and the transmitting and receiving ends must strictly adhere to time synchronization, such as the Global Positioning System (GPS) and wireless sensor networks. Deviceless sensing, on the other hand, achieves positioning by analyzing the echo signals reflected from the target, without the need for target-side transceivers, such as radar. Furthermore, GPS positioning performance is susceptible to obstruction, especially in environments with dense obstacles. Therefore, radar sensing is more practical and efficient in high-density, high-mobility vehicle-to-everything (V2X) scenarios. Dedicated radar equipment is primarily used in the military field and is not widely deployed in urban environments. Besides radar, cellular networks, due to the continuous expansion of their communication frequency bands, possess the potential for target detection and distance measurement, and with their wide coverage and widespread deployment in cities, leveraging cellular networks to improve target positioning performance has attracted increasing attention.

[0003] With the increasing demand for centimeter-level positioning accuracy for L4 / L5 autonomous vehicles to achieve precise lane keeping and obstacle avoidance navigation, and the requirement for millisecond-level response speed in vehicle-to-everything (V2X) systems, traditional cellular base station (BS) positioning technology is struggling to meet the reliability requirements in complex urban environments. Specifically, (1) Measurement association complexity: In multi-target scenarios, echo signals are difficult to match the target correctly, leading to ghost target interference; (2) Weak echo signals: The radar cross-sectional area of ​​moving targets (such as vehicles) is small, resulting in insufficient signal echo strength.

[0004] To address this issue, a method has been proposed to deploy receiving sensors on intelligent reflective surfaces (IRS) to reduce echo signal path loss and improve the positioning accuracy of weak targets. However, this approach focuses on deploying IRS at fixed locations and does not achieve high-resolution, high-reliability positioning performance of cellular networks in high-mobility scenarios, nor does it reduce interference between different measurements. Therefore, it is necessary to propose a novel multi-target positioning method and system that deploys IRS on vehicle surfaces to solve the problems faced by cellular network base station (BS) positioning. Summary of the Invention

[0005] The purpose of this invention is to provide a multi-target localization method and system based on intelligent reflective surfaces and cooperative cellular networks. This method can solve the problems of "ghost target" interference caused by chaotic multi-target signal association in complex urban environments and localization failure caused by weak vehicle echo signals faced by traditional cellular base stations. Specifically, by dynamically optimizing the unique matching association between the base station and the vehicle-mounted intelligent reflective surface, combined with beam flattening technology and joint resource allocation, high-precision, low-interference multi-target localization is achieved, meeting the requirements of centimeter-level accuracy and millisecond-level response for autonomous driving.

[0006] To solve the above-mentioned technical problems, the present invention is implemented using the following technical solution.

[0007] On one hand, the present invention provides a multi-target localization method based on a smart reflective surface cooperative cellular network, comprising: Obtain the vehicle's prior two-dimensional location information; Based on the vehicle's prior two-dimensional location information, optimize the matching relationship between the cellular network base station and the smart reflective surface deployed on the vehicle's roof; The optimized matching relationship is input into the pre-built radar perception model to obtain the echo signal received by the cellular network base station; Using the echo signal, the estimated distance between the target vehicle and the cellular network base station is estimated; Based on the estimated distance between the target vehicle and the cellular network base station, obtain the Cramero bound of the position estimation error; The maximum value of the Cramer-Rao bound of the position estimation error is used as the objective function to minimize and optimize, and the optimized target vehicle position information is output.

[0008] In conjunction with the first aspect, the optimized matching relationship between the cellular network base station and the smart reflective surface deployed on the vehicle roof further includes: Based on the vehicle's prior two-dimensional position information, the preset time Δ T Divided into N time slots, and the vehicle position is in the Δ T The internal temperature remains constant; Introducing binary variable set Through the set of binary variables A unique matching association is established between the cellular network base station and the smart reflective surface deployed on the roof of the vehicle; wherein, when When, it indicates that a unique matching association is established between the m-th cellular network base station in the n-th time slot and the smart reflective surface deployed on the top of the vehicle; when When, it means that no unique matching relationship has been established between the m-th cellular network base station in the n-th time slot and the smart reflective surface deployed on the top of the vehicle; Based on the established unique matching association, the distance information received by the cellular network base station is matched and associated with the location information of the target vehicle.

[0009] In conjunction with the first aspect, further, through the aforementioned binary variable set The unique matching association between the cellular network base station and the smart reflective surface deployed on the roof of the vehicle satisfies the following conditions: ; Where M represents the number of cellular network base stations; m represents the index value of the cellular network base station; This represents a set of binary variables; k represents the vehicle index value; K represents the number of vehicles; n represents the time slot index value; and N represents the number of time slots to be divided.

