A scene adaptive channel modeling method based on 6g full coverage scene classification

Abstract

The application discloses a scene adaptive channel modeling method based on 6G full coverage scene classification. By optimizing the channel model parameters using measurement or simulation data, a new architecture design and adaptive channel modeling for 6G full coverage scene conversion are realized. Including 1) based on the physical definition of 6G space-air-ground-sea full coverage scene, using the combination scheme of environmental parameters and statistical characteristics for scene conversion classification. 2) according to the scene classification result, a scene model parameter database is constructed. 3) input the model parameters corresponding to the scene into the 6G universal channel model to generate the real-time channel model of scene conversion. 4) according to the measured data, the parameters of the universal channel model are adjusted by adopting closed-loop feedback, and the optimized model of scene conversion is obtained by iteration. 5) verify the accuracy of specific scene adaptive channel modeling. For complex and diverse 6G communication environment, adaptive construction of high-precision channel model lays an important foundation for wireless communication research and development.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, and in particular relates to a scenario-adaptive channel modeling method based on 6G full coverage scenario classification. Background Technology

[0002] Driven by artificial intelligence and big data, 6G wireless communication will construct the nerve center connecting the physical and digital worlds, ushering in an era of ubiquitous sensing, interconnection, and intelligence. Integrated sensing based on 6G wireless communication and artificial intelligence will become key core technologies for this era of ubiquitous sensing, interconnection, and intelligence.

[0003] The network-level communication performance of 6G full coverage scenarios places higher demands on channel real-time performance and accuracy. However, the scenario categories defined by traditional channel standards (such as 3GPP and WINNER) are relatively coarse, failing to accurately, quickly, and effectively define many real-world application scenarios. They rely solely on subjective human judgment to broadly classify scenarios based on their characteristics, lacking universality and intelligence. Many standardized channel scenario models typically include no more than 18 scenario categories, while actual channel applications involve far more than this number. For example, vertical industries such as mining, manufacturing, public security, and transportation have different requirements for corresponding mobile communication user-level performance (rate, latency, etc.). Therefore, traditional standardized channel models can no longer accurately and in real-time describe the channel characteristics of different specific scenarios. Consequently, research on wireless channel modeling based on new scenario classification rules and frameworks has become the cornerstone of 6G wireless communication and sensing integration.

[0004] The requirements for 6G full-coverage communication networks differ significantly from previous communication methods, exhibiting three main characteristics: First, the need for seamless communication scenario transitions. This demands extremely high efficiency when switching between various communication environments. For the same person at different times, they might be in a low-scattering environment one minute, but in a densely populated, high-scattering environment or a high-density network environment the next. This stark contrast requires the channel to complete network and signal conversion within a limited timeframe. Second, the need for high-density communication scenarios. 6G network communication terminals and base stations will be more efficient and integrated. Third, the need for mobile communication scenarios. 6G networks will not only connect extensively to various mobile terminals but also maintain real-time data sharing with personal cloud platforms. In summary, for these new 6G scenarios, how can we select appropriate channel models for effective fitting, how should channel parameters be set within these models, and can scenario parameter matching be automated?

[0005] 6G full coverage encompasses scenarios such as terrestrial mobile communication, satellite communication, drone communication, and marine communication. Among these, factors such as Doppler frequency shift due to rapid satellite movement, rainfall fading, and ionospheric effects, as well as the impact of drone altitude on arbitrary 3D motion and large-scale parameters, all need to be considered. In terms of full application, the Doppler frequency shift and time-domain non-stationary characteristics of channels in vehicle-to-everything (V2X) applications at speeds exceeding 500 km / h, the vacuum tube waveguide effect in ultra-high-speed trains operating in vacuum tubes, the spherical wave characteristics and spatial non-stationary characteristics of VMI (Very Large Scale Infrared) antenna array channels, the distribution and mobility of ultra-dense scatterers in industrial IoT channels, and the accurate modeling of wireless channels using reconfigurable smart surface technology are all areas requiring new consideration and research.

[0006] Considering that the mixed application of various new technologies will lead to the cross-application of channel characteristics in 6G full coverage scenarios, a major challenge in 6G full coverage scenario channel modeling is how to comprehensively consider a variety of new channel characteristics and propose an adaptive channel model suitable for full coverage scenarios. Summary of the Invention

[0007] The purpose of this invention is to provide a scenario-adaptive channel modeling method based on 6G full coverage scenario classification, so as to solve the technical problems of accuracy, complexity and real-time performance of wireless channel modeling in 6G full coverage scenarios.

[0008] To solve the above-mentioned technical problems, the specific technical solution of the present invention is as follows:

[0009] An adaptive channel modeling method based on 6G full coverage scenario classification includes the following steps:

[0010] S1. Based on the physical parameter definition of the 6G full coverage scenario, and according to the influence of different features on the classification of multiple scenarios, a combination scheme of physical environment and statistical features is selected for cross-scenario classification.

