A method, system, device and storage medium for modeling a communication channel

Abstract

The application relates to a communication channel modeling method, system, device and storage medium, and particularly relates to a communication channel modeling method, which comprises the following steps: setting at least one transmitting end and at least one receiving end, the transmitting end is used to generate a signal, and the receiving end is used to receive the signal; generating a 'tracking light ray' from the receiving end, the 'tracking light ray' forms a 'light ray tracking path' after interacting with an object in a communication scene; recording the propagation information of the 'light ray tracking path'; and calculating a channel impulse response based on the propagation information to obtain the communication channel characteristics. Based on the above steps, the application introduces a light ray tracking operator into the communication channel modeling, so that the communication channel modeling can rely on existing GPU accelerated rendering technology and light ray tracking algorithm and other related technologies to complete a large amount of light ray tracking within a millisecond level time, and the efficiency and precision of the communication channel modeling are improved.

Description

Technical Field

[0001] This invention relates to the field of communication technology, and in particular to a communication channel modeling method, system, device, and storage medium. Background Technology

[0002] With the rapid development of mobile communication and information applications, communication systems are placing increasingly higher demands on the accuracy and efficiency of channel models. In the design and optimization of optical wireless communication systems, channel modeling is a fundamental step, and its results directly determine link budget, modulation design, bit error rate performance prediction, and system deployment scheme.

[0003] Currently, common channel modeling methods include: Deterministic methods: such as recursive algorithms, the Ceiling Bounce Model (CBM), and ray tracing methods in commercial optical simulation software like Zemax. These methods typically achieve high accuracy, but their computational complexity is high and simulation efficiency is low, especially in complex multipath scenarios where they are too time-consuming to support rapid modeling. Stochastic methods: such as the Geometric Stochastic Model (GBSM) and Monte Carlo simulation methods. These methods approximate channel characteristics through statistical means and have high computational efficiency, but they often lack accurate characterization of complex reflections, multipath effects, and non-line-of-sight links, easily leading to deviations between the modeling results and the real environment.

[0004] The above analysis reveals several common technical problems with existing channel modeling methods: First, they suffer from low computational efficiency while maintaining computational accuracy, failing to simultaneously balance efficiency and accuracy, and resulting in excessive time consumption. Second, most existing methods only consider light intensity distribution, neglecting the emission spectrum of the light source and the wavelength dependence of materials, leading to insufficient spectral dependence. Third, researchers often need to repeatedly implement light source definitions, scene modeling, and sampling methods based on existing methods, resulting in redundant development from light source definition to scene modeling. The lack of unified verification standards and open platforms leads to poor comparability and reproducibility of research results, exhibiting insufficient scalability and reproducibility. Fourth, in complex scenarios, NLoS (non-line-of-sight path) contributes significantly, but existing methods either ignore or underestimate NLoS in complex scenarios, exhibiting limited ability to capture higher-order reflection paths, resulting in significant modeling bias and insufficient characterization of non-line-of-sight paths.

[0005] It should be noted that the information disclosed in the background section above is only for understanding the background of this application, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0006] This application provides a communication channel modeling method, system, device, and storage medium to solve the technical problems of low computational efficiency, insufficient spectral dependence, insufficient scalability and reproducibility, and insufficient characterization of non-line-of-sight paths in existing communication channel modeling technologies.

[0007] The technical solution adopted in this application to solve the above-mentioned technical problems is as follows.

[0008] This application provides a ray tracing-based communication channel modeling method in its first aspect, comprising the following steps:

[0009] Set up at least one transmitter and at least one receiver. The transmitter is used to generate signals and the receiver is used to receive signals.

[0010] The ray tracing is generated from the receiving end, and the ray tracing interacts with objects in the communication scene to form a ray tracing path.

[0011] Record the propagation information of the ray tracing path;

[0012] Based on the propagation information, the channel impulse response is calculated to obtain the communication channel characteristics.

[0013] In some embodiments, when the tracking ray interacts with an object, a two-way scattering distribution function (BSDF) is used to model the relationship between the incident and outgoing directions of the tracking ray, taking into account the influence of the wavelength characteristics of the object's material on the interaction.

[0014] In some embodiments, the transmitter includes a light source, and when forming a ray tracing path, a sampling optimization strategy is employed, which includes Next Event Estimation (NEE), which can sample the light source.

[0015] In some embodiments, the sampling optimization strategy includes Monte Carlo sampling and / or Russian roulette strategy, which can randomly terminate the tracking ray.

[0016] In some embodiments, the sampling optimization strategy includes multiple importance sampling (MIS), which can reduce the estimation variance during ray tracing path formation.

[0017] In some embodiments, the tracking ray has a wavelength attribute, which is determined based on the probability distribution of the emission spectrum from the transmitter.

[0018] In some embodiments, when forming a ray tracing path, a bidirectional path connection strategy is adopted to connect the forward path generated by the receiver with the backward path sampled from the transmitter.

[0019] In some embodiments, the communication channel modeling method is implemented using a parallel computing architecture, which includes a graphics processing unit (GPU), a tensor processing unit (TPU), or a multi-core central processing unit (CPU).

[0020] In some embodiments, the calculated channel impulse response can be used for received signal power estimation, signal-to-noise ratio analysis, bit error rate prediction, and beamforming and resource allocation.