[0010] In conjunction with the first aspect, the expression for matching and associating the distance information received by the cellular network base station with the location information of the target vehicle is as follows: ; in, This represents the distance information between the m-th cellular network base station and the k-th smart reflective surface; This indicates the position information of a cellular network base station along the x-axis. This indicates the position information of the cellular network base station along the y-axis. This indicates the vehicle's position along the x-axis. This indicates the vehicle's position along the y-axis. Indicates the height of the cellular network base station; This indicates the height of the intelligent reflective surface.

[0011] In conjunction with the first aspect, further, the acquisition of the Cramer-Rao bound of the position estimation error includes: Based on the radar sensing and measurement model, the echo signal received by the m-th cellular network base station in the n-th time slot is obtained; Using the matching and association relationships identified by the binary variable set, the target vehicle location information corresponding to the echo signal is identified; The echo signal is processed using the maximum likelihood estimation method to obtain the estimated distance between the m-th cellular network base station and the k-th target vehicle. ; Based on the estimated distance The Cramer-Rao boundary of the position estimation error of the intelligent reflective surface on the k-th vehicle is calculated using the Fisher information matrix. .

[0012] In conjunction with the first aspect, the expression for the echo signal received by the m-th cellular network base station in the n-th time slot is further as follows: , in, This represents the echo signal received by the m-th base station in the n-th time slot; Represents a set of binary variables; This represents the cascaded channel from base station m to smart reflector k and back to base station m; Indicates the transmission power; This represents the transmitted signal of base station m; t represents time. The round-trip time of the echo signal received by the m-th cellular base station is represented by m; M represents the number of cellular base stations; m and This represents the index value of different cellular network base stations; k and K represents different vehicle index values; K represents the number of vehicles. Indicates base station The relationship with vehicle k; Indicates base station The cascaded channel from the intelligent reflector k to the base station m; Indicates base station The transmitted signal; Indicates from the first The transmission delay from the k-th smart reflective surface to the m-th cellular network base station; This indicates that the m-th cellular network base station and the... Channels between cellular network base stations; Indicates base station The transmission delay to base station m; Indicates the distance from base station m to the smart reflector. Then, the cascaded channel of base station m; This indicates additive interference.

[0013] In conjunction with the first aspect, the expression for the Cramer-Rao bound of the position estimation error of the intelligent reflective surface mounted on the k-th vehicle is as follows: ; in, The value represents the Cramero bound; M represents the number of cellular network base stations; m represents the index value of the cellular network base station. This represents the elevation angle of the m-th cellular network base station relative to the k-th vehicle; This represents the variance of the ranging error of the m-th cellular network base station; This represents the azimuth angle of the m-th cellular network base station relative to the k-th vehicle.

[0014] In conjunction with the first aspect, further, the maximum value of the Cramer-Rao bound of the position estimation error is taken as the expression of the objective function as: ; in, This represents the reflection unit coefficient matrix of the intelligent reflective surface mounted on the k-th vehicle in the n-th time slot; This represents the time slot allocation ratio variable; Indicates the Clamello boundary; This indicates time slot allocation; M represents the number of cellular network base stations; m represents the index value of the cellular network base station. This represents a set of binary variables; k represents the vehicle index value; K represents the number of vehicles; n represents the time slot index value; and N represents the number of time slots divided. Represents the set of positive integers; This represents the reflection phase offset of the l-th reflection unit of the k-th smart reflective surface.

[0015] In conjunction with the first aspect, further optimization of the target vehicle location information is output, including: Based on the error range of the vehicle's prior position, determine the spatial angle range between the intelligent reflective surface mounted on vehicle k and the m-th cellular network base station; The intelligent reflective surface is divided into multiple sub-arrays according to the spatial angle range, and the coefficients of the reflective units in each sub-array are set. Based on the coefficients of the reflection unit, the optimized matching relationship between the cellular network base station and the vehicle, as well as the time slot allocation ratio variable, are obtained through a two-step segmented method. Based on the coefficients of the reflection unit, the optimized matching relationship between the cellular network base station and the vehicle, and the time slot allocation ratio variable, the optimal target vehicle location information is obtained.