[0011] S2. Construct a 6G full coverage cross-scenario model database based on the cross-scenario classification results, including physical scenarios and their corresponding model statistical characteristic parameter values, such as root mean square delay spread, root mean square angle spread, etc.

[0012] S3. Based on the 6G full coverage cross-scenario model database constructed in S2, and according to the results of cross-scenario classification and identification, input the corresponding model statistical characteristic parameter values ​​into the universal channel model to generate a channel model for a single scenario or cross-fusion scenario.

[0013] S4. For the 6G full coverage scenario, the statistical characteristics of the model in S2 are logically operated on with the statistical characteristic parameter values ​​obtained from wireless channel simulation or measured data. The parameter value weights of the statistical characteristics of the universal channel model are adjusted according to the closed-loop feedback, and the optimized scenario adaptive channel model is obtained iteratively.

[0014] S5. Simulation was used to verify and confirm the accuracy of the adaptive channel modeling for this scenario.

[0015] Furthermore, the specific steps of step S1 are as follows:

[0016] Step S101: The classification of 6G full coverage scenarios follows the principle of macro to micro. First, they are divided into four major categories: air, space, land, and sea. Then, they are further divided according to the layer height or depth. For scenarios at the same layer, they are further subdivided according to the actual physical environment and signal propagation characteristics. For each subdivided scenario, detailed physical parameter definitions are given, including length, width, height, moving speed, transmit and receive distance range, and line-of-sight or non-line-of-sight propagation.

[0017] Step S102: By evaluating the impact of various scene environmental feature parameters on classification performance, in addition to achieving classification at line of sight and non-line of sight, as well as indoor and outdoor, multi-feature parameter combination is used to classify multi-target scenes. This is a classification method using environmental physical parameters.

[0018] Step S103: Using three methods—original data, probability distribution shape, and threshold classification—statistical characteristic classification is performed on the 6G full coverage scenario. The statistical characteristic parameters corresponding to the channel of each scenario are calculated in detail, and the parameter intervals are divided to classify the scenarios.

[0019] The time-domain channel impulse response (CIR) and frequency-domain channel transfer function (CTF) contain rich information in their raw data feature parameters.

[0020] First, various scenarios are classified based on the shape of the probability distribution. The characteristic parameters for classifying scenarios based on the probability distribution shape include the probability density function, the number of peaks in the time-delay power spectral density, skewness, and kurtosis / peak value. Considering that line-of-sight scenarios follow a Ricean distribution while non-line-of-sight scenarios follow a Rayleigh distribution, the probability density function is used to distinguish between line-of-sight and non-line-of-sight scenarios. The number of peaks in the time-delay power spectral density is used to distinguish rich scattering environments; peaks with different time delays represent multipath propagation from different scatterers. Skewness indicates the degree of asymmetry in the probability distribution; the skewness of line-of-sight scenarios is less than that of non-line-of-sight scenarios. Kurtosis indicates the steepness of the probability distribution; the kurtosis in the line-of-sight case is greater than that in the non-line-of-sight case.

[0021] Secondly, threshold comparisons are used for classification. Path loss (PL) is an important indicator of large-scale fading, influenced by transmission distance, and is used to classify communication scenarios with significantly different propagation distances. The Rice K-factor represents the ratio of line-of-sight path power to non-line-of-sight path power. Additionally, some time-domain, spatial-domain, and frequency-domain related characteristic parameters can also be used for multi-scenario classification. For example, in time-varying scenarios with high mobility, the time autocorrelation function decreases rapidly, and the stationary interval is small. The root mean square delay spread reflects the dispersion of the wireless channel in the time delay domain. For scenarios with abundant scatterers, this value is usually large.

[0022] Step S104: Using a combination of classification methods, namely S102 environmental physical parameter classification and S103 statistical characteristic classification, output the classification results for the 6G full coverage scenario.

[0023] Furthermore, the specific steps of step S2 are as follows:

[0024] Step S201: Perform channel measurement under scenario and parameter configuration to improve the diversity of the wireless channel database; construct aerospace coverage scenarios including satellites, drones, etc., outdoor scenarios such as typical cities, suburbs, vehicle-to-vehicle, high-speed rail, etc., and typical indoor scenarios such as office areas, corridors, residential areas, shopping malls, industrial IoT, etc.

[0025] Step S202: Using Wireless Insite software supplemented by ray tracing simulation data, the system configuration parameters are set by reconstructing the real propagation environment, generating simulation path parameters and channel impulse response, further increasing the number of channel model database samples for each communication scenario, and reducing the cost and complexity of actual measurement.

[0026] Step S203: Optimize and adjust the measurement and simulation data respectively to build a 6G full coverage scenario model database; the database contains the basic physical parameters of various scenarios, the calculated characteristic parameters, and the complete channel impulse response;

[0027] Step S204: Input the scene classification results from step S1 into the scene model database and output the statistical characteristic parameter set of the corresponding scene.

