Non-cellular system sensing method and device, product, equipment and medium

Abstract

The invention discloses a non-cellular system sensing method and device, a product, equipment and a medium, and the method comprises the steps: building a probability model of system parameters in a non-cellular system, deploying a plurality of distributed transmitting APs, a plurality of receiving APs and a CPU in the non-cellular system, and enabling the system parameters to comprise a receiving signal of any receiving AP, the probability model for receiving the received signal of the AP is determined based on the angle and the time delay corresponding to the path passing through the sampling point; local information is calculated based on a probability model through each receiving AP, and the local information comprises the probability that the receiving AP judges that scatterers exist on different sampling points; local information sent by the receiving APs is received through the CPU, global information is calculated based on the local information, and the global information comprises the probability that the CPU judges that scatterers exist on different sampling points; the perceived position of the scatterer is determined based on the global information. In this way, the position sensing precision can be improved through multi-AP cooperation, and the sensing cost is reduced.

Description

Technical Field

[0001] The present application relates to the field of wireless communication transmission technology, and in particular to a non-cellular system perception method, device, product, equipment and medium. Background Art

[0002] With the continuous development of communication technology, the resolution in various signal domains is constantly improving. At the same time, considering improving the utilization of frequency bands and reducing hardware requirements, designing integrated communication and perception technologies that can share frequency bands and hardware has become a research hotspot. At present, when communicating and perceiving, for the method of angle domain collaborative perception that mainly uses high angle resolution provided by a large number of antennas, it may be necessary to configure an array of fully digital beamforming for each AP, which will result in excessive costs in non-cellular systems. Therefore, how to develop a perception method that is applicable to non-cellular systems with a small number of antennas per AP, improve perception flexibility, and reduce perception costs is a problem that needs to be solved urgently. Summary of the invention

[0003] In view of this, the purpose of this application is to provide a non-cellular system perception method, device, product, equipment and medium to improve the flexibility of perception and reduce the perception cost. The specific scheme is as follows:

[0004] In a first aspect, the present application discloses a non-cellular system perception method, comprising:

[0005] Establishing a probability model of system parameters in a non-cellular system, wherein the non-cellular system is deployed with a plurality of distributed APs and a CPU connected to the plurality of distributed APs, the plurality of distributed APs include a transmitting AP and a plurality of receiving APs, the system parameters include a received signal of any receiving AP, and the probability model of the received signal of the receiving AP is determined based on an angle and a delay corresponding to a path passing through a sampling point;

[0006] By each of the receiving APs, local information is calculated based on the probability model, wherein the local information includes the probability that the receiving AP determines that a scatterer exists at different sampling points;

[0007] Receiving, by the CPU, local information sent by each receiving AP and calculating global information based on the local information, wherein the global information includes the probability that the CPU determines that there are scatterers at different sampling points;

[0008] The locations of perceived scatterers are determined based on the global information.

[0009] Optionally, also include:

[0010] Execute the first iteration logic, wherein the first iteration logic includes: each of the receiving APs estimates the probability distribution information of the local latent variables according to the variational Bayes principle, the local latent variables include: the target gain and the accuracy of the target gain, the target gain is a vector composed of the gain of the LOS path and the propagation path formed by different sampling points, and the accuracy of the target gain is the accuracy of the gain corresponding to the LOS path and the propagation path formed by different sampling points; the CPU estimates the probability distribution information of the common noise accuracy;

[0011] When the first iteration logic satisfies a first preset convergence condition, a second iteration logic is executed, and when the second iteration logic satisfies a second preset convergence condition, the step of determining the position of the perceived scatterer based on the global information is triggered;

[0012] Among them, the second iterative logic includes calculating local information based on the estimated result of the accuracy of the target gain through each of the receiving APs; receiving the local information sent by each of the receiving APs through the CPU and calculating global information and a first probability based on the local information, wherein the first probability is the probability of the presence of a scatterer at the position of the sampling point determined by the CPU and transmitted to the receiving AP; calculating a second probability based on the first probability and the local information through each receiving AP, wherein the second probability represents the probability of the presence of a scatterer at the position of the sampling point determined by the receiving AP, and the second probability is used for estimating the probability distribution information of the accuracy of the target gain when executing the first iterative logic in the next round; and updating the position of the sampling point based on the estimated result of the target gain and the estimated result of the common noise accuracy.

[0013] Optionally, updating the position of the sampling point based on the estimation result of the target gain and the estimation result of the common noise precision includes:

[0014] Constructing a proxy function based on the estimation result of the target gain and the estimation result of the common noise precision;

[0015] The partial derivatives of the proxy function with respect to the position of the sampling point in different dimensions are calculated, and the position of the sampling point is updated based on the partial derivatives.

[0016] Optionally, determining the position of the perceived scatterer based on the global information includes:

[0017] Extracting a target sampling point based on the global information, wherein the target sampling point is a sampling point where the CPU determines that the probability of a scatterer on the sampling point is greater than a preset threshold;

[0018] The position of the target sampling point is determined as the position of the perceived scatterer.

[0019] Optionally, the first preset convergence condition is that a first preset number of iterations is reached or the change in the estimation results of two adjacent iterations is less than a first preset threshold, and the second preset convergence condition is that a second preset number of iterations is reached or the estimation results of two adjacent iterations are less than a second preset threshold.

[0020] Optionally, the probability model of the received signal of the receiving AP is determined based on the time delay and the arrival angle corresponding to the path passing through the sampling point and the time delay and the arrival angle corresponding to the LOS path.

[0021] Optionally, also include:

[0022] Arbitrarily select two receiving APs, estimate the delay using the received signals of the two receiving APs, and obtain the estimated delay;

[0023] An initial position of a sampling point is determined based on the estimated time delay.

[0024] In a second aspect, the present application discloses a non-cellular system sensing device, comprising:

[0025] A probability model building module, used to build a probability model of system parameters in a non-cellular system, wherein the non-cellular system is deployed with multiple distributed APs and a CPU connected to the multiple distributed APs, the multiple distributed APs include a sending AP and multiple receiving APs, the system parameters include a received signal of any of the receiving APs, and the probability model of the received signal of the receiving AP is determined based on the angle and delay corresponding to the path passing through the sampling point;

[0026] A local information calculation module, configured to calculate local information based on the probability model through each of the receiving APs, wherein the local information includes the probability that the receiving AP determines that a scatterer exists at different sampling points;

[0027] A global information calculation module, configured to receive, through the CPU, local information sent by each of the receiving APs and calculate global information based on the local information, wherein the global information includes the probability that the CPU determines that there are scatterers at different sampling points;

[0028] A scatterer position determination module is used to determine the position of the perceived scatterer based on the global information.

