Sensitivity and inductance integrated physical environment reconstruction method

Abstract

The invention belongs to the technical field of wireless communication and positioning, and particularly relates to a communication and inductance integrated physical environment reconstruction method. Electromagnetic wave propagation characteristics are used as information carriers, and multi-dimensional information such as environmental physical parameters is captured in real time through a multi-mode sensing means; key environment features are extracted through a multi-source information fusion algorithm, and a parameterized digital twinborn model is constructed; through electromagnetic wave propagation characteristic calculation, evaluation and adjustment, an optimized ultra-fast electric field calculation method is adopted, an environment-channel association database is combined, electromagnetic propagation behaviors are deduced in real time in a reconstruction scene, and path parameters are output; a sensing-prediction-behavior closed-loop autonomous system is formed, a prediction result is converted into an optimization strategy, sensing and prediction parameters are reversely adjusted, and global and local two-stage closed-loop feedback is achieved. According to the invention, through OFDM waveform multiplexing, multi-mode perception fusion and carrier phase ranging, the problems of large ranging error, large channel prediction deviation and high deployment cost are solved.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication and positioning technology, and specifically relates to a method for reconstructing a physical environment that integrates sensing and communication. Background Technology

[0002] Wireless communication and positioning technologies mainly include traditional time-of-arrival (TOA) positioning and sensing technologies and positioning and sensing technologies based on classical channel models. Among them, traditional TOA-based positioning and sensing technologies are limited by multipath interference and non-line-of-sight propagation, resulting in generally low ranging accuracy both indoors and outdoors, making it difficult to meet the application requirements of high-precision and complex scenarios such as autonomous driving and industrial robots. On the other hand, classical channel models rely only on statistical parameters and lack explicit modeling of the geometry, material, and motion state of the real scattering bodies. This leads to high prediction errors in high-dynamic scenarios, resulting in beam mismatch, link interruption, and resource waste.

[0003] The shortcomings of existing technologies are summarized as follows: (1) Low ranging accuracy: Due to multipath interference and non-line-of-sight propagation, the first path signal in complex environments is often submerged by strong reflection paths, resulting in a time delay estimation deviation of several sampling periods. The ranging accuracy is low and it is difficult to meet the requirements of high dynamic and high security scenarios.

[0004] (2) Large channel prediction error: The lack of explicit modeling of the scatterer geometry, material dielectric properties and real-time motion trajectory leads to large channel prediction error, resulting in beam prepointing failure, frequent handover interruption and inefficient resource scheduling.

[0005] (3) Difficulty in large-scale deployment: It requires dedicated sensing signals or additional hardware support, resulting in high deployment costs and limited coverage, making it difficult to scale up in existing cellular networks. Summary of the Invention

[0006] To address the problems existing in the prior art, this invention provides a method for reconstructing a physical environment that integrates sensing and perception.

[0007] The technical solution adopted by this invention to solve its technical problem is as follows: A method for reconstructing a physical environment integrating sensing and perception, using the propagation characteristics of electromagnetic waves as an information carrier, includes the following steps: S1. Environmental perception: Real-time capture of multi-dimensional information on environmental physical parameters, obstacles, and moving scatterers through multi-modal sensing. S2. Virtual physical environment construction: Key environmental features are extracted through multi-source information fusion algorithms to construct a parameterized digital twin model, so that the shape, position, material and motion state of each object in the virtual physical environment are mapped with high precision to the real world. S3. Electromagnetic wave propagation prediction: Through the calculation, evaluation and adjustment of electromagnetic wave propagation characteristics, an optimized ultrafast electric field calculation method is adopted, combined with an environment-channel correlation database, to infer electromagnetic propagation behavior in real time in the reconstructed scenario and output path parameters, including but not limited to propagation direction, amplitude, time delay and phase. S4. Form a closed-loop autonomous system of perception-prediction-behavior, transform the prediction results into optimization strategies, and adjust the perception and prediction parameters in reverse to achieve a two-level closed-loop feedback of global and local levels.

[0008] Preferably, the physical environment reconstruction method is implemented based on a wireless environment prediction and reconstruction system, which adopts a five-layer architecture: infrastructure layer, resource layer, network function layer, application layer, and prediction plane spanning all layers; the prediction plane embeds an environment perception module, an electromagnetic wave propagation prediction module, and a behavior module, forming a two-level closed-loop feedback.