[0028] Furthermore, the specific steps of step S3 are as follows:

[0029] Step S301: Based on the 6G ubiquitous channel model, perform model statistical parameter matching according to the scenario classification results in S2; the parameters in the 6G ubiquitous channel model used here include user-defined parameters and wireless channel-related parameters; among them, user-defined parameters include the center frequency / 2 wavelength and bandwidth of the system configuration, array configuration-related parameters, and moving speed and trajectory parameters; the latter includes large-scale fading-related parameters, scattering clusters, and multipath-related parameters;

[0030] The channel matrix of the 6G ubiquitous geometric random channel model is represented as follows:

[0031] H = [PL·SH·BL·WE·AL] 1 / 2 ·H s

[0032] Where PL, SH, BL, WE, and AL represent large-scale fading, PL represents path loss, SH represents shadowing fading, BL represents blocking effect, AL represents atmospheric absorption loss, WE represents weather-related loss, and H represents path loss. s This is a small-scale fading phenomenon;

[0033] The small-scale fading H s as follows:

[0034]

[0035] Among them, M T M represents the number of antenna elements in the transmitting antenna array. R This refers to the number of antenna elements in the receiver's antenna array. For transmitting antenna array elements With receiving antenna array elements The channel impulse response between them is expressed as the line-of-sight LoS component. With non-line-of-sight NLoS components Superposition:

[0036]

[0037] Among them, K R (t) is the Rice factor. and They are as follows:

[0038]

[0039]

[0040] in,{*} T f represents transpose c Indicates the carrier frequency. and Antenna elements in different frequency bands Corresponding to the radiation patterns of vertical and horizontal polarization, The cross-polarization power ratio is given by μ, which characterizes the joint polarization imbalance. and It is from time t to The azimuth departure angle and pitch departure angle corresponding to the LosS path. and It is from time t to The azimuth and pitch angles of arrival corresponding to the LosS path. and It is a random phase that follows a uniform distribution in (0, 2π]. ψ l,m =108 / f c 2 f is the Faraday rotation angle, which is calculated here. c The unit is GHz. Under NLoS conditions arrive The power of the m-th sub-path in the n-th path, It is the delay of the Loss path at time t. At time t, the transmitting antenna array element With receiving antenna array elements The vector distance between them, where c is the speed of light; At time t and The time delay of the m-th sub-path of the n-th path between them. At time t and The power of the m-th sub-path of the n-th path between them; all of the above parameters are time-varying parameters.

[0041] Step S302: Substitute the statistical characteristic parameter data of the scene classification in step S2 into the universal channel model to obtain a channel model suitable for scene switching, i.e., the channel impulse response function or the channel transmission matrix; set the number of base stations to N. BS The number of users is N MS The channel transmission matrix of the multi-link channel model in scene transition is shown in the following formula:

[0042]

[0043] Each link corresponds to

[0044] Furthermore, the specific steps of step S4 are as follows:

[0045] Step S401: Directly calculate the relevant statistical characteristics of the channel using measured or simulated data from typical scenarios, including the time autocorrelation function, spatial cross-correlation function, and root mean square delay / angle spread.

[0046] Step S402: Based on the channel impulse function or channel transmission matrix output in step S302, calculate the channel correlation statistics and make a logical comparison with the results obtained from actual measurement or simulation in S401.

[0047] Step S403: Adaptively adjust the statistical characteristic parameter matrix of the classification based on the feedback results, and obtain an optimized scenario-switching channel model through feedback and iteration in different scenarios.

[0048] Furthermore, the specific steps of step S5 are as follows:

[0049] Step S501: Based on the iteratively optimized scene adaptive channel model, verify the performance of the integrated mechanism of channel scene classification and model parameter matching; generate channel data under several scenarios by measurement or simulation, and classify the scenarios by combining the physical environment parameters and statistical characteristics in step S1.

[0050] Step S502: Based on the scene statistical characteristics obtained from the classification, adaptively match the parameter matrix of the universal channel model, and logically compare the statistical quantities obtained from the model simulation with the results calculated from the measured data.

[0051] Step S503: By performing steps S501 and S502 multiple times, verify the accuracy of the integrated adaptive modeling solution based on 6G full coverage scenario classification.

[0052] This invention presents a scenario-adaptive channel modeling method based on 6G full coverage scenario classification, which has the following advantages: By constructing a comprehensive and clear classification architecture for 6G full coverage scenarios, applying a scenario feature combination classification method, and optimizing a universal channel model using measurement and simulation data, this invention proposes a scenario-adaptive channel modeling method based on 6G full coverage scenario classification. This method can adaptively establish scenario-transition channel models, reducing the complexity of traditional channel modeling and laying an important foundation for future research on wireless communication systems. Attached Figure Description

[0053] Figure 1 This is a schematic diagram of the process of the present invention;

[0054] Figure 2 This is a diagram illustrating the adaptive modeling process and parameter matching of the present invention.

[0055] Figure 3 This is a comparison chart of the root mean square delay spread between model simulation and measurement in a typical industrial IoT scenario of this invention.

