Radar channel and communication channel estimation method of communication and inductance integrated system

Abstract

The invention provides a radar channel and communication channel estimation method and device of a communication and inductance integrated system, and belongs to the technical field of wireless communication. The method comprises the following steps: determining a joint sparse vector of a radar channel and a communication channel based on a sparse vector of the radar channel and a sparse vector of the communication channel; determining a channel dynamic sparsity probability expression based on the joint sparse vector and a space-time Markov model; solving the channel dynamic sparsity probability expression based on a three-layer Bayesian reasoning framework to obtain a channel dynamic sparsity probability; substituting the dynamic sparsity probability of the channel into a maximum likelihood estimation problem of the environmental perception parameters, and solving the maximum likelihood estimation problem to obtain the environmental perception parameters; and estimating a radar channel and a communication channel based on the environment perception parameters. According to the radar channel and communication channel estimation method and device of the communication and inductance integrated system, the estimation precision of the radar channel and the communication channel can be effectively improved.

Description

Technical Field

[0001] This invention relates to the field of wireless communication technology, and in particular to a method and apparatus for estimating radar channels and communication channels in an integrated sensing and communication system. Background Technology

[0002] For Integrated Sensing and Communications (ISAC) systems, with the deployment of numerous antennas and the application of high-frequency bands, data transmission rates are continuously increasing, but this process is accompanied by increased costs and weakened signal penetration. Simultaneously, the increased sensitivity of channels to dynamic propagation environments introduces more significant uncertainties during transmission. Obtaining accurate Channel State Information (CSI) in complex propagation scenarios becomes more difficult. The widespread application of sensing devices makes environmental data acquisition more convenient, enhancing the perception and understanding of the physical world. Given that the variability and complexity of wireless channels stem from dynamic propagation environments, incorporating environmental information into channel estimation can improve estimation performance and enhance adaptability to environmental changes compared to traditional methods.

[0003] Traditional single-slot localization methods do not consider the mobility of the target, leading to increased localization errors in high-mobility scenarios over time, thus limiting their application in dynamic environments. Furthermore, single-static ISAC systems can only acquire radial velocity for moving objects, further restricting their ability to perceive dynamic environments. Therefore, how to combine the partial common sparsity exhibited by radar and communication channels in the location domain within single-static ISAC systems to achieve joint radar and communication channel estimation has become a crucial problem that urgently needs to be solved. Summary of the Invention

[0004] This invention provides a method and apparatus for estimating radar and communication channels in a sensor-integrated system, for the purpose of joint radar and communication channel estimation.

[0005] In a first aspect, the present invention provides a method for estimating radar channels and communication channels in an integrated sensing system, the method comprising: The joint sparse vector of the radar channel and the communication channel is determined based on the sparse vector of the radar channel and the sparse vector of the communication channel. The probabilistic expression for channel dynamic sparsity is determined based on the joint sparse vector and the spatiotemporal Markov model. The channel dynamic sparsity probability is obtained by solving the channel dynamic sparsity probability expression based on the three-layer Bayesian inference framework. The channel dynamic sparsity probability is substituted into the maximum likelihood estimation problem of the environment perception parameters, and the maximum likelihood estimation problem is solved to obtain the environment perception parameters. The radar channel and the communication channel are estimated based on the environmental perception parameters.

[0006] In one embodiment, the method further includes: The position and speed of the sensing target, scatterer, and terminal are estimated based on the sensing parameters.

[0007] In one embodiment, determining the joint sparse vector of the radar channel and the communication channel based on the sparse vector of the radar channel and the sparse vector of the communication channel includes: A grid region is constructed by taking the base station as the pole, the preset distance to the base station as the radius, and the preset angle range of the pole as the central angle; the grid region contains multiple grids, and the radial length and central angle of each grid are equal. The sparse vectors of the radar channel and the communication channel are determined based on the grid-based expression. Based on the support vectors of the sparse vectors of the radar channel and the sparse vectors of the communication channel, the joint sparse vector of the radar channel and the communication channel is determined. The support vectors of the sparse vectors of the radar channel are used to characterize whether there is a sensing target in the grid, and the support vectors of the sparse vectors of the communication channel are used to characterize whether there is a scatterer in the grid.

[0008] In one embodiment, the joint sparse vector is the union of the support vectors of the sparse vectors of the radar channel and the support vectors of the sparse vectors of the communication channel.

[0009] In one embodiment, solving the maximum likelihood estimation problem to obtain the environment perception parameters includes: In the E-step of the Expectation Maximization (EM) algorithm, the estimated values ​​of the environmental perception parameters for the current iteration are determined based on the estimated values ​​of the environmental perception parameters obtained in the previous iteration, and the posterior probability distribution of the Bayesian hidden variable set and the posterior probability distribution of the joint sparsity are optimized based on the estimated values ​​of the environmental perception parameters for the current iteration. In the M-step of the EM algorithm, the estimated values ​​of the environmental perception parameters for the current iteration are optimized based on the posterior probability distribution of the Bayesian hidden variable set optimized in the E-step and the posterior probability distribution of the joint sparsity.