[0029] In a third aspect, the present application discloses a computer program product, which implements the aforementioned non-cellular system perception method when executed.

[0030] In a fourth aspect, the present application discloses an electronic device, including a memory and a processor, wherein:

[0031] The memory is used to store the computer program;

[0032] The processor is used to execute the computer program to implement the aforementioned non-cellular system perception method.

[0033] In a fifth aspect, the present application discloses a computer-readable storage medium for storing a computer program, wherein the computer program implements the aforementioned non-cellular system perception method when executed by a processor.

[0034] It can be seen from the above scheme that the present application provides a method for perceiving a cellular-free system, including: establishing a probability model of system parameters in a cellular-free system, wherein the cellular-free system is deployed with multiple distributed APs and a CPU connected to the multiple distributed APs, the multiple distributed APs include a sending AP and multiple receiving APs, the system parameters include a received signal of any receiving AP, and the probability model of the received signal of the receiving AP is determined based on the angle and delay corresponding to the path passing through the sampling point; through each of the receiving APs, calculating local information based on the probability model, wherein the local information includes the probability that the receiving AP determines that there are scatterers at different sampling points; through the CPU, receiving the local information sent by each of the receiving APs and calculating global information based on the local information, wherein the global information includes the probability that the CPU determines that there are scatterers at different sampling points; determining the position of the perceived scatterer based on the global information.

[0035] It can be seen that the beneficial effects of the present application are: using a non-cellular system composed of multiple distributed APs and CPUs for perception to improve perception performance. The probability model of the received signal is determined based on the angle and delay corresponding to the path passing through the sampling point. In the non-cellular system, joint multi-AP collaborative perception is performed based on angle and delay to avoid misjudgment of a single AP and improve estimation accuracy. The angle and delay have a geometric relationship with the position of the scatterer and can be used for positioning. The estimation accuracy of the delay is independent of the number of antennas, and the number of antennas on the AP is relatively low. It can improve perception flexibility (applicable to both non-cellular systems with a small number of antennas and non-cellular systems with a large number of antennas), reduce perception costs, and use angle information based on delay estimation to improve perception performance.

[0036] Correspondingly, the non-cellular system sensing device, product, equipment and readable storage medium provided by the present application also have the above-mentioned technical effects. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0038] Figure 1 A flow chart of a non-cellular system perception method provided in an embodiment of the present application;

[0039] Figure 2 A diagram of a cellular-free system architecture for a scene-aware implementation provided in an embodiment of the present application;

[0040] Figure 3 A system factor diagram of the non-cellular system perception method provided in an embodiment of the present application;

[0041] Figure 4 A graph showing the variation of the detection rate of the sensing method with different numbers of subcarriers with respect to the transmission power when the successful detection range is 1 meter provided in an embodiment of the present application;

[0042] Figure 5 A flow chart of a non-cellular system perception method based on Turbo-VBI provided in an embodiment of the present application;

[0043] Figure 6 A schematic diagram of a non-cellular system sensing device provided in an embodiment of the present application;

[0044] Figure 7 A structural diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0045] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0046] With the continuous development of communication technology, the resolution in various signal domains is constantly improving. At the same time, considering improving the utilization of frequency bands and reducing hardware requirements, designing communication and perception integration technology that can share frequency bands and hardware has become a research hotspot. At present, synaesthesia integration has become one of the key technologies of the 6th generation mobile communication technology (6G). Based on the communication capabilities and communication waveforms of mobile communication networks, it can perceive surrounding drones, cars and other objects while communicating, which can be applied to a variety of scenarios such as low-altitude economy, Internet of Things, and telemedicine.

[0047] Currently, the performance evaluation criteria for synaesthesia signals include resolution, ambiguity function, Doppler sensitivity, Doppler tolerance, peak-to-average power ratio (PAPR), mutual information, data information rate, etc. The available waveforms include orthogonal frequency division multiplexing (OFDM), linear frequency modulation (LFM), orthogonal time-frequency-space (OTFS), etc. Signal optimization includes PAPR optimization, interference management, adaptive signal optimization, etc.

[0048] For the basic limits of the synaesthesia system, the performance criteria include estimation theory metrics of perception (mean square error (MSE), Cramer-Rao bound (CRB), Weiss-Wienstein bound (WWB), Ziv–Zakhai bound (ZZB), equivalent Fisher information matrix, normalized cross ambiguity function, multidimensional ambiguity function, detection probability and false alarm probability, radar capacity, etc.), information theory metrics of communication, estimated information rate method of synaesthesia, equivalent MSE method, capacity distortion function method, etc. Existing performance analysis on perception usually focuses on the trade-off between communication and perception, and the number of perception targets set is usually small.

[0049] In traditional radar detection and perception systems, the number of cooperating nodes is usually small. In such a scenario, for the perception method based on delay, the same delay between the transmitter and receiver may correspond to multiple locations, resulting in ambiguous perception results. In addition, for the perception method based on delay or angle, false detection may occur on the LOS path. In a non-cellular system, by setting up multiple APs at different geographical locations and controlling the collaboration through the CPU (Central Processing Unit), macro-diversity gain can be obtained, providing more uniform coverage, and realizing collaborative perception of larger-scale nodes. At the same time, considering that the angle domain collaborative perception method mainly uses high angle resolution provided by a large number of antennas, it may be necessary to configure a fully digital beamforming array for each AP, resulting in excessive cost. Taking the above into consideration, this application chooses a multi-AP collaborative perception method that is mainly based on delay and is also applicable to low antenna numbers.

[0050] See also Figure 1 As shown, the embodiment of the present application discloses a non-cellular system perception method, including:

[0051] Step S11: Establish a probability model of system parameters in a cellular-free system, wherein the cellular-free system is deployed with multiple distributed APs and a CPU connected to the multiple distributed APs, the multiple distributed APs include a sending AP and multiple receiving APs, the system parameters include a received signal of any receiving AP, and the probability model of the received signal of the receiving AP is determined based on the angle and delay corresponding to the path passing through the sampling point.

[0052] Among them, distributed can be understood as each AP is placed in a different position, and the system parameters may include the local variables of each AP. The local variables of each AP include the directly observable received signal of the AP and the local hidden variables of each AP. The local hidden variables of each AP include the vector composed of the gain of the LOS path and the propagation path formed through different sampling points, the accuracy of the gain of the corresponding LOS path and the propagation path formed through different sampling points, and the local support variables that reflect whether a single AP determines whether there is a scatterer at each sampling point. The system parameters can also include global support variables that reflect whether the CPU determines whether there is a scatterer at different sampling points. By setting the prior distribution of the system parameters, the probability model of the system parameters is obtained. The sampling point can be understood as a hypothetical scatterer.

