ISAC angular domain energy imaging method and system based on actual measurement channel response calibration ray tracking

Abstract

The invention discloses an ISAC angular domain energy imaging method and an ISAC angular domain energy imaging system based on actual measurement channel response calibration ray tracing, and mainly solves the problem that an imaging model faces serious channel mismatch and cross-scene generalization limitation due to the fact that simulation data lacks physical consistency constraint in the prior art. According to the scheme, the method comprises the steps that geometric features are extracted from public map data or laser point cloud data, and an ISAC scene environment geometric model is constructed; acquiring an actual measurement channel impulse response containing multipath information by using the communication and sensing integrated equipment; performing path matching on the actual measurement channel impulse response and the ray tracing simulation channel response, inverting environment propagation parameters and correcting scattering parameters; and inputting the calibrated parameters into a ray tracking model, counting energy contribution in each space direction under the condition of a single antenna, obtaining angular domain energy distribution and confidence coefficient, and mapping the angular domain energy distribution and confidence coefficient to a spherical grid to form an imaging result. According to the method, physical mismatch between simulation and real environments is effectively eliminated, low-cost and high-precision angular domain energy imaging and confidence evaluation are realized under the condition of a single antenna, and the method can be used for imaging reconstruction of an ISAC system, environmental perception and generation of a neural network training data set.

Description

Technical Field

[0001] This invention belongs to the field of integrated sensing and communication (ISAC), wireless channel modeling and electromagnetic imaging, and specifically relates to an ISAC angular domain energy imaging method and system, which can be used for ISAC system imaging reconstruction, environmental perception and generation of neural network training datasets. Background Technology

[0002] ISAC, by sharing spectrum and hardware resources, endows communication systems with environmental awareness and high-resolution imaging capabilities. In the field of ISAC imaging, existing technologies mainly follow two routes: one is traditional imaging algorithms based on idealized point targets or statistical channel models such as Rayleigh and Rician distributions; the other is data-driven imaging methods based on deep learning.

[0003] Traditional model-driven ISAC imaging methods typically establish echo models based on point scatterers, extended targets, and statistical channel assumptions such as Rayleigh and Rician, and invert scene reflectivity through techniques such as range compression, range-Doppler processing, and array beamforming or sparse reconstruction. Under conditions of slow time-varying, long coherence times, and multiple snapshots, these methods can achieve high resolution and interpretability. However, in scenarios with short burst communication frames, fast-moving targets, or strong multipath clutter, they often face problems such as insufficient available snapshots, model mismatch, and rank deficiency and sidelobe leakage caused by coherent multipath, resulting in a significant deterioration in angle, time delay, Doppler estimation, and imaging quality. Common improvement measures include spatial smoothing, coherent multipath suppression, and robust estimation, but these usually come with a loss of effective array elements, bandwidth, or coherent accumulation gain, thus limiting further improvements in imaging performance.

[0004] Data-driven deep learning-based ISAC imaging methods do not rely on strict point target or statistical channel priors. They can directly use raw IQ data from multiple antennas, multiple carriers, and multiple frames, or intermediate spectrograms such as range-angle maps, as input. Through end-to-end networks, or by unfolding the iterative inversion process into a trainable network, they achieve denoising and sidelobe removal, undersampling completion, and super-resolution reconstruction under the combined effect of data consistency constraints and learnable priors. These methods are generally more robust under complex multipath conditions, non-ideal hardware, and low snapshot speeds, and can achieve near-full-sampling imaging results with fewer time-frequency and pilot resources. However, their performance is highly dependent on the coverage and domain consistency of the training data, resulting in limited cross-scene generalization and insufficient interpretability. Furthermore, they may introduce reconstruction biases under low signal-to-noise ratio or out-of-distribution conditions. Therefore, in engineering implementation, it is usually necessary to combine physical constraints, confidence assessment, and online calibration mechanisms to ensure stable and reliable imaging output.

[0005] Patent application CN202510300859.2 discloses an imaging method and system based on an integrated sensing and communication device. It utilizes compressed sensing to acquire initial images and then performs image enhancement through a GAN network. While this approach leverages deep learning to alleviate the imaging challenges under sparse observations to some extent, its core logic lies in image-level post-processing enhancement rather than channel-level physical calibration. Furthermore, it fails to consider how to use measured CIR to correct environmental parameters and scattering coefficients in the ray-tracing model, resulting in a lack of strict physical consistency constraints in the generated data. Once there is a distributional shift between the training data and the actual scene, this type of method struggles to guarantee imaging reliability.

[0006] Therefore, the key to breaking through the bottleneck of existing imaging technology lies in how to perform parameter inversion and correction of the ray tracing model based on measured data, eliminate channel mismatch between simulation and measurement, and construct an ISAC dataset with strong physical consistency, low cost and realism. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of the prior art by proposing an ISAC angular domain energy imaging method and system based on measured channel response calibration ray tracing. This method uses measured channel impulse response to perform parameter inversion calibration of the ray tracing model and performs angular domain energy imaging under single-antenna conditions, thereby reducing system complexity and construction cost, and improving the reliability and performance of imaging results.

[0008] The key technology of this invention is to construct a calibration objective function by path matching the measured CIR and the simulated channel response using ray tracing, inverting the environmental propagation parameters and scattering coefficients, and thus obtaining high-fidelity channel response and angular domain energy imaging results that are more consistent with the real physical environment. The implementation scheme includes the following:

[0009] 1. An ISAC angular domain energy imaging method based on measured channel response calibration ray tracing, characterized in that it includes:

[0010] (1) Extract the geometric feature data of buildings, ground and scatterers in the target area from public map data or laser point cloud data, construct the geometric model of the ISAC scene environment, and set the initial electromagnetic parameters of the environment and the scattering model parameters;

[0011] (2) Using a communication sensing integrated device, a known detection signal is transmitted in the scene environment, and the received signal is matched and filtered and correlated at the receiving end to obtain the measured channel impulse response containing multipath delay, amplitude and phase information;

[0012] (3) Path matching is performed between the measured channel impulse response and the simulated channel response of ray tracing, a calibration objective function is constructed, and environmental propagation calibration parameters are calculated through parameter inversion algorithm to update the environmental electromagnetic parameters in the ray tracing model.