[0056] Figure 4 This is a comparison chart of the voltage level passthrough rate between the model simulation and measurement in a typical high-speed rail scenario according to the present invention;

[0057] Figure 5 This is a comparison chart of the average fading time between the model simulation and measurement in a typical high-speed rail scenario according to the present invention. Detailed Implementation

[0058] To better understand the purpose, structure, and function of this invention, the following detailed description of a scenario adaptive channel modeling method based on 6G full coverage scenario classification is provided in conjunction with the accompanying drawings.

[0059] (1) Under the new scene classification framework, a multi-feature combination scheme is adopted for scene classification.

[0060] First, adhering to the principles of comprehensive coverage, reasonable architecture, and clear logic, the channel scenarios are roughly divided into 11 major scenarios and 144 sub-scenarios based on the existing scenario classification model and various standardized documents:

[0061] A1: Small indoor space

[0062] A2: From indoors to outdoors

[0063] B1: Typical Urban Micro-Community

[0064] B2: Ultra-complex urban micro-community

[0065] B3: Large Indoor Hall

[0066] B4: Outdoor to Indoor Mini-Community

[0067] C1: Suburban Hongxiao District

[0068] C2: Urban Macrocell

[0069] C3: Urban Macro-Community from Outdoor to Indoor

[0070] D1: Rural Hongxiao District

[0071] D2: Rural Mobile Network

[0072] Each major scene category is divided into three modules: residential, consumer, and industrial, based on its scene attributes. Each module is further divided into low-speed and high-speed sub-scenes based on the velocity of scattering objects within the scene. Sub-scenes are further divided into local application scenarios based on LoS and NLoS (and OLoS) paths. Finally, they are subdivided into specific scenes containing detailed physical environment parameters, other measurement parameters, and channel-related parameters. Furthermore, the optimization and supplementation of the scene classification parameter table includes the following aspects: ① Filling in blank scenes: Some sub-scenes lack actual measurement values ​​for verification. A default initial value (defined according to the building parameter reference table) is set to facilitate subsequent classification continuation. Default values ​​affect the computational complexity and accuracy of later scene classification. ② Determining the input range of physical environment parameters: For example, building specifications and materials, density or mobility of scattering objects such as people and vehicles, etc., are given reference range values ​​based on summaries of different literature and engineering experience. ③ The data in the parameter table conforms to common sense.

[0073] Secondly, a multi-parameter feature combination scheme was adopted based on the environment. Environmental physical parameters such as target environment specifications, height, traffic flow, pedestrian flow, vegetation density, and water systems were extracted from electronic maps or the real physical environment for scene transformation. Scene classification was then performed based on the combined characteristics of the parameter datasets.

[0074] Furthermore, when considering mobile communication, the user's location is constantly changing. Therefore, when determining the communication scenarios of the electronic map, it is necessary to accurately divide and segment the scenarios to obtain more sub-classes of communication scenarios, thereby improving the precision and accuracy of channel modeling. Then, communication scenarios are identified and matched based on the real physical environment. During the user's mobile communication process, it is necessary to determine when the scenario changes, thereby solving the problem of channel changes during scenario segmentation.

[0075] (2) Construct a scene model database based on the results of scene classification.

[0076] First, based on the definitions of physical parameters of the scenario environment, the standardized channel model parameter configuration table, and channel measurements, a channel model parameter database is constructed, establishing a mapping relationship between the ranges of physical environment parameters (PEFs) and the corresponding channel feature factors (CFFs). Second, a standard library of channel model parameter configurations is established for each typical scenario.

[0077] (3) Input the classified scene statistical characteristic parameter set into the universal channel model to generate a real-time channel model for scene transformation.

[0078] The channel matrix of the 6G ubiquitous geometric random channel model is represented as follows:

[0079] H = [PL·SH·BL·WE·AL] 1 / 2 ·H s

[0080] Where PL, SH, BL, WE, and AL represent large-scale fading, PL represents path loss, SH represents shadowing fading, BL represents blocking effect, AL represents atmospheric absorption loss, WE represents weather-related loss, and H represents path loss. s This is a small-scale fading.

[0081] The small-scale fading H s as follows:

[0082]

[0083] Among them, M T M represents the number of antenna elements in the transmitting antenna array. RThis refers to the number of antenna elements in the receiver's antenna array. For transmitting antenna array elements With receiving antenna array elements The channel impulse response between them is expressed as the line-of-sight LoS component. With non-line-of-sight NLoS components Superposition:

[0084]

[0085] Among them, K R (t) is the Rice factor. and They are as follows:

[0086]

[0087]

[0088] in,{*} T f represents transpose c Indicates the carrier frequency. and Antenna elements in different frequency bands Corresponding to the radiation patterns of vertical and horizontal polarization, The cross-polarization power ratio is given by μ, which characterizes the joint polarization imbalance. and It is from time t to The azimuth departure angle and pitch departure angle corresponding to the LosS path. and It is from time t to The azimuth and pitch angles of arrival corresponding to the LosS path. and It is a random phase that follows a uniform distribution in (0, 2π]. ψ l,m =108 / f c 2 f is the Faraday rotation angle, which is calculated here. c The unit is GHz. Under NLoS conditions arrive The power of the m-th sub-path in the n-th path, It is the delay of the Loss path at time t. At time t, the transmitting antenna array element With receiving antenna array elements The vector distance between them, where c is the speed of light. At time t and The time delay of the m-th sub-path of the n-th path between them. At time t and The power of the m-th sub-path along the n-th path is given. All parameters are time-varying. The channel impulse response and channel matrix for the sub-scene are output using the above channel model formula.

