WO2023169590A1 - 一种适用于全频段全场景的6g普适信道建模方法 - Google Patents

一种适用于全频段全场景的6g普适信道建模方法 Download PDF

Info

Publication number
WO2023169590A1
WO2023169590A1 PCT/CN2023/082380 CN2023082380W WO2023169590A1 WO 2023169590 A1 WO2023169590 A1 WO 2023169590A1 CN 2023082380 W CN2023082380 W CN 2023082380W WO 2023169590 A1 WO2023169590 A1 WO 2023169590A1
Authority
WO
WIPO (PCT)
Prior art keywords
cluster
channel
time
clusters
path
Prior art date
Application number
PCT/CN2023/082380
Other languages
English (en)
French (fr)
Inventor
王承祥
吕振
Original Assignee
东南大学
网络通信与安全紫金山实验室
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 东南大学, 网络通信与安全紫金山实验室 filed Critical 东南大学
Publication of WO2023169590A1 publication Critical patent/WO2023169590A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/391—Modelling the propagation channel
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/391—Modelling the propagation channel
    • H04B17/3911—Fading models or fading generators
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Definitions

  • the present invention relates to a 6G universal channel modeling method suitable for all frequency bands and all scenarios, and belongs to the field of wireless communication technology.
  • 6G wireless channels can be summarized as full frequency bands (sub-6 GHz/millimeter wave/terahertz/optical wireless bands), full coverage scenarios (integration of air, space, ground and sea, including satellites, drones, land and ocean communication channels) and full coverage scenarios.
  • Application scenarios Internet of Vehicles, high-speed rail channels, large-scale antenna arrays, reconfigurable smart surfaces, industrial Internet of Things, etc.
  • 6G full-band and full-scenario channels also exhibit many new channel characteristics, bringing new challenges to 6G channel modeling work.
  • wireless channels In terms of full frequency bands, due to the application of high-frequency bands such as millimeter waves and terahertz, wireless channels exhibit characteristics such as large bandwidth, non-stationary frequency domain, diffuse scattering, large path loss, blocking effect and atmospheric absorption; in the visible light band, the channel will It no longer has small-scale fading, and exhibits negligible Doppler effect and frequency domain non-stationary characteristics.
  • full coverage in addition to land mobile communication scenarios, it also includes satellite communication, drone communication and marine communication scenarios. Among them, satellite communication channels need to consider the Doppler frequency shift, rainfall fading, and ionospheric effects caused by the rapid movement of satellites.
  • the Internet of Vehicles channel exhibits Doppler frequency shift and time-domain non-stationary characteristics due to the multi-mobility of both transmitting and receiving ends and clusters; in higher-speed mobile scenarios exceeding 500km/h, the channel Experience stronger Doppler frequency shift and more obvious time domain non-stationarity; in the ultra-high-speed train scenario running in a vacuum tube, the influence of the vacuum tube waveguide effect also needs to be considered; the very large-scale antenna array channel exhibits spherical wave characteristics and airspace non-stationary characteristics; ultra-dense scatterer distribution and multi-mobility need to be considered in industrial IoT channels; in addition, accurate modeling of wireless channels using reconfigurable smart surface technology must also be studied.
  • Channel model For example, when millimeter wave/terahertz frequency bands and large-scale antenna technology are applied simultaneously in high-speed mobile scenarios, wireless channels will simultaneously exhibit space-time-frequency non-stationary characteristics and spatial consistency (i.e., in multi-user scenarios, similar users’ Channel coefficients have correlation or a single user has spatial correlation at different trajectory points) and multi-band correlation, etc.
  • 5G standardized channel models such as B5GCM, 3GPP TR 38.901, IMT-2020 and QuaDRiGa have made attempts on this issue, but they cannot accurately and comprehensively describe all the aforementioned characteristics.
  • 3GPP TR 38.901 and IMT -2020 ignores the frequency non-stationary characteristics of high-frequency bands; in terms of full coverage, these channel models are targeted at land mobile communication channels and cannot be applied to satellite, drone and ocean communication scenarios; in terms of full application, they are not Supports modeling of hyperloop channels, reconfigurable smart surfaces and industrial IoT channels, and 3GPP TR 38.901 and IMT-2020 do not consider the spherical waves and airspace non-stationary characteristics of (very) large-scale antenna arrays.
  • the present invention proposes a universal channel modeling theory, applies it to a geometric random channel model, and proposes and discloses a 6G universal channel modeling method suitable for all frequency bands and all scenarios.
  • the purpose of this invention is to propose a 6G universal channel modeling method suitable for all frequency bands and all scenarios.
  • This method is a method that is universally applicable to all frequency bands (sub-6 GHz/millimeter wave/terahertz/visible light frequency band). ), full coverage scenarios (air, space, ground and sea integration, including satellites, drones, land and ocean communication scenarios) and full application scenarios (Internet of Vehicles, high-speed rail channels, large-scale antenna arrays, reconfigurable smart surfaces and industrial Internet of Things Scenario) 6G universal geometric random channel modeling method.
  • the present invention is a 6G pervasive channel modeling method suitable for all frequency bands and all scenarios.
  • both the transmitter and the receiver adopt large-scale uniform linear arrays, and the model is multi-hop propagation.
  • N qp (t) is when time t arrive
  • the number of paths corresponds to N qp (t) cluster pairs in the dual-cluster model.
  • the first-hop cluster and the last-hop cluster correspond to one-to-one.
  • N qp (t) clusters In the single-cluster model, it corresponds to N qp (t) clusters.
  • PL, SH, BL, WE, AL are large-scale fading
  • PL path loss
  • SH shadow fading
  • BL blocking effect
  • AL atmospheric absorption loss
  • WE weather influence loss
  • H s small-scale fading
  • the small-scale fading H s is as follows:
  • M T is the number of antenna elements in the transmitting end antenna array
  • MR is the number of antenna elements in the receiving end antenna array
  • K R (t) is the Rice factor, and They are as follows:
  • ⁇ * ⁇ T represents transposition
  • f c represents carrier frequency
  • Patterns corresponding to vertical polarization and horizontal polarization is the cross-polarization power ratio
  • ⁇ represents the joint polarization imbalance
  • ⁇ represents the joint polarization imbalance
  • ⁇ represents the joint polarization imbalance
  • ⁇ is time t from to The azimuth departure angle and pitch departure angle corresponding to the LoS path
  • ⁇ is time t from to The azimuth arrival angle and pitch arrival angle corresponding to the LoS path
  • the unit of f c is GHz.