[0013] (4) Based on the energy distribution characteristics of the non-mirror multipath component in the measured channel impulse response, the scattering parameters of different objects and regions in the scene environment are corrected.

[0014] (5) Input the updated environmental electromagnetic parameters and the corrected scattering coefficient into the ray tracing model. Based on the propagation effects such as reflection, transmission, scattering and diffraction, statistically analyze the energy contribution of each spatial direction under single antenna conditions to obtain the spatial angular domain energy distribution and confidence level.

[0015] (6) Map the angular energy distribution onto a preset spherical grid to form an angular energy imaging result.

[0016] Furthermore, the correction of scattering parameters for different objects and regions in the scene environment in step (4) includes the following:

[0017] (4a) Based on the final set of matching pairs The measured multipath observation set Corresponding reflection and diffraction paths of the middle mirror

[0018] Set of matching pairs The energy is removed from the measured total energy to obtain the measured scattering residual energy. :

[0019] (4b) Define the scattering parameters as the set of scattering coefficients. and scattering direction lobe parameter set :

[0020] (4c) Calculate the simulated scattering energy generated by the scattering path based on the scattering parameters. :

[0021] (4d) The simulated scattering energy Decomposed according to surface type and spatial partition, resulting in each Simulated scattering energy components corresponding to the element ;

[0022] (4e) According to each Simulated scattering energy components of the unit Construct residual assignment weights ;

[0023] (4f) The measured scattering residual energy components The allocation is as follows: ;

[0024] (4g) Based on the measured scattering energy components of each unit With simulated scattering energy components Differences, iteratively update scattering coefficients ;

[0025] (4h) Scattering direction lobe parameters Update again;

[0026] (4i) The simulation error limit between the measured and simulated scattered energy components. Maximum number of iterations of the algorithm To determine whether the scattering parameter correction is complete: if the measured scattering energy components... With simulated scattering energy components The difference is less than the simulation error limit Or the number of algorithm iterations reaches If the condition is met, the scattering parameter correction is terminated; otherwise, return to step (4g).

[0027] Furthermore, in step (6), the angular energy distribution is mapped onto a preset spherical grid to form an angular energy imaging result, which is implemented by:

[0028] (6a) Divide the corner domain into spherical grid unit According to the energy distribution of the angular domain Define the initial angular domain energy pixel. ;

[0029] (6b) Improve the initial angular domain energy pixels Spatial continuity, replacing indicator functions with angular domain kernel functions The mapped angular domain energy pixels after neighborhood diffusion are obtained. ;

[0030] (6c) Set the energy weighting coefficient Calculate the maximum energy pixel among all spherical grid cells. Energy pixels in the mapped angular domain Selecting from those that meet the requirements The high-energy grid cells are locally subdivided to form an adaptive spherical grid, thereby improving the corner energy pixels. The angular resolution in the target direction is used to obtain the final angular domain energy imaging results. .

[0031] 2. An ISAC angular domain energy imaging system based on measured channel response calibration ray tracing, characterized in that it comprises:

[0032] The initialization module is used to extract geometric feature data of buildings, ground and scatterers in the target area from public map data or laser point cloud data, construct the geometric model of the ISAC scene environment, and set the initial electromagnetic parameters of the environment and the scattering model parameters.

[0033] The impulse response module is used to transmit a known detection signal in the scene environment using the integrated communication and sensing device, and to perform matched filtering or correlation processing on the received signal at the receiving end to obtain the measured channel impulse response. The measured channel impulse response includes at least multipath delay, amplitude and phase information.

[0034] The ray tracing module is used to calculate the simulated ray tracing channel response and output a set of propagation paths under the geometric model of the scene environment and the initial electromagnetic parameters and scattering model parameters of the environment.

[0035] The propagation parameter calibration module is used to perform path matching between the measured channel impulse response and the ray tracing simulated channel response, construct a calibration objective function, and solve the environmental propagation calibration parameters through a parameter inversion algorithm to update the environmental electromagnetic parameters in the ray tracing model.

[0036] The scattering parameter calibration module is used to invert and correct the scattering parameters of different objects or regions in the scene environment based on the energy distribution characteristics of the non-mirror multipath component in the measured channel impulse response and the scattering path information obtained by ray tracing. The scattering parameters include at least scattering coefficients set according to category and spatial partition.

[0037] The imaging module is used to perform weighted accumulation and statistics on the energy contribution of each spatial direction under single-antenna conditions based on propagation effects such as reflection, transmission, scattering and diffraction under updated environmental electromagnetic parameters and corrected scattering parameters, to obtain the spatial angular domain energy distribution, and to map the angular domain energy distribution onto a preset spherical grid to form a confidence distribution map of the angular domain energy imaging result and the angular domain energy distribution.

[0038] Compared with the prior art, the present invention has the following advantages:

[0039] Firstly, by introducing measured channel impulse response to calibrate the ray tracing model, this invention significantly reduces environmental modeling errors and channel mismatch problems.

[0040] Secondly, this invention improves the characterization accuracy of complex multipath propagation scenarios by jointly inverting environmental propagation parameters and scattering coefficients.

[0041] Thirdly, this invention effectively reduces hardware costs and system complexity by using ray tracing for angular domain energy imaging under single-antenna conditions.

[0042] Fourth, by mapping the propagation effect weights of the ray-tracking output, this invention enables confidence assessment of the angular domain energy distribution, providing a reliability metric and engineering usability for imaging results.