[0021] In some embodiments, a ray tracing can be generated from the transmitter, and the ray tracing can interact with objects in the communication scene to form a ray tracing path.

[0022] Furthermore, in a second aspect, this application provides a communication channel modeling system 10, comprising:

[0023] The ray generation module is used to generate tracking rays from the receiving end in communication scenarios;

[0024] The interactive computing module is used to form ray tracing paths and record the propagation information of ray tracing paths when tracing rays interact with the surfaces of objects in the scene.

[0025] The channel calculation module is used to calculate the channel impulse response and obtain the corresponding communication channel characteristics based on the propagation information of the ray tracing path.

[0026] In some embodiments, a path optimization module is also included for implementing a sampling optimization strategy during light propagation. The sampling optimization strategy includes next event estimation (NEE), multiple importance sampling (MIS), Monte Carlo sampling, and / or Russian roulette strategy.

[0027] In some embodiments, the interactive computing module models the relationship between the incident and outgoing directions of the tracking light rays based on the two-way scattering distribution function (BSDF) when the light interacts with the surface.

[0028] In some embodiments, the light generation module assigns a wavelength attribute to each tracking ray and records the wavelength characteristics of the scene object materials during the interaction.

[0029] In some embodiments, the path optimization module employs a bidirectional path construction strategy.

[0030] In some embodiments, the communication channel modeling system 10 is implemented on a parallel computing architecture, which includes a graphics processor, a tensor processing unit, and a multi-core central processing unit.

[0031] Furthermore, in a third aspect, this application provides a terminal device, which includes a memory, a processor, and program instructions stored in the memory and executable on the processor. When the program instructions are executed by the processor, they implement the steps of the communication channel modeling method of this application.

[0032] Furthermore, in a fourth aspect, this application provides a computer-readable storage medium having program instructions stored thereon, which, when executed by a processor, implement the steps of the communication channel modeling method of this application.

[0033] The present invention has the following beneficial effects:

[0034] The first aspect of this invention provides a ray tracing-based communication channel modeling method. By setting up at least one transmitter and at least one signal receiver, the ray tracing generated by the signal receiver interacts with objects in the communication scene to form a ray tracing path. Then, based on the propagation information of the ray tracing path, the channel impulse response is calculated to obtain the communication channel characteristics. This achieves a combination of efficient ray tracing operators and communication modeling, offering advantages over traditional communication channel modeling such as unlicensed spectrum, high bandwidth capacity, no electromagnetic interference, and integration with lighting. Furthermore, because this application introduces ray tracing operators from the field of computer vision into communication channel modeling, it enables the modeling of communication channels to achieve massive-scale ray tracing within milliseconds using existing GPU-accelerated rendering technology and ray tracing algorithms, greatly improving the efficiency and accuracy of communication channel modeling. It simultaneously addresses the technical problem of prior art where efficiency and accuracy are difficult to balance. Therefore, this application provides a method for quickly and accurately simulating the complex characteristics of real-world communication channels, providing a more reliable data foundation for communication system design.

[0035] Furthermore, because the method in this application considers the wavelength characteristics of the material when modeling the relationship between the incident and exit directions, and assigns wavelength attributes to the ray tracing rays during ray tracing path generation, it enables spectrum-selective channel modeling by combining the wavelength-dependent reflection or absorption characteristics of the object's material when the ray interacts with objects in the communication scene. In implementation, the attenuation rate can be found by matching the light wavelength to the absorption spectrum of the object's material, and corresponding attenuation calculations can be performed. This solves the technical problem of existing methods, where most models only consider light intensity distribution and ignore the emission spectrum of the light source and the wavelength-dependent effect of the material, resulting in insufficient spectral dependence. This reduces the deviation from the actual measurement results and improves the accuracy of the measurement results.

[0036] Furthermore, since the method of this application adopts the next event estimation (NEE) approach when constructing ray tracing paths, the next event estimation (NEE) can directly sample the light source at each effective scattering point to explicitly construct the ray connection path, thereby improving the generation probability and sampling effectiveness of effective ray tracing paths, and further improving the efficiency and accuracy of communication channel modeling.

[0037] Furthermore, because the method in this application employs Monte Carlo sampling and / or Russian roulette strategy, some light rays can be randomly terminated when the number of path reflections is large, thereby reducing computational complexity.

[0038] Furthermore, because the method in this application employs a sampling optimization strategy, including Next Event Estimation (NEE), Monte Carlo sampling and / or Russian Roulette strategy, and Multiple Importance Sampling (MIS), in the case of multiple sampling strategies coexisting, this method uses Multiple Importance Sampling (MIS) to fuse the estimates from NEE and BSDF sampling. This suppresses the variance amplification effect of each strategy within their respective strengths, ensuring that the overall estimate remains unbiased and approximately minimizes variance. This improves the accuracy of communication channel modeling. Compared to traditional optical communication channel modeling schemes that only consider the line-of-sight path or at most the first-order reflection, the method in this application can capture the line-of-sight path and multi-order reflection paths in complex scenarios by performing BSDF sampling, NEE sampling, and MIS strategies on the reflecting surface, and by using existing open-source platforms for one-click import into complex scenarios. This allows for the construction of a more accurate channel model, solving the technical problem of insufficient characterization of non-line-of-sight paths in existing technologies.