[0010] In one embodiment, the E-step of the EM algorithm is solved using the Turbo framework; In the Turbo framework's 2D-MM message passing process, the support vectors of the sparse vectors of the radar channel and the sparse vectors of the communication channel are coupled through the joint sparse vector.

[0011] In a second aspect, the present invention provides a radar channel and communication channel estimation device for an integrated sensing system, the device comprising: The joint sparse determination module is used to determine the joint sparse vector of the radar channel and the communication channel based on the sparse vector of the radar channel and the sparse vector of the communication channel. The dynamic sparse representation module is used to determine the probabilistic expression of the channel's dynamic sparsity based on the joint sparse vector and the spatiotemporal Markov model. The dynamic sparsity solving module is used to solve the channel dynamic sparsity probability expression based on a three-layer Bayesian inference framework to obtain the channel dynamic sparsity probability. The sensing parameter solving module is used to substitute the channel dynamic sparsity probability into the maximum likelihood estimation problem of the environment sensing parameters, and solve the maximum likelihood estimation problem to obtain the environment sensing parameters. The channel estimation module is used to estimate the radar channel and the communication channel based on the environmental perception parameters.

[0012] Thirdly, the present invention provides an electronic device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described in the first aspect above.

[0013] Fourthly, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the method described in the first aspect above.

[0014] Fifthly, the present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the method described in the first aspect above.

[0015] The radar channel and communication channel estimation method and apparatus for the integrated sensing system provided by this invention proposes a sparse representation of the radar channel and communication channel in the location domain by utilizing the joint sparsity of the radar channel and communication channel in the location domain. Thus, the estimation of the radar channel and communication channel is realized based on the location domain sparsity. Therefore, compared with the existing single-slot positioning method, it fully considers the positioning error in high migration scenarios, thereby effectively improving the estimation accuracy of the radar channel and communication channel. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0017] Figure 1 This is a flowchart illustrating the radar channel and communication channel estimation method for the integrated sensing system provided by the present invention.

[0018] Figure 2 This is a schematic diagram of the framework of the D-SCVBI algorithm provided by the present invention.

[0019] Figure 3 This is the joint distribution provided by the present invention. Factor plot.

[0020] Figure 4 A schematic diagram of module A and module B of the D-SCVBI E step provided by this invention.

[0021] Figure 5 This is a schematic diagram of Algorithm 1 provided by the present invention.

[0022] Figure 6 This is a schematic diagram of the radar channel and communication channel estimation device of the integrated sensing system provided by the present invention.

[0023] Figure 7 This is a schematic diagram of the structure of the electronic device provided by the present invention. Detailed Implementation

[0024] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0025] The following describes one of the application scenarios of this invention: Consider a Time Division Duplexing (TDD) massive MIMO orthogonal Frequency Division Multiplexing (OFDM) ISAC system, where the base station (BS) is equipped with a uniform linear array of antennas spaced at half-wavelength intervals, containing... Root antenna.

[0026] This base station provides services to a single-antenna user equipment (UE) and simultaneously acts as a sensing target (hereinafter referred to as "sensing target") in a single static radar scattering environment. Specifically, the BS transmits downlink (DL) pilots to sense far-field point targets (sensing targets) in the scattering environment, and then the UE transmits uplink (UL) pilots to estimate the communication channel. This is done in a continuous two-dimensional scattering environment. There exists Individual perception targets and There are communication scatterers (scattering between the UE and the BS). Except for some overlap between the sensing targets and the communication scatterers, they are assumed to be independent and well separated from each other.

[0027] In the polar coordinate system, BS is located at the pole, and the polar axis is aligned with the horizontal line. In the... t Each time slot, the UE location is , No. k The location of the sensing target is , No. l The positions of the scatterers are Furthermore, assuming the BS is based on GPS or a 2D-DFT positioning algorithm, it has some prior information about the location of the user, radar target, and communication scattering objects.

[0028] in, Indicates that the UE is in the time slot t The position of the object in the polar coordinate system at any given time. Indicates time slot t The distance between the UE and the base station at any given time. Indicates time slot t The angle formed by the line connecting the UE and the base station with the polar axis at any given moment; Representing the perceived target k In the time slot t The position of the object in the polar coordinate system at any given time. Indicates time slot t Be aware of the target at all times k Distance to base station Indicates time slot t Be aware of the target at all times k The angle formed between the line connecting the base station and the polar axis; Represents a scatterer l In the time slot t The position of the object in the polar coordinate system at any given time. Indicates time slot t Time scatterer l Distance to base station Indicates time slot t Time scattererl The angle formed between the line connecting the base station and the polar axis.