[0009] Preferably, the first-level closed-loop feedback runs through the entire physical environment: the infrastructure and resource layer provides raw perception samples, the perception module extracts knowledge and reconstructs the virtual environment, the prediction module calculates and evaluates electromagnetic propagation characteristics, the behavior module generates optimization strategies, and feeds them back to the network function layer and application layer to achieve global adaptation; The second-level closed-loop feedback exists between adjacent modules: the perception module and the prediction module achieve rapid fine-tuning through model refresh signals, and the prediction module and the behavior module achieve immediate response and local fine-tuning through demand-policy conversion.

[0010] Preferably, the electromagnetic wave propagation prediction module employs a ranging method based on orthogonal frequency division multiplexing (OFDM) time delay, with the following steps: S301, symbol coarse synchronization and pilot detection, adopts a delay correlation algorithm based on cyclic prefix, directly using the peak position to represent the integer sample level arrival time of the direct path; after coarse synchronization, the time domain signal is converted into a frequency domain subcarrier sequence, and equally spaced comb-shaped pilot sequences are extracted in the frequency domain to provide input for subsequent channel state information acquisition; S302, Multipath Acquisition and First Path Extraction: Using a multi-frequency domain multipath acquisition method, effective multipath information is extracted from the pilot sequence; first path information is extracted from the effective multipath to provide initial values ​​for subsequent fine synchronization; S303, multipath tracking and coarse symbol synchronization, through differential channel power balance delay estimation algorithm, corrects the first path position to obtain high-precision delay estimation; and combined with carrier phase ranging algorithm, projects the received pilot signal onto the ideal first path reference signal generated based on high-precision delay estimation, and extracts the phase of the projection coefficient as the carrier phase observation value.

[0011] Preferably, in step S302, a small range of sampling points are selected to form a capture window, centered on the time delay estimate obtained from coarse synchronization. For each candidate delay within the acquisition window, a frequency domain phase rotation vector is constructed, the projected power of the received pilot on this vector is calculated, all candidate delays are traversed, and the paths with the strongest power are selected as effective multipaths. The effective multipaths obtained in the multipath acquisition stage are analyzed, the earliest arriving or the strongest path is selected as the initial estimate of the first path, and the initial position of the first path is output to provide accurate initial values ​​for subsequent fine synchronization.

[0012] Preferably, in step S303, based on the initial first path estimate, the frequency domain channel estimate is converted into a time domain power delay spectrum using IFFT; A differential channel power balance delay estimation algorithm is used to correct the first path position and obtain the accurate arrival time.

[0013] Preferably, the parametric digital twin model automatically extracts the position, size, orientation, and material properties of the scatterer through semantic segmentation, instance segmentation, and 3D reconstruction algorithms, thereby achieving millisecond-level environment capture and automatic digital twin reconstruction of complex dynamic scenes.

[0014] Preferably, the optimized electromagnetic propagation calculation method combines a parallel ray tracing algorithm with an FDTD acceleration engine, while maintaining an environment-channel correlation database to support rapid mapping between empirical models and full-wave simulation results to improve computational efficiency.

[0015] Preferably, it also includes data processing and visualization steps: generating and dynamically updating indicators including channel impulse response, power delay spectrum, angular power spectrum and Doppler power spectrum, and intuitively presenting the panoramic view of electromagnetic propagation through a high-performance rendering engine.

[0016] Preferably, multimodal sensing methods include drone aerial photography, camera vision, millimeter-wave radar, and traditional electromagnetic measurement.

[0017] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention, based on OFDM communication waveforms, utilizes multi-frequency domain multipath acquisition, differential power-balanced precision synchronization, and carrier phase projection ranging to exhibit superior robustness against strong multipath and noise without adding RF hardware. Based on a bottom-up environment reconstruction technology enhanced by multimodal perception, it integrates visual, millimeter-wave radar, and electromagnetic measurement data to achieve channel prediction and reduce errors. Together, these technologies construct a complete path from perception to cognition, enabling the reconstruction and simulation of the physical environment.

[0018] 2. This invention proposes a high-precision time delay estimation ranging method based on OFDM: candidate first path is locked by multi-frequency domain multipath acquisition, time delay estimation is achieved by combining differential power balancing, and phase observation is extracted by carrier phase projection.