[0043] Fifth, the present invention improves imaging performance and stability by adaptively performing spherical mesh and neighborhood diffusion mapping. Attached Figure Description

[0044] Figure 1 This is a flowchart illustrating the implementation of the ISAC angular domain energy imaging method based on measured channel response calibration ray tracing according to the present invention.

[0045] Figure 2 This is a block diagram of the ISAC angular domain energy imaging system based on measured channel response calibration ray tracing according to the present invention;

[0046] Figure 3 This is a schematic diagram illustrating the use of modeling files in an embodiment of the method of the present invention;

[0047] Figure 4 This is a schematic diagram of the ray tracing output after optimizing the simulation conditions in an embodiment of the method of the present invention;

[0048] Figure 5 These are angular domain energy imaging results generated by the present invention under different simulation parameters. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0050] This invention is designed for ISAC scenarios. It uses measured channel impulse response to calibrate environmental propagation parameters and scattering coefficients, and then realizes single-antenna angular domain energy imaging based on calibrated ray tracing. It is also used to construct datasets for ISAC imaging reconstruction and neural network training.

[0051] Example 1: ISAC angular domain energy imaging method based on measured channel response calibration ray tracing.

[0052] Reference Figure 1 The implementation steps of this embodiment include the following:

[0053] Step 1: Build the ISAC scene environment model and set system parameters.

[0054] In this embodiment, an environmental model of the target ISAC scene is first constructed, which includes, but is not limited to, buildings, ground and other potential scatterers. Its geometric information can be derived from publicly available map data, laser point cloud data or artificial modeling data.

[0055] During the environmental modeling phase, initial electromagnetic parameters are assigned to each type of object, including dielectric constant, conductivity, and initial scattering model parameters. These parameters can be derived from empirical models or publicly available literature and are used as initial inputs for ray tracing simulation.

[0056] The specific implementation of this step includes:

[0057] 1.1) Preprocessing multi-source geometric feature data:

[0058] Extract geometric feature data of buildings, ground and scatterers in the target area from public map data, laser point cloud data or artificial modeling data, convert them to the same spatial coordinate system, and complete scale consistency and registration to ensure spatial alignment between different data sources;

[0059] 1.2) Constructing instantiated geometry:

[0060] The building geometric feature data is instantiated into a building geometry, the ground geometric feature data is instantiated into a ground geometry, and the scatterer geometric feature data is instantiated into a scatterer geometry. The building geometry includes at least the facade and the roof, the ground geometry includes at least the ground patches, and the scatterer geometry is a set of parameterized geometries that can be used for ray propagation. All geometries are uniformly represented as boundary representations composed of a set of vertices and a set of patches.

[0061] 1.3) Constructing scene boundaries:

[0062] Based on the building's plan outline and height information, wall and roof panels are generated to form a closed boundary structure on the building surface. A continuous ground surface is generated based on the ground boundary and terrain height field, and spliced ​​and trimmed with the building's bottom boundary when necessary to avoid overlap and interweaving.

[0063] 1.4) Generate a meshed scene:

[0064] Based on the spatial location, size range, and principal direction of the scatterer, the scatterer is constructed as a ray-traced computable geometry to reduce the number of facets while ensuring that the scattering characteristics can be expressed.

[0065] The geometric boundaries of buildings, ground, and scattering bodies are triangulated with the fitted surface to generate a triangular mesh containing vertex coordinates, triangular patch indices, and normal vectors.

[0066] 1.5) Constructing the environmental geometric model:

[0067] Assign object IDs and material IDs to different types of facets to form an ISAC scene environment geometry model that allows the ray tracing engine to perform intersection calculations between rays and triangle facets, as well as calculations of reflection, diffraction, and scattering directions. ;

[0068] The geometric model of the scene environment , represents a collection containing multiple types of surfaces and objects. ;

[0069] 1.6) Initialize environmental material information:

[0070] Set collection The initial electromagnetic parameters of the environment and the scattering model parameters are used to obtain the set of material parameters. ;

[0071] Each type of surface is obtained based on the material electromagnetic parameters given in relevant ITU recommendations. relative permittivity Equivalent conductivity ;

[0072] The surface roughness of the scatterer is estimated based on the original laser point cloud data, and the surface roughness of each type of surface is obtained. equivalent roughness Finally, the set of material parameters is obtained. :

[0073] .

[0074] Step 2: Perform ISAC channel measurements and obtain the channel impulse response (CIR).

[0075] After completing the environment modeling, channel measurements are performed in the environment using the ISAC prototype system or a communication sensing integrated device. In this embodiment, the ISAC system adopts a single-antenna transceiver structure, and the antenna pattern can be obtained through calibration. The channel impulse response is obtained by transmitting a known probe signal and performing matched filtering at the receiving end, which includes:

[0076] 2.1) Set the carrier frequency of the ISAC device ,bandwidth Sampling period Transmission power ;

[0077] 2.2) Processing ISAC received data:

[0078] Receive signal With the known detection signal Utilizing impulse response Perform matched filtering to obtain the measured channel impulse response output. :

[0079] ,

[0080] in, For complex conjugate, For integration variables, This refers to the amount of sliding;

[0081] right Sampling is performed to obtain the discrete channel impulse response. :

[0082] ,

[0083] In the formula, To distinguish the number of multipaths, , Distribution is the first The amplitude and phase of the multipath, For delay index, Noise term;

[0084] 2.3) Constructing a multipath observation set:

[0085] To facilitate alignment with ray tracing, the discrete channel impulse response is used. The time delay, amplitude, and phase information of the resolvable multipath component are extracted, and the receiver obtains the direction-of-arrival unit vector through beam scanning and angle spectrum estimation. Based on these parameters, the measured multipath observation set is obtained. :

[0086]

[0087] in, Let be the propagation delay of the k-th multipath.