[0039] Furthermore, since the method of this application is based on a sampling optimization strategy, including next event estimation (NEE), Monte Carlo sampling and / or Russian roulette strategy, and multiple importance sampling (MIS), the method of this application has the characteristics of standardization and high efficiency. In addition, since the method of this application uses parallel computing to achieve fast simulation, it greatly reduces the cost for different researchers to repeatedly implement the underlying model, and provides a reproducible and scalable unified modeling platform for academic research and engineering applications.

[0040] Therefore, this application introduces ray tracing operators from the field of computer vision into communication channel modeling, and combines next event estimation, multiple importance sampling, wavelength-dependent modeling, and parallel acceleration architecture. This not only improves the modeling accuracy and efficiency, but also enhances the scalability and reproducibility of the modeling, making it widely applicable to various optical wireless communication scenarios such as visible light communication, infrared communication, and free space optical communication.

[0041] Other beneficial effects of the present invention will be further described below. Attached Figure Description

[0042] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein:

[0043] Figure 1 This is a flowchart of a communication channel modeling method and system based on ray tracing provided in an embodiment of this application;

[0044] Figure 2 This is a schematic diagram illustrating the principle of ray tracing-based channel modeling according to an embodiment of this application;

[0045] Figure 3 A schematic diagram of a channel modeling system based on ray tracing provided in an embodiment of this application;

[0046] Figure 4 This is a schematic diagram of the structure of an electronic device provided according to an embodiment of this application.

[0047] Figure 5 A two-dimensional top-view diagram of the conference room experimental scenario for the example;

[0048] Figure 6 The power measured at some measurement points in the experimental example is compared with the proposed scheme and the results obtained from MATLAB simulation.

[0049] Figure 7 This section compares the computational efficiency of the proposed scheme in the experimental example with that of the optical simulation platform Zemax. Detailed Implementation

[0050] The embodiments of the present invention will be described in detail below. It should be emphasized that the following description is merely exemplary and not intended to limit the scope and application of the present invention.

[0051] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of embodiments of the present invention, "a plurality of" means two or more, unless otherwise explicitly specified.

[0052] The embodiments of this application will be further described below with reference to the accompanying drawings.

[0053] like Figure 1 , 2 As shown, the first aspect of this application provides a communication channel modeling method based on ray tracing, including the following steps:

[0054] S1: Construct a communication scenario including at least one transmitter and at least one signal receiver 11;

[0055] S2: The signal receiver 11 generates a "tracing ray", which interacts with objects in the communication scene to form a "ray tracing path", and records the propagation information of the "ray tracing path".

[0056] S3: Calculate the channel impulse response based on the propagation information to obtain the communication channel characteristics.

[0057] In some embodiments, the transmitting end in step S1 can be a light source 12 to generate a modulated light signal, and the receiving end can receive the light signal and convert it into an electrical signal. Optionally, the light source 12 can be a light-emitting diode (LED), a laser, or other device capable of emitting light radiation, and the receiving end can be a photodetector (PD) or other photosensitive element.

[0058] In some embodiments, such as Figure 2 As shown, in step S2, a large number of "tracing rays," such as line-of-sight rays 14, can be generated from the receiving end. These rays are then sampled using a field-of-view grid 16 calculated in matrix form, allowing random initial directions within the field of view and propagating the rays within a connected scene field constructed in the 3D virtual environment. When the rays interact with the surface of scene objects 13, scattering or transmission occurs. At this point, the propagation direction and energy information can be updated according to preset scattering or transmission rules, forming a "ray tracing path." Optionally, the scattering or transmission rules may include specular reflection, diffuse reflection, refraction, or a combination thereof.

[0059] In some embodiments, the propagation information in steps S2 and S3 includes the propagation distance, propagation delay, and power information of the "ray tracing path". The channel impulse response is formed by accumulating the propagation distance, propagation delay, and power information, and then the impulse response is calculated to obtain the corresponding communication channel characteristics.

[0060] In some embodiments, communication channel characteristics include, but are not limited to, parameters such as received power distribution, power delay profile (PDP), delay spread, and channel frequency response, which are used to characterize the propagation characteristics of signals and system performance in an optical wireless communication environment.

[0061] Based on the above design, the method of this application constructs a communication scene, setting at least one light source 12 and at least one signal receiver 11 within the scene. The "tracing ray" generated by the signal receiver 11 interacts with objects in the communication scene to form a "ray tracing path." Then, based on the propagation information of the "ray tracing path," the channel impulse response is calculated to obtain the communication channel characteristics. This achieves a combination of efficient ray tracing operators and communication modeling. Compared to traditional communication channel modeling, the method of this application has advantages such as unlicensed spectrum, high bandwidth capacity, no electromagnetic interference, and integration with lighting. Furthermore, because this application introduces ray tracing operators from the field of computer vision into communication channel modeling, it enables communication channel modeling to achieve massive-scale ray tracing within milliseconds using existing GPU-accelerated rendering technology and ray tracing algorithms, breaking the bottleneck of difficulty in balancing efficiency and accuracy in existing technologies and improving the efficiency and accuracy of communication channel modeling. Therefore, this application provides a method that can quickly and accurately simulate the complex characteristics of real-world communication channels, providing a more reliable data foundation for communication system design.

