Channel knowledge map construction method and system based on nerve ray tracing

Abstract

The invention provides a channel knowledge map construction method and system based on nerve ray tracing, and the method comprises the steps: collecting the actual measurement data of electromagnetic wave propagation through deploying channel detection equipment; the actually measured data are preprocessed; training a deep learning model by using the preprocessed data and a material library, so that the deep learning model can learn a complex nonlinear relationship of interaction between electromagnetic waves and the environment; through a trained ray tracing frame based on a neural network, after a three-dimensional map and resolution are given, nerve ray tracing is carried out on each position, according to obtained multipath channel information, related channel parameters are predicted, a channel knowledge map in the current environment is generated, and proper values are set to visualize the channel knowledge map. According to the method, the fitting capability of the neural network to the nonlinear relationship is used for calculating the interaction effect of the ray and the environment, the electromagnetic property of the material in an actual scene can be better predicted, and the method has excellent migration capability.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method and system for constructing a channel knowledge map based on neural ray tracing. Background Art

[0002] In the field of wireless communications, especially with the development of 5G and future communication technologies, Multiple Input Multiple Output (MIMO) technology has become the key to improving data transmission rate and system capacity. The successful implementation of MIMO technology relies on the accurate acquisition of channel state information (CSI), which is essential for beamforming and signal preprocessing to maximize signal gain and reduce interference. However, traditional channel acquisition methods mainly rely on the training process of pilot signals, which leads to increased time and energy overhead as the number of antennas increases, limiting the scalability and energy efficiency of the system.

[0003] In modern wireless communication systems, the concept of channel knowledge maps is widely used to describe and predict the propagation characteristics of wireless signals in a specific environment. This map not only includes the path loss and multipath effects of wireless signals, but also covers the electromagnetic characteristics of the environment, such as reflection, refraction, diffraction, and scattering. Channel knowledge maps are crucial to improving the performance of wireless communication systems, enabling the system to perform accurate channel prediction, beamforming, network planning and optimization, and provide high-precision positioning services.

[0004] Traditional methods for constructing channel knowledge maps include interpolation, black-box neural networks, and traditional ray tracing. The advantage of the interpolation method is that it is simple and easy to use, and can quickly generate continuous channel knowledge maps. However, its disadvantage is that it depends on the quality and density of the measured data, and the accuracy of the prediction will decrease when the data is sparse or the environment changes greatly. The advantage of the black-box neural network is that it can capture the complex relationship between channel characteristics and environmental characteristics, and is suitable for modeling nonlinear and high-dimensional data. However, the disadvantage of this method is that the model is poorly interpretable and requires a large amount of training data. Secondly, traditional methods for constructing channel knowledge maps face many challenges. First, these methods usually rely on large-scale field measurements, which is not only time-consuming and labor-intensive, but also costly. Moreover, with changes in the environment, such as the reconstruction of buildings and the emergence of new obstacles, traditional channel knowledge maps need to be updated frequently, which further increases the complexity and cost of maintenance. In addition, traditional channel modeling methods are often difficult to adapt to dynamically changing environments, resulting in a decrease in the accuracy of channel prediction.

[0005] As for the channel knowledge map, ray tracing is the most accurate method. Ray tracing technology is a traditional method used in the field of wireless communications to simulate the propagation of electromagnetic waves in complex environments. It predicts the characteristics of wireless channels by simulating the propagation paths of electromagnetic waves, including direct radiation, reflection, refraction, and scattering. This method can provide a channel model close to the actual physical world, so it has been widely used in the planning, optimization, and performance evaluation of wireless communication systems. The core of ray tracing technology is to build a three-dimensional model that includes the geometric structure and material properties of the environment, and then use this model to calculate the interaction between electromagnetic waves and various objects in the environment. These interactions include reflection, refraction, and scattering of waves when encountering obstacles. Each interaction affects the amplitude, phase, and polarization state of the wave. By tracing all possible paths from the transmitter to the receiver, the response of the entire channel can be synthesized, providing important information for the design and analysis of wireless communication systems.

[0006] Although ray tracing technology can provide highly accurate channel models in theory, it still faces some challenges and limitations in practical applications. First, traditional ray tracing technology is often based on idealized physical models when simulating the interaction between electromagnetic waves and the environment. These models may not fully take into account the complexity and variability in the actual environment, resulting in deviations between the simulation results and the actual situation. In order to achieve accurate ray tracing simulations, detailed environmental parameters are required, including the geometric structure and electromagnetic properties of the objects. In practice, obtaining these parameters is time-consuming and expensive, especially for large or complex environments. In addition, traditional ray tracing models are often optimized for specific environments and lack the ability to migrate between different environments. This means that in a new environment, complex simulations and parameter adjustments may need to be performed from scratch. The real-world communication environment is constantly changing, and traditional ray tracing technology does not perform well in adapting to these changes. Any changes in the environment, such as the movement or addition of objects, the emergence of new materials, may affect the channel characteristics, and the model needs to be able to quickly adapt to these changes.

[0007] With the development of deep learning technology, neural networks have been introduced into wireless channel modeling to address the limitations of traditional methods. Neural networks are able to learn complex nonlinear relationships from data, providing new possibilities for channel modeling. In particular, the Neural Radiant Field (NeRF) technology has made breakthroughs in the field of computer vision, demonstrating the ability to synthesize new perspective images from sparse samples, which provides a new perspective for the modeling and prediction of wireless channels.