[0089] (4) Model Adjustment and Optimization. Based on the scenario classification results, the parameter configuration documents for the corresponding scenarios are input into the 6G ubiquitous geometric random channel model. The channel model outputs the channel impulse response matrix, channel statistical characteristics, etc., as needed. At the same time, the weights of the channel model configuration parameter values ​​are fine-tuned based on the corresponding measurement data or simulation data, and the results of the channel model are iteratively optimized.

[0090] This invention provides a scenario-adaptive channel modeling method based on 6G full coverage scenario classification, comprising the following steps:

[0091] S1. Based on the physical parameter definition of the 6G full coverage scenario, and according to the influence of different features on the classification of multiple scenarios, a combination scheme of physical environment and statistical features is selected for cross-scenario classification.

[0092] S2. Construct a 6G full-coverage cross-scenario model database based on the cross-scenario classification results, which includes physical scenarios and their corresponding model statistical characteristic parameter values;

[0093] S3. Referring to the 6G full coverage cross-scenario model database constructed in S2, based on the results of cross-scenario classification and identification, input the corresponding model statistical characteristic parameter values ​​into the universal channel model to generate a channel model for a single scenario or cross-fusion scenario.

[0094] S4. For the 6G full coverage scenario, the statistical characteristics of the model in S2 are logically operated on with the statistical characteristic parameters obtained from wireless channel simulation or measured data. The parameter weights of the statistical characteristics of the universal channel model are adjusted according to the closed-loop feedback, and the optimized scenario adaptive channel model is obtained iteratively.

[0095] S5. Simulation was used to verify and confirm the accuracy of the adaptive channel modeling for this scenario.

[0096] The specific steps of step S1 are as follows:

[0097] Step S101: The classification of 6G full coverage scenarios follows a macro-to-micro principle, initially dividing them into four main categories: air, space, land, and sea. These are then further subdivided based on layer height / depth. For scenarios within the same layer, further subdivisions are made according to the actual physical environment and signal propagation characteristics. Detailed physical parameter definitions are provided for each subdivided scenario, including length, width, height, movement speed, transceiver distance range, and line-of-sight / non-line-of-sight propagation.

[0098] Step S102: Evaluate the impact of various scene environmental feature parameters on classification performance. In addition to achieving line-of-sight and non-line-of-sight classification, as well as indoor and outdoor classification, a method that combines multiple feature parameters can be used to classify multi-target scenes. Table 1 shows the typical parameters of each sub-scene (fourth-level classification) in the categories of land scene (first-level classification), typical city (second-level classification), and open area (third-level classification).

[0099] Table 1 Typical parameters of sub-scenes in land scenes.

[0100]

[0101] Step S103 and Table 2 provide the statistical characteristic parameters in the full coverage scene classification, including three methods: using raw data, probability distribution shape, and threshold classification.

[0102] Table 2 Commonly Used Feature Parameters for Scene Classification

[0103]

[0104] The time-domain channel impulse response (CIR) and frequency-domain channel transfer function (CTF) contain rich information about the raw data characteristics. In time-varying environments, it is important to consider the challenges that large amounts of channel measurement data pose to real-time decision-making.

[0105] First, various scenarios can be classified based on the shape of the probability distribution. Relevant characteristic parameters include the probability density function, the number of peaks in the time-delay power spectral density, skewness, and kurtosis / peak value. Considering that line-of-sight scenarios generally follow a Ricean distribution while non-line-of-sight scenarios follow a Rayleigh distribution, the probability density function can be used to distinguish between line-of-sight and non-line-of-sight scenarios. The number of peaks in the time-delay power spectral density can be used to distinguish rich scattering environments; peaks with different time delays represent multipath propagation from different scatterers. Skewness indicates the degree of asymmetry in the probability distribution. The skewness of line-of-sight scenarios is generally less than that of non-line-of-sight scenarios. Kurtosis indicates the steepness of the probability distribution. Existing research shows that the kurtosis is greater in the line-of-sight case than in the non-line-of-sight case.

[0106] Secondly, threshold comparisons are used for classification, and the available thresholds are shown in Table 2. Path loss (PL) is an important indicator of large-scale fading, mainly affected by transmission distance, and can classify communication scenarios with significantly different propagation distances. Rice's K-factor represents the ratio of line-of-sight path power to non-line-of-sight path power. In addition, some time-domain, spatial-domain, and frequency-domain related characteristic parameters can also be used for multi-scenario classification. For example, for time-varying scenarios with high mobility, the time autocorrelation function decreases rapidly, and the stationary interval is small. The root mean square delay spread reflects the dispersion of the wireless channel in the time delay domain. For scenarios with abundant scatterers, this value is usually large.