[0016] Secondly, a multi-target positioning system based on a smart reflective surface cooperative cellular network includes: The location acquisition module is used to acquire the vehicle's prior two-dimensional location information; The association optimization module is used to optimize the matching association between the cellular network base station and the smart reflective surface deployed on the top of the vehicle based on the vehicle's prior two-dimensional location information. The echo signal module is used to input the optimized matching correlation into the pre-built radar perception model in order to obtain the echo signal received by the cellular network base station. The distance estimation module is used to estimate the distance between the target vehicle and the cellular network base station using the echo signal; The error boundary analysis module is used to obtain the Cramer-Rao boundary of the position estimation error based on the estimated distance between the target vehicle and the cellular network base station. The joint optimization decision module is used to minimize the maximum value of the Cramer-Rao bound of the position estimation error as the objective function and output the optimized target vehicle position information.

[0017] Compared with the prior art, the beneficial effects achieved by the present invention are as follows: This invention dynamically optimizes the unique matching association between the base station (BS) and the inter-station (IRS) based on the vehicle's prior position (each IRS in a time slot is associated with only one BS). Combined with beam flattening technology, it designs the IRS phase offset matrix, significantly reducing the signal association complexity and "ghost target" interference in multi-target scenarios. It uses a radar perception model to capture enhanced echo signals, accurately calculates the distance between the vehicle and the base station through maximum likelihood estimation, and derives the Cramer-Rao bound (CRLB) for position estimation error. Finally, with the goal of minimizing the maximum CRLB, it jointly optimizes the association relationship, phase offset, and time resource allocation. While reducing the base station requirement and the number of time slots, it effectively improves the positioning accuracy and reliability in weak signal environments, meeting the centimeter-level positioning and millisecond-level response requirements of autonomous driving. Attached Figure Description

[0018] Figure 1 The image shows a vehicle-to-everything (V2X) application scenario for IRS-assisted cellular network positioning provided in an embodiment of the present invention. Figure 2 The diagram shown is a schematic of the IRS cooperative cellular network positioning scheme provided in an embodiment of the present invention. Figure 3 The image shown is an illustration of the IRS-assisted cellular network positioning scheme for single-target scenarios provided by an embodiment of the present invention. Figure 4 The image shown is an illustration of the IRS cooperative cellular network positioning scheme for multi-target scenarios provided by an embodiment of the present invention. Figure 5 The diagram shown is a flowchart of a multi-target localization method based on a smart reflective surface cooperative cellular network provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical solution of the present invention, rather than limitations thereof. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0020] The term "and / or" simply describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A alone, A and B simultaneously, or B alone. Additionally, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0021] Considering vehicle-to-everything (V2X) applications, the IRS (Infrastructure Reference Controller) is deployed on the top of the vehicle to assist the BS (Base Station) in providing location services. See also... Figure 1 M BSs are used to provide location services for K vehicles. Example 1

[0022] This embodiment introduces a multi-target localization method based on a smart reflective surface cooperative cellular network, including the following steps: Step S1: Obtain the vehicle's prior two-dimensional location information; Step S2: Based on the vehicle's prior two-dimensional location information, determine the matching association between the BS and the IRS deployed on the vehicle's roof; wherein, the association relationship must satisfy the following constraints: a. Within each time slot, a single IRS can be associated with at most one BS; b. A single BS can serve at most one IRS within each time slot; c. A single IRS-BS pair is associated only once.

[0023] Step S3: Input the matching correlation into the pre-built radar perception model to obtain the echo signal received by the BS; Step S4: Using the echo signal, estimate the distance between the target vehicle and the BS; Step S5: Obtain the Cramer-Rao boundary of the position estimation error based on the estimated distance between the target vehicle and the BS; Step S6: Minimize the maximum value of the Cramer-Rao bound of the position estimation error as the objective function, and output the optimized two-dimensional position information of the vehicle.

[0024] In summary, this embodiment completely eliminates "ghost target" interference in multi-target scenarios by matching association constraints, ensuring the unique correspondence between the echo signal and the vehicle; it utilizes the IRS phase offset matrix to directionally enhance the reflected signal power of the target base station, effectively overcoming the problem of weak vehicle echo; and it combines beam flattening technology to cover uncertain location intervals, achieving centimeter-level positioning accuracy through joint optimization of association relationships and time slot allocation. This significantly improves performance compared to the traditional Closest scheme, thus meeting the stringent requirements of L5 autonomous driving for high reliability and low latency. Example 2

[0025] Step S1: Obtain the vehicle's prior two-dimensional location information; In the vehicle-to-everything (V2X) application scenario of this invention, each BS is equipped with a single transmitting antenna and a single receiving antenna. Since the relative position of the IRS to the target vehicle is constant, the position coordinates of the IRS can be equated to the position coordinates of the vehicle. While deploying multiple IRSs on the target vehicle can acquire its attitude information, this invention only considers the case of deploying a single IRS on the top of the target vehicle.