[0029] Because the transmission interval is very short, the physical environment remains constant within each time slot (containing one uplink transmission and one downlink transmission), but varies between different time slots. To avoid interference, a guard interval exists between the uplink and downlink pilots. In this configuration, the BS can receive UL communication signals from the user terminal and perform DL tracking in the same time slot, enabling coordinated processing of uplink and downlink signals, thereby enhancing communication and sensing capabilities.

[0030] In addition, it has G Subcarriers and N The OFDM signal with one symbol is used for both UL communication and DL sensing. It is the subcarrier spacing. T b Represents the duration of the basic symbol. This includes the duration of the cyclic prefix (CP). T cp The total duration of the symbols, including B It is the total bandwidth of the system.

[0031] Figure 1 This is a flowchart illustrating the radar channel and communication channel estimation method for the integrated sensing system provided by this invention, as shown below. Figure 1 As shown, the method may include the following steps: Step 110: Determine the joint sparse vector of the radar signal and the communication channel based on the sparse vector of the radar channel and the sparse vector of the communication channel. Step 120: Determine the probabilistic expression for the channel dynamic sparsity based on the joint sparse vector and the spatiotemporal Markov model; Step 130: Solve the channel dynamic sparsity probability expression based on the three-layer Bayesian inference framework to obtain the channel dynamic sparsity probability; Step 140: Substitute the channel dynamic sparsity probability into the maximum likelihood estimation problem of the environment sensing parameters, and solve the maximum likelihood estimation problem to obtain the environment sensing parameters. Step 150: Estimate the radar channel and communication channel based on environmental perception parameters.

[0032] It should be noted that the entity executing the radar channel and communication channel estimation method of the aforementioned integrated sensing system can be a network-side device, such as a base station. The technical solution of this invention will be described in detail below using the example of a base station executing the radar channel and communication channel estimation method of the integrated sensing system provided by this invention.

[0033] Target perception aims to detect the presence of targets and estimate their location. To achieve this goal, in time slots... t The BS sends DL pilot signals to the desired corner sector, where in the first... g The subcarrier n One DL pilot symbol The received signal reflected by the target can be expressed as: (1) in, Represents the radar channel matrix; For having variance Additive white Gaussian noise with a mean of 0.

[0034] Considering the region There exists K A moving sensing target, time slot t The radar channel is: (2) in, , and They represent the first k The first goal in t The radar cross section (RCS), round-trip delay, and angle of arrival (AOA) for each time slot.

[0035] in k =0 indicates a mobile user. Round-trip time is... Doppler frequency shift is , It is the carrier frequency. It's the speed of light. and The goal k The magnitude and direction of the speed.

[0036] The response vector of the antenna array on the BS side (the time delay phase interval between antennas) is: (3) UE in time slot t Send UL pilot to BS In the first g The subcarrier n The received signal corresponding to each OFDM symbol is represented as follows: (4) in, Represents the communication channel vector. The mean is 0 and the variance is Additive white Gaussian noise.

[0037] Assume the communication channel has a line-of-sight (LoS) path and L A single-reflection non-line of sight (NLoS).

[0038] In this context, the LoS path is defined as the 0th channel path, and the UE is considered the 0th scatterer, with the location... Then the time slot t The communication channel model is as follows: (5) in, Indicates the first l Complex gain of a channel path. , and Let AoA, departure angle, and relative delay be represented by , respectively, for the l-th path.

[0039] for , , , leave the corner Relative delay The Doppler frequency shift of the LoS path is The NLoS path consists of the UE and the scatterer. l The resulting overall Doppler frequency shift is . and It is a scatterer l The magnitude and direction of the speed.

[0040] This invention calculates the actual velocity of overlapping radar targets (the sensing target and the scatterer overlap) by combining the different radial velocities between the radar channel and the communication channel. In this case, the Doppler effect is not seen as a hindrance, but rather as an opportunity to enhance sensing capabilities.

[0041] In one embodiment, step 110 may include: Step 1101: Using the base station as the pole, the preset distance to the base station as the radius, and the preset angle range of the pole as the central angle, construct a grid region; the grid region contains multiple grids, and the radial length and central angle of each grid are equal. Step 1102: Determine the sparse vectors of the radar channel and the communication channel based on the grid-based expression; Step 1103: Based on the support vectors of the sparse vectors of the radar channel and the sparse vectors of the communication channel, determine the joint sparse vector of the radar channel and the communication channel. Among them, the support vectors of the sparse vectors of the radar channel are used to characterize whether there is a sensing target in the grid, and the support vectors of the sparse vectors of the communication channel are used to characterize whether there is a scatterer in the grid.

[0042] Specifically, since the optimization problem is non-convex and requires the simultaneous optimization of multiple variables, it is easy to get trapped in local optima. Therefore, it is difficult to directly estimate the positions of the sensing target, scatterer and UE using the Maximum A Posteriori (MAP) method.

[0043] To address this problem, this invention introduces a grid for solving to achieve better perception and estimation performance.