[0088] Step 3: Calibrate the environmental propagation parameters based on the measured CIR.

[0089] 3.1) Set delay threshold With angular domain threshold ;

[0090] 3.2) Constructing a set of ray tracing paths:

[0091] Using the ISAC scene environment geometry model and ray tracing program, a set of ray tracing paths containing time delay and arrival unit direction information is generated. :

[0092] ,

[0093] in, It is the effective multipath quantity. For the first Propagation delay of multiple paths, , The first The amplitude and phase of the multipath. The unit vector representing the direction of arrival;

[0094] 3.3) Constructing constraints for the set of matching pairs:

[0095] Based on measured multipath observation sets and ray tracing path set Construct matching pair set constraints :

[0096] ,

[0097] in, For ray tracing Path delay, For the actual measurement Multipath delay, , These represent the path arrival directions of ray tracing and actual measurement, respectively.

[0098] Based on the candidate matching pair set Construct matching costs or matching probabilities, and use optimal assignment to determine the final set of matching pairs. To obtain the matching path corresponding to each path, the unassigned measured multipath and ray tracing path are used as residual paths for subsequent scattering parameter inversion;

[0099] 3.4) Constructing the calibration objective function:

[0100] Based on the set of material parameters to be optimized Construct global parameters to be calibrated :

[0101]

[0102] in, , These represent the unified reflection loss scaling factor and transmission loss scaling factor, respectively.

[0103] Based on the final set of matched pairs and global parameters The calibration objective function is constructed using the regularized least squares criterion. :

[0104] ,

[0105] in, , In the parameters respectively The first ray tracing calculation Path amplitude and phase, These are the initial parameters; These are regularization weights used to avoid overfitting. To match the weights;

[0106] 3.5) Solve the calibration objective function The problem of minimizing the problem yields the optimal calibration parameters. :

[0107] ,

[0108] in, For the first The physically feasible range of electromagnetic parameters for similar materials is: similar materials;

[0109] 3.6) Based on the optimal calibration parameters Update the environmental propagation calibration parameters in the ray tracing model.

[0110] Step 4: Correct the scattering coefficient based on the measured CIR.

[0111] 4.1) Constructing the measured scattering residual energy:

[0112] Based on the final set of matching pairs The measured multipath observation set Set of matching pairs corresponding to the specular reflection path and diffraction path The energy is removed from the measured total energy to obtain the measured scattering residual energy. :

[0113]

[0114] 4.2) Define scattering parameters:

[0115] According to the scattering coefficient The scattering parameters are defined as the set of scattering coefficients. :

[0116] ,

[0117] in, For the first Materials, spatial zoning The corresponding scattering coefficient, For the total number of material categories, This represents the total number of spatial partitions.

[0118] Based on the scattering lobe intensity parameters The main lobe of the scattering lobe points to Scattering lobe sharpness parameter Define the set of scattering direction lobe parameters. :

[0119] ,

[0120] in, For the first Materials, spatial zoning The corresponding scattering lobe intensity parameter is used to adjust the overall amplitude of the scattering direction distribution; For the first Materials, spatial zoning The corresponding scattering lobe main lobe direction is used to characterize the dominant emission direction of the scattered energy; For the first Materials, spatial zoning The corresponding scattering lobe sharpness parameter is used to control the scattering lobe width;

[0121] 4.3) Constructing simulated scattering energy components:

[0122] 4.3.1) Based on the scattering parameter set of scattering coefficients and scattering direction lobe parameter set Calculate the simulated scattering energy generated by the scattering path. :

[0123] ,

[0124] in, This is the set of scattering paths obtained by ray tracing. and Scattering paths The corresponding surface category index and spatial partition index, The direction of incidence. The direction of scattering emission; Here is the scattering angle distribution function, used to characterize the scattering angle distribution in the incident direction. Under certain conditions, the scattered energy is directed towards the emission direction. The allocation of scattering lobes can be parameterized using a directional scattering lobe model;

[0125] 4.3.2) The simulated scattering energy Decomposed according to surface type and spatial partition, resulting in each Simulated scattering energy components corresponding to the element :

[0126] ,

[0127] This expression should satisfy: ;

[0128] 4.4) Constructing the measured scattering residual energy components:

[0129] 4.4.1) According to each Simulated scattering energy components of the unit Construct residual assignment weights :

[0130]

[0131] in, To prevent constants with a denominator of zero;

[0132] 4.4.2) Assign weights based on residuals and measured scattering residual energy Calculate the measured scattering residual energy components : ;

[0133] 4.5) Update the scattering direction lobe parameters:

[0134] 4.5.1) Based on the measured scattered energy components of each unit With simulated scattering energy components Differences, iteratively update scattering coefficients :

[0135] ,

[0136] In the formula, This indicates the projection onto the physically feasible region. Step size;

[0137] 4.5.2) Further analysis of the scattering direction lobe parameters Update:

[0138] ,

[0139] in, Update the step size for the directional lobe parameters;

[0140] 4.5.3) Based on the optimal calibration parameters Update the scattering direction lobe parameters in the ray tracing model.

[0141] Step 5: Generate spatial angular domain energy and confidence distribution data.

[0142] 5.1) Perform gain compensation on path energy based on the antenna pattern, and set weights for matching confidence weighting and path type weighting. ;

[0143] 5.2) The set of effective propagation paths obtained from ray tracing Obtain the destination direction for each path. and corresponding complex amplitude ;

[0144] The energy from each path is then assigned to a corner domain unit according to its direction of arrival and weighted and accumulated to obtain the corner domain energy distribution. :

[0145] ,

[0146] in, For corner domain units, For indicator functions;

[0147] 5.3) Energy distribution in the aforementioned angular domain Normalized scaling correction is performed to obtain the normalized angular domain energy. :

[0148] ,

[0149] in, To prevent constants with a denominator of zero;

[0150] 5.4) Weighted according to matching confidence Accumulate the weights within the angular domain cells and output the corresponding angular domain energy confidence distribution. :

[0151] .