[0053] In the embodiment of the present application, in a non-cellular system, multiple APs with several antennas can be deployed for coverage to provide services to users. All APs are connected to the CPU, and the CPU can control different APs to achieve collaboration. One AP is selected as the sending AP, and the other APs are selected as the receiving APs. The sending AP sends a sensing signal on one antenna. Through resource scheduling, the orthogonality of the sensing signal and the resources occupied by the signal used for other functions in the system is ensured to eliminate interference. The receiving AP uses the received sensing signal to estimate the position of the scatterer. The sensing signal can be a downlink data signal. Sampling at different locations in the coverage area forms several sampling points.

[0054] In this embodiment, the probability model of the received signal of the receiving AP is determined based on the delay and arrival angle corresponding to the path passing through the sampling volume and the delay and arrival angle corresponding to the LOS path. That is, in this embodiment, multi-AP collaborative perception is performed based on the delay information and angle information under the non-cellular system architecture. Through multi-AP collaboration, the LOS path energy leakage perceived by a single AP can be overcome, and the location perception accuracy can be improved by using multi-AP collaboration, which improves the flexibility under different configurations.

[0055] This embodiment can establish the received signal of the receiving AP on each subcarrier based on the delay and arrival angle corresponding to the path passing through the scatterer and the delay and arrival angle corresponding to the LOS path; determine the received signal of the receiving AP based on the received signal of the receiving AP on all subcarriers. The delay and angle correspond to the changes in the received signals between subcarriers and between antennas respectively. The change information between subcarriers can be used according to the delay, and the change information between antennas can be used according to the angle. The combination of the two can provide more samples for estimation and further improve the estimation accuracy.

[0056] In the embodiment of the present application, two receiving APs can be arbitrarily selected, and the received signals of the two receiving APs can be used to estimate the delay to obtain the estimated delay; the initial position of the sampling point is determined based on the estimated delay. In this way, the initial sampling point position set is calculated, and compared with the gridding method, the number of sampling points can be reduced in the case of large areas and small scatterers, thereby reducing the amount of calculation of subsequent algorithms. Specifically, the delay estimation can use the global least squares-estimation of signal parameters by rotation invariant technique (TLS-ESPRIT) method.

[0057] Furthermore, a probability model of system parameters in the non-cellular system may be established based on the initial positions of the sampling points.

[0058] In an optional embodiment, the cellular-free system is deployed with multiple distributed APs and a CPU connected to the multiple APs, the multiple APs include a sending AP and multiple receiving APs. Furthermore, two receiving APs can be arbitrarily selected to estimate the delay using the received signal, and collaboratively solve the position that meets the delay to obtain a set of sampling point positions (that is, an initial position set of the sampling points), and a probability model of the system parameters in the cellular-free system is established based on the sampling point positions.

[0059] Step S12: calculating local information based on the probability model through each of the receiving APs, wherein the local information includes the probability that the receiving AP determines that there are scatterers at different sampling points.

[0060] In this embodiment, local information may be calculated by each receiving AP based on an estimation result of the accuracy of the target gain.

[0061] Step S13: receiving, by the CPU, local information sent by each receiving AP and calculating global information based on the local information, wherein the global information includes the probability of the CPU determining that scatterers exist at different sampling points.

[0062] The embodiment of the present application can: execute the first iterative logic, wherein the first iterative logic includes: each receiving AP estimates the probability distribution information of the local hidden variable according to the variational Bayes principle, and the local hidden variable includes: the target gain and the accuracy of the target gain, the target gain is a vector composed of the gain of the LOS path (i.e., Line-of-Sight path) and the propagation path formed through different sampling points, and the accuracy of the target gain is the accuracy of the gain corresponding to the LOS path and the propagation path formed through different sampling points; the CPU estimates the probability distribution information of the common noise accuracy; the common noise accuracy can be understood as the noise accuracy determined by the CPU. In this embodiment, the estimation of the noise accuracy is calculated by the CPU based on the results of each AP, which can improve the estimation accuracy.

[0063] When the first iterative logic satisfies the first preset convergence condition, the second iterative logic is executed. When the second iterative logic satisfies the second preset convergence condition, the step of determining the position of the perceived scatterer based on the global information is triggered. If the first iterative logic does not meet the first preset convergence condition, the step of estimating the probability distribution information of the local latent variable according to the variational Bayesian principle for each receiving AP is entered, and the first iterative logic is executed again. If the second iterative logic does not meet the second preset convergence condition, the step of estimating the probability distribution information of the local latent variable according to the variational Bayesian principle for each receiving AP is entered, and the first iterative logic is executed again.

[0064] Among them, the second iterative logic includes calculating local information based on the estimated result of the accuracy of the target gain through each of the receiving APs; receiving the local information sent by each of the receiving APs through the CPU and calculating global information and a first probability based on the local information, wherein the first probability is the probability of the presence of a scatterer at the position of the sampling point determined by the CPU and transmitted to the receiving AP; calculating a second probability based on the first probability and the local information through each receiving AP, wherein the second probability represents the probability of the presence of a scatterer at the position of the sampling point determined by the receiving AP, and the second probability is used for estimating the probability distribution information of the accuracy of the target gain when executing the first iterative logic in the next round; and updating the position of the sampling point based on the estimated result of the target gain and the estimated result of the common noise accuracy.

[0065] In a specific implementation, a proxy function may be constructed based on the estimated result of the target gain and the estimated result of the common noise precision; partial derivatives of the proxy function with respect to the position of the sampling point in different dimensions are calculated, and the position of the sampling point is updated based on the partial derivatives.

[0066] In this embodiment, the first preset convergence condition may be that a first preset number of iterations is reached or the change in the estimation results of two adjacent iterations is less than a first preset threshold, and the second preset convergence condition may be that a second preset number of iterations is reached or the estimation results of two adjacent iterations are less than a second preset threshold.

[0067] Step S14: Determine the position of the perceived scatterer based on the global information.

[0068] In an embodiment of the present application, a target sampling point can be extracted based on the global information, wherein the target sampling point is a sampling point where the CPU determines that the probability of a scatterer being present on the sampling point is greater than a preset threshold; and the position of the target sampling point is determined as the position of the perceived scatterer.

[0069] It can be seen that in the embodiment of the present application, the probability model of the received signal of the receiving AP is determined based on the angle and delay corresponding to the path passing through the sampling point. In the cellular-free system, joint multi-AP collaborative perception is performed based on the delay to avoid misjudgment of a single AP. The delay has a geometric relationship with the position of the scatterer and can be used for positioning. The estimated accuracy of the delay is independent of the number of antennas, and the number of antennas on the AP is relatively low, which can reduce the perception cost.