[0008] Therefore, inspired by the NeRF concept, the present invention proposes a channel knowledge map construction method and system for neural ray tracing. Summary of the invention

[0009] In view of the defects in the prior art, the object of the present invention is to provide a channel knowledge map construction method and system based on neural ray tracing.

[0010] A channel knowledge map construction method based on neural ray tracing provided by the present invention includes:

[0011] Step S1: collecting measured data of electromagnetic wave propagation by deploying channel detection equipment in a variety of typical environments;

[0012] Step S2: preprocessing the measured data;

[0013] Step S3: using the preprocessed data and material library to train a deep learning model, so that the deep learning model can learn the complex nonlinear relationship between the interaction between electromagnetic waves and the environment;

[0014] Step S4: Through the trained neural network-based ray tracing framework, after a given three-dimensional map and resolution, neural ray tracing is performed for each position, and the relevant channel parameters are predicted based on the obtained multipath-level channel information, and the channel knowledge map under the current environment is generated and set to appropriate values ​​to visualize it.

[0015] Preferably, the measured data is the basis for the model to learn the electromagnetic wave propagation characteristics in the actual environment, including detailed channel state information;

[0016] The channel state information includes a channel impulse response;

[0017] The preprocessing includes data cleaning, normalization and feature extraction to ensure data quality and prepare for subsequent model training;

[0018] The relevant channel parameters include channel gain, delay spread, and angle spread.

[0019] Preferably, in step S3, signal propagation is modeled by a trained neural network, and the steps include:

[0020] Step S3.1: Find all reachable paths and all interaction points between the ray and the environment through the traditional ray tracing algorithm;

[0021] Step S3.2: For each interaction of each path, the material number and the auxiliary information are input into the material information embedding network to encode the electromagnetic properties of the material and obtain the material information embedding vector;

[0022] Step S3.3: Input the material embedding vector and the incident wave information into the shared layer and the reflection / scattering loss prediction network to obtain the corresponding loss of this interaction;

[0023] Step S3.4: Sum up all interaction losses of each path to obtain the total path loss and the receiving power at the receiving end.

[0024] Preferably, the step S4 includes predicting a narrowband channel model of the multipath, and the narrowband channel model of the multipath is as follows:

[0025]

[0026] Where L represents the number of propagation paths, f represents the carrier frequency, and n r With n t Respectively represent the number of antenna units at the receiving end and the transmitting end, α l ∈C and τ l ∈R represent the amplitude loss and propagation delay of the lth path respectively;

[0027] According to the number of interactions between the lth path and the environment, from the transmitter to the receiver, the amplitude of the electric field gradually decreases due to the interaction with the object. Let E l,i represents the amplitude of the electric field of the i-th interactive incident wave, then the electric field amplitude at the receiving end is as follows:

[0028]

[0029] Among them, T l,i (·) represents the transfer function of the electric field of the ith interaction to the lth path, I l is the total number of interactions in the lth path.

[0030] Preferably, the method further includes a model evaluation and optimization step, wherein the prediction performance of the model is evaluated by comparing the channel knowledge map constructed by neural ray tracing with the actual channel knowledge map, and necessary optimization is performed;

[0031] The optimization includes adjusting the model structure, training strategy and loss function;

[0032] The optimization includes expressing the training objective of the neural network as an optimization problem of the following formula:

[0033]

[0034] Among them, ε represents the communication scenario, represents a sampling channel in this communication scenario, It represents the difference between the real channel and the predicted channel.

[0035] A channel knowledge map construction system based on neural ray tracing is provided according to the present invention, comprising:

[0036] Module M1: Collect measured data of electromagnetic wave propagation by deploying channel detection equipment in a variety of typical environments;

[0037] Module M2: preprocessing the measured data;

[0038] Module M3: Using the preprocessed data and material library, training a deep learning model so that the deep learning model can learn the complex nonlinear relationship between the interaction between electromagnetic waves and the environment;

[0039] Module M4: Through the trained neural network-based ray tracing framework, after a given 3D map and resolution, neural ray tracing is performed for each position, and the relevant channel parameters are predicted based on the obtained multipath-level channel information, generating a channel knowledge map in the current environment and setting appropriate values ​​to visualize it.

[0040] Preferably, the measured data is the basis for the model to learn the electromagnetic wave propagation characteristics in the actual environment, including detailed channel state information;

[0041] The channel state information includes a channel impulse response;

[0042] The preprocessing includes data cleaning, normalization and feature extraction to ensure data quality and prepare for subsequent model training;

[0043] The relevant channel parameters include channel gain, delay spread, and angle spread.

[0044] Preferably, in the module M3, signal propagation is modeled by a trained neural network, and the steps include:

[0045] Module M3.1: Find all reachable paths and all interaction points between rays and the environment through traditional ray tracing algorithms;

[0046] Module M3.2: For each interaction of each path, the material number and auxiliary information are input into the material information embedding network to encode the electromagnetic properties of the material and obtain the material information embedding vector;

[0047] Module M3.3: Input the material embedding vector and incident wave information into the shared layer and reflection / scattering loss prediction network to obtain the corresponding loss of this interaction;

[0048] Module M3.4: Sum up all interaction losses of each path to obtain the total path loss and the received power at the receiving end.