[0152] Step 6: Perform single-antenna angular domain energy imaging.

[0153] 6.1) Divide the corner domain into spherical grid unit And the angular domain energy distribution obtained in step 5 Define the initial angular domain energy pixel. ;

[0154] 6.2) Enhance corner energy pixels Spatial continuity, replacing indicator functions with angular domain kernel functions The mapped angular domain energy pixels after neighborhood diffusion are obtained. :

[0155] ,

[0156] in, The direction unit vector of the th effective propagation path, The unit vector is the direction of the center of the grid cell. Kernel function, used to describe the contribution of a path to neighboring corner pixels;

[0157] 6.4) Set the energy weighting coefficient Calculate the maximum energy pixel among all spherical grid cells. Energy pixels in the mapped angular domain Selecting from those that meet the requirements The high-energy grid cells are locally subdivided to form an adaptive spherical grid, thereby improving the corner energy pixels. The angular resolution in the target direction is used to obtain the final angular domain energy imaging results. .

[0158] It should be noted that the step numbers in the above embodiments are only used to facilitate the explanation of the technical solution of the present invention and are not intended to limit the execution order. Unless otherwise specified, the order of each step can be adjusted according to actual needs.

[0159] Example 2: ISAC angular domain energy imaging system based on measured channel response calibration ray tracing.

[0160] Reference Figure 2 This embodiment includes: initialization module 1, impulse response module 2, ray tracing module 3, propagation parameter calibration module 4, scattering parameter calibration module 5, imaging module 6, and output module 7. Module 4 includes path set submodule 41, matching cost submodule 42, and parameter inversion submodule 43; module 5 includes residual energy submodule 51, simulated scattering energy submodule 52, and scattering parameter iteration submodule 53.

[0161] The working principle of the entire system is as follows:

[0162] The initialization module 1 is used to extract geometric feature data of buildings, ground and scatterers in the target area from public map data or laser point cloud data, construct an ISAC scene environment geometric model, set the initial electromagnetic parameters of the environment and the scattering model parameters, and transmit the ISAC scene environment geometric model as output to module 3.

[0163] The impulse response module 2 is used to transmit a known detection signal in the scene environment using the integrated communication and sensing device, and to perform matched filtering or correlation processing on the received signal at the receiving end to obtain the measured channel impulse response. The measured channel impulse response includes at least multipath delay, amplitude and phase information, and the measured channel impulse response is transmitted as an output to module 4 and module 5.

[0164] The ray tracing module 3 is used to calculate the ray tracing simulation channel response and output the propagation path set under the scene environment geometric model constructed by the initialization module 1 and the initial electromagnetic parameters and scattering model parameters of the environment, and transmit the propagation path set as output to module 4 and module 5.

[0165] The propagation parameter calibration module 4 uses the measured channel impulse response output by the impulse response module 2 and the propagation path set output by the impulse response module 3. Using the path set submodule 41, it constructs the measured multipath observation set and the ray tracing path set, respectively, and generates a candidate matching pair set based on the time delay threshold and the angular domain threshold. The matching cost submodule 42 constructs a matching cost for the candidate matching pairs, including the time delay difference, angular domain difference, and amplitude difference, and uses optimal assignment to determine the final matching pair set to obtain a one-to-one matching relationship. The parameter inversion submodule 43 uses an iterative optimization algorithm to solve for the environmental propagation calibration parameters and update the environmental electromagnetic parameters in the ray tracing model, transmitting the updated environmental electromagnetic parameters as parameters to module 5.

[0166] The scattering parameter calibration module 5 uses the measured channel impulse response output by the impulse response module 2 and the calibrated environmental electromagnetic parameters output by the propagation parameter calibration module 4. Using the residual energy submodule 51, it obtains the calibrated scattering path information through ray tracing using the calibrated environmental electromagnetic parameters, and removes energy contributions matching specular reflection, transmission, and diffraction paths from the measured multipath observation set to obtain the scattering-related residual energy. The simulated scattering energy submodule 52 calculates the simulated scattering energy based on the scattering parameters in the ray tracing model and decomposes the simulated scattering energy by object or region to obtain the scattering energy components of each spatial partition. The scattering parameter iteration submodule 53 iteratively updates the scattering parameters based on the difference between the residual energy and the simulated scattering energy, projects the updated scattering parameters onto the physically feasible region, and obtains the updated scattering parameters. The updated environmental electromagnetic parameters and scattering parameters are then transmitted as output to module 6.

[0167] The imaging module 6 is used to perform weighted accumulation and statistics on the energy contribution of each spatial direction under the single-antenna condition based on propagation effects such as reflection, transmission, scattering and diffraction, under the updated environmental electromagnetic parameters and corrected scattering parameters of the propagation parameter calibration module 4 and the scattering parameter calibration module 5, to obtain the spatial angular domain energy distribution, and to map the angular domain energy distribution to a preset spherical grid to form angular domain energy imaging results and confidence distribution map for output.

[0168] It should be noted that the above functional modules can be implemented, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, they can be implemented, in whole or in part, as program instruction products. A program instruction product includes one or a set of program instructions. When the program instructions are loaded and executed on a computer, the described process or function is generated, in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The program instructions can be stored in a computer-readable and writable storage medium, or transferred from one computer's readable and writable storage medium to another.

[0169] In this embodiment, the direct coupling or communication connection between the modules can be achieved through indirect coupling or communication connection via interfaces, devices, or modules. The functional modules and sub-modules in this embodiment can dynamically reside within a single processing unit, or each module can exist physically independently, or two or more modules can dynamically reside within a single processing unit. When these dynamic components are implemented as software functional modules and sold or used as independent products, they can also be stored in a computer-readable and writable storage medium. This storage medium can be a memory, disk, or optical disc, etc.