[0062] In some embodiments, when generating the "ray tracing path" in step S2, the relationship between the incident direction and the outgoing direction of the "tracing ray" can be modeled based on the wavelength characteristics of the material of the object in the communication scene to achieve material scattering sampling. Specifically, the material and reflectivity of the object surface in the communication scene can be set according to actual needs to simulate the real communication environment. The material and reflectivity are parameters of the model surface, which can be obtained in the sampling.

[0063] For example, this embodiment can use a two-way scattering distribution function (BSDF) to model the relationship between the incident and outgoing directions of the "tracing ray". First, non-normalized selection weights for each component are calculated, taking into account factors such as the object's material properties, metallicity, and the ratio of diffuse to specular transmission. Next, a component is selected for sampling using a uniformly random number. Assuming the diffuse component is selected, the diffuse reflection sampling function is called to obtain the outgoing direction, the probability density of that component, and its weight. Here, the diffuse reflection sampling is a cosine-weighted hemispherical sampling. The two-way scattering distribution function (BSDF) is specifically expressed as:

[0064]

[0065]

[0066] in, These are the incident and exit directions, respectively. For the weight of mirror transmission, For diffuse transmission weight, For diffuse reflection, For diffuse transmission, For specular reflection, This is a specular transmission term.

[0067] Specifically, when detailing the engineering implementation, it can be further written as:

[0068]

[0069]

[0070]

[0071]

[0072] in, The normal vector of the reflection point; The diffuse reflectance coefficient; For Fresel items; Represented as ; The normal distribution function; For geometric occlusion terms; Let be the transmission function that can be expressed by Snell's law; This represents the diffuse transmittance ratio; This is an indicator function.

[0073] In some embodiments, during the generation of the "ray tracing path" in step S2, the tracing rays can be assigned a wavelength attribute. The wavelength attribute is determined based on the probability distribution of the emission spectrum of the light source 12. Optionally, the light source 12 can provide an emission spectrum, with wavelength as the horizontal axis and relative probability as the vertical axis. During ray generation, a wavelength attribute can be assigned to each ray, and the wavelength is determined based on the probability distribution of the emission spectrum of the light source 12, with wavelengths sampled probabilistically from the emission spectrum.

[0074] Based on the above design, the method of this application considers the wavelength characteristics of the material when modeling the relationship between the incident direction and the outgoing direction, and assigns the wavelength attribute to the ray tracing ray when the "ray tracing path" is generated, so that when the ray interacts with the object 13 in the communication scene, the wavelength-dependent reflection or absorption characteristics of the object material can be combined to achieve spectrum-selective channel modeling.

[0075] During implementation, the attenuation rate can be found by matching the light wavelength to the absorption spectrum of the material, and the corresponding attenuation can be calculated. This solves the technical problem of insufficient spectral dependence in existing methods, which mostly consider only the light intensity distribution and ignore the emission spectrum of the light source and the wavelength dependence effect of the material. This reduces the deviation from the actual measurement results and improves the accuracy of the measurement results.

[0076] In some embodiments, when constructing the "ray tracing path" in step S2, a sampling optimization strategy may be introduced to improve simulation efficiency and accuracy. Optionally, the sampling optimization strategy includes the Next Event Estimation (NEE) method, which directly samples the light source 12 at each effective scattering point to explicitly construct the ray connection path, thereby increasing the generation probability of the effective "ray tracing path".

[0077] Specifically, assuming the set of light sources is If sampling points are selected on the surface of the light source The radiation emitted from that point will then scatter at the scattering point. The direct contribution can be expressed as:

[0078]

[0079] in, Represents the BSDF scattering function; For the light source in direction Radiance on the surface; This is a geometric term used to characterize angular attenuation; This is a visibility function used to determine whether light is blocked; The probability of selecting a light source; This represents the conditional probability density for selecting sampling points on the surface of the light source. It is the square of the Eulerian distance between y and x; It is the normal vector at the intersection point with the surface of the light source. This estimator is unbiased and can significantly improve the effectiveness of path sampling.

[0080] Based on the above design, the method of this application adopts the next event estimation (NEE) approach when constructing the "ray tracing path". By directly sampling the light source at each effective scattering point, the ray connection path is explicitly constructed, thereby improving the generation probability and sampling effectiveness of the effective "ray tracing path", and thus improving the efficiency and accuracy of communication channel modeling. In addition, optionally, a bidirectional path connection strategy can be adopted when constructing the "ray tracing path", that is, connecting the forward path generated from the receiver with the backward path sampled from the light source, to ensure that each scattering point can form at least one complete path.

[0081] In some embodiments, the sampling optimization strategy further includes Monte Carlo sampling and / or Russian roulette strategies. Specifically, a probability threshold can be set. and the threshold for the number of reflections When the number of reflections k exceeds season Generate a random number rand between [0, 1]. If rand > 0, ... If the calculation fails, the calculation will terminate; otherwise, the accumulated losses will be normalized again.

[0082]

[0083] In subsequent experimental examples, in, Take 7. Take 0.95.

[0084] After each reflection, path transmittance The update can be performed using the following formula:

[0085]

[0086] in, This represents the scattering function value. This represents the sampling probability density.