[0070] See also Figure 2 As shown, Figure 2 A schematic diagram of a non-cellular system provided in an embodiment of the present application. The non-cellular system perception method provided in an embodiment of the present application includes the following steps:

[0071] Step 1: If Figure 2 As shown, multiple configurations N are deployed in a non-cellular system. E The APs with the root antennas provide coverage and provide services to users. All APs are connected to the CPU, which can control different APs to achieve collaboration. Select an AP as the sending AP, N A The APs are used as receiving APs, and the transmitting APs send sensing signals on one antenna. Through resource scheduling, the orthogonality of the sensing signals and the signals used for other functions in the system is ensured to eliminate interference. The receiving AP uses the received sensing signals to estimate the scatterer position. The sensing signals can be downlink data signals. Samples are collected at different locations in the coverage area to form N G sampling points, establish the nth A The approximate received signal of a receiving AP on the kth subcarrier:

[0072]

[0073] Where X(k) is the transmitted signal of the transmitting antenna on the kth subcarrier, For nth A The noise vector of all antennas of the receiving AP on the kth subcarrier, where the elements have a mean of 0 and a variance of σ 2 The delay and angle correspond to the changes in the received signals between subcarriers and antennas, respectively. The delay can be used to make use of the change information between subcarriers, and the angle can be used to make use of the change information between antennas. The combination of the two can provide more samples for estimation and further improve the estimation accuracy. Specifically, the channel vector is:

[0074]

[0075] in, are the gain, angle of arrival (AOA) and delay of the LOS path, f sc is the subcarrier spacing, For nth A The receiving APs are about the angle The array vector of j is an imaginary symbol, and the transmitting antenna on the kth subcarrier passes through the nth G The scatterer at the sampling point reaches the nth A Channels of APs:

[0076]

[0077] in The transmitting antenna passes through the nth G The scatterer at the sampling point reaches the nth A The gain, AOA and delay of the propagation path of each AP, specifically:

[0078]

[0079] Where c is the speed of light, is the location of the transmitting antenna of the transmitting AP, where and are the coordinates of the transmitting antenna of the transmitting AP in the first dimension and the second dimension respectively. In this embodiment, the position is represented by two-dimensional coordinates, which are respectively called the first dimension and the second dimension. For nth A The center position of all antennas of a receiving AP. The center position can be understood as the average position of all antennas. and The nth A The center positions of all antennas of a receiving AP are coordinates in the first and second dimensions. For nth G The locations of the sampling points, and The nth G The positions of the sampling points are in the coordinates of the first dimension and the second dimension. In this embodiment, the positions of the sampling points can be obtained by any two receiving APs respectively estimating the delay of the received signal and cooperating to obtain the positions. Specifically, each AP roughly estimates the delay component contained in the received signal based on the received signal on its own antenna, and extracts the positions that meet the delay estimation results of the two APs as the sampling point positions for subsequent processing. Array vector The nth E Elements:

[0080]

[0081] in, For nth A Receiving AP nth E Antenna relative to the nth A The relative positions of the centers of all antennas of the receiving AP, where and The nth A Receiving AP nth E Antenna relative to the nth A The first and second dimension offsets of the centers of all antennas of a receiving AP, and λ is the system wavelength. is the subcarrier set used, the nth A All subcarriers used for sensing by APs The mathematical model of the received signal on the receiving AP, that is, the received signal of the receiving AP is:

[0082]

[0083] in, k i is the number of the i-th subcarrier used for sensing, where To include the nth A The sensing signal matrix affected by the angle and delay corresponding to the path passing through all sampling points received by all antennas of the receiving AP on the kth subcarrier, To include the nth A The perception signal matrix affected by the angle and delay corresponding to the path passing through all sampling points received by all antennas of a receiving AP on all subcarriers. Specifically,

[0084]

[0085] Based on the SBL principle, according to the above expression of the received signal, combined with the Turbo-VBI framework, a probability model of the system parameters is established, and each parameter is initialized. The corresponding factor graph is as follows: Figure 3 As shown, except for the received signal In addition, circles represent hidden variables and black squares represent the set prior distribution. Specifically, each variable (i.e., the system parameters in the proposed non-cellular system perception method) is set as follows:

[0086] Among them, the variables include: A The received signal vector of the receiving AP Send AP and nth A The vector composed of the LOS path between the receiving APs and the gain of the propagation path formed by the scatterers at different sampling points Send AP and nth A The precision vector corresponding to the gain vector of the LOS path between the receiving APs and the propagation path formed by the scatterers at different sampling points n A =0,…,N A . nth A AP determines the nth GWhether there is a local support variable of the scatterer at the sampling point The common noise accuracy γ. And the common support vector

[0087] This embodiment can establish a probability distribution model for distributed processing based on the sampling point location set and angle and delay information. The structure of the receiver further affects the distribution of the received signal.

[0088] Step 11: Set the prior distribution of the local variables of each AP in the system, where the local variables of each AP include the nth directly observable A The received signal of each AP and the local hidden variables of each AP. The local hidden variables of each AP include the vector composed of the LOS path and the gain of the propagation path formed by different sampling points The accuracy of the gain corresponding to the LOS path and the propagation path formed through different sampling points n G =0,…,N G , and reflects the nth A AP determines whether there is a local support variable of the scatterer at each sampling point

[0089] The probability model of the received signal of the receiving AP is determined based on the angle and delay corresponding to the path passing through the sampling point; specifically, for the nth A The received signal of each AP Distribution in represents a random vector x that follows a complex Gaussian distribution with mean μ and covariance matrix Σ. Representative size is The unit matrix of γ is the common noise accuracy. For the vector composed of the gain of the LOS path and the propagation path formed by different sampling points No.n G Sampling points (n G = 0 represents the LOS path) corresponding gain Distribution For the gain accuracy For LOS diameter, distribution in Indicates that the random variable x follows a Gamma distribution with parameters a and b. For the NLOS path, the distribution is:

[0090]

[0091] Where δ(·) represents the impulse function. A AP determines the nth GWhether there is a local support variable of the scatterer at the sampling point distributed:

[0092]

[0093] in For the CPU to n G The supporting variables for the joint determination of whether there is a scatterer at each sampling point. is the conditional probability of the local support variable. The local support variable reflects the nth A AP determines the nth G The local support variable is obtained by each receiving AP, and the public support variable is obtained by the CPU, reflecting the nth sampling point. A AP determines the nth G A variable that determines whether there is a scatterer at the sampling point. The value is 1 or -1, indicating whether it exists or not.

[0094] Step 12: Set the distribution of global hidden variables in the system, including the CPU for the nth G Support variables for joint determination of whether there is a scatterer at each sampling point The common support vector distributed:

[0095]

[0096] Where Z(ζ) is the partition function that makes the probability distribution satisfy the integral of 1. For the common noise accuracy γ, the distribution is the common support vector The elements in the common support vector are the common support variables. Calculating the common support variables is calculating the common support vector.