[0170] The effects of this invention can be further illustrated by the following simulation results:

[0171] I. Simulation Conditions

[0172] In the simulation implementation, to simulate the actual measurement process, an equivalent measured channel impulse response is first generated under a high-fidelity ray tracing propagation model; then, a ray tracing model is established with the environmental electromagnetic parameters and scattering parameters initialized with the deviation, and the propagation parameters and scattering parameters are inverted and calibrated based on the equivalent measured channel impulse response; finally, angular domain energy imaging is performed on the calibrated model.

[0173] The environment model used in the simulation is as follows: Figure 3 As shown.

[0174] Using ray tracing programs Figure 3 The parameters of the ISAC device are simulated in the environment shown, and the channel impulse response (CIR) is obtained to simulate the CIR measurement data of the real ISAC device.

[0175] II. Simulation Content

[0176] Simulation 1: Under the above simulation conditions, using steps 3.1) to 3.6) of this invention, under the matching relationship constraints, the electromagnetic parameters related to the medium are solved by parameter inversion, and the propagation parameters in the ray tracing model are updated. Then, following steps 4.1) to 4.5) of this invention, the scattering parameters are iteratively updated based on the difference between the scattering residual energy and the simulated scattering energy. The updated scattering parameters are then projected onto the physically feasible region to obtain the calibrated scattering parameters. The results are as follows: Figure 4 , where 4(a) is a schematic diagram of ray tracing multipath distribution, Figure 4 (b) is a schematic diagram of the output channel response for ray tracing.

[0177] Depend on Figure 4 (b) As can be seen, under the optimized simulation conditions, the difference between the channel impulse response CIR B obtained by ray tracing and the true channel impulse response CIR A is small, indicating that the ray tracing model's fit to multipath delay, amplitude, and phase characteristics is improved after parameter optimization. Furthermore, through the propagation parameter inversion and scattering parameter calibration process described in this invention, the electromagnetic parameters and scattering parameters of the medium in the scene environment can be effectively corrected, thereby accurately restoring the environmental material properties and scattering characteristics.

[0178] Simulation 2: Under the above simulation conditions, following steps 5) and 6) of this invention, angular domain energy imaging processing is performed on the indoor scene using the true value simulation parameters, the unoptimized initial simulation parameters, and the calibrated simulation parameters, respectively. The results are as follows: Figure 5 .in, Figure 5 (a) is the angular domain energy imaging result generated by the true value simulation parameters, 5(b) is the angular domain energy imaging result generated by the unoptimized initial simulation parameters, and 5(c) is the angular domain energy imaging result generated by the optimized parameters.

[0179] Depend on Figure 5 It is evident that when using uncalibrated initial simulation parameters to perform angular domain energy imaging of an indoor scene, the location and shape of the energy concentration area are significantly different from the true value imaging results. When imaging is performed using the calibrated and optimized parameters described in this invention, the angular domain energy imaging results are more consistent with the true value imaging results in terms of the main energy concentration direction and energy distribution shape, thus verifying the effectiveness of the present invention in jointly calibrating propagation and scattering parameters based on equivalent measured channel impulse response.

[0180] In summary, this embodiment calibrates the ray tracking propagation parameters and scattering parameters by measuring the equivalent channel impulse response under single-antenna conditions. Then, it performs angular domain energy statistics and spherical mesh mapping on the calibrated model, which can output angular domain energy imaging and realize the visualization and reliability indication of the energy contribution in the spatial direction of the scene.

Claims

1. A method of ISAC angular-energy imaging based on measured channel response calibration of ray tracing, characterized in that, The method comprises the following steps: (1) extracting the geometric feature data of buildings, ground and scatterers in the target area from public map data or laser point cloud data, constructing an ISAC scene environment geometric model, and setting initial electromagnetic parameters and scatter model parameters of the environment; (2) transmitting a known probe signal in the scene environment by using a communication and perception integrated device, and performing matched filtering and correlation processing on the received signal at the receiving end to obtain a measured channel impulse response containing multipath time delay, amplitude and phase information; (3) path matching the measured channel impulse response with the ray tracing simulation channel response, constructing a calibration target function and calculating the environment propagation calibration parameters by a parameter inversion algorithm to update the environment electromagnetic parameters in the ray tracing model; (4) based on the energy distribution characteristics of the non-specular multipath components in the measured channel impulse response, the scattering parameters of different objects and regions in the scene environment are inverted and corrected; (5) inputting the updated environment electromagnetic parameters and the corrected scattering coefficients into the ray tracing model, and based on the reflection, transmission, scattering and diffraction propagation effects, the energy contribution of each spatial direction under single antenna condition is counted to obtain the spatial angular energy distribution and confidence; (6) mapping the angular energy distribution to a preset spherical grid to form an angular energy imaging result.