  • the power of the m-th subpath in the n-th path is the delay of the LoS path at time t, is the transmitting end antenna array element at time t and receiving antenna array element
  • the vector distance between , c is the speed of light. is at time t and The delay of the m-th sub-path of the n-th path between is at time t and The power of the m-th sub-path of the n-th path between. All the above parameters are time-varying parameters.
  • the model models the three parts of the LoS path, the rough ocean surface, and the multipath propagation of the evaporative waveguide above the sea surface as and And using the power coefficient, S 1 and S 2 control the disappearance and appearance of the corresponding part as the distance between the two ships changes, that is,
  • the 6G ubiquitous channel modeling method divides the channel into the sub-channel H TI from the transmitting end to the reconfigurable intelligent surface, the sub-channel H IR from the reconfigurable intelligent surface to the receiving end and For the sub-channel H TR from the transmitting end to the receiving end, the three sub-channels are modeled separately and the phase shift diagonal matrix ⁇ is introduced to realize intelligent control of the channel environment.
  • the calculation methods of H IR , H TI and H TR are the same as H s . Only the parameter values and cluster distributions are different.
  • the wavelength of the optical signal is extremely short, and the size of the receiving end is usually several million wavelengths, so rapid signal fading on several wavelengths will not occur; on the other hand,
  • the LED lights in the visible light communication system emit incoherent light, the optical signal has no phase information.
  • the superposition of the real multipath signals at the receiving end will not cause fast fading, but will appear as slowly changing shadow fading. Therefore, although now
  • the channel transmission matrix of the multi-link channel model is as follows:
  • channel matrix H is divided into the following steps in detail:
  • delay expansion DS In addition to SH, the other relevant large-scale parameters include delay expansion DS, azimuth arrival angle expansion ASA, azimuth departure angle expansion ASD, pitch arrival angle expansion ESA, pitch departure angle expansion ESD, Rice factor KR and cross-polarization ratio XPR.
  • delay expansion DS azimuth arrival angle expansion ASA, azimuth departure angle expansion ASD, pitch arrival angle expansion ESA, pitch departure angle expansion ESD, Rice factor KR and cross-polarization ratio XPR.
  • X DS (P) is The sine wave superposition method generates a normally distributed variable with a spatial consistency of mean 0 and variance 1. Represents the mean value of DS in the f c band, Indicates the variance of DS in the f c band.
  • the configuration values can be divided into three types; for the land mobile communication scenario 1.5m ⁇ h UT ⁇ 22.5m, you can refer to the values in Table 7.5-6 in the 3GPPTR38.901 standardization document; for the UAV scenario 22.5m ⁇ h UT ⁇ 300m, the value refers to the value in Table B1.2 in the 3GPPTR36.777 standardization document; for the low-orbit satellite communication scenario, the value refers to the value in Table 6.7-2 in the 3GPPTR38.811 standardization document, urban macro cell Under UMa scenario NLoS conditions, the carrier frequency is between 2-4GHz is calculated as follows
  • the generation process of other large-scale parameters is the same as the generation process of delay spread DS. After all eight large-scale parameters are generated, multiplying by the cross-correlation matrix between large-scale parameters can obtain all large-scale parameters with spatial consistency in the logarithm Then, the values in the logarithmic domain need to be converted into the linear domain. At this point, the large-scale parameters of the channel can be obtained;
  • the space-time-frequency non-stationarity of the model is mainly reflected in two aspects. On the one hand, it is the space-time-frequency non-stationarity.
  • N qp (t) is the number of clusters
  • N surv (t) is the number of surviving clusters, which is determined by the cluster survival probability P surv ( ⁇ t, ⁇ r, ⁇ f)
  • N new (t) is the number of new clusters, Obeying the Poisson distribution with mean E[N new (t)], define ⁇ G as the birth rate of the cluster, and ⁇ R as the combination rate of the cluster, that is, the death rate.
  • the step S4 is specifically:
  • Step S401 Use the ellipsoidal Gaussian scattering distribution to obtain the position of the scatterer, that is, use
  • the scatterers in the nth cluster centered on the standard deviation on the three coordinate axes are and Gaussian distribution, after obtaining the positions of the scatterers, they can be converted into spherical coordinates, relative to the position of the first transmitting antenna and the location of the first receiving antenna
  • the position of the scatterer within the nth cluster of and It can be expressed as and in, and Respectively represent the distance, azimuth angle and pitch angle from the m-th sub-path of the n-th cluster to the antenna array of X (X ⁇ T,R ⁇ , representing the transmitting end and receiving end respectively);
  • Step S402 Calculate the intra-cluster sub-path delay at the initial moment.
  • the intra-cluster sub-path delay can be calculated by Calculate, where, express and The delay of the virtual link between is time t 0 arrive the distance between, is time t 0 arrive the distance between. is the distance between the first hop cluster and the last hop cluster, ⁇ link is a non-negative random variable obeying exponential distribution;
  • Step S403. In a large-scale antenna array, the power of the sub-paths within the cluster will vary along the time axis and the array axis, and is usually modeled as a lognormal process varying over time and a lognormal process varying along the array, non-normalized intra-cluster sub-radius power for:
  • DS is the root mean square delay spread
  • Z n is the shadow term of each cluster in dB
  • r ⁇ is the delay distribution scaling factor
  • ⁇ n (p, q) is a two-dimensional space lognormal process , used to simulate smooth power changes on the antenna array
  • the power of the final sub-path within the cluster can be obtained by normalizing the power of all clusters. If the cluster is newly generated, by Replace with get and The initial power of the m-th sub-path in the n-th cluster;
  • Step S404 For the surviving clusters, it is necessary to update the power, delay and other small-scale parameters of the sub-paths in the cluster at different times.
  • the p-th root transmits antenna For the trajectory segment at time t 1 , that is, at the next time after the cluster is generated, the p-th root transmits antenna The coordinates are Among them, the coordinates of the p-th transmitting antenna at the initial time pass Calculate, at time t 1 , the coordinates of the m-th scatterer in the n-th first-hop cluster pass calculate.