[0087] Based on the above design, the method of this application can randomly terminate some light rays when the number of path reflections is large, thereby reducing the computational complexity by using Monte Carlo sampling and / or Russian roulette strategy.

[0088] In some embodiments, during the light propagation process in step S2, the sampling optimization strategy further includes a Multiple Importance Sampling (MIS) strategy, which is used to weight and fuse the sampling contributions from the light source and the scattering contributions from the material.

[0089] Specifically, to achieve a combination effect with minimal variance among different sampling strategies, a multiple importance sampling (MIS) strategy is adopted to weight and fuse the results from NEE sampling and BSDF sampling. Let the probability densities of the two sampling methods be respectively... and Then its weight can be defined according to the balance heuristic:

[0090]

[0091] The weights for light source sampling, The weights for BSDF sampling.

[0092] In another implementation, a power heuristic can also be used, which is defined as:

[0093]

[0094] in This is an empirical parameter, taking a value of 1 or 2, typically 2. The weighting method described above effectively reduces variance and improves the stability of path estimation, thus achieving good convergence performance under different scenarios and material conditions.

[0095] After MIS processing, the conversion relationship between the power of each light ray and the calculated radiance is as follows:

[0096]

[0097] in The angle between the incident ray and the surface normal is... For receiving area, This represents the sampling probability density of the receiver for the incident direction.

[0098] Based on the above design, the method in this application can systematically capture multi-order reflection paths at the physical interaction level through BSDF modeling and recursive propagation. Whenever a ray intersects a surface, the incident-outgoing directions are sampled and energy is weighed according to a preset BSDF: the specular component ensures a high probability of hitting the "main lobe" direction, while the diffuse reflection / transmission component covers a large number of scattering directions on the rough surface through cosine-weighted importance sampling. Through this recursive process of "intersection-sampling-propagation," light naturally generates first-order, second-order, and even higher-order reflection / transmission paths. Simultaneously, combined with the Russian roulette termination strategy, long paths with minimal contribution can be adaptively truncated while maintaining unbiasedness. This statistically preserves the main energy contribution while avoiding the variance and computational cost of invalid long chains, thus improving computational efficiency and accuracy.

[0099] Furthermore, to avoid slow convergence due to relying solely on random hits, this method explicitly employs Next Event Estimation (NEE) to connect the light source at each scattering point: directly sampling the light source surface element or its importance distribution at the scattering point, projecting a "shadow ray 15" for visibility detection, and accumulating its measurable contribution to the receiver. This step is equivalent to forcibly attempting a direct connection to the light source at all possible scattering points, which can significantly increase the probability of discovering high-contribution paths without increasing bias, effectively reducing estimation variance and accelerating convergence. Compared to relying solely on roaming path tracing, NEE ensures systematic coverage of the "main energy path," ensuring that direct light source components in first-order and higher-order reflections are not missed.

[0100] In situations where multiple sampling strategies coexist, this method employs Multiple Importance Sampling (MIS) to fuse estimates from source element emission (NEE) and bottom-ground scattering (BSDF) sampling (recursive scattering). MIS uses a balanced / power heuristic as weights to suppress the variance amplification effect of each strategy within their respective strengths, ensuring that the overall estimate remains unbiased and approximately minimizes variance. This improves the accuracy of communication channel modeling. The logic here is based on the fact that source element emission sampling excels at capturing strong direct components, while BSDF sampling excels at exploring complex indirect components; a trade-off between the two avoids the systematic instability caused by focusing solely on direct components or relying solely on random walks.

[0101] Furthermore, compared to traditional optical communication channel modeling schemes that only consider the line-of-sight path or at most the first-order reflection, the method of this application embodiment can capture the line-of-sight path and multi-order reflection paths in complex scenes by performing BSDF sampling, NEE sampling, and MIS strategies on the reflecting surface, and by importing the model into the Nvidia Falcor open-source platform with one-click compatibility for complex scenes. This allows for the construction of a more accurate channel model, solving the technical problem of insufficient characterization of non-line-of-sight paths in existing technologies. By introducing sampling optimization strategies and parallel computing architecture, this application significantly improves modeling efficiency while ensuring accuracy, enabling real-time modeling of millions of rays.

[0102] In some embodiments, in step S3, after all “tracing ray” paths have propagated, their propagation distance, propagation delay, and power information can be recorded, and the impulse response of the communication channel can be calculated based on this.

[0103] For a propagation length of The propagation delay of the light ray path can be expressed as:

[0104]

[0105] in, It is the speed of light.

[0106] Based on power channel impulse response It can be represented as:

[0107]

[0108] in, Let N be the time, and N be the total number of paths. Let be the received power of the i-th path. Let be the impulse function. Therefore, the total received power at the receiver is... It can be represented as:

[0109]

[0110] Furthermore, the channel impulse response calculation results of this application can also be used for communication system optimization such as received signal power estimation, signal-to-noise ratio analysis, bit error rate prediction, and beamforming and resource allocation.

[0111] In some embodiments, the method of this application can be implemented using a parallel computing architecture, including but not limited to a graphics processing unit (GPU), a tensor processing unit (TPU), and a multi-core central processing unit (CPU), to achieve real-time modeling of millions of rays.