[0107] Step S104: Using a combination of classification methods, namely S102 environmental physical parameter classification and S103 statistical characteristic classification, output the classification results for the 6G full coverage scenario.

[0108] The specific steps of step S2 are as follows:

[0109] Step S201: Conduct channel measurements under typical scenarios and parameter configurations to improve the diversity of the wireless channel database. Construct scenarios including aerospace coverage such as satellites and drones, outdoor scenarios such as typical cities, suburbs, vehicle-to-vehicle, and high-speed rail, and typical indoor scenarios such as office areas, corridors, residential areas, shopping malls, and industrial IoT.

[0110] Step S202: Using Wireless Insite software supplemented by ray tracing simulation data, the system configuration parameters are set to reconstruct the real propagation environment, generate simulation path parameters and channel impulse response, further increase the number of channel model database samples for each communication scenario, and reduce the cost and complexity of actual measurement.

[0111] Step S203: Optimize and adjust the measurement and simulation data respectively to construct a 6G full coverage scenario model database. This database contains basic physical parameters for various scenarios, calculated characteristic parameters, and complete channel impulse responses. Table 3 shows typical values ​​of statistical characteristics in 3GPP 38.901.

[0112] Table 3 Model Database Characteristic Parameters

[0113]

[0114]

[0115] Step S204: Input the scene classification results from step S1 into the scene model database and output the statistical characteristic parameter set of the corresponding scene.

[0116] The specific steps of step S3 are as follows:

[0117] Step S301: This invention optimizes the modeling based on a geometrically random universal channel model. This approach considers both the high accuracy of the geometrically random channel model and its flexibility in application to different scenarios, thus making it suitable for 6G full coverage scenarios. The relevant parameters in the model are summarized and categorized as follows: Figure 2 As shown, the parameters mainly include user-defined parameters and wireless channel-related parameters. The former includes the system configuration's center frequency / wavelength and bandwidth, array configuration parameters, and parameters such as movement speed and trajectory. The latter includes large-scale fading-related parameters, scattering clusters, and multipath-related parameters.

[0118] The channel matrix of the 6G ubiquitous geometric random channel model is represented as follows:

[0119] H = [PL·SH·BL·WE·AL] 1 / 2 ·H s

[0120] Where PL, SH, BL, WE, and AL represent large-scale fading, PL represents path loss, SH represents shadowing fading, BL represents blocking effect, AL represents atmospheric absorption loss, WE represents weather-related loss, and H represents path loss. s This is a small-scale fading.

[0121] The small-scale fading H s as follows:

[0122]

[0123] Among them, M T M represents the number of antenna elements in the transmitting antenna array. R This refers to the number of antenna elements in the receiver's antenna array. For transmitting antenna array elements With receiving antenna array elements The channel impulse response between them is expressed as the line-of-sight LoS component. With non-line-of-sight NLoS components Superposition:

[0124]

[0125] Among them, K R (t) is the Rice factor. and They are as follows:

[0126]

[0127]

[0128] in,{*} T f represents transpose c Indicates the carrier frequency. and Antenna elements in different frequency bands Corresponding to the radiation patterns of vertical and horizontal polarization, The cross-polarization power ratio is given by μ, which characterizes the joint polarization imbalance. and It is from time t to The azimuth departure angle and pitch departure angle corresponding to the LosS path. and It is from time t to The azimuth and pitch angles of arrival corresponding to the LosS path. and It is a random phase that follows a uniform distribution in (0, 2π]. ψ l，m =108 / f c 2 f is the Faraday rotation angle, which is calculated here. c of

[0129] The unit is GHz. Under NLoS conditions arrive The power of the m-th sub-path in the n-th path, It is the delay of the Loss path at time t. At time t, the transmitting antenna array element With receiving antenna array elements The vector distance between them, where c is the speed of light. At time t and The time delay of the m-th sub-path of the n-th path between them. At time t and The power of the m-th sub-path of the n-th path between them; all of the above parameters are time-varying parameters.

[0130] Step S302: Substitute the statistical characteristic parameter data of the scene classification in step S2 into the universal channel model to obtain a channel model suitable for scene switching. Assume the number of base stations is N. BS The number of users is N MS The channel transmission matrix of the multi-link channel model in scene transition is shown in the following formula:

[0131]

[0132] Each link corresponds to

[0133] Step S303: Table 4 of this invention provides the basic model parameters for typical satellite, drone, industrial IoT, (ultra) high-speed rail and marine communication scenarios.

[0134] Table 4 Model parameter matching for typical scenarios

[0135]

[0136]

[0137] The specific steps of step S4 are as follows:

[0138] Step S401: Directly calculate the relevant statistical characteristics of the channel using measured or simulated data from typical scenarios, including time autocorrelation function, spatial cross-correlation function, root mean square delay / angle spread, etc.