  • arrive distance can be passed Calculated, it can be obtained in the same way arrive distance
  • step S5 In order to more accurately model the space-time-frequency evolution process of the cluster, two sampling intervals are introduced, one is the time domain sampling interval ⁇ t, the frequency domain sampling interval ⁇ f and the spatial domain (array domain) sampling.
  • the channel parameters are continuously updated at an interval ⁇ r; the other is an integer multiple of ⁇ t, ⁇ f and ⁇ r, which are ⁇ t BD , ⁇ f BD and ⁇ r BD respectively.
  • the birth and death evolution process of clusters occurs at these sampling points.
  • the joint survival probability of the sender and receiver clusters is:
  • the average number of new clusters is:
  • the survival probability of clusters on the frequency axis is:
  • F( ⁇ f BD ) and Can be determined through channel measurements represents the scene-dependent correlation factor on the frequency axis
  • the average number of new clusters is:
  • the average number of new clusters is:
  • D qp (t) is the straight-line distance between the sending and receiving ends at time t
  • D is the initial distance between the sending and receiving ends
  • ⁇ s is the scattering coefficient of the pipe wall
  • ⁇ h 0.
  • the present invention proposes a universal channel modeling theory, and applies the theory to the geometric random channel model, using cluster-based geometric random channel modeling methods and frameworks, and using a unified channel impulse response expression, Model the channel characteristics of 6G full-band and all-scenarios, and propose a channel that is basically suitable for full-band channels such as sub-6GHz, millimeter wave, terahertz, and visible light, as well as full-coverage scenario channels such as low-orbit satellites, drones, and ocean communications. , as well as the 6G universal geometric random channel model for channels in all application scenarios such as ultra-large-scale antenna arrays, industrial Internet of Things, and reconfigurable smart surfaces.
  • the 6G universal channel model by adjusting the parameters of the 6G universal channel model, it can be simplified into a dedicated channel model for specific frequency bands and specific scenarios.
  • the 6G universal geometric random channel model is crucial for 6G channel model standardization, 6G common theory and technology research, and system integration construction.
  • Figure 1 is a flow chart in an embodiment of the present invention.
  • Figure 2 is a schematic diagram of the 6G wireless channel in the present invention.
  • Figure 3 is a schematic diagram of the theory of ubiquitous channel modeling in the present invention.
  • Figure 4 is a schematic diagram of the 6G universal geometric random channel model in the present invention.
  • the present invention proposes a universal channel modeling theory, and proposes a 6G universal geometric random channel model based on this theory. Therefore, the present invention mainly includes the universal channel model modeling theory and the 6G universal geometric random channel There are two parts to model building:
  • the universal channel modeling theory uses a unified channel modeling method and framework, a unified channel impulse response expression, and comprehensively considers the statistical characteristics of the 6G full-band and full-scenario channel to construct a model that is universally applicable to all 6G frequency bands and scenarios, and
  • the 6G ubiquitous channel model that can accurately reflect its channel characteristics is shown in Figure 3.
  • the 6G universal channel model can be simplified into a dedicated channel model suitable for specific frequency bands and specific scenarios by adjusting the parameters of the channel model.
  • the complex mapping relationship between channel model parameters, channel characteristics and communication system performance can be studied.
  • As a unified channel model framework it can conduct research on 6G channel model standardization and 6G commonality theory and technology. and system integration construction are crucial.
  • the schematic diagram of the 6G universal geometric random channel model is shown in Figure 4.
  • the antenna types in the model can be uniform linear array, uniform area array and other antenna array types, supporting any antenna polarization type.
  • both the transmitter and the receiver adopt large-scale uniform linear arrays.
  • the model in the schematic diagram is multi-hop propagation, where, is the p-th array element of the transmitter antenna array, The distance between the q-th element of the receiving end antenna array and the transmitting end (receiving end) antenna element is ⁇ T ( ⁇ R ). is the azimuth angle of the transmitter (receiver) antenna array in the xy plane, is the pitch angle of the transmitter (receiver) antenna array.
  • the propagation path between the two clusters is modeled as a virtual link.
  • the model degenerates into a single-hop model.
  • N qp (t) is when time t arrive
  • the number of paths corresponds to N qp (t) cluster pairs in the dual-cluster model.
  • the first-hop cluster and the last-hop cluster correspond to one-to-one.
  • the single-cluster model corresponds to N qp (t) clusters.
  • M n (t) scatterers in the cluster express The mth scatterer in , express The mth scatterer in ; from the perspective of the path, understood as arrive The m-th sub-diameter connected scatterers, understood as to The m-th sub-diameter connected scatterers; therefore, and is time t from to The azimuth departure angle and pitch departure angle corresponding to the mth sub-radius, and is time t from to The azimuth arrival angle and pitch arrival angle corresponding to the mth sub-path of Represents the movement speed of the sending end, receiving end, first hop cluster and last hop cluster respectively, represent the azimuth angles of the movement direction of the sender, receiver, first-hop cluster and last-hop cluster respectively, Represent the pitch angles of the movement directions of the transmitter, receiver, first-hop cluster, and last-hop cluster respectively
  • PL, SH, BL, WE, AL are large-scale fading
  • PL is path loss
  • SH shadow fading
  • BL is blocking effect
  • AL is atmospheric absorption loss, such as oxygen absorption loss in millimeter wave frequency band and terahertz frequency band.
  • WE is weather-affected loss, such as rain fading in satellite communication scenarios. This invention mainly focuses on the calculation of small-scale fading H s . The method is as follows:
  • MT is the number of antenna elements in the transmitter (receiver) antenna array, is the transmitter antenna array element and receiving antenna array element
  • NoS non-line-of-sight
  • K R (t) is the Rice factor.
  • ⁇ * ⁇ T represents transposition
  • f c represents carrier frequency
  • Patterns corresponding to vertical polarization and horizontal polarization is the cross-polarization power ratio
  • ⁇ represents the joint polarization imbalance
  • ⁇ represents the joint polarization imbalance
  • ⁇ represents the joint polarization imbalance
  • ⁇ represents the joint polarization imbalance
  • ⁇ is time t from to
  • the LoS path corresponds to the azimuth arrival angle and pitch arrival angle. and is a random phase obeying the (0,2 ⁇ ] uniform distribution
  • the unit of f c is GHz.