[0019] 3. This invention proposes a bottom-up physical environment perception reconstruction method based on enhanced perception: by capturing information such as environmental physical parameters in real time through multimodal perception, a virtual physical environment is constructed, and the propagation characteristics of electromagnetic waves are predicted and intelligently optimized to form a perception-prediction-behavior model.

[0020] 4. This invention achieves simple hardware deployment by reusing the OFDM waveform of the 5G NR standard. It can be applied to existing cellular networks without the need for dedicated signals, resulting in low deployment costs and meeting the application requirements of highly dynamic and complex scenarios. This invention integrates visual, millimeter-wave radar, and electromagnetic measurement data, introducing multimodal sensing dynamic digital modeling to overcome the limitations of traditional models and reduce channel prediction errors.

[0021] In summary, this invention solves the problems of large ranging error, large channel prediction deviation, and high deployment cost by reusing OFDM waveforms, multimodal sensing fusion, and carrier phase ranging. Attached Figure Description

[0022] Figure 1 This is a flowchart of a high-precision time delay estimation algorithm.

[0023] Figure 2 This is a schematic diagram of a perception-enhanced wireless environment prediction and reconstruction platform. Detailed Implementation

[0024] To facilitate understanding of the present invention, it will be described in more detail below with reference to the accompanying drawings and specific embodiments. However, the present invention can be implemented in many different forms and is not limited to the embodiments described in this specification. Rather, these embodiments are provided to provide a more thorough and complete understanding of the disclosure of the present invention.

[0025] Example 1: A sensor-integrated physical environment reconstruction method. Sensor-integration endows cellular networks with a "sixth sense," enabling centimeter-level positioning, speed perception, and behavior recognition of users and objects. Physical environment reconstruction, through high-fidelity environment modeling, overcomes the limitations of traditional statistical channel models, supporting advanced channel prediction, beam pre-tracking, and precise interference avoidance. The deep integration of these two methods provides deterministic guarantees for key 6G scenarios such as ultra-reliable low-latency communication, vehicle-to-everything (V2X) collaborative sensing, and high-precision control in the industrial internet, demonstrating significant value and effectiveness. The specific solution is as follows: ① A high-precision time delay estimation and ranging method based on orthogonal frequency division multiplexing (OFDM) technology This solution, without altering the existing 5G / 6G cellular network RF front-end and antenna hardware, embeds a series of high-precision sensing algorithm modules into the baseband digital signal processing link. It reuses standard OFDM communication waveforms to achieve deep integration of communication and sensing, forming a comprehensive sensing capability covering the entire base station. The system simultaneously supports downlink data broadcasting and uplink sensing echo reception. The base station can insert dedicated or multiplexed pilot OFDM sensing symbols into regular communication frames, enabling real-time detection, tracking, and distance estimation of non-cooperative targets while completing user data interaction.

[0026] The OFDM sensing signal transmitted by the base station maintains a frame structure completely consistent with the communication signal, including a cyclic prefix (CP) to effectively combat multipath interference. After the signal is reflected by the target, it forms an echo. The receiver first performs matched filtering to improve the signal-to-noise ratio, then removes the cyclic prefix and performs a discrete Fourier transform to convert the time-domain signal into a frequency-domain subcarrier sequence. In the frequency domain, an equally spaced comb-shaped pilot sequence is extracted for subsequent high-precision time delay and phase estimation. Because the echo signal introduces a phase rotation caused by transmission delay and Doppler shift relative to the original transmitted signal, the received pilot sequence naturally carries target range and velocity information.

[0027] The data preprocessing stage first performs coarse synchronization using a delay correlation algorithm based on cyclic prefixes. ; in For CP length, The number of FFT points.

[0028] The peak position is the approximate starting sampling point of the OFDM symbol, which directly represents the integer sample level arrival time of the direct path. After coarse synchronization, the subcarrier frequency offset is estimated and compensated, followed by removal of the cyclic prefix. A zero-forcing receiver is then used to demodulate the time-domain signal using OFDM. The frequency-domain receive pilot vector is extracted based on the known pilot positions. This provides reliable input for subsequent acquisition of channel state information.