[0049] Preferably, the module M4 includes predicting a narrowband channel model of a multipath, and the narrowband channel model of a multipath is as follows:

[0050]

[0051] Where L represents the number of propagation paths, f represents the carrier frequency, and n r With n t Respectively represent the number of antenna units at the receiving end and the transmitting end, α l ∈C and τ l ∈R represent the amplitude loss and propagation delay of the lth path respectively;

[0052] According to the number of interactions between the lth path and the environment, from the transmitter to the receiver, the amplitude of the electric field gradually decreases due to the interaction with the object. Let E l,i represents the amplitude of the electric field of the i-th interactive incident wave, then the electric field amplitude at the receiving end is as follows:

[0053]

[0054] Among them, T l,i (·) represents the transfer function of the electric field of the ith interaction to the lth path, I l is the total number of interactions in the lth path.

[0055] Preferably, a model evaluation and optimization module is also included, which evaluates the prediction performance of the model by comparing the channel knowledge map constructed by neural ray tracing with the actual channel knowledge map, and performs necessary optimization;

[0056] The optimization includes adjusting the model structure, training strategy and loss function;

[0057] The optimization includes expressing the training objective of the neural network as an optimization problem of the following formula:

[0058]

[0059] Among them, ε represents the communication scenario, represents a sampling channel in this communication scenario, It represents the difference between the real channel and the predicted channel.

[0060] Compared with the prior art, the present invention has the following beneficial effects:

[0061] 1. Based on the actual channel measurement data, the present invention uses the excellent fitting ability of neural networks for nonlinear relationships to calculate the interaction effect between rays and the environment, including reflection and scattering. Compared with traditional ray tracing schemes, this method can better predict the electromagnetic properties of materials in actual scenes and has excellent migration capabilities.

[0062] 2. The neural ray tracing technology solution of the present invention effectively improves the accuracy and environmental adaptability of wireless channel modeling, reduces dependence on field measurements, and saves time and cost.

[0063] 3. The model of the present invention can quickly adapt to environmental changes and improve the reliability of channel prediction.

[0064] 4. The neural ray tracing method of the present invention enhances the model's ability to migrate between different environments, providing a powerful tool for the design, optimization and performance evaluation of wireless communication systems, and has important theoretical and practical significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0066] Figure 1 To associate the digital twin with the real scene for ray tracing;

[0067] Figure 2 The electric field changes as the signal interacts with the object in the ray tracing path.

[0068] Figure 3 The network architecture used by neural ray tracing to predict the outgoing electric field;

[0069] Figure 4 Taking concrete as an example, the prediction effect of neural ray tracing on reflection loss is demonstrated;

[0070] Figure 5 Taking concrete as an example, the prediction effect of neural ray tracing on scattering loss is demonstrated;

[0071] Figure 6 Comparison of the channel knowledge map constructed for neural ray tracing and the real channel knowledge map;

[0072] Figure 7 NMSE of the received power channel knowledge map constructed by different methods at different sampling rates;

[0073] Figure 8 The received power prediction effects of two neural ray tracing network structures for different types of paths. DETAILED DESCRIPTION

[0074] The present invention is described in detail below in conjunction with specific embodiments. The following embodiments will help those skilled in the art to further understand the present invention, but are not intended to limit the present invention in any form. It should be noted that, for those of ordinary skill in the art, several changes and improvements can also be made without departing from the concept of the present invention. These all belong to the protection scope of the present invention.

[0075] The present invention utilizes neural networks to learn the complex nonlinear relationship between electromagnetic waves and the environment to achieve high-precision channel knowledge map construction. By learning the electromagnetic properties of different materials in the environment, it can not only improve the accuracy of model predictions, but also enhance the model's ability to migrate between different environments and its adaptability to dynamically changing environments.

[0076] A channel knowledge map construction method based on neural ray tracing provided by the present invention includes:

[0077] Step S1: By deploying channel detection equipment in a variety of typical environments, the measured data of electromagnetic wave propagation are collected. The measured data include detailed channel state information, such as channel impulse response, etc., which are the basis for the model to learn the electromagnetic wave propagation characteristics in the actual environment.

[0078] Step S2: preprocessing the measured data, including data cleaning, normalization and feature extraction, to ensure data quality and prepare for subsequent model training.

[0079] Step S3: Use the preprocessed data and material library to train the deep learning model so that the deep learning model can learn the complex nonlinear relationship between electromagnetic waves and the environment. During the model training process, the present invention adopts advanced optimization algorithms and strategies to ensure that the model can effectively learn from the data, learn the electromagnetic properties of different materials in an end-to-end training framework, and accurately predict the channel characteristics. Through the trained neural network, we model the signal propagation in the following sub-steps:

[0080] Step S3.1: Using a traditional ray tracing algorithm, find all reachable paths and all interaction points between the ray and the environment.

[0081] Step S3.2: For each interaction of each path, the material number and auxiliary information are input into the material information embedding network to encode the electromagnetic properties of the material and obtain a material information embedding vector.

[0082] Step S3.3: Input the material embedding vector and the incident wave information into the shared layer and the reflection / scattering loss prediction network to obtain the corresponding loss of this interaction.