[0112] Based on the above design, it can be understood that the method of this application introduces ray tracing operators from the field of computer vision into communication channel modeling, and combines next event estimation, multiple importance sampling, Monte Carlo sampling, Russian roulette strategy, wavelength-dependent modeling and parallel acceleration architecture design, which not only improves the modeling accuracy and efficiency, but also enhances the scalability and reproducibility of modeling, and can be widely applied to various optical wireless communication scenarios such as visible light communication, infrared communication, and free space optical communication.

[0113] Furthermore, it is understood that the method of this application, based on the above sampling optimization strategy, has the characteristics of standardization and high efficiency, and utilizes parallel computing to achieve rapid simulation, which greatly reduces the cost for different researchers to repeatedly implement the underlying model, and provides a reproducible and scalable unified modeling platform for academic research and engineering applications.

[0114] Furthermore, in some embodiments, the method of the present invention can also be applied to the case where a "tracing ray" is generated from the transmitting end. Specifically, a "tracing ray" is generated from the transmitting end, and the "tracing ray" interacts with objects in the communication scene to form a "ray tracing path". The specific steps can be referred to the above-disclosed embodiments, and will not be repeated here.

[0115] like Figure 3 As shown, a second aspect of this application provides a communication channel modeling system for implementing the above-described ray tracing-based communication channel modeling method. The system may include a ray generation module 100, an interactive calculation module 200, a path optimization module 300, and a channel calculation module 400.

[0116] In some embodiments, the ray generation module 100 is used to generate line-of-sight rays originating from the receiver in a three-dimensional virtual environment. The ray generation module 100 can emit a large number of rays according to the field of view of the receiver to ensure that potential propagation paths are statistically covered. In some embodiments, the ray generation module 100 can also assign a wavelength attribute to each ray, the wavelength being determined based on the emission spectrum probability distribution of the light source, thereby supporting subsequent spectral dependence modeling.

[0117] In some embodiments, the interactive computing module 200 updates the propagation direction and energy information of light rays based on preset scattering or transmission rules when light interacts with the surface of objects in the scene. Specifically, the interactive computing module 200 can call the two-way scattering distribution function (BSDF) to determine the outgoing direction based on the incident direction, surface normal vector, and material properties, and calculate energy attenuation. In some embodiments, the interactive computing module 200 can further consider the wavelength-dependent characteristics of materials, i.e., the absorption and reflection characteristics of light of different wavelengths on the surface are different. This difference is related to specific materials and colors, such as objects of a certain color absorbing less light of the corresponding color, thus more realistically depicting the spectral effects of the channel. Specifically, an initial wavelength can be assigned to each ray, the probability of which is determined by the spectrum of the light source. During the tracking process, this wavelength remains unchanged. BSDF queries and refractive index calculations both use this specific wavelength. Finally, by classifying and integrating all paths according to their wavelengths, a spectrally rich channel model is naturally obtained.

[0118] In some implementations, the path optimization module 300 introduces a sampling optimization strategy during light propagation to increase the probability of generating effective paths and reduce estimation variance. This module can implement Next Event Estimation (NEE), which directly samples the light source position at scattering points, ensuring that every effective scattering point can connect to the light source, thereby reducing the number of invalid paths. Simultaneously, the path optimization module 300 can also combine a Multiple Importance Sampling (MIS) strategy to weightedly fuse the light source sampling and BSDF sampling results, allowing the advantages of different sampling methods to complement each other, thereby improving convergence speed and reducing variance. Furthermore, this module can combine Monte Carlo sampling with a Russian roulette strategy to terminate some paths based on path transmittance and a set probability when the number of path reflections is excessive, thereby reducing computational overhead while maintaining unbiased estimation.

[0119] In some embodiments, the channel calculation module 400 records and processes the propagation information of each effective path after the light propagation is complete, and finally calculates the channel impulse response. Specifically, the channel calculation module 400 can obtain the overall channel impulse response by superimposing the propagation delay and received power of the path, and further calculate the total received power at the receiver. In some embodiments, the channel calculation module 400 can also output derived indicators such as signal-to-noise ratio, frequency domain response, or bit error rate, thereby providing a basis for subsequent communication system design and optimization.

[0120] In some embodiments, the modules of the system in this application can be implemented using independent hardware units or integrated on the same hardware platform through software programming. In one optional approach, the entire system significantly improves computational efficiency through a parallel computing architecture, including but not limited to a graphics processing unit (GPU), a tensor processing unit (TPU), and a multi-core central processing unit (CPU), achieving speedups of tens to hundreds of times compared to traditional CPU serial implementations, making high-precision, large-scale ray tracing modeling of complex scenes feasible. In another optional approach, almost all computationally intensive tasks are executed in parallel on the GPU. These include pixel-by-pixel and sample-by-sample ray generation, hit point shading calculation (covering BSDF weight calculation and normalization, importance sampling, BSDF evaluation, and combined calculation of multiple probabilities), shadow ray testing, next event estimation (NEE) sampling, path iteration for multiple bounces, and writing the final result to the buffer. Statistical and data export functions, such as parallel buffer writing in the CIR system and subsequent parallel reduction operations, are also performed on the GPU.