[0029] In the channel acquisition phase, the delay estimate obtained from coarse synchronization is used as the center, and small sampling points are extracted before and after to form an acquisition window, thereby improving computational efficiency and focusing on areas of concentrated energy. To address the severe multipath effect indoors, a multi-frequency domain multipath acquisition method is proposed, utilizing the received pilot subcarrier sequence for signal decomposition. This method fully leverages the time-shifting characteristics of the Fourier transform, transforming time-domain multipath delay estimation into a frequency-domain phase estimation problem. Specifically, for each candidate delay τl within the acquisition window, a corresponding frequency-domain phase rotation vector is constructed: ; Calculate the projected power of the received pilot on this vector: ; Iterate through all candidate delays and select the strongest paths as effective multipaths. Among these, the earliest arriving path or the path with the strongest power is used as the initial estimate of the first path. This method can reliably capture the position of the first path under strong multipath interference, providing accurate initial values ​​for subsequent fine synchronization.

[0030] After obtaining a relatively accurate first-path position, a differential channel power balance delay estimation algorithm is used to further refine the coarse estimation result. First, the frequency domain channel estimate is transformed into a time domain power delay spectrum using IFFT: Finally, the arrival time was obtained. .

[0031] After obtaining a relatively accurate Time of Arrival (TOA) during the channel tracking phase, to further improve accuracy, a carrier phase ranging algorithm is proposed. This algorithm projects the received pilot signal onto an ideal first-path reference signal generated based on high-precision time delay estimation. This ideal first-path reference signal is free from multipath interference, and the phase of the projection coefficients is extracted as the carrier phase observation. Compared to time delay estimation, carrier phase ranging theoretically improves accuracy by two orders of magnitude and exhibits extremely strong robustness to noise and residual multipath.

[0032] First, the local reference pilot XP[k] is pre-compensated using a precise TOA to generate the ideal first path: ; Will receive pilot YP projection: ; ; The final distance is estimated to be ; The flowchart of the entire high-precision time delay estimation algorithm is as follows: Figure 1 As shown.

[0033] ② Bottom-up Physical Environment Perception Reconstruction Method Based on Perception Enhancement This solution addresses the challenges of complex and highly dynamic 6G wireless environments, specifically the challenges posed by the large number, diverse types, and dramatic time-varying nature of scatterers. It proposes a bottom-up, perception-enhanced physical environment reconstruction technology. By employing multimodal sensing to capture environmental physical parameters, obstacle information, and moving scatterer information in real time, a high-fidelity virtual physical environment is constructed. Based on this, advanced prediction and intelligent optimization of electromagnetic wave propagation characteristics are achieved, ultimately forming a closed-loop autonomous system of perception, prediction, and behavior. This provides core support for channel prediction, intelligent transmission design, and network performance optimization.

[0034] The solution uses the propagation characteristics of electromagnetic waves as the information carrier, rather than the electromagnetic waves themselves, focusing on the environmental features carried during propagation to achieve deep integration between the physical environment and core services such as upper-level communication transmission and precise positioning. The system supports a hybrid sensing model, integrating multiple sensing resources such as UAV aerial photography, camera vision, millimeter-wave radar, and traditional electromagnetic measurement to simultaneously collect multi-dimensional samples of the environment's geometric structure (position, shape), material properties (dielectric constant, roughness), and motion trajectory. Key environmental features are extracted through multi-source information fusion algorithms to construct a parametric digital twin model, ensuring that the shape, position, material, and motion state of each object in the virtual physical environment maintain a high-precision mapping to the real world. Within this virtual environment, electromagnetic wave propagation characteristics, including multipath delay, amplitude, phase, angle, and Doppler distribution, can be rapidly generated at any time, space, and frequency point.

[0035] The core of the solution consists of three major functional modules: perception, prediction, and behavior. These three modules work together to drive the reconstruction and optimization of the physical environment. The perception module continuously collects multimodal data from the infrastructure and resource layers, constructing and dynamically updating an environmental knowledge base and digital twin model. The prediction module comprises three sub-modules: electromagnetic wave propagation characteristic calculation, evaluation, and adjustment. It employs optimized ultrafast electric field calculation methods (such as ray tracing and full-wave simulation acceleration algorithms) combined with an environment-channel correlation database to simulate electromagnetic propagation behavior in real time within the reconstructed scenario, outputting path parameters such as propagation direction, amplitude, delay, and phase. The behavior module transforms the prediction results into optimization strategies for the network functional layer and application layer, while simultaneously translating upper-layer service requirements into adjustment instructions for perception and prediction parameters, achieving bidirectional intelligent interaction.