[0139] Step S402: Based on the typical scenario impulse response function and channel matrix output in step S3, calculate the channel correlation statistics and logically compare them with the results obtained from measurement or simulation in step S401. This invention... Figure 3-5 The paper presents the differences in measurement and model statistical characteristics between typical industrial IoT and high-speed rail scenarios.

[0140] Step S403: Adaptively adjust the statistical characteristic parameter matrix of the classification based on the feedback results, and obtain an optimized scenario-switching channel model through feedback and iteration in different scenarios.

[0141] The specific steps of step S5 are as follows:

[0142] Step S501: Based on the iteratively optimized scenario-adaptive channel model, verify the performance of the integrated mechanism of channel scenario classification and model parameter matching. Channel data for several scenarios are generated through measurement or simulation. Scenario classification is performed by combining the physical environment parameters and statistical characteristics from step S1.

[0143] Step S502: Based on the scene statistical characteristics obtained from the classification, adaptively match the parameter matrix of the universal channel model, and logically compare the statistical quantities obtained from the model simulation with the results calculated from the measured data.

[0144] Step S503: By performing steps S501 and S502 multiple times, the accuracy of the adaptive modeling integrated scheme based on 6G full coverage scenario classification of the present invention is verified.

[0145] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. A scenario-adaptive channel modeling method based on 6G full coverage scenario classification, characterized in that, Includes the following steps: S1. Based on the physical parameter definition of the 6G full coverage scenario, and according to the influence of different features on the classification of multiple scenarios, a combination scheme of physical environment and statistical features is selected for cross-scenario classification. S2. Construct a 6G full-coverage cross-scenario model database based on the cross-scenario classification results, which includes physical scenarios and their corresponding model statistical characteristic parameter values; S3. Based on the 6G full coverage cross-scenario model database constructed in S2, and according to the results of cross-scenario classification and identification, input the corresponding model statistical characteristic parameter values ​​into the universal channel model to generate a channel model for a single scenario or cross-fusion scenario. S4. For the 6G full coverage scenario, the statistical characteristics of the model in S2 are logically operated with the statistical characteristic parameter values ​​obtained from wireless channel simulation or the statistical characteristic parameter values ​​obtained from measured data. The parameter weights of the universal channel model statistical characteristics are adjusted according to the closed-loop feedback, and the optimized scenario adaptive channel model is obtained iteratively. S5. Simulation was used to verify and confirm the accuracy of the adaptive channel modeling for this scenario; The specific steps of step S1 are as follows: Step S101: The classification of 6G full coverage scenarios follows the principle of macro to micro. First, they are divided into four major categories: air, space, land, and sea. Then, they are further divided according to the layer height or depth. For scenarios at the same layer, they are further subdivided according to the actual physical environment and signal propagation characteristics. For each subdivided scenario, detailed physical parameter definitions are given, including length, width, height, moving speed, transmit and receive distance range, and line-of-sight or non-line-of-sight propagation. Step S102: By evaluating the impact of various scene environmental feature parameters on classification performance, in addition to achieving classification at line of sight and non-line of sight, as well as indoor and outdoor, multi-feature parameter combination is used to classify multi-target scenes. This is a classification method using environmental physical parameters. Step S103: Using three methods—original data, probability distribution shape, and threshold classification—statistical characteristic classification is performed on the 6G full coverage scenario. The statistical characteristic parameters corresponding to the channel of each scenario are calculated in detail, and the parameter intervals are divided to classify the scenarios. Step S104: Using a combination of classification methods, namely S102 environmental physical parameter classification and S103 statistical characteristic classification, output the classification results for the 6G full coverage scenario.

2. The scene adaptive channel modeling method based on 6G full coverage scene classification according to claim 1, characterized in that, In step S103, the time-domain channel impulse response (CIR) and frequency-domain channel transfer function (CTF) contain rich information in the original data feature parameters. First, various scenarios are classified based on the shape of the probability distribution. The characteristic parameters for classifying scenarios based on the probability distribution shape include the probability density function, the number of peaks in the time-delay power spectral density, skewness, and kurtosis. Considering that line-of-sight scenarios follow a Ricean distribution while non-line-of-sight scenarios follow a Rayleigh distribution, the probability density function is used to distinguish between line-of-sight and non-line-of-sight scenarios. The number of peaks in the time-delay power spectral density is used to distinguish rich scattering environments; peaks with different time delays represent multipath propagation from different scatterers. Skewness indicates the degree of asymmetry in the probability distribution; the skewness of line-of-sight scenarios is less than that of non-line-of-sight scenarios. Kurtosis indicates the steepness of the probability distribution; the kurtosis in the line-of-sight case is greater than that in the non-line-of-sight case. Secondly, threshold comparison is used for corresponding classification; path loss PL is an important indicator of large-scale fading, which is affected by transmission distance, and communication scenarios with different propagation distances are classified; Rice K factor represents the ratio of line-of-sight path power to non-line-of-sight path power.