[0121] Through the above design, the communication channel modeling system of this application can achieve efficient and accurate optical channel modeling in complex environments. It not only ensures the ability to characterize non-direct-view paths and multi-level reflection paths, but also significantly improves modeling efficiency. It is suitable for visible light communication, infrared communication and other optical wireless communication scenarios.

[0122] like Figure 4 As shown, the third embodiment of this application also provides a terminal device for implementing the aforementioned ray tracing-based communication channel modeling method. The terminal device may include a memory 401, a processor 402, and program instructions stored in the memory 401 and executable on the processor 402.

[0123] The memory 401 stores computer-executable instructions. When executed by the processor 402, these instructions implement the steps of the ray tracing-based communication channel modeling method described in the previous embodiment. Specifically, when the program is executed by the processor 402, it can perform functions such as communication scene construction, ray generation and propagation, path optimization, and channel impulse response calculation, thereby achieving automated modeling and calculation of the communication channel.

[0124] In one optional embodiment, the terminal device may further include a communication interface 403 for data interaction between the memory 401 and the processor 402. The communication interface 403 may be implemented using an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. The bus may include an address bus, a data bus, and a control bus to ensure high-speed data transmission and reliable interaction between modules.

[0125] In one alternative implementation, the terminal device's memory 401, processor 402, and communication interface 403 can be integrated onto the same chip, with the modules communicating via an internal bus. The processor 402 can be a central processing unit (CPU), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices. For scenarios requiring large-scale parallel ray tracing computation, the processor 402 may also include a graphics processing unit (GPU) or a tensor processing unit (TPU) to support parallel computation, thereby significantly improving the system's modeling speed and simulation accuracy.

[0126] The fourth embodiment of this application also provides a computer-readable storage medium storing program instructions thereon. When executed by a processor, the program instructions implement the steps of the above-described ray tracing-based communication channel modeling method. The computer-readable storage medium may include, but is not limited to, read-only memory (ROM), random access memory (RAM), flash memory, disk storage, optical disk storage, or other non-volatile media capable of storing computer program code.

[0127] The fifth embodiment of this application also provides a computer program that, when loaded into a computer device and run, performs the aforementioned ray tracing-based communication channel modeling method. This program can be distributed as software or embedded in a hardware device for direct execution in an embedded system or high-performance computing platform.

[0128] Through the above design, the terminal device, computer-readable storage medium, and computer program of this application embodiment can automate the ray tracing-based communication channel modeling process at the hardware level, achieving both high precision and high efficiency. The system can call GPU acceleration libraries as needed to complete ray tracing tasks and store and analyze the generated propagation path data, thereby achieving real-time modeling and channel characteristic reconstruction in complex environments, further expanding the feasibility and applicability of the method of this invention in scientific research and engineering applications.

[0129] In addition, to verify the effectiveness of the ray tracing-based communication channel modeling method proposed in this invention, experiments were conducted in typical scenarios to evaluate the modeling accuracy and computational efficiency of the method.

[0130] The experiment was conducted in a real conference room, and the scene was reconstructed 1:1 using UE5 (Unreal Engine 5).

[0131] like Figure 5As shown, measurements were taken using an optical power meter at nine desktop locations (T1–T9), with a photodetector (PD) fixed to the lectern. At each measurement point, the PD rotated horizontally in eight directions (0°–315°, with a step size of 45°), obtaining a total of 72 samples. The measurement data were compared with traditional MATLAB simulation results and the framework proposed in this invention. Traditional MATLAB methods ignore the complex details of the room, considering only the line-of-sight (LoS) and first-order non-line-of-sight (NLoS) components: the LoS component uses a Lambertian model, and the first-order NLoS is calculated by dividing the room surface into small elements, calculating the reflected power of each element, and summing their contributions.

[0132] For comparison, 12 representative samples were randomly selected, and the results are as follows: Figure 6 As shown in the figure. Experimental results show that the proposed channel modeling scheme is closer to the actual measurement data than the traditional MATLAB simulation results at most sampling points, and successfully captures multiple NLoS paths ignored by the traditional method. By comparing all 72 sampling points, the proposed method reduces the root mean square error (NMSE) of the measured values ​​by 20.18% compared with the traditional scheme, which significantly improves the accuracy of signal strength modeling in complex environments.

[0133] Besides achieving high accuracy in complex scenes, the main advantage of this invention lies in its computational efficiency. On the one hand, while maintaining the diversity of Monte Carlo simulations, a NEE strategy is adopted to maximize the contribution of each ray; on the other hand, the solution fully utilizes GPU acceleration, performing parallel computation at the "per pixel × per sample" granularity per thread, significantly improving simulation speed.

[0134] For comparison, the method of this invention was compared with Zemax, which is widely used for optical wireless channel modeling. The previously constructed room model was imported into Zemax, and the computation time required for single-point ray tracing was measured under different ray counts, ranging from 10,000 to 2,500,000 rays. The results are as follows: Figure 7 As shown, even with one million rays, the proposed method maintains millisecond-level computation time, with the speedup becoming more pronounced as the number of rays increases. In high-precision simulations, when the number of rays reaches 2,500,000, the proposed method achieves over 400 times the speedup compared to Zemax. These results demonstrate that the framework of this invention offers a significant improvement in computational efficiency and can serve as a powerful tool for future visible light communication (VLC) channel modeling and simulation.