[0097] Step 13: Initialize parameters, including the location of each sampling point Common support vector Parameters of the distribution Conditional probability of local support variables and nth A AP determines the nth G The probability that there is a scatterer at the sampling point The parameters a and b of the two Gamma distributions are Distribution parameters of common noise accuracy γ d, parameters related to gain accuracy Common noise accuracy related parameters <γ q(γ) .

[0098] Step 2: Definition is the set of received signals of all antennas of all receiving APs on all subcarriers, and V is the set of received signals of all antennas of all receiving APs on all subcarriers. Figure 3 The set of all variables in excluding the received signal, To include all sampling point location parameters The perception problem can be summarized in the form of EM framework.

[0099]

[0100] Where t is the number of iterations. In step E, the distribution of all variables is updated, and the objective function is updated according to the updated distribution In the M step, according to the new objective function, optimize Maximize the objective function. In this application, for distributed processing, define the nth A The set of local latent variables corresponding to AP in is the gain vector The corresponding precision vector, based on the Turbo principle, adds the internal iteration of each receiving AP on the basis of the EM framework. In one iteration, each receiving AP iterates The CPU iterates the distribution of γ until convergence, and then calculates the nth decision value passed to the CPU by each receiving AP. G The probability that there is a scatterer at each sampling point, the CPU calculates the global probability by combining the probabilities of each AP, and calculates the nth judgment probability transmitted to each receiving AP. G Each receiving AP obtains the probability from the CPU and calculates the probability of the existence of a scatterer at each sampling point, thus completing an iteration of step E. First, each AP performs a variational Bayesian (VB) The update of the approximate marginal probability distribution of the latent variables in is as follows:

[0101] Step 21: Update the corresponding nth A Gain of all LOS and NLOS paths of an AP The estimate of the distribution of is obtained by initialization or by step 22, step 23 and γ obtained in step 24 q(γ) ,definition in diag(x) is a diagonal matrix with x as the diagonal, The estimated results of the distribution of :

[0102]

[0103] in,<f(x)>q(x) =∫f(x)q(x)dx, Re{·} is the real part of the complex number, ∝ represents proportionality, exp{·} is the exponential function, and according to the estimation result, The mean is The covariance is The complex Gaussian distribution of .

[0104] Step 22: Update for the nth A The accuracy of the LOS path gain corresponding to each AP The estimate of the distribution of is obtained from step 21 and The estimated results of the distribution of :

[0105]

[0106] in, for a+1-1 times, for The first element of According to the estimation results, Obey a+1 and is the parameter of the Gamma distribution. According to the characteristics of the Gamma distribution, we have:

[0107]

[0108] Among them, Ψ(·) is the digamma function.

[0109] Step 23: Update for the nth A AP corresponds to the nth G The accuracy of the path gain at each sampling point n G >0, obtained by initialization or step 44 And step 21 obtained and Estimated results:

[0110]

[0111] Among them, the nth A AP determines the nth G The probability that there is a scatterer at the sampling point for The nth G +1 element, According to the estimation results, Obey and is the parameter of the Gamma distribution. According to the characteristics of the Gamma distribution, we have:

[0112]

[0113] Step 24: Update the estimate of the distribution of the common noise precision γ obtained in step 21 and The distribution estimation results of γ are:

[0114]

[0115] According to the estimation results, γ obeys and is the parameter of the Gamma distribution. According to the characteristics of the Gamma distribution, we have:

[0116]

[0117] Step 3: Determine whether the iteration in step 2 has reached the convergence requirement. If not, execute step 2; otherwise, execute step 4. Specifically, the convergence requirement is to reach the number of iterations or The change in the estimated results between two adjacent iterations is small enough.

[0118] Step 4: Perform external iteration. According to the principle of VB and message passing (MP), each receiving AP calculates the message about the local support vector based on the result of local iteration (the judgment result of each receiving AP on whether there is a scatterer at different sampling points) and passes it to the CPU. The CPU jointly estimates the statistical information of the public support vector based on the messages transmitted by all APs (the judgment result of the CPU on whether there is a scatterer at different sampling points), and passes the message to each AP. Each AP estimates the distribution of the local support vector based on the received message. The specific steps are as follows:

[0119] Step 41: Update the nth A The common support vectors passed to the CPU by the APs The message is initialized by and initialized or obtained in step 23 and No.n A The AP sends the CPU the information about the nth G The probability distribution of whether there is a scatterer at each sampling point is:

[0120]

[0121] in,

[0122]

[0123] Among them, the intermediate variables are:

[0124]

[0125] The message refers to Can be Complete description. It can be observed that the nth A The AP passes the decision to the CPU. G The probability that there is a scatterer at each sampling point The probability distribution can be fully described, so the transmission The message can be delivered completely.

[0126] Step 42: Update the CPU to the common support vector The estimate of the distribution of is initialized by and the value obtained in step 41 Estimated results:

[0127]

[0128] in,

[0129]

[0130] Among them, the intermediate variable

[0131]

[0132] It can be observed that the CPU determines the nth G The probability that there is a scatterer at each sampling point The probability distribution can be fully described.

[0133] Step 43: Update CPU pass to nth A AP on local support variables The message is initialized by and the value obtained in step 42 Estimated results:

[0134]

[0135] in,

[0136]

[0137] It can be observed that the CPU is passed to the nth A The determination of the nth AP G The probability that there is a scatterer at each sampling point (that is, the aforementioned first probability) can fully describe the probability distribution.

[0138] Step 44: Update the nth A AP determines the nth G Whether there is a local support variable of the scatterer at the sampling point The estimate of the distribution of is obtained according to step 41 and the value obtained in step 43 Estimated results:

[0139]

[0140] in,

[0141]

[0142] It can be observed that the nth A AP determines the nth G The probability that there is a scatterer at the sampling point (that is, the aforementioned second probability) can fully describe the probability distribution.

[0143] Step 5: Update the position of each sampling point based on the value obtained in step 21 and Constructing a proxy function As the objective function of M-step optimization,

[0144]

[0145] Where C is a constant, Tr(·) is the trace of the matrix, and the partial derivative of the proxy function with respect to the coordinates of each sampling point is calculated and the position of each sampling point is updated. The specific steps are as follows:

[0146] Step 51: Calculate the surrogate function pair The partial derivative of

[0147]

[0148] when hour,

[0149]

[0150] in, for The element in row i and column j in hour,

[0151]

[0152]

[0153] when hour,

[0154]

[0155] Step 52: Calculate the surrogate function pair The partial derivative of

[0156] Among them hour,

[0157]

[0158] when hour,

[0159]

[0160]

[0161] when hour,

[0162]

[0163] Step 53: Update Location

[0164]

[0165] Among them, Δζ is the step size.