[0083] Step S3.4: Sum up all interaction losses of each path to obtain the total path loss and the receiving power at the receiving end.

[0084] Step S4: After a given three-dimensional map and resolution, neural ray tracing is performed at each position through a trained neural network-based ray tracing framework, and channel gain, delay spread, angle spread and other related channel parameters are predicted based on the obtained multipath-level channel information, and a channel knowledge map under the current environment is generated and set to appropriate values ​​to visualize it. Step S4 includes predicting the wireless channel, and the formula is as follows:

[0085]

[0086] in, represents a neural network that encodes the electromagnetic characteristics of the environment, where Θ ε represents the trainable network parameters, represents a neural network that computes interaction effects, represents the trainable network parameters, Represents all geometric information in the environment, Indicates the relevant hardware devices. Represents the reachable path information between the sender and receiver found by the ray tracing algorithm.

[0087] The step S4 includes predicting a narrowband channel model of the multipath, wherein the narrowband channel model of the multipath is as follows:

[0088]

[0089] Where L represents the number of propagation paths, f represents the carrier frequency, and n r With n t Respectively represent the number of antenna units at the receiving end and the transmitting end, α l ∈C and τ l ∈R represent the amplitude loss and propagation delay of the lth path respectively.

[0090] According to the number of interactions between the lth path and the environment, from the transmitter to the receiver, the amplitude of the electric field gradually decreases due to the interaction with the object. Let E l,i represents the amplitude of the electric field of the i-th interactive incident wave, then the electric field amplitude at the receiving end is as follows:

[0091]

[0092] Among them, T l,i (·) represents the transfer function of the electric field of the ith interaction to the lth path, I l is the total number of interactions in the lth path.

[0093] In addition, the present invention compares the channel knowledge map constructed by neural ray tracing with the actual channel knowledge map to evaluate the prediction performance of the model and perform necessary optimization. The optimization includes adjusting the model structure, training strategy and loss function to further improve the accuracy and robustness of the model. This step is iterative to ensure that the model can maintain optimal performance under various environments and conditions. The optimization includes expressing the training objective of the neural network as the following optimization problem:

[0094]

[0095] Among them, ε represents the communication scenario, represents a sampling channel in this communication scenario, It represents the difference between the real channel and the predicted channel.

[0096] The present invention aims to solve the challenges faced by existing ray tracing technology in wireless channel modeling, especially the problem of deviation between simulation results and actual conditions caused by idealized physical models in practical applications. When simulating the interaction between electromagnetic waves and the environment, the existing technology often fails to fully capture the complexity and variability in the actual environment, and the accurate acquisition of environmental parameters is time-consuming and expensive. In addition, traditional models lack the ability to migrate between different environments and have poor adaptability to dynamic environmental changes, which limits the applicability and accuracy of the model in diverse and changing environments. Therefore, the present invention seeks to propose a new wireless channel modeling method to improve the accuracy, adaptability and migration ability of the model, so that it can better reflect the electromagnetic wave propagation characteristics in the actual communication environment and effectively respond to the challenges brought by environmental changes.

[0097] Furthermore, the theoretical basis and implementation process of this solution are described in detail. In actual scenarios, the channel is composed of the communication environment ε, which includes the location, orientation, shape and electromagnetic characteristics of the base station, users and other objects, and the signal propagation law g(·) and related hardware equipment. (such as the antenna pattern of the transmitting and receiving ends, etc.). Therefore, the communication channel can be expressed as:

[0098]

[0099] The ray tracing method is a manifestation of the signal propagation law g(·). It uses the electromagnetic characteristics of the geometry and various interaction modes in the environment to find feasible propagation paths to simulate the channel between the transmitter and the receiver. The interaction modes include reflection, diffraction, and scattering. Ray tracing can calculate the loss, delay, and angle of each path. The relevant parameters of these paths are used to extract channel information. Further, ray tracing is divided into two parts: (1) path tracing Calculate all reachable paths between the sender and receiver, where is all the geometric information in the environment. (2) Loss calculation It is used to model the loss of electromagnetic waves on a given path and integrate the information of all paths into the transceiver channel.

[0100] However, it is very difficult to obtain accurate electromagnetic properties information ε of objects in the environment. First of all, the traditional method of measuring the electromagnetic properties of objects requires specific equipment and environment (such as in a darkroom). Therefore, it is often difficult to measure the electromagnetic properties of materials in actual scenarios. In addition, the commonly used electromagnetic parameters such as conductivity and magnetic permeability are often insufficient to accurately characterize the electromagnetic properties of materials (the electrical properties of materials are also related to ambient temperature, humidity, etc.).

[0101] At the same time, When modeling, traditional ray tracing often uses ready-made theoretical models, such as Fresnel's law to calculate reflection loss. However, these theoretical formulas often cannot accurately describe the effects of more complex interactions on radio wave propagation, such as scattering and diffraction. In addition, such calculations are more complicated when the electromagnetic properties of the material are highly nonlinear.

[0102] In response to the above challenges, the present invention uses a neural network to model the electromagnetic properties of various materials and the effects of various interaction modes on signal propagation. represents a neural network that encodes the electromagnetic properties of the environment, where Θ ε represents the trainable network parameters. A neural network representing a computational interaction effect can predict the corresponding loss based on the electromagnetic properties of the corresponding material (represented by the material embedding vector). Based on the modeling of the electromagnetic properties encoded by the neural network and the corresponding interaction effects, the wireless channel can be predicted:

[0103]

[0104] Therefore, the training objective of the neural network of the present invention can be expressed as the following optimization problem:

[0105]

[0106] Where ε represents the communication scenario, represents a sampling channel in this communication scenario, Represents the difference between the real channel and the predicted channel.