[0026] After obtaining the prior two-dimensional location information of the vehicle, in order to solve the problem of the association between the IRS and BS on the target vehicle, this invention proposes a multi-target localization method based on IRS cooperative cellular network. By designing IRS phase offset, the direction of the echo signal is actively controlled, thereby promoting the association between IRS and BS.

[0027] Step S2: Based on the prior two-dimensional location information of the vehicle, determine the matching association between the BS and the IRS deployed on the top of the vehicle; See Figure 2 Based on the vehicle's prior two-dimensional location information, the matching and association relationship between the BS and the IRS deployed on the vehicle's roof is optimized, specifically including the following steps: Step S21: Set the preset time Δ T Divided into N time slots, and the vehicle position is within Δ T The time slot remains constant within Δ; where the nth time slot is within Δ T The proportion in is denoted as ,satisfy Given the given conditions, the dwell time in the nth time slot is... .

[0028] Step S22: Introduce a set of binary variables Through binary variable set A unique matching association is established between the identifier BS and the IRS deployed on the top of the vehicle; where, when When, it indicates that a unique matching association is established between the m-th BS in the n-th time slot and the IRS deployed on the top of the vehicle; when The time interval indicates that the m-th BS in the n-th time slot has not established a unique matching relationship with the IRS deployed on the top of the vehicle, but it can still be associated with other IRSs.

[0029] The established unique matching association can effectively reduce the complexity of distance estimation and target localization. Furthermore, to establish a unique association between the BS and the measured distance and to avoid mutual interference between the BS and the IRS, this unique matching association satisfies the following three matching constraints: (1) Where M represents the number of BS; m represents the index value of the BS; This represents a set of binary variables; k represents the vehicle index value; K represents the number of vehicles; n represents the time slot index value; and N represents the number of time slots to be divided.

[0030] Step S23: Based on the established unique matching association, match and associate the distance information received by the BS with the location information of the target vehicle.

[0031] The matching expression for distance information and the location information of the target vehicle is as follows: (2) in, This represents the distance information between the m-th cellular network base station and the k-th smart reflective surface; This indicates the position information of a cellular network base station along the x-axis. This indicates the position information of the cellular network base station along the y-axis. This indicates the vehicle's position along the x-axis. This indicates the vehicle's position along the y-axis. Indicates the height of the cellular network base station; This indicates the height of the intelligent reflective surface.

[0032] By adopting the inequality constraint of the above formula (1), the distance signal received by the BS as represented by formula (2) can be matched and associated with a specific target vehicle, thereby effectively improving the positioning accuracy and avoiding the complexity of target association.

[0033] Step S3: Input the matching relationship into the pre-built radar perception model to obtain the echo signal received by the BS; Based on the matching and association scheme in step 2, a radar sensing and measurement model is constructed. Let the radar signal transmitted by the m-th BS be... Furthermore, the radar signal is normalized, and its expression is: (3) Where W represents the total number of symbols processed during the signal processing time; Indicates the time required for processing a single symbol; This represents the radar signal transmitted by the m-th BS.

[0034] Furthermore, the first k The IRS relative to the first m The direction information of each BS is provided by { φ k,m , ϕ k,m} provides, among which φ k,m , ϕ k,m They respectively represent the connection of the firstk The IRS and the m The azimuth and elevation angles of the geometric path of each BS. k The IRS and the m The channel power gain between two base stations (BS) can be expressed as... ,in It is the channel power gain at a reference distance of 1 meter. and They are the first k The IRS and the m The downlink and uplink channel vectors between BSs are defined as follows: (5) (6) in , and The first k The IRS and the m BS between edges x shaft and y Spatial frequency angle of the axis For the set of complex numbers, and These are the KR product and transpose operators, respectively.

[0035] At this point, the echo signal received by the m-th BS in the n-th time slot consists of five parts: the echo signal reflected by the matched BS and then reflected by the IRS + the echo signals transmitted by other BSs and then reflected by the IRS + the interference signals from other BSs + the interference signals from other IRSs + Gaussian noise, and its expression is as follows: (4) in, This represents the echo signal received by the m-th base station in the n-th time slot; Represents a set of binary variables; This represents the cascaded channel from base station m to smart reflector k and back to base station m; Indicates the transmission power; This represents the transmitted signal of base station m; t represents time. The round-trip time of the echo signal received by the m-th cellular base station is represented by m; M represents the number of cellular base stations; m and This represents the index value of different cellular network base stations; k and K represents different vehicle index values; K represents the number of vehicles. Indicates base station The relationship with vehicle k; Indicates base station The cascaded channel from the intelligent reflector k to the base station m; Indicates base station The transmitted signal; Indicates from the first The transmission delay from the k-th smart reflective surface to the m-th cellular network base station; This indicates that the m-th cellular network base station and the... Channels between cellular network base stations; Indicates base station The transmission delay to base station m; Indicates the distance from base station m to the smart reflector. Then, the cascaded channel of base station m; This indicates additive interference.