[0166] Step 6: Determine whether the iterations in step 4-step 5 meet the convergence requirement, if not, execute step 2, otherwise execute step 7. Specifically, the convergence requirement is to reach the number of iterations or the change between two adjacent iteration results is small enough.

[0167] Step 7: Extract π snG For sampling points with a value greater than 0.5, the corresponding position of the sampling point is the position of the perceived scatterer. In the case of a single scatterer, four receiving APs, and four receiving antennas per AP, a successful detection is considered when the detected scatterer position is within 1 meter of the actual scatterer. The detection rates under different transmit powers and subcarrier numbers are as follows: Figure 4 shown.

[0168] It can be seen that the beneficial effect of the embodiment of the present application is that: the cellular-free system composed of multiple distributed APs and CPUs is used for perception to improve the perception performance. By using the cellular-free system for joint multi-AP collaborative perception, it can be prevented that when a single AP is sensing, the energy of the LOS path leaks to the adjacent sampling point position, causing the sampling point position close to the LOS path without a scatterer to be misjudged as having a scatterer, resulting in a false alarm.

[0169] Joint multi-AP collaborative positioning is mainly based on delay. The delay is the time for the signal to propagate between the transmitter and the receiver. Multiplying it by the speed of light equals the distance the signal has traveled. In this application, it is assumed that the channel mainly contains single-hop paths. For single-hop NLOS paths, the corresponding distance is the distance from the transmitter to the scatterer plus the distance from the scatterer to the receiver. It has a geometric relationship with the position of the scatterer, so it can be used for positioning. The estimation accuracy of the delay can be adjusted by the parameters of OFDM modulation, which is independent of the number of antennas and can be achieved with a single antenna. In contrast, the angular interval that can be resolved for angle domain signals is basically inversely proportional to the number of antennas, so multiple antennas are required to achieve high angular resolution. Therefore, this application has a low requirement for the number of antennas on the AP, which helps to reduce the deployment cost of the non-cellular system. Considering the situation where the AP is configured with multiple antennas, more received signal samples brought by multiple antennas can be used, and the angle information can be used to improve the perception performance based on the delay-based estimation. Therefore, the proposed scheme has more flexible adaptability to different AP configurations. The local latent variable distribution estimation with higher complexity is completed by each receiving AP in a distributed manner, which helps to ensure the scalability of the non-cellular system.

[0170] It can be understood that the perception method provided in the embodiment of the present application is a perception method for a non-cellular system based on Turbo-VBI. Figure 5 This is a flow chart of a Turbo-VBI-based non-cellular system perception method according to an embodiment of the present application. A probability model of system parameters is established, and the system parameters are initialized. The perception process includes two layers of iterations. When the inner iteration meets the convergence condition, the outer iteration is entered. If the outer iteration converges, the perceived scatterer and the corresponding position are extracted according to the edge probability distribution of the estimated common support vector. By establishing a probability model of system parameters, and solving it based on the expectation maximization (EM) framework. Two receiving APs are arbitrarily selected to estimate the delay using the received signal, and the initial sampling point position set is obtained collaboratively. Based on the perception signal received by the receiving AP and the sampling point position, a system probability model is established to initialize the system parameters. Each receiving AP estimates the probability distribution information of the local hidden variables corresponding to each AP separately, and the estimated channel gain distribution information is passed to the CPU to estimate the statistical information of the common noise accuracy, and it is judged whether it converges. If it converges, each receiving AP calculates the distribution information of the local support vector separately, and passes it to the CPU to jointly estimate the distribution information of the common support vector. The CPU adjusts the sampling point position, and determines whether it converges. If it converges, the perceived scatterer position is output. Specifically, the following steps may be included:

[0171] Step 1: Deployment-based N A Configuration N E The receiving AP of the root antenna senses the signal received and uses the subcarrier set The transmitted signal on the kth subcarrier is X(k), which is sampled at different locations in the coverage area to form NG Based on the principle of sparse Bayesian learning (SBL), a probability model of system parameters is established and each parameter is initialized.

[0172] Step 2: Define the nth A The set of local latent variables corresponding to AP Each AP is based on the variational Bayes (VB) principle. The approximate marginal probability distribution of the latent variables in the update is updated and transmitted to the central processing unit (CPU) to update the approximate marginal probability distribution of the common noise accuracy;

[0173] Step 3: Determine whether the iteration in step 2 meets the convergence requirement. If not, execute step 2, otherwise execute step 4. Specifically, the convergence requirement is that the number of iterations is reached or the change between two adjacent iteration results is small enough;

[0174] Step 4: Perform external iteration. According to the VB and message passing (MP) principles, each receiving AP calculates a message about the local support vector based on the result of the local iteration and passes it to the CPU. The CPU jointly estimates the statistical information of the public support vector based on the messages passed by all APs, and passes the message to each AP. Each AP estimates the distribution of the local support vector based on the received message.

[0175] Step 5: Update the position of each sampling point, construct a proxy function, calculate the partial derivative of the proxy function with respect to the coordinates of each sampling point and update the position of each sampling point; this application provides an accurate expression for position adjustment.

[0176] Step 6: Determine whether the iterations in step 4-step 5 meet the convergence requirement. If not, execute step 2, otherwise execute step 7. Specifically, the convergence requirement is that the number of iterations is reached or the change between two adjacent iteration results is small enough;

[0177] Step 7: Extract π snG For sampling points greater than 0.5, the corresponding positions of the sampling points are the positions of the perceived scatterers.

[0178] The embodiment of the present application can realize multi-AP collaborative perception based on delay information as the main and angle information as the auxiliary in a cellular-free system architecture when the number of antennas per access point (AP) is small. Through multi-AP collaboration, the misjudgment caused by the direct (LOS) path energy leakage between the transceiver during single AP perception can be corrected, the misjudgment rate can be reduced, and the perception performance can be improved. The embodiment of the present application provides a cellular-free system perception method based on Turbo-VBI, which mainly performs collaborative perception of multiple APs in a cellular-free system based on delay, does not rely on the high spatial resolution brought by a large number of antennas, can be used in situations with any number of antennas, and improves the perception performance.