[0107] Specifically, the present invention considers a multipath narrowband channel model, and the channel can be written as:

[0108]

[0109] Where L represents the number of propagation paths, f represents the carrier frequency, and n r With n t Represents the number of antenna units at the receiving end and the transmitting end, α l ∈C and τ l ∈R represent the amplitude loss and propagation delay of the lth path respectively. The present invention can be achieved by the electric field E on the lth path. l Iterations are used to characterize the amplitude loss α of the path. l .exist Figure 2 In the present invention, it is assumed that the kth path has two interactions with the environment. From the transmitting end to the receiving end, the amplitude of the electric field gradually decreases due to the interaction with the object. The electric field is as follows:

[0110]

[0111] in Represent two orthogonal polarization directions. s ,E p ∈C represent the amplitude components of the electric field E in these two directions. Let E l,i represents the amplitude of the electric field of the incident wave in the i-th interaction, then the electric field of the i+1-th interaction (or the electric field strength at the receiver end) is:

[0112] E l,i+1 =T l,i (E l,i )

[0113] (6) Where T l,i (·) represents the transfer function of the electric field of the ith interaction to the lth path. Therefore, the electric field at the receiver is expressed as:

[0114]

[0115] Among them I l is the total number of interactions in the lth path. As shown above, the total transfer function of the electric field in the lth path can be written as a cascade of several interaction transfer functions. Specifically, T l,i Depending on the type of interaction (reflection, scattering, etc.), the electromagnetic properties of the material at the interaction point and the geometric relationship of the electromagnetic wave path relative to the interaction surface. Come to E l,i Calculate, that is:

[0116]

[0117] Substituting (6), (7), (8) into (4), the present invention can train the network by solving the following optimization problem:

[0118]

[0119] Figure 3 Shown In every interaction, E l,i Predicted emission electric field E l,i+1 The neural network architecture used. The network for predicting the loss of the shared intermediate layer and reflection / scattering is based on encoding the electromagnetic properties of the material, and then further calculates the loss caused by this interaction on the radio wave. The specific training is done through equation (9), that is, it is hoped that the electric field strength received by the final receiver predicted by the network is consistent with the actual situation. Figure 3 As shown, the electric field prediction for each interaction can be calculated as follows:

[0120] The first step is to select the one-hot encoding vector of the corresponding material according to the location of the interaction point.

[0121] In the second step, based on the auxiliary information (carrier frequency, environmental climate, etc.), the material information embedding network is used to obtain the material information embedding vector of the material in the current communication environment.

[0122] The third step is to combine the incident direction, incident polarization mode, interaction surface normal vector, and obtain the intermediate vector through the shared intermediate layer of scattering and reflection.

[0123] The fourth step is to determine whether there is a LoS path between the interaction point and the receiving end to decide whether to use the reflection loss prediction network or the scattering loss prediction loss, and calculate the outgoing electric field (if it is scattering, add the offset angle of the scattering direction compared to the reflection direction as input).

[0124] By using the neural network to calculate the attenuation of each interaction electric field, combined with By searching all reachable paths, we can use (2) to obtain the channel predicted by neural ray tracing In actual scenarios, the present invention can obtain the actual channel through measurement By minimizing the difference between the predicted channel and the actual channel, the neural network of the present invention can learn end-to-end.

[0125] Figure 4Taking concrete as an example, the performance of the neural network model developed in this study in predicting the reflection loss of concrete materials in parallel and perpendicular polarization directions at different incident angles is presented. The model uses a complex network architecture combining multilayer perceptron (MLP) and Transformer for training and prediction under the conditions of a carrier frequency of 3.6GHz and a sampling rate of 15%. By comparing the model prediction values ​​with the true values ​​measured in the experiment, it can be clearly observed that the neural network model has a high accuracy in simulating the electromagnetic wave reflection characteristics of concrete materials. In the parallel polarization state, the model prediction values ​​and the true values ​​maintain good consistency over a wide range of incident angles. However, near the Brewster angle (about 67 degrees), there is a certain degree of deviation between the model prediction values ​​and the true values. This phenomenon may be due to the fact that at the Brewster angle, the electromagnetic properties of concrete materials undergo a sharp change, resulting in a decrease in the prediction accuracy of the model at this specific angle. In addition, the comparative analysis of the predicted values ​​and the true values ​​in the figure further verifies the stability and reliability of the model in dealing with the reflection characteristics of concrete materials. Although there are certain prediction errors at specific angles, overall the prediction performance of the model is quite outstanding, which has important theoretical and practical significance for the study of the propagation characteristics of electromagnetic waves in various materials and related engineering applications.