[0036] Step S4: Using the echo signal, estimate the distance between the target vehicle and the BS; Step S5: Obtain the Cramer-Rao boundary of the position estimation error based on the estimated distance; Specifically, calculating the Cramer-Rao bound of the position estimation error includes the following steps: Based on the radar sensing and measurement model constructed in step S3, the echo signal received by the m-th BS in the n-th time slot is obtained. By utilizing the matching relationships identified by binary variable sets, the location information of the target vehicle corresponding to the echo signal can be identified; The maximum likelihood estimation method is used to estimate and process the echo signal to obtain the estimated distance between the m-th BS and the k-th target vehicle. Its expression is: (5) in, This represents the estimated distance between the m-th BS and the k-th target vehicle; This represents the actual distance between the m-th BS and the k-th target vehicle; This indicates the distance estimation error.

[0037] Based on the estimated distance The Cramer-Rao boundary of the IRS position estimation error of the k-th vehicle is calculated using the Fisher information matrix. .

[0038] The Cramer-Rao boundary of the IRS position estimation error carried by the k-th vehicle. The expression is: (6) in, The value represents the Cramero bound; M represents the number of cellular network base stations; m represents the index value of the cellular network base station. This represents the elevation angle of the m-th cellular network base station relative to the k-th vehicle; This represents the variance of the ranging error of the m-th cellular network base station; This represents the azimuth angle of the m-th cellular network base station relative to the k-th vehicle.

[0039] Step S6: Minimize the maximum value of the Cramer-Rao boundary of the position estimation error as the objective function, and output the optimized two-dimensional position information of the vehicle.

[0040] This invention achieves this by jointly optimizing the BS-IRS correlation and the phase offset matrix of the IRS. and time allocation The objective function for minimizing the maximum Cramer-Rao bound of the position estimation error is as follows: (7) in, This represents the reflection unit coefficient matrix of the IRS mounted on the k-th vehicle in the n-th time slot; This represents the time slot allocation ratio variable; Indicates the Clamello boundary; This indicates time slot allocation; M represents the number of BSs; m represents the index value of the BS; This represents a set of binary variables; k represents the vehicle index value; K represents the number of vehicles; n represents the time slot index value; and N represents the number of time slots divided. Represents the set of positive integers; This represents the reflection phase offset of the l-th reflection unit of the k-th IRS.

[0041] Since formula (7) is non-convex and the variables to be solved are tightly coupled, it is difficult to solve directly. Therefore, beam flattening technology is used to solve it in different scenarios, as follows: (1) Single target localization scenario Based on the error range of the vehicle's prior position, determine the spatial angle range between the IRS mounted on the vehicle and the m-th BS. The IRS is divided into multiple sub-arrays according to the spatial angle range, and the coefficients of the reflection units in each sub-array are set. Substitute the reflection unit coefficients into the objective function of the above formula (7), and use the Polyblock-based algorithm to solve the objective function optimally to obtain the optimized time slot allocation ratio variable; The optimal target vehicle position information is obtained based on the coefficients of the reflection unit and the optimized time slot allocation ratio variable.

[0042] Specifically, when only a single vehicle is located, i.e., K=1, the solution for the IRS phase offset and time allocation design is obtained. If more than M time slots are used, the BS may be allocated multiple time slots. In this case, merging time slots matched and associated with the same BS will not affect sensing performance. Therefore, the number of time slots is set to N=M (number of base stations) and variables are ignored. Meanwhile, the IRS is matched and associated with the BS according to the BS index order, thus simplifying formula (7) to: (8) in, Indicates intelligent reflection unit Phase shift in time slot m; Indicates the index value of the reflective element; L represents the total number of IRS reflective elements; This represents the proportion allocated to the m-th time slot.

[0043] Furthermore, due to the uncertainty of position addressed by beam flattening technology, the vehicle's actual position fluctuates around its prior position, leading to variations from the first... m The spatial resolution angle from BS to IRS (i.e.) and Located within two angular ranges, respectively represented as and ,in , , , .