  • the power of the m-th sub-path of the n-th path, and the delay of the LoS path at time t is the transmitting end antenna array element at time t and receiving antenna array element
  • the vector distance between , c is the speed of light. is at time t and The delay of the m-th sub-path of the n-th path between All the above parameters are time-varying parameters.
  • the specular reflection component and dense multipath component are respectively modeled as and in and modeling methods and The same, only the parameter values and the distribution of clusters are different.
  • the channel is divided into a sub-channel H TI from the transmitter to the reconfigurable smart surface, a sub-channel H IR from the reconfigurable smart surface to the receiver, and a sub-channel H TR from the sender to the receiver.
  • the three sub-channels are modeled separately and the phase shift diagonal matrix ⁇ is introduced to achieve intelligent control of the channel environment.
  • the calculation method of H IR , H TI and H TR is the same as that of H s , except that the parameter values and cluster distribution are different.
  • the channel transmission matrix of the multi-link channel model is as follows:
  • the other relevant large-scale parameters include delay spread (DS), azimuth spread of arrival (ASA), azimuth spread of departure (ASD), pitch angle of arrival spread ( elevation spread of arrival (ESA), elevation spread of departure (ESD), Rice factor KR and cross-polarization ratio (XPR).
  • DS delay spread
  • ASSD azimuth spread of arrival
  • ASD azimuth spread of departure
  • ESA elevation spread of arrival
  • ESD elevation spread of departure
  • XPR cross-polarization ratio
  • the method of generating large-scale parameters is the same.
  • delay extension DS as an example, as shown in the following formula:
  • P (P T , P R ) consists of the transmitter and receiver position vectors
  • P T (t) (x T (t), y T (t), z T (t))
  • P R (t) (x R (t), y R (t), z R (t)) respectively represent the coordinate vector of the sending end and the coordinate vector of the receiving end at time t, and their initial values are generated according to the simulation environment and requirements.
  • X DS (P) is a spatially consistent normally distributed variable with a mean of 0 and a variance of 1 generated by the sine wave superposition method. Represents the mean value of DS in the f c frequency band, Indicates the variance of DS in the f c band.
  • the configuration values can be divided into three types.
  • land mobile communication scenarios 1.5m ⁇ h UT ⁇ 22.5m
  • UAV scenarios (22.5m ⁇ h UT ⁇ 300m)
  • the values Refer to the values in Table B1.2 in the 3GPPTR36.777 standardization document
  • low-orbit satellite communication scenarios refer to the values in Table 6.7-2 in the 3GPPTR38.811 standardization document.
  • the carrier frequency is between 2-4GHz The calculation is as follows
  • Multiplying the cross-correlation matrix between large-scale parameters can obtain the values of all large-scale parameters with spatial consistency in the logarithmic domain. Then, the values in the logarithmic domain are needed. Convert to linear domain. At this point, the large-scale parameters of the channel can be obtained.
  • Step S401 Use the ellipsoidal Gaussian scattering distribution to obtain the position of the scatterer, that is, use
  • the scatterers in the nth cluster centered on the standard deviation on the three coordinate axes are and Gaussian distribution, after obtaining the positions of the scatterers, they can be converted into spherical coordinates, relative to the position of the first transmitting antenna and the location of the first receiving antenna
  • the position of the scatterer within the nth cluster of and It can be expressed as and in, and Respectively represent the distance, azimuth angle and pitch angle from the m-th sub-path of the n-th cluster to the antenna array of X (X ⁇ T,R ⁇ , representing the transmitting end and receiving end respectively);
  • Step S402 Calculate the intra-cluster sub-path delay at the initial moment.
  • the intra-cluster sub-path delay can be calculated by Calculate, where, express and The delay of the virtual link between is time t 0 arrive the distance between, is time t 0 arrive the distance between. is the distance between the first hop cluster and the last hop cluster, and ⁇ link is a non-negative random variable obeying exponential distribution.
  • Step S403. In a large-scale antenna array, the power of the sub-paths within the cluster will vary along the time axis and the array axis, and is usually modeled as a lognormal process varying over time and a lognormal process varying along the array, non-normalized intra-cluster sub-radius power for:
  • DS is the root mean square delay spread
  • Z n is the shadow term of each cluster in dB
  • r ⁇ is the delay distribution scaling factor
  • ⁇ n (p, q) is a two-dimensional space lognormal process , used to simulate smooth power changes on the antenna array.
  • the power of the final sub-path within the cluster can be obtained by normalizing the power of all clusters. If the cluster is newly generated, you can use Replace with get and The initial power of the m-th sub-path in the n-th cluster.
  • Step S404 For the surviving clusters, small-scale parameters such as power and delay of the sub-paths within the cluster need to be updated at different times.
  • the trajectory segment at time t 1 that is, at the next time after cluster generation, the p-th transmitting antenna
  • the coordinates are Among them, the coordinates of the p-th transmitting antenna at the initial time able to pass calculate, able to pass calculate.
  • to first hop cluster distance can be passed Calculated, it can be obtained in the same way arrive distance
  • the delay of the sub-path within the cluster at time t 1 Using the geographical locations of the sender, receiver and scatterer at the previous moment, we can get and
  • the space-time-frequency non-stationarity of the model is mainly reflected in two aspects. On the one hand, it is the parameters that change in space-time-frequency, and on the other hand, it is the birth and death process of clusters in the space-time-frequency domain.
  • N qp (t) is the number of clusters
  • N surv (t) is the number of surviving clusters, which is determined by the cluster survival probability P surv ( ⁇ t, ⁇ r, ⁇ f)
  • N new (t) is the number of new clusters, It obeys the Poisson distribution with mean E[N new (t)].
  • ⁇ G is the birth rate of the cluster
  • ⁇ R is the combination rate (death rate) of the cluster.
  • the joint survival probability of the sender and receiver clusters is:
  • the average number of new clusters is:
  • the survival probability of a cluster on the frequency axis is:
  • F( ⁇ f BD ) and Can be determined through channel measurements Represents scene-dependent correlation factors on the frequency axis.
  • the survival probability of the cluster is:
  • the average number of new clusters is:
  • the average number of new clusters is:
  • D qp (t) is the straight-line distance between the sending and receiving ends at time t
  • D is the initial distance between the sending and receiving ends
  • ⁇ s is the scattering coefficient of the pipe wall
  • ⁇ h 0.
  • the 6G universal channel model can be simplified into multiple dedicated channel models. As shown in table 2.