[0135] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0136] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0137] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0138] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0139] The background section of this invention may include background information about the problems or environment in which the invention is being developed, and is not necessarily a description of prior art. Therefore, the content included in the background section does not constitute an admission of prior art by the applicant.

[0140] The above description provides a further detailed explanation of the present invention in conjunction with specific / preferred embodiments, and it should not be construed that the specific implementation of the present invention is limited to these descriptions. For those skilled in the art, various substitutions or modifications can be made to these described embodiments without departing from the concept of the present invention, and all such substitutions or modifications should be considered within the scope of protection of the present invention. In the description of this specification, the reference to terms such as "an embodiment," "some embodiments," "preferred embodiment," "example," "specific example," or "some examples," etc., indicates that the specific features, structures, materials, or characteristics described in connection with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in any suitable manner in one or more embodiments or examples. Without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification and the features of different embodiments or examples. Although the embodiments of the present invention and their advantages have been described in detail, it should be understood that various changes, substitutions, and modifications can be made herein without departing from the scope of protection of the patent application.

Claims

1. A communication channel modeling method based on ray tracing, characterized in that, Includes the following steps: The system includes at least one transmitter and at least one receiver, wherein the transmitter is used to generate a signal and the receiver is used to receive the signal. A ray of light is generated from the receiving end, and the ray of light interacts with objects in the communication scene to form a ray tracing path. When forming the ray tracing path, the forward path generated by the receiver is connected to the backward path sampled from the transmitter; When the tracking ray interacts with the object, the relationship between the incident and outgoing directions of the tracking ray is modeled using a two-way scattering distribution function (BSDF), taking into account the influence of the wavelength characteristics of the object's material on the interaction. Record the propagation information of the ray tracing path; Based on the propagation information, the channel impulse response is calculated to obtain the communication channel characteristics.

2. The communication channel modeling method according to claim 1, characterized in that, The transmitter includes a light source. When forming the ray tracing path, a sampling optimization strategy is adopted. The sampling optimization strategy includes next event estimation (NEE), which can sample the light source.

3. The communication channel modeling method according to claim 2, characterized in that, The sampling optimization strategy includes Monte Carlo sampling and / or Russian roulette strategy, which can randomly terminate the tracking ray.

4. The communication channel modeling method according to claim 2 or 3, characterized in that, The sampling optimization strategy includes multiple importance sampling (MIS), which can reduce the estimation variance during ray tracing path formation.

5. The communication channel modeling method according to any one of claims 1-3, characterized in that, The tracking ray has a wavelength attribute, which is determined based on the probability distribution of the emission spectrum of the transmitting end.

6. The communication channel modeling method according to claim 1, characterized in that, A bidirectional path connection strategy is adopted to connect the forward path generated by the receiver with the backward path sampled from the transmitter.

7. The communication channel modeling method according to claim 1, characterized in that, The communication channel modeling method is implemented using a parallel computing architecture, which includes a graphics processing unit, a tensor processing unit, and a multi-core central processing unit.

8. The communication channel modeling method according to claim 1, characterized in that, The calculated channel impulse response can be used for received signal power estimation, signal-to-noise ratio analysis, bit error rate prediction, and beamforming and resource allocation.

9. The communication channel modeling method according to claim 1, characterized in that, The transmitting end can also generate a tracking ray, which interacts with objects in the communication scene to form a ray tracing path.

10. A communication channel modeling system, characterized in that, include: A ray generation module (100) is used to generate a tracking ray from the receiving end in a communication scenario; The interactive computing module (200) is used to form a ray tracing path and record the propagation information of the ray tracing path when the ray interacts with the surface of the scene object. When the ray interacts with the surface, the interactive computing module (200) models the relationship between the incident direction and the outgoing direction of the ray based on the two-way scattering distribution function (BSDF). The channel calculation module (400) is used to calculate the channel impulse response and obtain the corresponding communication channel characteristics based on the propagation information of the ray tracing path; The path optimization module (300) is used to connect the forward path generated by the receiver with the backward path sampled from the transmitter.

11. The communication channel modeling system according to claim 10, characterized in that, The path optimization module (300) is also used to implement a sampling optimization strategy during light propagation, which includes next event estimation (NEE), multiple importance sampling (MIS), Monte Carlo sampling, and / or Russian roulette strategy.

12. The communication channel modeling system according to claim 10, characterized in that, The light generation module (100) assigns wavelength attributes to each tracking ray and records the wavelength characteristics of scene object materials during the interaction process.

13. The communication channel modeling system according to claim 10, characterized in that, The path optimization module (300) connects the forward path generated by the receiver with the backward path sampled from the transmitter through a bidirectional path construction strategy.

14. The communication channel modeling system according to claim 10, characterized in that, The communication channel modeling system is implemented using a parallel computing architecture, which includes a graphics processor, a tensor processing unit, and a multi-core central processing unit.

15. A terminal device, characterized in that, The terminal device includes a memory (401), a processor (402), and program instructions stored in the memory (401) and executable on the processor (402). When the program instructions are executed by the processor (402), they implement the steps of the communication channel modeling method as described in any one of claims 1-9.

16. A computer-readable storage medium, characterized in that, It stores program instructions, which, when executed by a processor, implement the steps of the communication channel modeling method as described in any one of claims 1-9.

Images