[0044] To solve the above problem, the following method is adopted: Divide the IRS along the x-axis and y-axis into Each subarray; wherein the subarray size is determined by the range of angular uncertainty: (9) (10) in, Indicates the number of subarrays along the x-axis; Indicates the number of reflecting units along the x-axis; This represents the maximum azimuth angle between base station m and the target; This represents the minimum azimuth angle between base station m and the target; Indicates the number of subarrays along the y-axis; This indicates the number of reflecting units along the y-axis. This represents the maximum elevation angle between base station m and the target; This represents the minimum elevation angle between base station m and the target.

[0045] Each subarray contains along the x-axis The elements and along the y-axis Each element.

[0046] Coverage beamwidth along the x-axis and along the y-axis The coverage angle range of each subarray beam can be represented as: (11) (12) in, Represents the first in the x-axis direction The maximum azimuth angle between each subarray and the base station; This represents the maximum azimuth angle between the first subarray and the base station along the x-axis. This indicates the number of reflection units in each subarray along the x-axis. Represents the first in the x-axis direction The maximum elevation angle between the subarray and the base station; This represents the maximum elevation angle between the first subarray and the base station along the x-axis. This indicates the number of reflection units in each subarray along the y-axis.

[0047] It should be noted that if the error range of the target vehicle's position is less than the beamwidth, all beams will be aligned with the vehicle's prior position.

[0048] The optimal solution for interactive two-slot allocation when M=2 and Its expression is: (13) (14) in, This indicates the optimal time allocation ratio for base station 1; This represents the signal-to-noise ratio of the echo signal received by base station 1; This represents the signal-to-noise ratio of the echo signal received by base station 2; This indicates the optimal time allocation ratio for base station 2.

[0049] It should be noted that formula (8) is a monotonic optimization problem relative to time allocation. Therefore, if M > 2, the Polyblock algorithm is used to solve the monotonic optimization problem. The optimal solution is obtained by using the Polyblock-based algorithm to obtain the parameters to be estimated. See [link to relevant documentation]. Figure 3 The specific system parameters are as follows: , , , , , , The results reflect the power of the two-dimensional echo signal to illustrate the effectiveness of the echo signal control.

[0050] in, Figure 3 The first figure shows the locations of 10 BS and IRS, and combined with the remaining three figures, it can be seen that only the 2nd, 4th and 8th BSs that are matched and associated were assigned an equal proportion of time.

[0051] (2) Multi-target localization scenario Based on the error range of the vehicle's prior position, determine the spatial angle range between the intelligent reflective surface mounted on vehicle k and the m-th cellular network base station. and ,in and These are the upper and lower bounds of the spatial upper azimuth angle, respectively. and These are the upper and lower bounds of the spatial pitch angle, respectively. Based on the spatial angle range, the intelligent reflective surface is divided into multiple sub-arrays along the x-axis and y-axis, and the coefficients of the reflective units in each sub-array are set; specifically, when the IRS is matched and associated with m BSs, the number and size of the sub-arrays along the x-axis are respectively... and The number and size of the subarrays along the y-axis are respectively and By using beam-flattening technology and incorporating , , , Parameters, aligned with the spatial angle range and To set the beam direction of the subarray; Based on the coefficients of the reflection unit, the optimized matching relationship between the cellular network base station and the vehicle, as well as the time slot allocation ratio variable, are obtained through a piecewise two-step method. ; The optimal target vehicle location information is obtained based on the coefficients of the reflective unit, the optimized matching relationship between the cellular network base station and the vehicle, and the time slot allocation ratio variable.

[0052] Specifically, when expanding the beam to multiple targets, the problem of unmatched IRS interference needs to be solved. This invention adds the following conditions to form a new non-convex problem and proposes a segmented two-step method to solve the target matching association and time allocation problem. This method selects only two BSs with the minimum CRLB value to locate each vehicle, thereby reducing the total number of time slots and algorithm complexity.

[0053] The new conditions are as follows: a. Add new interference constraints: (15) in, Indicates the nth time slot The IRS and the first The relationships between individual business units (BS); Indicates whether the reflected signal of the IRS affects the k-th IRS when it is associated with the m-th BS. The BS generates interference.

[0054] b. Add intermediate variables

[0055] Integrating the constraints of a and b with formula (7), the updated objective function expression is: (16) in, Indicates an intermediate variable.

[0056] Further, see Figure 5 The process of reducing complexity using a two-step, segmented approach is as follows: Step 1: Initialize the total number of time slots N, and input the base station location and the prior location information of K vehicles; Step 2: Based on the initial information, select two base stations with the minimum CRLB for each vehicle, and calculate the optimal time slot allocation ratio according to formulas (13) and (14), where the time slot allocation ratio for vehicle k is denoted as... and ; Step 3: If N < 2K, then select Then input the correlation result into the time slot; otherwise, proceed to step 5. Step 4: If satisfied If yes, continue inputting the associated results; otherwise, let... N = N +1, add a new time slot, sort the BS and IRS matching associations according to the normalized time ratio, and re-execute Step 2; Step 5: Solve for the optimal time slot allocation using the MO algorithm.