  • PL is the path loss
  • SH is the shadow fading
  • BL is the blocking effect
  • AL is the atmospheric absorption loss.
  • MT MR
  • LoS non-line-of-sight
  • K R (t) is the Rice factor.
  • ⁇ * ⁇ T represents transposition
  • f c represents carrier frequency
  • Patterns corresponding to vertical polarization and horizontal polarization is the cross-polarization power ratio
  • ⁇ represents the joint polarization imbalance
  • ⁇ represents the joint polarization imbalance
  • ⁇ represents time t from to The azimuth arrival angle and pitch arrival angle corresponding to the m-th sub-radius
  • the azimuth departure angle and pitch departure angle corresponding to the LoS path and is time t from to
  • the azimuth arrival angle and pitch arrival angle corresponding to the LoS path and is a random phase obeying the (0,2 ⁇ ] uniform distribution
  • is under NLoS conditions arrive
  • the power of the m-th sub-path of the n-th path is the delay of the LoS path at time t
  • the other relevant large-scale parameters include delay spread (DS), azimuth spread of arrival (ASA), azimuth spread of departure (ASD), pitch angle of arrival spread ( elevation spread of arrival (ESA), elevation spread of departure (ESD), Rice factor KR and cross-polarization ratio (XPR).
  • DS delay spread
  • ASSD azimuth spread of arrival
  • ASD azimuth spread of departure
  • ESA elevation spread of arrival
  • ESD elevation spread of departure
  • XPR cross-polarization ratio
  • the method of generating large-scale parameters is the same.
  • delay extension DS as an example, as shown in the following formula:
  • P (P T , P R ) consists of the transmitter and receiver position vectors
  • P T (t) (x T (t), y T (t), z T (t))
  • P R (t) (x R (t), y R (t), z R (t)) respectively represent the coordinate vector of the sending end and the coordinate vector of the receiving end at time t, and their initial values are generated according to the simulation environment and requirements.
  • X DS (P) is a spatially consistent normally distributed variable with a mean of 0 and a variance of 1 generated by the sine wave superposition method. Represents the mean value of DS in the f c frequency band, Indicates the variance of DS in the f c band.
  • the configuration values can be divided into three types.
  • the values in this embodiment can refer to Table 7.5-6 in the 3GPP TR 38.901 standardization document.
  • the eight large-scale parameters are independently generated using this method, and multiplied by the cross-correlation matrix between the large-scale parameters can obtain spatial consistency.
  • the values of all large-scale parameters are in the logarithmic domain. Then, the values in the logarithmic domain need to be converted into the linear domain. At this point, the large-scale parameters of the channel can be obtained.
  • Step S401 Use the ellipsoidal Gaussian scattering distribution to obtain the position of the scatterer, that is, use
  • the scatterers in the nth cluster centered on the standard deviation on the three coordinate axes are and Gaussian distribution, after obtaining the positions of the scatterers, they can be converted into spherical coordinates, relative to the position of the first transmitting antenna and the location of the first receiving antenna
  • the position of the scatterer within the nth cluster of and It can be expressed as and in, and Respectively represent the distance, azimuth angle and pitch angle from the m-th sub-path of the n-th cluster to the antenna array of X (X ⁇ T,R ⁇ , representing the transmitting end and receiving end respectively);
  • Step S402 Calculate the intra-cluster sub-path delay at the initial moment.
  • the intra-cluster sub-path delay can be calculated by Calculate, where, express and The delay of the virtual link between is time t 0 arrive the distance between, is time t 0 arrive the distance between. is the distance between the first hop cluster and the last hop cluster, and ⁇ link is a non-negative random variable obeying exponential distribution.
  • Step S403. In a large-bandwidth large-scale antenna array, the power of the sub-paths within the cluster It will change along the time axis, frequency axis and array axis, and is usually modeled as a lognormal process changing with time and a lognormal process changing along the array, non-normalized intra-cluster sub-radius power for:
  • DS is the root mean square delay spread
  • Z n is the shadow term of each cluster in dB
  • r ⁇ is the delay distribution scaling factor
  • ⁇ n (p, q) is a two-dimensional space lognormal process , used to simulate smooth power changes on the antenna array.
  • Step S404 For the surviving clusters, small-scale parameters such as power and delay of the sub-paths within the cluster need to be updated at different times.
  • the p-th transmitting antenna For the trajectory segment at time t 1 , that is, at the next time after cluster generation, the p-th transmitting antenna The coordinates are Among them, the pth transmitting antenna at the initial time coordinate of able to pass calculate, able to pass calculate.
  • to first hop cluster distance can be passed Calculated, it can be obtained in the same way arrive distance
  • the delay of the sub-path within the cluster at time t 1 Using the geographical locations of the sender, receiver and scatterer at the previous moment, we can get and
  • the space-time-frequency non-stationarity of the model is mainly reflected in two aspects. On the one hand, it is the parameters that change in space-time-frequency, and on the other hand, it is the birth and death process of clusters in the space-time-frequency domain.
  • N qp (t) is the number of clusters
  • N surv (t) is the number of surviving clusters, which is determined by the cluster survival probability P surv ( ⁇ t, ⁇ r, ⁇ f)
  • N new (t) is the number of new clusters, It obeys the Poisson distribution with mean E[N new (t)].
  • ⁇ G is the birth rate of the cluster
  • ⁇ R is the combination rate (death rate) of the cluster.
  • the joint survival probability of the sender and receiver clusters is:
  • the average number of new clusters is:
  • the survival probability of a cluster on the frequency axis is:
  • F( ⁇ f BD ) and Can be determined through channel measurements Represents scene-dependent correlation factors on the frequency axis.
  • the survival probability of the cluster is:
  • the average number of new clusters is:
  • the small-scale parameters of different antenna pairs can be obtained. At this point, all parameter values in the channel matrix can be obtained.