[0179] It can be seen that the embodiment of the present application uses a non-cellular system composed of multiple distributed APs and CPUs for perception to improve perception performance. By using a non-cellular system for joint multi-AP collaborative perception, it can be prevented that when a single AP is perceived, the energy of the LOS path leaks to the adjacent grid, causing the grid without scatterers near the LOS path to be misjudged as having scatterers, resulting in false alarms. The delay estimation results of two APs are used to form an initial sampling point position set. Compared with gridded position sampling, in the case of large areas and small scatterers, the number of sampling points can be effectively reduced, and the complexity of subsequent calculations can be reduced. The CPU makes a joint judgment based on the target gain estimation results of each AP, which can improve the estimation accuracy of the common noise accuracy; the position is adjusted by multi-AP collaboration, local information is calculated based on the probability model, and the local information sent by each receiving AP is received by the CPU and the global information is calculated based on the local information. The position of the perceived scatterer is determined based on the global information, and the gain of the scatterer position estimation accuracy can be obtained. The received signal in the receiving AP is a signal determined based on the angle and delay corresponding to the path through the scatterer. The estimation accuracy of the delay is independent of the number of antennas, and there is no requirement for the number of antennas on the AP, which can reduce the hardware requirement cost of the perception device. Considering the case where the AP is configured with multiple antennas, more received signal samples brought by multiple antennas can be used to improve the perception performance by using angle information based on delay-based estimation. Therefore, the proposed solution has more flexible adaptability to different AP configurations.

[0180] For further information, see Figure 6 As shown, the present application implements and discloses a non-cellular system sensing device, including:

[0181] A probability model building module 11 is used to build a probability model of system parameters in a non-cellular system, wherein the non-cellular system is deployed with multiple distributed APs and a CPU connected to the multiple distributed APs, the multiple distributed APs include a sending AP and multiple receiving APs, the system parameters include a received signal of any of the receiving APs, and the probability model of the received signal of the receiving AP is determined based on the angle and delay corresponding to the path passing through the sampling point;

[0182] A local information calculation module 12, configured to calculate local information based on the probability model through each of the receiving APs, wherein the local information includes the probability that the receiving AP determines that a scatterer exists at different sampling points;

[0183] A global information calculation module 13, configured to receive, through the CPU, local information sent by each of the receiving APs and calculate global information based on the local information, wherein the global information includes the probability that the CPU determines that there are scatterers at different sampling points;

[0184] The scatterer position determination module 14 is configured to determine the position of the perceived scatterer based on the global information.

[0185] The device also includes:

[0186] The first iterative logic execution module is used to execute the first iterative logic, wherein the first iterative logic includes: each receiving AP estimates the probability distribution information of the local hidden variable according to the variational Bayes principle, and the local hidden variable includes: the target gain and the accuracy of the target gain, the target gain is a vector composed of the gain of the LOS path and the propagation path formed by different sampling points, and the accuracy of the target gain is the accuracy of the gain corresponding to the LOS path and the propagation path formed by different sampling points; the CPU estimates the probability distribution information of the common noise accuracy;

[0187] The second iteration logic execution trigger module is used to trigger the execution of the second iteration logic when the first iteration logic meets the first preset convergence condition; wherein the second iteration logic includes calculating local information based on the estimated result of the accuracy of the target gain through each of the receiving APs; receiving local information sent by each of the receiving APs through the CPU and calculating global information and a first probability based on the local information, the first probability being the probability of the existence of a scatterer at the position of the sampling point determined by the CPU transmitted to the receiving AP; calculating a second probability based on the first probability and the local information through each of the receiving APs, the second probability representing the probability of the existence of a scatterer at the position of the sampling point determined by the receiving AP, the second probability being used for the probability distribution information of the accuracy of the target gain estimated in the next round of execution of the first iteration logic; updating the position of the sampling point based on the estimated result of the target gain and the estimated result of the common noise accuracy. The device may also include a second iteration logic execution module, the second iteration logic execution module is used to execute the second iteration logic, and the second iteration logic execution module may include a local information calculation module 12, a global information calculation module 13, a second probability calculation module and a sampling point position update module. The local information calculation module 12 is specifically used to calculate local information based on the estimated result of the accuracy of the target gain through each of the receiving APs. The global information calculation module 13 is specifically used to receive the local information sent by each receiving AP through the CPU and calculate the global information and the first probability based on the local information. The second probability calculation module is used to calculate the second probability based on the first probability and the local information through each receiving AP. The sampling point position update module is used to update the position of the sampling point based on the estimation result of the target gain and the estimation result of the common noise accuracy.

[0188] a scatterer position determination triggering module, configured to trigger the scatterer position determination module to execute the step of determining the position of the perceived scatterer based on the global information when the second iterative logic satisfies a second preset convergence condition;

[0189] The sampling point position update module may specifically include:

[0190] A proxy function construction module, used to construct a proxy function based on the estimation result of the target gain and the estimation result of the common noise precision;

[0191] The sampling point updating module is used to calculate the partial derivatives of the proxy function with respect to the position of the sampling point in different dimensions, and update the position of the sampling point based on the partial derivatives.

[0192] The scatterer position determination module 14 is specifically used to extract a target sampling point based on the global information, wherein the target sampling point is a sampling point where the CPU determines that the probability of a scatterer on the sampling point is greater than a preset threshold; and the position of the target sampling point is determined as the position of the perceived scatterer.

[0193] The first preset convergence condition is that a first preset number of iterations is reached or the estimation result of two adjacent iterations changes less than a first preset threshold, and the second preset convergence condition is that a second preset number of iterations is reached or the estimation result of two adjacent iterations changes less than a second preset threshold.

[0194] The probability model of the received signal of the receiving AP is determined based on the time delay and arrival angle corresponding to the path passing through the sampling point and the time delay and arrival angle corresponding to the LOS path.

[0195] The device also includes a sampling point initial position determination module, which is used to arbitrarily select two receiving APs, estimate the delay using the received signals of the two receiving APs, and obtain the estimated delay; and determine the initial position of the sampling point based on the estimated delay.

[0196] It can be seen that the embodiment of the present application uses a non-cellular system composed of multiple distributed APs and CPUs for perception to improve perception performance. The probability model of the received signal is determined based on the angle and delay corresponding to the path passing through the sampling point. In the non-cellular system, joint multi-AP collaborative perception is performed based on angle and delay to avoid misjudgment of a single AP and improve estimation accuracy. The angle and delay have a geometric relationship with the position of the scatterer and can be used for positioning. The estimation accuracy of the delay is independent of the number of antennas, and the number of antennas on the AP is relatively low, which can reduce the perception cost. The angle information is used to improve the perception performance based on the estimation of the delay.

[0197] Furthermore, an embodiment of the present application also discloses a computer program product, which implements the aforementioned non-cellular system perception method when executed.

[0198] For the specific process of the above-mentioned non-cellular system perception method, reference may be made to the corresponding contents disclosed in the above-mentioned embodiments, which will not be described in detail here.

[0199] See also Figure 7 As shown, an embodiment of the present application discloses an electronic device 20, including a processor 21 and a memory 22; wherein the memory 22 is used to store a computer program; the processor 21 is used to execute the computer program, the non-cellular system perception method disclosed in the above embodiment.