[0036] The system adopts a five-layer architecture: infrastructure layer, resource layer, network function layer, application layer, and a prediction plane spanning all layers. The prediction plane embeds three modules: perception, prediction, and behavior, forming a two-level closed-loop feedback workflow. The first-level closed loop permeates the entire physical environment: the infrastructure and resource layers provide raw perception samples → the perception module extracts knowledge and reconstructs the virtual environment → the prediction module calculates and evaluates electromagnetic propagation characteristics → the behavior module generates optimization strategies → feedback is sent to the network function layer and application layer, achieving global adaptation. The second-level closed loop exists between adjacent modules: perception and prediction achieve rapid fine-tuning through model refresh signals, and prediction and behavior achieve instant response through demand-policy conversion. This two-level closed-loop structure significantly enhances the system's adaptability to environmental dynamics, allowing for parallel long-cycle global optimization and short-cycle local fine-tuning, ensuring both superior prediction accuracy and response speed.

[0037] The specific implementation is divided into three core modules, such as Figure 2 As shown: The environmental perception module introduces deep learning-assisted computer vision technology to drive multiple sensors to collect scene images and point cloud data in real time. Through semantic segmentation, instance segmentation and 3D reconstruction algorithms, it automatically extracts the position, size, orientation and material properties of scatterers, realizing millisecond-level environmental capture and automatic digital twin reconstruction of complex dynamic scenes.

[0038] The electromagnetic wave propagation prediction module, based on a reconstructed high-fidelity virtual environment and combined with an optimized parallel ray tracing and FDTD acceleration engine, rapidly calculates the reflection, refraction, diffraction, and scattering paths of electromagnetic waves, outputting accurate parameters such as propagation direction, amplitude, time delay, phase, and arrival / departure angles. Simultaneously, it maintains an environment-channel correlation database, supporting rapid mapping between empirical models and full-wave simulation results, further improving computational efficiency.

[0039] The data processing and visualization module is responsible for real-time prediction of channel fading in time-varying scenarios, generating and dynamically updating key indicators such as channel impulse response (CIR), power delay spectrum (PDP), angular power spectrum (PAS), and Doppler power spectrum (DPS). It uses a high-performance rendering engine to intuitively present the electromagnetic propagation panorama in the current environment, supporting real-time decision-making for network planning, beam management, and resource allocation.

[0040] This invention designs an integrated sensing and communication physical environment reconstruction method. Unlike traditional methods that rely on dedicated channel models, it solves the problem of low ranging accuracy caused by first-path flooding under strong multipath interference by embedding a time delay phase estimation algorithm into the existing cellular network baseband. Through environmental reconstruction, it addresses the issue of high prediction errors, providing assurance for physical environment reconstruction decisions in highly dynamic and complex scenarios. This invention achieves simple hardware deployment by reusing 5G NR standard OFDM waveforms, requiring no dedicated signals and applicable to existing cellular networks, resulting in low deployment costs and meeting the application requirements of highly dynamic and complex scenarios. Furthermore, this invention integrates visual, millimeter-wave radar, and electromagnetic measurement data, introducing multimodal sensing dynamic digital modeling to overcome the limitations of traditional models and reduce channel prediction errors.

[0041] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

Claims

1. A method for reconstructing a physical environment using integrated sensory perception, characterized in that, Using the propagation characteristics of electromagnetic waves as an information carrier includes the following steps: S1. Environmental perception: Real-time capture of multi-dimensional information on environmental physical parameters, obstacles, and moving scatterers through multi-modal sensing. S2. Virtual physical environment construction: Key environmental features are extracted through multi-source information fusion algorithms to construct a parameterized digital twin model, so that the shape, position, material and motion state of each object in the virtual physical environment are mapped with high precision to the real world. S3. Electromagnetic wave propagation prediction: Through the calculation, evaluation and adjustment of electromagnetic wave propagation characteristics, an optimized ultrafast electric field calculation method is adopted, combined with an environment-channel correlation database, to infer electromagnetic propagation behavior in real time in the reconstructed scenario and output path parameters, including but not limited to propagation direction, amplitude, time delay and phase. S4. Form a closed-loop autonomous system of perception-prediction-behavior, transform the prediction results into optimization strategies, and adjust the perception and prediction parameters in reverse to achieve a two-level closed-loop feedback of global and local levels.