[0044] Specifically, in step 1101, the two-dimensional space where the UE is located can be divided into... Non-overlapping regions. and Define a uniformly spaced grid within the range of (i.e., the preset angle range of the poles). and uniform angle grid Among them, the base station serves as the pole. It is a preset distance, used as the radius.

[0045] Understandable, The specific size can be determined based on the sensing target, scatterer, and UE furthest from the base station, i.e. It should be able to cover the sensing target, scatterer, and UE that are furthest from the base station.

[0046] In practice, dynamic position grids are introduced to eliminate the interval error between grids. This represents a dynamic position grid.

[0047] In this scenario, there always exists a grid set that covers the true locations of all sensed targets, scatterers, and UEs. In the algorithm, a uniform grid can be chosen as the initial point to efficiently approximate the optimal solution to the MAP estimation problem.

[0048] In step 1102, for ease of representation (index t omitted), the sparse basis of the dynamic location grid of the radar channel and communication channel (determined by the central angle of the grid and the distance to the base station) is defined as follows: (6) in Then formulas (2) and (5) correspond to the radar channel and communication channel, respectively: (7) (8) in, For the sparse vector of the location-domain radar channel, It is a sparse vector of the location-domain communication channel.

[0049] The q The elements are , indicating that it is located in the region Complex reflection coefficient of the target in the middle. The q The elements are , representing the complex channel gain of the communication channel, with the corresponding scatterer located in region . and It is a diagonal matrix, and the 0th diagonal element is... and .for Then there is the first q The diagonal elements are respectively and .

[0050] Using the sparse position-domain representation of radar and communication channels, the uplink and downlink received signals of all subcarriers can be represented as: (9) (10) in, It is the Kronecker product; It is the Khatri-Rao product. For simplicity, it is defined as follows: In the time slot t The g On each subcarrier, ,but G Subcarriers are stacked and . This represents the radar measurement matrix, while This represents the communication measurement matrix.

[0051] The radar signal and communication signal models can be combined into a single linear observation model: (11) make , , , , .

[0052] in, Represents environmental perception parameters. r t express t Time slot location domain grid (angle and polar radius representation).v t express t The velocity magnitude of the sensing target / scatterer / UE in the time slot location domain grid. φ t express t The velocity direction corresponding to the sensed target / scatterer / UE in the time slot location domain grid. express t UE location in time slot express t Slotted UE speed magnitude express t Time slot UE velocity direction.

[0053] Because the mobile channel evolves over time, the sparse vector of the radar channel... Sparse vectors of communication channels It exhibits dynamic sparsity. Therefore, this invention proposes the D-SCVBI (Dynamic Subspace-Constrained Variational Bayesian Inference) framework, which uses a three-layer hierarchical sparse prior model to capture the spatiotemporal two-dimensional structural sparsity under the uncertain perception matrix in formula (11).

[0054] Specifically, this invention employs a three-layer sparse prior model to characterize the joint sparsity of the location domain radar channel and communication channel in the same time slot.

[0055] In step 1103, let and express and The accuracy, of which and They are and The variance of . The corresponding support vectors are respectively and The joint support vectors are .

[0056] make This represents the joint sparse vector of the radar and communication channels, used to characterize the positions of the sensed target and the scatterer. This represents the logical operator "OR"; express s r The Middle q One element, express s c The Middle q One element, .

[0057] It should be noted that, s r Used to represent Does the target channel coefficient exist at the location in the grid? (The coefficient is generated if a sensed target is present.) s c Used to represent Does the location in the grid have a scattering channel coefficient (the coefficient is generated if a scatterer is present)?

[0058] For example, if the grid If there is a perceived target, then If the grid If there is a scatterer in it, then If the grid If there is no perceived target, then If the grid If there is no scatterer, then .in, This indicates that the base station can "see" the UE through radar echo signals. This indicates that the Loss path exists.

[0059] The radar channel and communication channel estimation method for the integrated sensing system provided by this invention introduces a grid to express the sparse vectors of the radar channel, the sparse vectors of the communication channel, and the joint sparse vectors of the radar channel and the communication channel. This simplifies the complex optimization problem and improves the estimation efficiency of the radar channel and the communication channel.

[0060] The joint distribution of all the above random variables can be expressed as: (12) Suppose a complex Gaussian distribution is The prior knowledge. With As a condition, If the elements are independent, then, Layer III (13) Precision vector Represented by the Bernoulli-Gamma distribution, then Layer II (14) in, , and , They are respectively right and The conditional gamma prior parameters.

[0061] To effectively represent The parameter must be zero or non-zero, and the parameter selection must satisfy the following conditions: ,and .

[0062] Support Vectors , and The joint distribution is expressed as: Layer I (15) The conditional distribution is as follows: (16) for based on The probability, The degree of overlap between the target and the scatterer.

[0063] It should be noted that Layer III, Layer II, and Layer I in formulas (13) to (15) represent the third, second, and first layers of the three-layer Bayesian inference framework, respectively.