[0200] For the specific process of the above-mentioned non-cellular system perception method, reference may be made to the corresponding contents disclosed in the above-mentioned embodiments, which will not be described in detail here.

[0201] Furthermore, the memory 22 as a carrier for storing resources may be a read-only memory, a random access memory, a magnetic disk or an optical disk, etc., and the storage method may be temporary storage or permanent storage.

[0202] In addition, the electronic device 20 also includes a power supply 23, a communication interface 24, an input / output interface 25 and a communication bus 26; wherein the power supply 23 is used to provide working voltage for each hardware device on the electronic device 20; the communication interface 24 can create a data transmission channel between the electronic device 20 and an external device, and the communication protocol it follows is any communication protocol that can be applied to the technical solution of the present application, and is not specifically limited here; the input / output interface 25 is used to obtain external input data or output data to the outside world, and its specific interface type can be selected according to specific application needs, and is not specifically limited here.

[0203] Furthermore, an embodiment of the present application also discloses a computer-readable storage medium for storing a computer program, wherein the computer program, when executed by a processor, implements the non-cellular system perception method disclosed in the aforementioned embodiment.

[0204] For the specific process of the above-mentioned non-cellular system perception method, reference may be made to the corresponding contents disclosed in the above-mentioned embodiments, which will not be described in detail here.

[0205] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0206] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein may be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module may be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

[0207] The above is a detailed introduction to a non-cellular system perception method, device, product, equipment and medium provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

Claims

1. A method for sensing a non-cellular system, characterized in that: include: Establishing a probability model of system parameters in a non-cellular system, wherein the non-cellular system is deployed with a plurality of distributed APs and a CPU connected to the plurality of distributed APs, the plurality of distributed APs include a transmitting AP and a plurality of receiving APs, the system parameters include a received signal of any of the receiving APs, and the probability model of the received signal of the receiving AP is determined based on an angle and a delay corresponding to a path passing through a sampling point; By each of the receiving APs, local information is calculated based on the probability model, wherein the local information includes the probability that the receiving AP determines that a scatterer exists at different sampling points; Receiving, by the CPU, local information sent by each receiving AP and calculating global information based on the local information, wherein the global information includes the probability that the CPU determines that there are scatterers at different sampling points; The locations of perceived scatterers are determined based on the global information.

2. The non-cellular system sensing method according to claim 1, characterized in that: Also includes: Execute the first iteration logic, wherein the first iteration logic includes: each of the receiving APs estimates the probability distribution information of the local latent variables according to the variational Bayes principle, the local latent variables include: the target gain and the accuracy of the target gain, the target gain is a vector composed of the gain of the LOS path and the propagation path formed by different sampling points, and the accuracy of the target gain is the accuracy of the gain corresponding to the LOS path and the propagation path formed by different sampling points; the CPU estimates the probability distribution information of the common noise accuracy; When the first iteration logic satisfies a first preset convergence condition, a second iteration logic is executed, and when the second iteration logic satisfies a second preset convergence condition, the step of determining the position of the perceived scatterer based on the global information is triggered; Among them, the second iterative logic includes calculating local information based on the estimated result of the accuracy of the target gain through each of the receiving APs; receiving the local information sent by each of the receiving APs through the CPU and calculating global information and a first probability based on the local information, wherein the first probability is the probability of the presence of a scatterer at the position of the sampling point determined by the CPU and transmitted to the receiving AP; calculating a second probability based on the first probability and the local information through each receiving AP, wherein the second probability represents the probability of the presence of a scatterer at the position of the sampling point determined by the receiving AP, and the second probability is used for estimating the probability distribution information of the accuracy of the target gain when executing the first iterative logic in the next round; and updating the position of the sampling point based on the estimated result of the target gain and the estimated result of the common noise accuracy.

3. The method for sensing a non-cellular system according to claim 2, characterized in that: The updating of the position of the sampling point based on the estimation result of the target gain and the estimation result of the common noise precision comprises: Constructing a proxy function based on the estimation result of the target gain and the estimation result of the common noise precision; The partial derivatives of the proxy function with respect to the position of the sampling point in different dimensions are calculated, and the position of the sampling point is updated based on the partial derivatives.

4. The method for sensing a non-cellular system according to claim 3, characterized in that: The determining the position of the perceived scatterer based on the global information comprises: Extracting a target sampling point based on the global information, wherein the target sampling point is a sampling point where the CPU determines that the probability of a scatterer on the sampling point is greater than a preset threshold; The position of the target sampling point is determined as the position of the perceived scatterer.

5. The method for sensing a non-cellular system according to claim 2, characterized in that: The first preset convergence condition is that a first preset number of iterations is reached or the estimation result of two adjacent iterations changes less than a first preset threshold, and the second preset convergence condition is that a second preset number of iterations is reached or the estimation result of two adjacent iterations changes less than a second preset threshold.

6. The non-cellular system sensing method according to claim 1, characterized in that: The probability model of the received signal of the receiving AP is determined based on the time delay and arrival angle corresponding to the path passing through the sampling point and the time delay and arrival angle corresponding to the LOS path.

7. The method for sensing a non-cellular system according to any one of claims 1 to 6, characterized in that: Also includes: Arbitrarily select two receiving APs, estimate the delay using the received signals of the two receiving APs, and obtain the estimated delay; An initial position of a sampling point is determined based on the estimated time delay.

8. A non-cellular system sensing device, characterized in that: include: A probability model building module, used to build a probability model of system parameters in a non-cellular system, wherein the non-cellular system is deployed with multiple distributed APs and a CPU connected to the multiple distributed APs, the multiple distributed APs include a sending AP and multiple receiving APs, the system parameters include a received signal of any of the receiving APs, and the probability model of the received signal of the receiving AP is determined based on the angle and delay corresponding to the path passing through the sampling point; A local information calculation module, configured to calculate local information based on the probability model through each of the receiving APs, wherein the local information includes the probability that the receiving AP determines that a scatterer exists at different sampling points; A global information calculation module, configured to receive, through the CPU, local information sent by each of the receiving APs and calculate global information based on the local information, wherein the global information includes the probability that the CPU determines that there are scatterers at different sampling points; A scatterer position determination module is used to determine the position of the perceived scatterer based on the global information.

9. A computer program product, characterized in that When the computer program product is executed, the non-cellular system perception method according to any one of claims 1 to 7 is implemented.

10. An electronic device, characterized in that: comprising a memory and a processor, wherein: The memory is used to store the computer program; The processor is used to execute the computer program to implement the non-cellular system perception method as described in any one of claims 1 to 7.

11. A computer-readable storage medium, characterized in that: Used to store a computer program, wherein when the computer program is executed by a processor, the non-cellular system perception method according to any one of claims 1 to 7 is implemented.

Images