Landscapes

  • Physics & Mathematics (AREA)
  • Electromagnetism (AREA)
  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Variable-Direction Aerials And Aerial Arrays (AREA)
  • Mobile Radio Communication Systems (AREA)

Abstract

本发明公开了一种适用于全频段全场景的6G普适信道建模方法。具体包括以下步骤：S1、设置传播场景和传播条件，确定载波频率、天线类型和收发端布局等；S2、生成路径损耗、阴影衰落以及阻挡效应等大尺度衰落；S3、生成具有空间一致性的大尺度参数；S4、生成服从椭球高斯散射分布的散射***置，并根据收发端和散射体的位置计算簇的时延、角度和功率，生成信道系数；S5、根据收发端运动和簇的生灭过程，进行大、小尺度参数的更新，生成新的信道系数。本发明中公开的6G普适几何随机信道模型是目前业界唯一一个可以普遍适用于全频段、全覆盖场景和全应用场景的信道模型，对于6G信道模型标准化、6G共性理论技术研究及***融合构建至关重要。

Description

一种适用于全频段全场景的6G普适信道建模方法 技术领域
本发明涉及一种适用于全频段全场景的6G普适信道建模方法，属于无线通信技术领域。
背景技术
6G无线信道可以总结为全频段(sub-6 GHz/毫米波/太赫兹/光无线频段)、全覆盖场景(空天地海一体化，包含卫星、无人机、陆地和海洋通信信道)和全应用场景(车联网、高铁信道、大规模天线阵列、可重构智能表面、工业物联网等)信道，如下图2所示。与此同时，6G全频段全场景信道也展现出了众多新的信道特性，为6G信道建模工作带来了新的挑战。
在全频段方面，由于毫米波和太赫兹等高频段的应用，无线信道展示出大带宽、频域非平稳、漫散射、大路径损耗、阻挡效应和大气吸收等特性；在可见光波段，信道将不再具有小尺度衰落，并展现出可忽略的多普勒效应和频域非平稳特性。在全覆盖方面，除陆地移动通信场景外，还包括卫星通信、无人机通信和海洋通信场景。其中，卫星通信信道需要考虑由于卫星的快速移动带来的多普勒频移、降雨衰落，以及电离层效应。在无人机通信***中，主要考虑无人机的三维(3D)任意运动和大尺度参数受无人机高度的影响。在全应用方面，车联网信道展示出多普勒频移，以及由于收发两端和簇的多移动性带来的时域非平稳特性；在时速超过500km/h的更高速移动场景下，信道经历更强的多普勒频移和更明显的时域非平稳；在真空管道中运行的超高速列车场景中，还需考虑真空管道波导效应的影响；超大规模天线阵列信道展示出球面波特性和空域非平稳特性；工业物联网信道中需要考虑超密集的散射体分布和多移动性；此外，应用可重构智能表面技术的无线信道精确建模也是必须研究的内容。
考虑到多种不同新技术的混合应用会带来不同信道特性的结合，6G信道建模的一个重要挑战就在于如何综合考虑多种多样的信道特性，提出一个适用于全频段全场景的普适信道模型。例如，在高速移动的场景中同时应用毫米波/太赫兹频段和大规模天线技术，无线信道会同时展示出空间-时间-频率非平稳特性、空间一致性(即多用户场景下，相近用户的信道系数具有相关性或单用户在不同的轨迹点具有空间相关性)以及多频段相关性等。
综上所述，建立一个准确的、普适的、灵活的6G信道模型刻不容缓。B5GCM、3GPP TR 38.901、IMT-2020和QuaDRiGa等5G标准化信道模型已在该问题上做出尝试，但它们都不能准确和全面的描述前述的所有特性。在全频谱方面，这些模型都不适用于可见光频段，并且或多或少的忽略了毫米波/太赫兹频段的一些特性，例如QuaDRiGa忽略了对大气吸收和阻挡效应建模，3GPP TR 38.901和IMT-2020忽略了高频段的频率非平稳特性；在全覆盖方面，这些信道模型都是针对的陆地移动通信信道，不能适用于卫星、无人机和海洋通信场景；在全应用方面，他们都不支持对超高铁信道、可重构智能表面和工业物联网信道建模，并且，3GPP TR 38.901和IMT-2020未考虑(超)大规模天线阵列的球面波和空域非平稳特性。综上所述，这些模型仍旧缺少普适性，并没有考虑之前提到的所有信道特性。为了弥补这一研究空白，本发明提出普适信道建模理论，并将其应用到几何随机信道模型中，提出并公开了一种适用于全频段全场景的6G普适信道建模方法。
发明内容
技术问题：本发明的目的是提出一种适用于全频段全场景的6G普适信道建模方法，该方法是一种普遍适用于全频段(sub-6 GHz/毫米波/太赫兹/可见光频段)、全覆盖场景(空天地海一体化，包含卫星、无人机、陆地和海洋通信场景)和全应用场景(车联网、高铁信道、大规模天线阵列、可重构智能表面和工业物联网场景)的6G普适几何随机信道建模方法。
技术方案：本发明的一种适用于全频段全场景的6G普适信道建模方法，6G普适几何随机信道模型中发送端和接收端均采用大规模均匀线阵，模型为多跳传播，其中，为发送端天线阵列的第p个阵元，为接收端天线阵列的第q个阵元，发送端和接收端天线阵元间的距离为δT(δR)；为xy平面内发送端和接收端天线阵列的方位角，为发送端和接收端天线阵列的俯仰角；到的第n条传播路径，n＝1,2,3,...,Nqp(t)，其中为第n条路径上靠近发送端的首跳簇，为靠近接收端的末跳簇，将两簇之间的传播路径建模为虚拟的链路；当首跳簇和末跳簇间的虚拟链路时延为零时，模型退化为单跳模型。此外，Nqp(t)为时刻t时到的路径数量，在双簇模型中对应着Nqp(t)个簇对，首跳簇和末跳簇一一对应，在单簇模型中对应着Nqp(t)个簇；微观上，对第n条路径上的簇和进行分析，能见簇内存在Mn(t)个散射体，表示中的第m个散射体， 表示中的第m个散射体；从路径的角度来看，理解为到的第m个子径连接的散射体,理解为至的第m个子径连接的散射体；因此，和是时刻t时从至的第m个子径对应的方位离开角和俯仰离开角，和是时刻t时从至的第m个子径对应的方位到达角和俯仰到达角，此外，模型对发送端、接收端以及簇的运动情况分别建模，支持收发端及簇的任意速度和轨迹的三维运动，其中分别表示发送端、接收端，首跳簇以及末跳簇的运动速度，分别表示发送端，接收端，首跳簇以及末跳簇的运动方向的方位角，分别表示发送端，接收端，首跳簇以及末跳簇的运动方向的俯仰角；