[0064] Gradually changing propagation environments lead to similar scattering structures in adjacent time slots, resulting in time-dependent channel support. Furthermore, the clustering characteristics of scatterers on the base station side cause… The distribution of non-zero element clusters creates spatial sparsity.

[0065] In step 120, a spatiotemporal Markov model (2D-MM) is used to model the spatiotemporal correlation of channel support.

[0066] Spatial transition probability as well as Time transition probability as well as and , .at the same time, .

[0067] The dynamic sparsity probabilistic modeling of the spatiotemporal Markov model (2D-MM) of latent support vectors is as follows: (17) in ,in This refers to sparsity. To automatically learn noise accuracy, shape parameters are assumed. and rate parameters for The prior, namely .

[0068] The goal of the maximum likelihood estimation problem for environmental sensing parameters is to track environmental sensing parameters. , Given and The time slots are obtained through maximum likelihood estimation. t Unknown environmental perception parameters , can be represented as: (18) in, This represents the estimated values ​​of environmental perception parameters. Let... , , , , and This represents the hidden variables in the Bayesian model.

[0069] Obtain environmental perception parameter estimates Next, the conditional marginal posterior needs to be calculated. and You can get and .

[0070] The basic idea of ​​the D-SCVBI framework proposed in this invention is to utilize the spatiotemporal Markov model prior while simultaneously approximating the marginal posterior, and in the 1st... t Each time slot Maximize the log-likelihood .

[0071] This invention decomposes the joint optimization problem (i.e., formula (18)) into multiple sub-problems, time slots t The subproblem only involves the current observations. and posterior of and , The optimization avoids the computational complexity and memory requirements from increasing. t The problem of rapidly increasing growth.

[0072] In the time slot t The joint probability distribution is as follows: (19) Because of the exact posterior Since it is difficult to obtain, an approximate probability density function is used. and Approximation, i.e. .

[0073] The joint probability distribution of formula (18) based on the fusion of the prior information from the previous time slot using sequence Bayesian inference is as follows: (20) in .

[0074] As can be seen from formulas (18) to (20), to solve the maximum likelihood estimation problem of environmental perception parameters, it is necessary to obtain the channel dynamic sparsity probability.

[0075] Therefore, in step 130, the present invention first solves the channel dynamic sparsity probability expression based on a three-layer Bayesian inference framework to obtain the channel dynamic sparsity probability. That is, formula (17) is solved sequentially using formulas (15), (14), and (13) to obtain the channel dynamic sparsity probability. .

[0076] In step 140, the channel dynamic sparsity probability can be substituted into the maximum likelihood estimation problem of the environment sensing parameters, and the maximum likelihood estimation problem can be solved to obtain the environment sensing parameters.

[0077] That is, to After calculations using formulas (20) and (19), the results are substituted into formula (18), and formula (18) is solved to obtain the environmental perception parameters. .

[0078] In one embodiment, solving the maximum likelihood estimation problem to obtain environmental perception parameters may include: In the E-step of the EM algorithm, the estimated values ​​of the environmental perception parameters for the current iteration are determined based on the estimated values ​​of the environmental perception parameters obtained in the previous iteration, and the posterior probability distribution of the Bayesian hidden variable set and the posterior probability distribution of the joint sparsity are optimized based on the estimated values ​​of the environmental perception parameters for the current iteration. In the M-step of the EM algorithm, the estimated values ​​of the environmental perception parameters for this iteration are optimized based on the posterior probability distribution of the Bayesian hidden variable set optimized in the E-step and the posterior probability distribution of the joint sparsity.

[0079] Specifically, this invention proposes that the D-SCVBI algorithm iteratively execute the following two steps until convergence: D-SCVBI-E step: Based on the estimated environmental perception parameters obtained from the previous iteration, determine the estimated environmental perception parameters for the current iteration. and based on calculate (The posterior probability distribution of the Bayesian hidden variable set) and (Posterior probability distribution of joint sparsity) Approximate marginal posterior and .

[0080] D-SCVBI-M step: Based on the D-SCVBI-E step and Build The substitution function, and then proceed. Maximize the substitution function.

[0081] For the D-SCVBI-M step, due to complex coupled variables and the lack of a closed-form formula, directly maximizing the likelihood function is not feasible. Very difficult. Therefore, we construct its alternative function and discuss... Maximize the substitution function.

[0082] set up For a certain fixed point The substitution function constructed above satisfies the following property: (21a) (21b) (21c) Based on the Expectation Maximization (EM) algorithm, the following alternative function is used, letting Indicates the first i Optimization variables at the start of the next iteration: (twenty two) However, formula (22) is non-convex and the exact posterior is difficult to obtain, so we use... To approximate the conditional marginal posterior According to the mean-field assumption , yes The j-th variable in , To approximate the conditional marginal posterior For a given point , No. i In the next iteration Updated as follows (twenty three) Use the following gradient updates to obtain a stable solution (twenty four) in, The step size is determined by Armijo's rules.