[0126] Figure 5 Taking concrete as an example, the performance of the neural network model developed in this study in predicting the scattering loss prediction of concrete materials at different scattering offset angles when the incident angle is 45 degrees is demonstrated. This simulation also combines the advantages of multi-layer perceptron (MLP) and Transformer at a carrier frequency of 3.6GHz and a sampling rate of 15%, forming a complex network architecture, and through a large amount of data training, in order to achieve accurate simulation of the scattering characteristics of concrete materials. It can be clearly seen from the figure that the scattering loss predicted by the model (gray points) and the actual measured scattering loss (black points) show a high degree of consistency at most scattering offset angles, which shows that the model can effectively capture the scattering behavior of concrete materials at specific incident angles. With the increase of the scattering offset angle, the electric field intensity of the scattered wave gradually weakens. This trend is accurately reflected in the model prediction, which further confirms the effectiveness of the model in simulating the scattering characteristics of concrete materials. In summary, the neural network model proposed in this study shows high accuracy and reliability in predicting the scattering loss of concrete materials. It can not only provide strong theoretical support for the analysis of electromagnetic properties of various materials, but also has important practical significance for the prediction of electromagnetic wave propagation in related engineering applications. Future research can further optimize the model on this basis to improve its applicability and prediction accuracy under a wider range of conditions.

[0127] Figure 6Two sub-figures (a) and (b) show the channel knowledge maps of user received power in a certain building environment constructed by the neural ray tracing technology proposed in this study under the condition of a sampling rate of 15%. These maps reflect the signal strength that users at different locations may receive under a specific transmitter position. Figure (a) represents the channel knowledge map constructed by the neural ray tracing technology, while Figure (b) represents the real channel knowledge map obtained by actual measurement. By comparing the two figures, the present invention can observe that there is a significant similarity between the map constructed by the neural ray tracing technology and the real measurement results. This similarity shows that the neural ray tracing technology can effectively simulate and predict the propagation characteristics of electromagnetic waves in the building environment, including direct radiation, reflection and occlusion. In Figures (a) and (b), the black area represents the area with low signal received power, which is usually caused by the occlusion of the building structure. The gray to white gradient area represents the area where the signal received power gradually increases. Through this color coding, the present invention can intuitively see that the neural ray tracing technology can not only identify the area where the signal is occluded, but also accurately depict the path and intensity changes of the signal propagation under reflection and diffraction. In addition, the construction of this channel knowledge map is of great significance for the design and optimization of wireless communication systems. It can help engineers better understand the propagation behavior of signals in actual environments, so as to make more reasonable decisions in network planning, base station deployment and signal coverage optimization.

[0128] Figure 7The changing trends of the normalized root mean square error (NMSE) of four methods for constructing the received power channel knowledge map at different sampling rates are shown. As the sampling rate increases from 0.05 to 0.30, the NMSE of all methods shows a downward trend, which indicates that increasing the sampling rate helps to reduce the prediction error and improve the accuracy of channel knowledge map construction. Under low sampling rate (0.05) conditions, RadioUNet showed the best received power prediction effect. This is because RadioUNet maintained a certain spatial understanding ability during the training process. Even if there is no measurement data or the measurement data is small, it still has a good channel knowledge map construction ability. Under low sampling rate conditions, the NeRT method based on MLP+Transformer shows a lower NMSE, and its advantage becomes more obvious as the sampling rate increases. Finally, it maintains the lowest NMSE among all methods, indicating that this method has good performance and generalization ability when dealing with channel knowledge map construction tasks. In addition, the NeRT based on MLP+Transformer shows more obvious superiority than the NeRT based on MLP when the sampling rate is low, while this superiority weakens when the sampling rate is high, which shows that the addition of the attention mechanism is significantly helpful for the network to build a channel knowledge map when the samples are insufficient. The NMSE of RadioUNet and the NeRT method based on MLP also gradually decreases with the increase of the sampling rate, showing a certain prediction accuracy, but the overall performance is slightly inferior to the NeRT method based on MLP+Transformer. The Kriging Interpolation method can also achieve a low NMSE at a high sampling rate, but its error is relatively high at a low sampling rate. In summary, the NeRT method based on MLP+Transformer can maintain a low NMSE at different sampling rates, showing its superior performance in channel knowledge map construction. This result provides an important reference for selecting a suitable channel knowledge map construction method in practical applications, especially when sampling resources are limited, the NeRT method based on MLP+Transformer may be a more reliable choice.

[0129] Figure 8The normalized root mean square error (NMSE) performance of the models with two network structures in predicting various path types is shown. These path types are distinguished based on the number of reflections and scatterings that the electromagnetic wave experiences during propagation. The figure includes models using only multilayer perceptron (MLP) and combining multilayer perceptron with Transformer (MLP+Transformer), and each model considers different sampling rates (10%, 20%, 30%). It can be observed from the figure that for a single reflection path, the NMSE of all models is relatively low, indicating that the model has good prediction performance when dealing with simple reflection paths. As the number of reflections increases, especially for three and four reflection paths, the NMSE of the MLP model increases significantly, showing a decrease in performance when dealing with complex reflection paths. In contrast, the MLP+Transformer model shows lower NMSE on all path types, especially when dealing with three and four reflection paths, where its performance advantage is more obvious. This may be because the Transformer structure can effectively capture long-range dependencies, thereby providing better performance in the prediction of complex paths. For scattered paths, the NMSE of the MLP model is relatively high, while the MLP+Transformer model is able to significantly reduce this error, especially at a sampling rate of 30%, where its NMSE is almost close to 0, showing superior performance when dealing with scattered paths. Overall, the model combining MLP and Transformer shows better prediction performance when dealing with various complex paths, especially under high sampling rate conditions. This result emphasizes the importance of considering model complexity and sampling rate in channel modeling, as well as the potential of the MLP+Transformer model in improving prediction accuracy.