6G普适几何随机信道模型的信道矩阵表示为：
H＝[PL·SH·BL·WE·AL]1/2·Hs
其中，PL,SH,BL,WE,AL为大尺度衰落，PL为路径损耗，SH为阴影衰落，BL为阻挡效应，AL为大气吸收损耗，WE为天气影响损耗，Hs为小尺度衰落。
所述小尺度衰落Hs如下：
其中，MT为发送端天线阵列中天线阵元数量，MR为接收端天线阵列中天线阵元数量，为发送端天线阵元与接收天线阵元之间的信道冲激响应，表示为视距LoS分量与非视距NLoS分量的叠加：
其中，KR(t)为莱斯因子，和分别如下：

其中，{*}T表示转置，fc表示载波频率，和表示在不同频段上天线单元对应垂直极化和水平极化的方向图，为交叉极化功率比，μ表征联合极化不均衡，和是时刻t时从至的LoS路径对应的方位离开角以及俯仰离开角,和是时刻t时从至的LoS路径对应的方位到达角以及俯仰到达角， 和是服从(0,2π]均匀分布的随机相位，ψl,m＝108/fc 2为法拉第旋转角，此处计算法拉第旋转角时fc的单位是GHz，是NLoS条件下到的第n条路径中的第m个子径的功率，是在时刻t时LoS路径的时延，是在时刻t时发送端天线阵元与接收天线阵元间的矢量距离，c为光速。是在时刻t时与之间的第n条路经的第m个子径的时延，是在时刻t时与之间的第n条路经的第m个子径的功率，以上所有参数均为时变参数。
所述6G普适信道建模方法，在海洋通信信道场景中，模型分别将LoS径、粗糙的海洋表面和海面上空蒸发波导的多径传播这三个部分建模为和并使用功率系数，S1和S2控制对应部分随两船距离变化的消失和出现，即将计算公式中的NLoS部分分为和两部分，其中S1+S2＝1，在工业物联网信道中将镜面反射分量和密集多径分量分别建模为和其中 和的建模方法与相同，只是参数值和簇的分布不同；
所述6G普适信道建模方法，在可重构智能表面场景中，将信道分为发送端到可重构智能表面的子信道HTI，可重构智能表面到接收端的子信道HIR和发送端到接收端的子信道HTR，对三个子信道分别建模并引入相移对角矩阵Φ实现对信道环境的智能调控，HIR,HTI和HTR与的计算方法与Hs相同，只是参数值和簇的分布不同。
所述6G普适信道建模方法，针对可见光频段信道建模时，一方面光信号波长极短，接收端尺寸通常为几百万个波长，不会发生几个波长上的信号快速衰落；另一方面由于可见光通信***中LED灯发出的是非相干光，光信号没有相位信息，在接收端的实数多径信号叠加后并不会引起快衰落，而表现为缓慢变化的阴影衰落，因此，虽然现在的可见光模型表示形式是多径叠加的信道冲激响应形式，其本质是建模了PL和SH的大尺度模型，所以Hs＝0， pH,pV为LED面阵的行和列数。
所述6G普适信道建模方法，在多链路场景中，假设基站数目为NBS，用户数为NMS，多链路信道模型的信道传输矩阵如下公式所示：
每条链路对应的i＝1,2…NBS,j＝1,2…NMS，即为前述的单链路信道模型H。
所述6G普适信道建模方法，信道矩阵H生成详细分为以下步骤：
S1、设置传播场景和传播条件，确定载波频率和天线类型、收发端布局以及收发端运动轨迹等；
S2、生成路径损耗、阴影衰落、氧气吸收以及阻挡效应的大尺度衰落；本方法主要聚焦于小尺度衰落的建模，此部分计算可参考标准化信道模型或业界经典模型；
S3、根据收发端位置和运动情况生成具有空间一致性的时延扩展、4个角度扩展的大尺度参数；
除SH外，其余相关大尺度参数有时延扩展DS、方位到达角扩展ASA、方位离开角扩展ASD、俯仰到达角扩展ESA、俯仰离开角扩展ESD、莱斯因子KR和交叉极化比XPR，时延扩展DS的生成如下公式所示：
其中，P＝(PT,PR)由收发端位置矢量组成，PT(t)＝(xT(t),yT(t),zT(t))和PR(t)＝(xR(t),yR(t),zR(t))分别表示t时刻发送端的坐标矢量和接收端的坐标矢量，其初始值根据仿真环境及要求生成；XDS(P)是用正弦波叠加法生成的服从均值为0、方差为1的空间一致性的正态分布变量，表示DS在fc频段的均值，表示DS在fc频段的方差。根据终端的高度hUT，的配置值可分为三种类型；对于陆地移动通信场景1.5m≤hUT≤22.5m，可以参考3GPPTR38.901标准化文档中的表7.5-6中的值；对于无人机场景22.5m≤hUT≤300m，取值参考3GPPTR36.777标准化文档中的表B1.2中的值；对于低轨卫星通信场景，取值参考3GPPTR38.811标准化文档中的表6.7-2中的值，城市宏小区UMa场景NLoS条件下，载波频率介于2-4GHz情况下的计算如下
其他的大尺度参数的生成与时延扩展DS的生成过程相同，8个大尺度参数都生成后，乘以大尺度参数间的互相关矩阵可以得到具有空间一致性的全部大尺度参数在对数域的值，然后，需要将对数域的值转化到线性域，至此，信道的大尺度参数都可得到；
S4、生成服从椭球高斯散射分布的散射体，并根据收发端和散射体的地理位置信息计算簇的时延、角度和功率，生成信道系数；
S5、根据收发端运动和簇的生灭过程，进行大、小尺度参数的更新，生成新的信道系数，模型的空-时-频非平稳性主要体现在两个方面，一方面是空-时-频变化的参数，另一个方面是簇在空-时-频域的生灭过程，t时刻簇数量计算如下：
Nqp(t)＝Nsurv(t)+Nnew(t)