2. The method of claim 1, wherein, In the step (1), the ISAC scene environment geometric model is constructed, and the initial electromagnetic parameters and the scatter model parameters of the environment are set, and the implementation comprises: (1a) converting the extracted geometric feature data of buildings, ground and scatterers in the target area to the same spatial coordinate system, and completing the scale unification and registration to ensure the spatial alignment between different data sources; (1b) instantiating the building geometric feature data as a building geometric body, instantiating the ground geometric feature data as a ground geometric body, and instantiating the scatterer geometric feature data as a scatterer geometric body, wherein the building geometric body at least includes an outer facade and a roof, the ground geometric body at least includes a ground patch, and the scatterer geometric body is a set of parameterized geometric bodies available for ray propagation, and each geometric body is uniformly represented as a boundary representation and a polygon mesh representation composed of a vertex set and a patch set; (1c) generating wall patches and roof patches according to the plane profile and height information of the building to form a closed boundary structure on the building surface, and generating a continuous ground surface according to the ground boundary and the terrain height field, and splicing and cutting with the building bottom boundary when necessary to avoid overlapping and penetration; (1d) constructing the scatterer into a geometric body that can be calculated by ray tracing according to the spatial position, size range and main direction of the scatterer to reduce the number of patches while ensuring the expressiveness of the scattering characteristics; (1e) triangulating the geometric boundaries of buildings, ground and scatterers with the fitting surface to generate a triangular mesh containing vertex coordinates, triangular patch indexes and normal vectors; (1f) attaching object IDs and material IDs to different categories of patches to form an ISAC scene environment geometry model that the ray tracing engine can perform ray-triangle intersection, reflection, diffraction, and scattering direction calculation ; (1g) the scene environment geometry model is constructed , representing a collection of multiple classes of surfaces and objects ; (1h) Set of settings The initial electromagnetic parameters of the environment and the scattering model parameters in the set of material parameters : each class of surfaces according to the material electromagnetic parameters given in the ITU related recommendations relative permittivity equivalent conductivity ; According to the original laser point cloud data, the surface roughness of the scattering body is estimated, and the equivalent roughness of each type of surface is obtained Finally, a set of material parameters is obtained :​ 。 3. The method of claim 1, wherein, In the step (2), the received signal is subjected to matched filtering and correlation processing at the receiving end, and the implementation comprises: (2a) receiving a signal corresponding to the known probe signal , performing matched filtering with the impulse response to obtain a measured channel impulse response matched filtering output : ， wherein is complex conjugate, is the integral variable, is the slip. (2b) to sampling to obtain a discrete channel impulse response : ， wherein is the number of resolvable multipaths, , are the amplitude and phase of the th multipath, respectively, is the delay index, is the noise term; (2c) extracting the time delay, amplitude and phase information of the distinguishable multipath components from the discrete channel impulse response , the direction of arrival unit vector can be obtained by beam scanning and angle spectrum estimation at the receiving end, and the real-time multipath observation set is constructed: ， wherein is the first propagation delay of the multipath.

4. The method of claim 1, wherein, In the step (3), the measured channel impulse response is path matched with the ray tracing simulation channel response, and the implementation comprises: (3a) Using the ISAC scenario environment geometry model and a ray tracing program, generate a set of ray tracing paths containing time delays and arrival unit direction information : ， wherein, is the number of effective multipaths, is the propagation delay of the th multipath, , are the amplitude and phase of the th multipath, respectively, is the direction of arrival unit vector; (3b) setting a time delay threshold with angular domain threshold ; (3c) constructing a set of matching pairs constraints from the set of observed multipath observations and the set of ray-traced paths , constructing a set of matching pairs constraints : ， wherein, is the ray-traced first path delay, is the measured first multipath delay, , are the ray-traced and measured path directions of arrival, respectively; (3d) determining a set of candidate matching pairs based on the set of candidate matching pairs constructing a matching cost or matching probability, and determining a final set of matching pairs using optimal assignment to obtain a matching path corresponding to each path, and unassigned measured multipath and ray-traced paths as residual paths for subsequent scattering parameter inversion.

5. The method of claim 1, wherein, The (3) includes constructing a calibration target function and calculating the environment propagation calibration parameter through a parameter inversion algorithm, and the implementation includes: (3e) constructing a set of global parameters to be calibrated depending on the set of material parameters to be optimized :​ ， wherein , respectively represent the uniform reflection loss scaling factor and the transmission loss scaling factor; (3f) based on the final matching pair set and global parameters , a calibration target function is constructed using a regularized least squares criterion : ， wherein, , are the path amplitudes and phases calculated by ray tracing under the parameters , are the path amplitudes and phases calculated by ray tracing under the parameters are the initial parameters; are the regularization weights to avoid overfitting, are the matching pair weights; (3g) obtaining optimal calibration parameters by solving a minimization problem of a calibration target function :​ ， wherein as the the physical feasible range of electromagnetic parameters of the materials; (3h) updating the environment propagation calibration parameters in the ray tracing model in dependence on the optimal calibration parameters .

6. The method of claim 1, wherein, The (4) includes correcting the scattering parameters of different objects and regions in the scene environment, and the implementation includes: (4a) according to the final matching pair set , the measured multipath observation set , the matching pair set corresponding to the specular reflection path and the diffraction path , the measured scattering residual energy is obtained by eliminating the energy of the matching pair set from the total measured energy : ， (4b) defining a set of scattering parameters as a set of scattering coefficients and a set of scattering direction lobe parameters : (4b1 ) Defining a set of scattering coefficients according to the scattering coefficient : ， wherein, is the number of material classes, is the number of spatial partitions, is the corresponding scattering coefficient, is the total number of material classes, is the total number of spatial partitions; (4b2) defining a set of scattering direction lobe parameters based on the scattering lobe intensity parameter , the main lobe of the scattering lobe is pointing , the scattering lobe sharpness parameter , defining a set of scattering direction lobe parameters : ， wherein, is a first class of materials, spatial partitions a corresponding scattering lobe intensity parameter, for adjusting the overall amplitude of the scattering directional distribution; is a first class of materials, spatial partitions a corresponding scattering lobe main lobe pointing, for characterizing the dominant exit direction of the scattered energy; is a first class of materials, spatial partitions a corresponding scattering lobe sharpness parameter, for controlling the scattering lobe width; (4c) calculating simulated scattered energy produced by the scattering path from the scattering parameters : ， wherein, is a set of scattering paths obtained by ray tracing, is a set of scattering paths obtained by ray tracing, is a set of scattering paths obtained by ray tracing, is a set of scattering paths obtained by ray tracing, is an incident direction, is a scattering exit direction; is a scattering angle distribution function, used to characterize the distribution of scattered energy from the incident direction to the exit direction under the condition that the scattering energy can be parameterized by a directional scattering lobe model. (4d) said simulated scattered energy decomposition by surface class and spatial partitioning to obtain a simulated scattered energy component corresponding to each cell : ， The formula shall satisfy: ; (4e) according to each of the simulated scattered energy components of the cells constructing residual allocation weights : ， wherein is a constant to prevent the denominator from being zero; (4f) residual allocation weights and measured scattered residual energy resulting in measured scattered residual energy components ; (4g) Iteratively updating the scattering coefficients based on the difference between the measured scattering energy component of each cell and the simulated scattering energy component :​ ， In the formula, denotes the projection onto the physically feasible interval, is the step size; (4h) further updating the scattering direction lobe parameter update: ， wherein, is the direction lobe parameter update step size; (4i) set simulated error limit of measured scattered energy component to simulated scattered energy component , maximum number of iterations of algorithm , determine whether the correction of the scattering parameters is finished: If the difference between the measured scattering energy component and the simulated scattering energy component is less than the simulation error limit or the number of algorithm iterations reaches , then the scattering parameter correction is terminated. Otherwise, return to step (4g).