[0083] In step 150, the environmental perception parameters for the last iteration are obtained using formula (24). Then, you can Substitute into formulas (2) to (11) to obtain the radar channel and the communication channel.

[0084] The radar and communication channel estimation method for the integrated sensing system provided by this invention proposes a sparse representation of the radar and communication channels in the location domain by utilizing the joint sparsity of the radar and communication channels in the location domain. Thus, the estimation of the radar and communication channels is realized based on the location domain sparsity. Therefore, compared with the existing single-slot positioning method, it fully considers the positioning error in high migration scenarios, thereby effectively improving the estimation accuracy of the radar and communication channels.

[0085] In one embodiment, the method further includes: The position and speed of the sensing target, scatterer, and terminal are estimated based on the sensing parameters.

[0086] It is understandable that, after obtaining environmental perception parameters Later, due to Specifically, it includes t Time slot location domain grid (angle and polar radius representation) t The velocity magnitude of the perceived target / scatterer / UE in the time slot location domain grid. t The velocity direction corresponding to the sensed target / scatterer / UE in the time slot location domain grid. t Time slot UE position, t Slotted UE speed size and t The UE velocity direction in the time slot can therefore be based on It can directly estimate the position and speed of the sensing target, scatterer, and terminal.

[0087] The radar channel and communication channel estimation method for the integrated sensing system provided by this invention estimates the position and moving speed of the sensing target, scatterer and terminal by sensing parameters, and can realize dynamic environmental parameter tracking in multiple time slots, thereby avoiding the increase of positioning error in high mobility scenarios and effectively improving practicality.

[0088] In one embodiment, the E-step of the EM algorithm can be solved using the Turbo framework; In the 2D-MM message passing process of the Turbo framework, the support vectors of the sparse vectors of the radar channel and the support vectors of the sparse vectors of the communication channel are coupled through joint sparse vectors.

[0089] Specifically, the D-SCVBI E-step can be based on the Turbo framework, which combines the SC-VBI estimator and 2D-MM message passing, such as... Figure 2 As shown.

[0090] Joint distribution The factor graph is as follows Figure 3 As shown in Table 1, the expressions for each factor node are as follows.

[0091] Table 1. Factors, Distributions, and Functional Forms

[0092] To overcome the performance limitations of cyclic factor graphs, such as Figure 4 Divide it into observation structures and support vector structure In the algorithm design, modules A and B perform inference separately. and .

[0093] definition , as well as , Module A output To module B, module B outputs information. Up to module A, both modules are executed iteratively until convergence.

[0094] For each Turbo iteration, module A is based on... and messages from module B and A sparse VBI estimator. Then module A obtains the posterior information... and Subtracting prior information (2D-MM prior) from the middle, the external information is passed to module B, that is... (25) Similarly, module B in The message passing is performed on the upper level, and the external message is passed to module A.

[0095] For Module A, the approximate conditional marginal posterior can be minimized. and The Kullback-Leibler (KL) divergence between them is determined, while taking into account The factorization structure therefore presents the following problems: (26) Based on the EM method, Each part can be obtained through alternating optimization, that is, for any given... The optimal solution that minimizes the KL divergence Given by the following formula: (27) in Specifically, , , , and The iterative updates are as follows, with time slot symbols omitted for simplicity. t : (1) Update make The mean and variance are expressed as follows: (28) The main computational burden in the algorithm is The high-dimensional matrix inversion operation is performed. To reduce the algorithm complexity, the Subspace Constrained Variational Bayesian Inference (SC-VBI) method is adopted.

[0096] (2) Update in, , As given below: (29) in, yes The posterior expectation.

[0097] (3) Update , , ,in, , , .

[0098] (4) Update ,in , .

[0099] For Module B message passing, update : exist In, with and The relevant factor plot is obtained through variable nodes. Coupled together, they can exchange information to gain access to... and A more accurate estimate. Derived from the sum-product rule. The information transmission algorithm on the surface. To simplify the expression, , , and They were used to represent , , and The abbreviation of, among which .

[0100] From factor nodes and To the variable node The messages are given by the following formulas: (30) in .

[0101] The channel support prior calculation for transmission to the next time slot is as follows: (31) in , , The formula is as follows: (32) information The following can be calculated: (33) in, The calculation summary is as follows Figure 5 In Algorithm 1 shown.

[0102] (34) in, .

[0103] Furthermore, for factor nodes Return to variable node (35) The radar channel and communication channel estimation device for the integrated sensing system provided by the present invention will be described below. The radar channel and communication channel estimation device for the integrated sensing system described below can be referred to in correspondence with the radar channel and communication channel estimation method for the integrated sensing system described above, and can achieve the same technical effect. It will not be repeated here.