3. The scene adaptive channel modeling method based on 6G full coverage scene classification according to claim 2, characterized in that, The specific steps of step S2 are as follows: Step S201: Perform channel measurement under scenario and parameter configuration to improve the diversity of the wireless channel database; construct air-space coverage scenarios, outdoor scenarios, and indoor scenarios; Step S202: Using software supplemented by ray tracing simulation data, by reconstructing the real propagation environment, setting system configuration parameters, generating simulation path parameters and channel impulse response, increasing the number of channel model database samples for each communication scenario, and reducing the cost and complexity of actual measurement; Step S203: Optimize and adjust the measurement and simulation data respectively to build a 6G full coverage scenario model database; the database contains the basic physical parameters of various scenarios, the calculated characteristic parameters, and the complete channel impulse response; Step S204: Input the scene classification results from step S1 into the scene model database and output the statistical characteristic parameter set of the corresponding scene.

4. The scene adaptive channel modeling method based on 6G full coverage scene classification according to claim 3, characterized in that, The specific steps of step S3 are as follows: Step S301: Based on the 6G ubiquitous channel model, perform model statistical parameter matching according to the scenario classification results in S2; the parameters in the 6G ubiquitous channel model used here include user-defined parameters and wireless channel-related parameters; among them, user-defined parameters include the center frequency, wavelength and bandwidth of the system configuration, array configuration-related parameters, and moving speed and trajectory parameters; the latter includes large-scale fading-related parameters, scattering clusters and multipath-related parameters; The channel matrix of the 6G ubiquitous geometric random channel model is represented as follows: H=[PL·SH·BL·WE·AL] 1 / 2 ·H s Where PL, SH, BL, WE, and AL represent large-scale fading, PL represents path loss, SH represents shadowing fading, BL represents blocking effect, AL represents atmospheric absorption loss, WE represents weather-related loss, and H represents path loss. s This is a small-scale fading phenomenon; The small-scale fading H s as follows: Among them, M T M represents the number of antenna elements in the transmitting antenna array. R This refers to the number of antenna elements in the receiver's antenna array. For transmitting antenna elements With receiving antenna unit The channel impulse response between them is expressed as the line-of-sight LoS component. With non-line-of-sight NLoS components Superposition: Among them, K R (t) is the Rice factor. and They are as follows: in,{*} T f represents transpose c Indicates the carrier frequency. Antenna elements in different frequency bands Corresponding to the vertical polarization pattern, Antenna elements in different frequency bands Corresponding to the vertical polarization pattern, Antenna elements in different frequency bands Corresponding to the horizontal polarization pattern, Antenna elements in different frequency bands Corresponding to the horizontal polarization pattern, The cross-polarization power ratio is given by μ, which characterizes the joint polarization imbalance. and It is from time t to The azimuth departure angle and pitch departure angle corresponding to the LosS path. and It is from time t to The azimuth and pitch angles of arrival corresponding to the LosS path. and It is a random phase that follows a uniform distribution in (0, 2π]. ψ l,m =108 / f c 2 f is the Faraday rotation angle, which is calculated here. c The unit is GHz. It is the delay of the Loss path at time t. At time t, the transmitting antenna array element With receiving antenna array elements The vector distance between them, where c is the speed of light; At time t and The time delay of the m-th sub-path of the n-th path between them. At time t and The power of the m-th sub-path of the n-th path between them; all of the above parameters are time-varying parameters. Step S302: Substitute the statistical characteristic parameter data of the scene classification in step S2 into the universal channel model to obtain a channel model suitable for scene switching, i.e., the channel impulse response function or the channel transmission matrix; set the number of base stations to N. BS The number of users is N MS The channel transmission matrix of the multi-link channel model in scene transition is shown in the following formula: Each link corresponds to 5. The scene adaptive channel modeling method based on 6G full coverage scene classification according to claim 4, characterized in that, The specific steps of step S4 are as follows: Step S401: Directly calculate the relevant statistical characteristics of the channel using measured or simulated data from typical scenarios, including the time autocorrelation function, spatial cross-correlation function, root mean square delay, and angle spread. Step S402: Based on the channel impulse function or channel transmission matrix output in step S302, calculate the channel correlation statistics and make a logical comparison with the results obtained from actual measurement or simulation in S401. Step S403: Adaptively adjust the statistical characteristic parameter matrix of the classification based on the feedback results, and obtain an optimized scenario-switching channel model through feedback and iteration in different scenarios.

6. The scene adaptive channel modeling method based on 6G full coverage scene classification according to claim 5, characterized in that, The specific steps of step S5 are as follows: Step S501: Based on the iteratively optimized scene adaptive channel model, verify the performance of the integrated mechanism of channel scene classification and model parameter matching; generate channel data under several scenarios by measurement or simulation, and classify the scenarios by combining the physical environment parameters and statistical characteristics in step S1. Step S502: Based on the scene statistical characteristics obtained from the classification, adaptively match the parameter matrix of the universal channel model, and logically compare the statistical quantities obtained from the model simulation with the results calculated from the measured data. Step S503: By performing steps S501 and S502 multiple times, verify the accuracy of the integrated adaptive modeling solution based on 6G full coverage scenario classification.

Images