2. The sensory integration physical environment reconstruction method according to claim 1, characterized in that, The physical environment reconstruction method is based on a wireless environment prediction and reconstruction system. The system adopts a five-layer architecture: infrastructure layer, resource layer, network function layer, application layer, and prediction plane that spans all layers. The prediction plane embeds an environment perception module, an electromagnetic wave propagation prediction module, and a behavior module, forming a two-level closed-loop feedback.

3. The sensory integration physical environment reconstruction method according to claim 2, characterized in that, The first-level closed-loop feedback runs through the entire physical environment: the infrastructure and resource layer provides raw perception samples, the perception module extracts knowledge and reconstructs the virtual environment, the prediction module calculates and evaluates electromagnetic propagation characteristics, the behavior module generates optimization strategies, and feeds them back to the network function layer and application layer to achieve global adaptation. The second-level closed-loop feedback exists between adjacent modules: the perception module and the prediction module achieve rapid fine-tuning through model refresh signals, and the prediction module and the behavior module achieve immediate response and local fine-tuning through demand-policy conversion.

4. The sensory integration physical environment reconstruction method according to claim 3, characterized in that, In the electromagnetic wave propagation prediction module, a ranging method based on orthogonal frequency division multiplexing (OFDM) time delay is adopted, and the steps are as follows: S301, symbol coarse synchronization and pilot detection, adopts a delay correlation algorithm based on cyclic prefix, directly using the peak position to represent the integer sample level arrival time of the direct path; after coarse synchronization, the time domain signal is converted into a frequency domain subcarrier sequence, and equally spaced comb-shaped pilot sequences are extracted in the frequency domain to provide input for subsequent channel state information acquisition; S302, Multipath Acquisition and First Path Extraction: Using a multi-frequency domain multipath acquisition method, effective multipath information is extracted from the pilot sequence; first path information is extracted from the effective multipath to provide initial values ​​for subsequent fine synchronization; S303, multipath tracking and coarse symbol synchronization, through differential channel power balance delay estimation algorithm, corrects the first path position to obtain high-precision delay estimation; and combined with carrier phase ranging algorithm, projects the received pilot signal onto the ideal first path reference signal generated based on high-precision delay estimation, and extracts the phase of the projection coefficient as the carrier phase observation value.

5. The sensory integration physical environment reconstruction method according to claim 4, characterized in that, In step S302, a small range of sampling points are selected to form a capture window, centered on the time delay estimate obtained from coarse synchronization. For each candidate delay within the acquisition window, a frequency domain phase rotation vector is constructed, the projected power of the received pilot on this vector is calculated, all candidate delays are traversed, and the paths with the strongest power are selected as effective multipaths. The effective multipaths obtained in the multipath acquisition stage are analyzed, the earliest arriving or the strongest path is selected as the initial estimate of the first path, and the initial position of the first path is output to provide accurate initial values ​​for subsequent fine synchronization.

6. The method for reconstructing a synesthetic physical environment according to claim 4, characterized in that, In step S303, based on the initial first path estimation, the frequency domain channel estimation is transformed into the time domain power delay spectrum using IFFT; A differential channel power balance delay estimation algorithm is used to correct the first path position and obtain the accurate arrival time.

7. The method for reconstructing a synesthetic physical environment according to claim 1, characterized in that, Parametric digital twin models automatically extract the position, size, orientation, and material properties of scatterers through semantic segmentation, instance segmentation, and 3D reconstruction algorithms, enabling millisecond-level environment capture and automatic digital twin reconstruction of complex dynamic scenes.

8. The method for reconstructing a sensory-integrated physical environment according to claim 1, characterized in that, The optimized electromagnetic propagation calculation method combines a parallel ray tracing algorithm with an FDTD acceleration engine, while maintaining an environment-channel correlation database to support rapid mapping between empirical models and full-wave simulation results to improve computational efficiency.

9. The method for reconstructing a synesthetic physical environment according to claim 1, characterized in that, It also includes data processing and visualization steps: generating and dynamically updating indicators including channel impulse response, power delay spectrum, angular power spectrum and Doppler power spectrum, and intuitively presenting the panoramic view of electromagnetic propagation through a high-performance rendering engine.

10. The method for reconstructing a synesthetic physical environment according to claim 1, characterized in that, Multimodal sensing methods include drone aerial photography, camera vision, millimeter-wave radar, and traditional electromagnetic measurement.

Images