[0130] The present invention also provides a channel knowledge map construction system based on neural ray tracing. The channel knowledge map construction system based on neural ray tracing can be implemented by executing the process steps of the channel knowledge map construction method based on neural ray tracing, that is, those skilled in the art can understand the channel knowledge map construction method based on neural ray tracing as a preferred implementation of the channel knowledge map construction system based on neural ray tracing.

[0131] A channel knowledge map construction system based on neural ray tracing is provided according to the present invention, comprising:

[0132] Module M1: By deploying channel detection equipment in a variety of typical environments, the measured data of electromagnetic wave propagation are collected. The measured data is the basis for the model to learn the electromagnetic wave propagation characteristics in the actual environment, including detailed channel state information. The channel state information includes the channel impulse response.

[0133] Module M2: Preprocessing the measured data, including data cleaning, normalization and feature extraction, to ensure data quality and prepare for subsequent model training.

[0134] Module M3: Use the preprocessed data and material library to train the deep learning model so that the deep learning model can learn the complex nonlinear relationship between the interaction between electromagnetic waves and the environment. In the module M3, the signal propagation is modeled through the trained neural network, including: Module M3.1: Use the traditional ray tracing algorithm to find all reachable paths and all interaction points between the rays and the environment. Module M3.2: For each interaction of each path, the material number and auxiliary information are input into the material information embedding network to encode the electromagnetic properties of the material and obtain the material information embedding vector. Module M3.3: The material embedding vector and the incident wave information are input into the shared layer and the reflection / scattering loss prediction network to obtain the corresponding loss of this interaction. Module M3.4: All interaction losses of each path are summarized to obtain the total path loss and the receiving power at the receiving end.

[0135] Module M4: Through the trained neural network-based ray tracing framework, after a given three-dimensional map and resolution, neural ray tracing is performed for each position, and the relevant channel parameters are predicted based on the obtained multipath-level channel information, and the channel knowledge map under the current environment is generated and set to appropriate values ​​to visualize it. The relevant channel parameters include channel gain, delay extension, and angle extension. The module M4 includes predicting the narrowband channel model of the multipath, and the narrowband channel model of the multipath is as follows:

[0136]

[0137] Where L represents the number of propagation paths, f represents the carrier frequency, and n r With n t Respectively represent the number of antenna units at the receiving end and the transmitting end, α l ∈C and τ l ∈R represent the amplitude loss and propagation delay of the lth path respectively. According to the number of interactions between the lth path and the environment, from the transmitter to the receiver, the amplitude of the electric field gradually decreases due to the interaction with the object. Let E l,i represents the amplitude of the electric field of the i-th interactive incident wave, then the electric field amplitude at the receiving end is as follows:

[0138]

[0139] Among them, T l,i (·) represents the transfer function of the electric field of the ith interaction to the lth path, I l is the total number of interactions in the lth path.

[0140] The present invention also includes a model evaluation and optimization step, by comparing the channel knowledge map constructed by neural ray tracing with the actual channel knowledge map, evaluating the prediction performance of the model, and performing necessary optimization. The optimization includes adjusting the model structure, training strategy and loss function. The optimization includes expressing the training objective of the neural network as an optimization problem of the following formula:

[0141]

[0142] Among them, ε represents the communication scenario, represents a sampling channel in this communication scenario, It represents the difference between the real channel and the predicted channel.

[0143] Those skilled in the art know that, in addition to realizing the system and its various devices, modules, and units provided by the present invention in a purely computer-readable program code, it is entirely possible to realize the same functions in the form of logic gates, switches, application-specific integrated circuits, programmable logic controllers, and embedded microcontrollers by logically programming the method steps. Therefore, the system and its various devices, modules, and units provided by the present invention can be considered as a hardware component, and the devices, modules, and units included therein for realizing various functions can also be regarded as structures within the hardware component; the devices, modules, and units for realizing various functions can also be regarded as both software modules for realizing the method and structures within the hardware component.

[0144] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art can make various changes or modifications within the scope of the claims, which does not affect the essence of the present invention. In the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other arbitrarily.

Claims

1. A channel knowledge map construction method based on neural ray tracing, characterized in that: include: Step S1: collecting measured data of electromagnetic wave propagation; Step S2: preprocessing the measured data; Step S3: using the preprocessed data and material library to train a deep learning model, so that the deep learning model can learn the complex nonlinear relationship between the interaction between electromagnetic waves and the environment; Step S4: Through the trained neural network-based ray tracing framework, after a given three-dimensional map and resolution, neural ray tracing is performed for each position, and the relevant channel parameters are predicted based on the obtained multipath-level channel information, and the channel knowledge map under the current environment is generated and set to appropriate values ​​to visualize it.

2. The method for constructing a channel knowledge map based on neural ray tracing according to claim 1, characterized in that: The measured data is the basis for the model to learn the electromagnetic wave propagation characteristics in the actual environment, including detailed channel state information; The channel state information includes a channel impulse response; The preprocessing includes data cleaning, normalization and feature extraction to ensure data quality and prepare for subsequent model training; The relevant channel parameters include channel gain, delay spread, and angle spread.