[0104] Figure 6 This is a schematic diagram of the radar channel and communication channel estimation device for the integrated sensing system provided by the present invention. Figure 6 As shown, the device may include: The joint sparse determination module 610 is used to determine the joint sparse vector of the radar channel and the communication channel based on the sparse vector of the radar channel and the sparse vector of the communication channel. The dynamic sparse representation module 620 is used to determine the probabilistic expression of channel dynamic sparsity based on the joint sparse vector and the spatiotemporal Markov model. The dynamic sparsity solving module 630 is used to solve the channel dynamic sparsity probability expression based on a three-layer Bayesian inference framework to obtain the channel dynamic sparsity probability. The sensing parameter solving module 640 is used to substitute the channel dynamic sparsity probability into the maximum likelihood estimation problem of the environment sensing parameters, and solve the maximum likelihood estimation problem to obtain the environment sensing parameters. The channel estimation module 650 is used to estimate the radar channel and the communication channel based on the environmental perception parameters.

[0105] In one embodiment, the apparatus further includes: The target determination module (not shown in the figure) is used to estimate the position and moving speed of the sensed target, scatterer and terminal based on the sensed parameters.

[0106] In one embodiment, determining the joint sparse vector of the radar channel and the communication channel based on the sparse vector of the radar channel and the sparse vector of the communication channel includes: A grid region is constructed by taking the base station as the pole, the preset distance to the base station as the radius, and the preset angle range of the pole as the central angle; the grid region contains multiple grids, and the radial length and central angle of each grid are equal. The sparse vectors of the radar channel and the communication channel are determined based on the grid-based expression. Based on the support vectors of the sparse vectors of the radar channel and the sparse vectors of the communication channel, the joint sparse vector of the radar channel and the communication channel is determined. The support vectors of the sparse vectors of the radar channel are used to characterize whether there is a sensing target in the grid, and the support vectors of the sparse vectors of the communication channel are used to characterize whether there is a scatterer in the grid.

[0107] In one embodiment, the joint sparse vector is the union of the support vectors of the sparse vectors of the radar channel and the support vectors of the sparse vectors of the communication channel.

[0108] In one embodiment, solving the maximum likelihood estimation problem to obtain the environment perception parameters includes: In the E-step of the Expectation Maximization (EM) algorithm, the estimated values ​​of the environmental perception parameters for the current iteration are determined based on the estimated values ​​of the environmental perception parameters obtained in the previous iteration, and the posterior probability distribution of the Bayesian hidden variable set and the posterior probability distribution of the joint sparsity are optimized based on the estimated values ​​of the environmental perception parameters for the current iteration. In the M-step of the EM algorithm, the estimated values ​​of the environmental perception parameters for the current iteration are optimized based on the posterior probability distribution of the Bayesian hidden variable set optimized in the E-step and the posterior probability distribution of the joint sparsity.

[0109] In one embodiment, the E-step of the EM algorithm is solved using the Turbo framework; In the 2D-MM message passing process of the Turbo framework, the support vectors of the sparse vectors of the radar channel and the support vectors of the sparse vectors of the communication channel are coupled through the joint sparse vector.

[0110] Figure 7 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 7 As shown, the electronic device may include: a processor 710, a communications interface 720, a memory 730, and a communication bus 740, wherein the processor 710, the communications interface 720, and the memory 730 communicate with each other via the communication bus 740. The processor 710 can call logical instructions in the memory 730 to execute the radar channel and communication channel estimation method of the integrated sensing system described in any of the above embodiments, for example including: The joint sparse vector of the radar channel and the communication channel is determined based on the sparse vector of the radar channel and the sparse vector of the communication channel. The probabilistic expression for channel dynamic sparsity is determined based on the joint sparse vector and the spatiotemporal Markov model. The channel dynamic sparsity probability is obtained by solving the channel dynamic sparsity probability expression based on the three-layer Bayesian inference framework. The channel dynamic sparsity probability is substituted into the maximum likelihood estimation problem of the environment perception parameters, and the maximum likelihood estimation problem is solved to obtain the environment perception parameters. The radar channel and the communication channel are estimated based on the environmental perception parameters.

[0111] Furthermore, the logical instructions in the aforementioned memory 730 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, essentially, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0112] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the radar channel and communication channel estimation method of the integrated sensing system described in any of the above embodiments, for example including: The joint sparse vector of the radar channel and the communication channel is determined based on the sparse vector of the radar channel and the sparse vector of the communication channel. The probabilistic expression for channel dynamic sparsity is determined based on the joint sparse vector and the spatiotemporal Markov model. The channel dynamic sparsity probability is obtained by solving the channel dynamic sparsity probability expression based on the three-layer Bayesian inference framework. The channel dynamic sparsity probability is substituted into the maximum likelihood estimation problem of the environment perception parameters, and the maximum likelihood estimation problem is solved to obtain the environment perception parameters. The radar channel and the communication channel are estimated based on the environmental perception parameters.