7. The method of claim 1, wherein, The (5) includes statistically analyzing the energy contribution of each spatial direction under the single antenna condition to obtain the spatial angular energy distribution and the confidence, and the implementation includes: (5a) the set of effective propagation paths obtained according to ray tracing , the arrival direction of each path and the corresponding path complex amplitude ; (5b) Gain compensate the path energy according to the antenna pattern, and set the weight for matching the confidence weighting and path type weighting ; (5c) attributing and weighting cumulating the path energies to the angular domain elements according to their arrival directions to obtain an angular energy distribution : ， wherein is an angular domain element, is an indicator function; (5d) normalizing the angular energy distribution performing a normalization scale correction to obtain a normalized angular energy : ； (5e) Weighting the confidence according to the match , accumulating the weights within the angular bins and outputting the corresponding angular energy confidence distribution : 。 8. The method of claim 1, wherein, The (6) includes mapping the angular energy distribution to the preset spherical grid to form the angular energy imaging result, and the implementation includes: (6a) dividing the angular domain into spherical grid cells , defining initial angular energy pixels according to the angular energy distribution ; (6b) boosting initial angular energy pixels with spatial continuity, replacing the indicator function with an angular kernel function , obtaining the mapped angular energy pixels after neighborhood diffusion : , wherein, a direction unit vector of the first effective propagation path, a direction unit vector of the center of the grid cell, a kernel function for describing the contribution of the path to the neighboring angular domain pixels; (6c) setting energy weight coefficient , calculating the maximum value in all spherical grid cell energy pixels , selecting high-energy grid cells satisfying in the angular energy pixel mapping to carry out local subdivision, forming an adaptive spherical grid, to improve the angular resolution of the target direction of the angular energy pixel , and finally obtaining the angular energy imaging result .

9. An ISAC angular-energy imaging system that calibrates ray tracing based on a measured channel response, characterized by, The initialization module is configured to extract geometric feature data of buildings, ground, and scatterers in a target region from public map data or laser point cloud data, construct an ISAC scene environment geometric model, and set initial electromagnetic parameters and scattering model parameters of the environment. The impulse response module is configured to transmit a known probe signal in the scene environment by using a communication and perception integrated device, and perform matched filtering or correlation processing on a received signal at a receiving end to obtain a measured channel impulse response, wherein the measured channel impulse response at least includes multipath time delay, amplitude, and phase information. The ray tracing module is configured to calculate a ray tracing simulation channel response and output a propagation path set under the scene environment geometric model and the initial electromagnetic parameters and scattering model parameters of the environment. The propagation parameter calibration module is configured to perform path matching on the measured channel impulse response and the ray tracing simulation channel response, construct a calibration target function, and solve the environment propagation calibration parameter by using a parameter inversion algorithm to update the electromagnetic parameters of the environment in the ray tracing model. The scattering parameter calibration module is configured to inversely analyze and correct the scattering parameters of different objects or regions in the scene environment based on the energy distribution characteristics of the non-specular multipath components in the measured channel impulse response and in combination with the scattering path information obtained by ray tracing, wherein the scattering parameters at least include scattering coefficients set according to categories and spatial partitions. The imaging module is configured to, under the updated electromagnetic parameters of the environment and the corrected scattering parameters, perform weighted accumulation and statistical analysis on the energy contribution of each spatial direction under the single antenna condition based on reflection, transmission, scattering, and diffraction propagation effects, obtain a spatial angular energy distribution, and map the angular energy distribution to a preset spherical grid to form an angular energy imaging result.

10. The system of claim 9, wherein: The propagation parameter calibration module includes: A path set sub-module configured to construct a measured multipath observation set and a ray tracing path set, respectively, and generate a candidate matching pair set based on a time delay threshold and an angular domain threshold; A matching cost module configured to construct a matching cost including a time delay difference, an angular domain difference, and an amplitude difference for the candidate matching pair, and determine a final matching pair set by using an optimal assignment to obtain a one-to-one matching relationship; A calibration target function sub-module configured to construct a calibration target function including a regularization term based on the final matching by using a weighted least squares form; A parameter inversion sub-module configured to solve the environment propagation calibration parameter by using an iterative optimization algorithm and update the electromagnetic parameters of the environment in the ray tracing model; The scattering parameter calibration module includes: ​ a residual energy sub-module configured to eliminate energy contributions matching the specular reflection path, the transmission path and the diffraction path from the measured multipath observation set to obtain residual energy related to scattering; a simulated scattering energy sub-module configured to calculate simulated scattering energy based on the scattering parameters in a ray tracing model, and decompose the simulated scattering energy by objects or regions to obtain scattering energy components of each spatial partition; a scattering parameter iteration sub-module configured to iteratively update the scattering parameters according to a difference between the residual energy and the simulated scattering energy, and project the updated scattering parameters to a physically feasible interval.

Images