3. The method for constructing a channel knowledge map based on neural ray tracing according to claim 1, characterized in that: In step S3, signal propagation is modeled using a trained neural network, and the steps include: Step S3.1: Find all reachable paths and all interaction points between the ray and the environment through the traditional ray tracing algorithm; Step S3.2: For each interaction of each path, the material number and the auxiliary information are input into the material information embedding network to encode the electromagnetic properties of the material and obtain the material information embedding vector; Step S3.3: Input the material embedding vector and the incident wave information into the shared layer and the reflection / scattering loss prediction network to obtain the corresponding loss of this interaction; Step S3.4: Sum up all interaction losses of each path to obtain the total path loss and the receiving power at the receiving end.

4. The method for constructing a channel knowledge map based on neural ray tracing according to claim 1, characterized in that: The step S4 includes predicting a narrowband channel model of the multipath, wherein the narrowband channel model of the multipath is as follows: Where L represents the number of propagation paths, f represents the carrier frequency, and n r With n t Respectively represent the number of antenna units at the receiving end and the transmitting end, α l ∈C and τ l ∈R represent the amplitude loss and propagation delay of the lth path respectively; According to the number of interactions between the lth path and the environment, from the transmitter to the receiver, the amplitude of the electric field gradually decreases due to the interaction with the object. Let E l,i represents the amplitude of the electric field of the i-th interactive incident wave, then the electric field amplitude at the receiving end is as follows: Among them, T l,i (·) represents the transfer function of the electric field of the ith interaction to the lth path, I l is the total number of interactions in the lth path.

5. The method for constructing a channel knowledge map based on neural ray tracing according to claim 1, characterized in that: It also includes model evaluation and optimization steps, which compares the channel knowledge map constructed by neural ray tracing with the actual channel knowledge map to evaluate the prediction performance of the model and make necessary optimizations; The optimization includes adjusting the model structure, training strategy and loss function; The optimization includes expressing the training objective of the neural network as an optimization problem of the following formula: Among them, ε represents the communication scenario, represents a sampling channel in this communication scenario, It represents the difference between the real channel and the predicted channel.

6. A channel knowledge map construction system based on neural ray tracing, characterized in that: include: Module M1: Collect measured data on electromagnetic wave propagation; Module M2: preprocessing the measured data; Module M3: Using the preprocessed data and material library, training a deep learning model so that the deep learning model can learn the complex nonlinear relationship between the interaction between electromagnetic waves and the environment; Module M4: Through the trained neural network-based ray tracing framework, after a given 3D map and resolution, neural ray tracing is performed for each position, and the relevant channel parameters are predicted based on the obtained multipath-level channel information, generating a channel knowledge map in the current environment and setting appropriate values ​​to visualize it.

7. The channel knowledge map construction system based on neural ray tracing according to claim 1 is characterized in that: The measured data is the basis for the model to learn the electromagnetic wave propagation characteristics in the actual environment, including detailed channel state information; The channel state information includes a channel impulse response; The preprocessing includes data cleaning, normalization and feature extraction to ensure data quality and prepare for subsequent model training; The relevant channel parameters include channel gain, delay spread, and angle spread.

8. The channel knowledge map construction system based on neural ray tracing according to claim 6 is characterized in that: In the module M3, signal propagation is modeled by a trained neural network, including: Module M3.1: Find all reachable paths and all interaction points between rays and the environment through traditional ray tracing algorithms; Module M3.2: For each interaction of each path, the material number and auxiliary information are input into the material information embedding network to encode the electromagnetic properties of the material and obtain the material information embedding vector; Module M3.3: Input the material embedding vector and incident wave information into the shared layer and reflection / scattering loss prediction network to obtain the corresponding loss of this interaction; Module M3.4: Sum up all interaction losses of each path to obtain the total path loss and the received power at the receiving end.

9. The channel knowledge map construction system based on neural ray tracing according to claim 6, characterized in that: The module M4 includes predicting a narrowband channel model of a multipath, wherein the narrowband channel model of the multipath is as follows: Where L represents the number of propagation paths, f represents the carrier frequency, and n r With n t Respectively represent the number of antenna units at the receiving end and the transmitting end, α l ∈C and τ l ∈R represent the amplitude loss and propagation delay of the lth path respectively; According to the number of interactions between the lth path and the environment, from the transmitter to the receiver, the amplitude of the electric field gradually decreases due to the interaction with the object. Let E l,i represents the amplitude of the electric field of the i-th interactive incident wave, then the electric field amplitude at the receiving end is as follows: Among them, T l,i (·) represents the transfer function of the electric field of the ith interaction to the lth path, I l is the total number of interactions in the lth path.

10. The channel knowledge map construction system based on neural ray tracing according to claim 6, characterized in that: It also includes a model evaluation and optimization module, which evaluates the prediction performance of the model and performs necessary optimization by comparing the channel knowledge map constructed by neural ray tracing with the actual channel knowledge map; The optimization includes adjusting the model structure, training strategy and loss function; The optimization includes expressing the training objective of the neural network as an optimization problem of the following formula: Among them, ε represents the communication scenario, represents a sampling channel in this communication scenario, It represents the difference between the real channel and the predicted channel.

Images