[0113] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the radar channel and communication channel estimation method for the integrated sensing system described in any of the above embodiments, for example including: The joint sparse vector of the radar channel and the communication channel is determined based on the sparse vector of the radar channel and the sparse vector of the communication channel. The probabilistic expression for channel dynamic sparsity is determined based on the joint sparse vector and the spatiotemporal Markov model. The channel dynamic sparsity probability is obtained by solving the channel dynamic sparsity probability expression based on the three-layer Bayesian inference framework. The channel dynamic sparsity probability is substituted into the maximum likelihood estimation problem of the environment perception parameters, and the maximum likelihood estimation problem is solved to obtain the environment perception parameters. The radar channel and the communication channel are estimated based on the environmental perception parameters.

[0114] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0115] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0116] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for estimating radar and communication channels in an integrated sensing system, characterized in that, The method includes: The joint sparse vector of the radar channel and the communication channel is determined based on the sparse vector of the radar channel and the sparse vector of the communication channel. The probabilistic expression for channel dynamic sparsity is determined based on the joint sparse vector and the spatiotemporal Markov model. The channel dynamic sparsity probability is obtained by solving the channel dynamic sparsity probability expression based on the three-layer Bayesian inference framework. The channel dynamic sparsity probability is substituted into the maximum likelihood estimation problem of the environment perception parameters, and the maximum likelihood estimation problem is solved to obtain the environment perception parameters. The radar channel and the communication channel are estimated based on the environmental perception parameters.

2. The radar channel and communication channel estimation method for the integrated sensing system according to claim 1, characterized in that, The method further includes: The position and speed of the sensing target, scatterer, and terminal are estimated based on the sensing parameters.

3. The radar channel and communication channel estimation method for the integrated sensing system according to claim 1, characterized in that, The determination of the joint sparse vector of the radar channel and the communication channel based on the sparse vector of the radar channel includes: A grid region is constructed by taking the base station as the pole, the preset distance to the base station as the radius, and the preset angle range of the pole as the central angle; the grid region contains multiple grids, and the radial length and central angle of each grid are equal. The sparse vectors of the radar channel and the communication channel are determined based on the grid-based expression. Based on the support vectors of the sparse vectors of the radar channel and the sparse vectors of the communication channel, the joint sparse vector of the radar channel and the communication channel is determined. The support vectors of the sparse vectors of the radar channel are used to characterize whether there is a sensing target in the grid, and the support vectors of the sparse vectors of the communication channel are used to characterize whether there is a scatterer in the grid.

4. The radar channel and communication channel estimation method for the integrated sensing system according to claim 3, characterized in that, The joint sparse vector is the union of the support vectors of the sparse vectors of the radar channel and the support vectors of the sparse vectors of the communication channel.

5. The radar channel and communication channel estimation method for the integrated sensing system according to claim 3, characterized in that, Solving the maximum likelihood estimation problem to obtain the environmental perception parameters includes: In the E-step of the Expectation Maximization (EM) algorithm, the estimated values ​​of the environmental perception parameters for the current iteration are determined based on the estimated values ​​of the environmental perception parameters obtained in the previous iteration, and the posterior probability distribution of the Bayesian hidden variable set and the posterior probability distribution of the joint sparsity are optimized based on the estimated values ​​of the environmental perception parameters for the current iteration. In the M-step of the EM algorithm, the estimated values ​​of the environmental perception parameters for the current iteration are optimized based on the posterior probability distribution of the Bayesian hidden variable set optimized in the E-step and the posterior probability distribution of the joint sparsity.

6. The radar channel and communication channel estimation method for the integrated sensing system according to claim 5, characterized in that, The E-step of the EM algorithm is solved using the Turbo framework. In the 2D-MM message passing process of the spatiotemporal Markov model of the Turbo framework, the support vectors of the sparse vectors of the radar channel and the support vectors of the sparse vectors of the communication channel are coupled through the joint sparse vector.

7. A radar channel and communication channel estimation device for an integrated sensing system, characterized in that, The device includes: The joint sparse determination module is used to determine the joint sparse vector of the radar channel and the communication channel based on the sparse vector of the radar channel and the sparse vector of the communication channel. The dynamic sparse representation module is used to determine the probabilistic expression of the channel's dynamic sparsity based on the joint sparse vector and the spatiotemporal Markov model. The dynamic sparsity solving module is used to solve the channel dynamic sparsity probability expression based on a three-layer Bayesian inference framework to obtain the channel dynamic sparsity probability. The sensing parameter solving module is used to substitute the channel dynamic sparsity probability into the maximum likelihood estimation problem of the environment sensing parameters, and solve the maximum likelihood estimation problem to obtain the environment sensing parameters. The channel estimation module is used to estimate the radar channel and the communication channel based on the environmental perception parameters.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the radar channel and communication channel estimation method for the integrated sensing system as described in any one of claims 1 to 6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the radar channel and communication channel estimation method for the integrated sensing system as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the radar channel and communication channel estimation method for the integrated sensing system as described in any one of claims 1 to 6.

Images