[0057] See Figure 4 The positioning performance was compared under different numbers of IRS elements. The specific system parameters are as follows: , , , , , The results show that Closest, Time Division, and the CRLB of the proposed solution increase with the number of targets. K The CRLB increases monotonically with the increase of the number of targets. For the Closest scheme, when the number of targets is small, the CRLB increases faster due to more severe interference and reduced time resource utilization efficiency. The proposed scheme can bring greater positioning performance gains than the Closest scheme when there are more targets. As the number of targets increases, the Closest scheme only considers the nearest BS, which inevitably increases the probability that different targets are associated with the same BS, thus requiring more time slots to achieve interference-free sensing between different targets. Example 3

[0058] A multi-target localization system based on a smart reflective surface cooperative cellular network includes: The location acquisition module is used to acquire the vehicle's prior two-dimensional location information; The association optimization module is used to optimize the matching association between the BS and the IRS deployed on the top of the vehicle based on the prior two-dimensional location information of the vehicle. The echo signal module is used to input the optimized matching correlation into the pre-built radar perception model in order to obtain the echo signal received by the BS. The distance estimation module is used to estimate the distance between the target vehicle and the BS using the echo signal; The error boundary analysis module is used to obtain the Cramer-Rao boundary of the position estimation error based on the estimated distance between the target vehicle and the BS. The joint optimization decision module is used to minimize the maximum value of the Cramer-Rao bound of the position estimation error as the objective function and output the optimized target vehicle position information.

[0059] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A multi-target localization method based on a smart reflective surface cooperative cellular network, characterized in that, include: Obtain the vehicle's prior two-dimensional location information; Based on the vehicle's prior two-dimensional location information, optimize the matching relationship between the cellular network base station and the smart reflective surface deployed on the vehicle's roof. The optimized matching relationship is input into the pre-built radar perception model to obtain the echo signal received by the cellular network base station; Using the echo signal, the estimated distance between the target vehicle and the cellular network base station is estimated; Based on the estimated distance between the target vehicle and the cellular network base station, obtain the Cramero bound of the position estimation error; The maximum value of the Cramer-Rao bound of the position estimation error is used as the objective function to minimize and optimize, and the optimized target vehicle position information is output.

2. The multi-target localization method based on intelligent reflective surface cooperative cellular network according to claim 1, characterized in that, The optimized matching relationship between cellular network base stations and smart reflective surfaces deployed on the vehicle roof includes: Based on the vehicle's prior two-dimensional position information, the preset time Δ T Divided into N time slots, and the vehicle position is in the Δ T The internal temperature remains constant; Introducing binary variable set Through the set of binary variables A unique matching association is established between the cellular network base station and the smart reflective surface deployed on the roof of the vehicle; wherein, when When, it indicates that a unique matching association is established between the m-th cellular network base station in the n-th time slot and the smart reflective surface deployed on the top of the vehicle; when When, it means that no unique matching relationship has been established between the m-th cellular network base station in the n-th time slot and the smart reflective surface deployed on the top of the vehicle; Based on the established unique matching association, the distance information received by the cellular network base station is matched and associated with the location information of the target vehicle.

3. The multi-target localization method based on intelligent reflective surface cooperative cellular network according to claim 2, characterized in that, Through the binary variable set The unique matching association between the cellular network base station and the smart reflective surface deployed on the roof of the vehicle satisfies the following conditions: ； Where M represents the number of cellular network base stations; m represents the index value of the cellular network base station; This represents a set of binary variables; k represents the vehicle index value; K represents the number of vehicles; n represents the time slot index value; and N represents the number of time slots to be divided.

4. The multi-target localization method based on intelligent reflective surface cooperative cellular network according to claim 2, characterized in that, The expression for matching and associating the distance information received by the cellular network base station with the location information of the target vehicle is as follows: ； in, This represents the distance information between the m-th cellular network base station and the k-th smart reflective surface; This indicates the position information of a cellular network base station along the x-axis. This indicates the position information of the cellular network base station along the y-axis. This indicates the vehicle's position along the x-axis. This indicates the vehicle's position along the y-axis. Indicates the height of the cellular network base station; This indicates the height of the intelligent reflective surface.

5. The multi-target localization method based on intelligent reflective surface cooperative cellular network according to claim 2, characterized in that, The Cramérod bound for obtaining the position estimation error includes: Based on the radar sensing and measurement model, the echo signal received by the m-th cellular network base station in the n-th time slot is obtained; By utilizing the matching and association relationships identified by the binary variable set, the location information of the target vehicle corresponding to the echo signal is identified; The echo signal is processed using the maximum likelihood estimation method to obtain the estimated distance between the m-th cellular network base station and the k-th target vehicle. ; Based on the estimated distance The Cramer-Rao boundary of the position estimation error of the intelligent reflective surface on the k-th vehicle is calculated using the Fisher information matrix. .

6. The multi-target localization method based on intelligent reflective surface cooperative cellular network according to claim 5, characterized in that, The expression for the echo signal received by the m-th cellular network base station in the n-th time slot is: ； in, This represents the echo signal received by the m-th base station in the n-th time slot; Represents a set of binary variables; This represents the cascaded channel from base station m to smart reflector k and back to base station m; Indicates the transmission power; This represents the transmitted signal of base station m; t represents time. The round-trip time of the echo signal received by the m-th cellular base station is represented by m; M represents the number of cellular base stations; m and This represents the index value of different cellular network base stations; k and K represents different vehicle index values; K represents the number of vehicles. Indicates base station The relationship with vehicle k; Indicates base station The cascaded channel from the intelligent reflector k to the base station m; Indicates base station The transmitted signal; Indicates from the first The transmission delay from the k-th smart reflective surface to the m-th cellular network base station; This indicates that the m-th cellular network base station and the... Channels between cellular network base stations; Indicates base station The transmission delay to base station m; Indicates the distance from base station m to the smart reflector. Then, the cascaded channel of base station m; This indicates additive interference.

7. The multi-target localization method based on intelligent reflective surface cooperative cellular network according to claim 5, characterized in that, The Cramero boundary of the intelligent reflective surface position estimation error on the k-th vehicle. The expression is: ; in, The value represents the Cramero bound; M represents the number of cellular network base stations; m represents the index value of the cellular network base station. This represents the elevation angle of the m-th cellular network base station relative to the k-th vehicle; This represents the variance of the ranging error of the m-th cellular network base station; This represents the azimuth angle of the m-th cellular network base station relative to the k-th vehicle.

8. The multi-target localization method based on intelligent reflective surface cooperative cellular network according to claim 1, characterized in that, The expression for the objective function is: taking the maximum value of the Cramer-Rao bound of the position estimation error as the objective function. ； in, This represents the reflection unit coefficient matrix of the intelligent reflective surface mounted on the k-th vehicle in the n-th time slot; This represents the time slot allocation ratio variable; Indicates the Clamello boundary; This indicates time slot allocation; M represents the number of cellular network base stations; m represents the index value of the cellular network base station. This represents a set of binary variables; k represents the vehicle index value; K represents the number of vehicles; n represents the time slot index value; and N represents the number of time slots divided. Represents the set of positive integers; This represents the reflection phase offset of the l-th reflection unit of the k-th smart reflective surface.

9. The multi-target localization method based on intelligent reflective surface cooperative cellular network according to claim 8, characterized in that, Output optimized target vehicle location information, including: Based on the error range of the vehicle's prior position, determine the spatial angle range between the intelligent reflective surface mounted on vehicle k and the m-th cellular network base station; The intelligent reflective surface is divided into multiple sub-arrays according to the spatial angle range, and the coefficients of the reflective units in each sub-array are set. Based on the coefficients of the reflection unit, the optimized matching relationship between the cellular network base station and the vehicle, as well as the time slot allocation ratio variable, are obtained through a two-step segmented method. Based on the coefficients of the reflection unit, the optimized matching relationship between the cellular network base station and the vehicle, and the time slot allocation ratio variable, the optimal target vehicle location information is obtained.

10. A multi-target positioning system based on a smart reflective surface cooperative cellular network, characterized in that, include: The location acquisition module is used to acquire the vehicle's prior two-dimensional location information; The association optimization module is used to optimize the matching association between the cellular network base station and the smart reflective surface deployed on the top of the vehicle based on the vehicle's prior two-dimensional location information. The echo signal module is used to input the optimized matching correlation into the pre-built radar perception model in order to obtain the echo signal received by the cellular network base station. The distance estimation module is used to estimate the distance between the target vehicle and the cellular network base station using the echo signal; The error boundary analysis module is used to obtain the Cramer-Rao boundary of the position estimation error based on the estimated distance between the target vehicle and the cellular network base station. The joint optimization decision module is used to minimize the maximum value of the Cramer-Rao bound of the position estimation error as the objective function and output the optimized target vehicle position information.

Images