其中，Nqp(t)为簇的数量，Nsurv(t)为幸存簇的数目，由簇的生存概率Psurv(Δt,Δr,Δf)决定，Nnew(t)为新生簇的数目，服从均值为E[Nnew(t)]的泊松分布，定义λG为簇的出生率、λR为簇的结合率即灭亡率。
所述步骤S4具体为：
步骤S401、使用椭球高斯散射分布，获取散射体的位置，即以为中心的第n个簇内的散射体服从标准差在三个坐标轴上分别为和的高斯分布，在获得散射体的位置后可将它们转换成球坐标，相对于第一根发送天线的位置和第一根接收天线的位置的第n个簇内的散射体的位置和可以表示为和其中，和分别表示第n个簇第m个子径到X(X∈{T,R}，分别表示发送端和接收端)天线阵列的距离、方位角和俯仰角；
步骤S402、初始时刻簇内子径时延计算，在多跳信道模型中，簇内子径的时延可通过计算，其中，表示和之间虚拟链路的时延，为t0时刻到之间的距离，为t0时刻到之间的距离。为首跳簇和末跳簇间的距离，τlink为服从指数分布的非负随机变量；
步骤S403、在大规模天线阵中，簇内子径的功率会沿着时间轴和阵列轴变化，通常将其建模为随时间变化的对数正态过程和沿阵列变化的对数正态过程，非归一化的簇内子径功率为：
其中，DS为均方根时延扩展，Zn是以dB为单位的每簇阴影项，rτ为时延分布比例因子，ξn(p,q)是一个二维空间对数正态过程，用于模拟天线阵列上平滑的功率变化；
在大带宽场景下，需要考虑频域非平稳特性，所以我们在频域上将功率值乘以项，其中，是依赖于频率的常数因子，最后，通过将所有簇的功率归一化即可得到最终簇内子径的功率如果簇是新产生的，通过将替换成 得到和之间第n簇内第m个子径的初始功率；
步骤S404、对于幸存的簇，需要在不同的时刻更新簇内子径的功率、时延等小尺度参数，对于时刻t1处的轨迹段，即在簇生成后的下一个时刻,第p根发送天线的坐标为其中，初始时刻第p根发送天线的坐标通过计算，在t1时刻，第n个首跳簇中第m个散射体的坐标通过计算。在t1时刻，到的距离可以通过计算得到，同理可得到的距离t1时刻的簇内子径的时延利用前一时刻发送端、接收端和散射体的地理位置可以得到和(t＝t2,t3,...)。
所述步骤S5中：为了更精确地建模簇的空-时-频演进过程，引入两种采样间隔，一种是时域采样间隔Δt，频域采样间隔Δf和空间域(阵列域)采样间隔Δr，信道参数连续更新；另一种是的Δt，Δf和Δr的整数倍，分别是ΔtBD，ΔfBD和ΔrBD，在这些采样点发生簇的生灭演进过程，发送端和接收端簇沿阵列轴和时间轴的生存概率：

其中，和分别表示发送和接收天线单元在阵列轴上以及时间轴上的位置差，和分别表示在阵列轴和时间轴依赖于场景的相关因子，发送端和接收端簇的联合生存概率为：
新生簇的平均数量为：
当研究大带宽时，在频率轴上也会存在簇的生灭过程，频率轴上簇的生存概率为：
其中，F(ΔfBD)和可以通过信道测量决定，表示在频率轴依赖于场景的相关因子，
综上所述，同时考虑空-时-频域簇的生灭过程时，簇的生存概率为：
新生簇的平均数量为：
超高速列车场景中，考虑真空管道超高速列车场景的波导效应和管道壁粗糙度对信道的影响，新生簇的平均数量为：

其中，Dqp(t)为t时刻收发端的直线距离，D为收发端的初始距离，ρs为管道壁散射系数，为粗糙度σh＝0时的散射系数。
有益效果：本发明提出了一种普适信道建模理论，并将该理论应用于几何随机信道模型,使用基于簇的几何随机信道建模方法和框架，使用统一的信道冲激响应表达式，对6G全频段全场景信道特性进行建模，提出了一种基本适用于sub-6GHz、毫米波、太赫兹和可见光等全频段信道，低轨卫星、无人机和海洋通信等全覆盖场景信道，以及超大规模天线阵列、工业物联网、可重构智能表面等全应用场景信道的6G普适几何随机信道模型。并且，通过调整6G普适信道模型的参数，可简化为特定频段、特定场景的专用信道模型。6G普适几何随机信道模型对于6G信道模型标准化、6G共性理论技术研究及***融合构建至关重要。
附图说明
图1为本发明实施例中的流程图。
图2为本发明中6G无线信道示意图。
图3为本发明中普适信道建模理论示意图。
图4为本发明中6G普适几何随机信道模型示意图。
具体实施方式
为了实现上述目的，本发明提出了普适信道建模理论，并基于该理论提出了6G普适几何随机信道模型，因此，本发明主要包含普适信道模型建模理论和6G普适几何随机信道模型构建两部分：
1.普适信道建模理论
普适信道建模理论即使用统一的信道建模方法和框架、统一的信道冲激响应表达式、综合考虑6G全频段全场景信道的统计特性，构建普遍适用于6G各个频段、各个场景，并能准确反应其信道特性的6G普适信道模型，如图3所示。同时，6G普适信道模型可通过调整信道模型的参数，简化为适用于特定频段、特定场景的专用信道模型。通过对6G普适信道模型进行分析，可以研究信道模型参数、信道特性和通信***性能之间的复杂映射关系，并且其作为一个统一的信道模型框架，对6G信道模型标准化、6G共性理论技术研究及***融合构建至关重要。
2.6G普适几何随机信道模型
6G普适几何随机信道模型示意图如图4示。模型中天线类型可以是均匀线阵、均匀面阵等天线阵列类型，支持任意天线极化类型，示意图中发送端和接收端均采用大规模均匀线阵。示意图中模型为多跳传播，其中，为发送端天线阵列的第p个阵元，接收端端天线阵列的第q个阵元，发送端(接收端)天线阵元间的距离为δT(δR)。为xy平面内发送端(接收端)天线阵列的方位角，为发送端(接收端)天线阵列的俯仰角。为了更易理解，我们在图中只描述了到的第n(n＝1,2,3,...,Nqp(t))条传播路径，其中为第n条路径上靠近发送端的首跳簇，为靠近接收端的末跳簇，将两簇之间的传播路径建模为虚拟的链路。当首跳簇和末跳簇间的虚拟链路时延为零时，模型退化为单跳模型。此外，Nqp(t)为时刻t时到的路径数量，在双簇模型中对应着Nqp(t)个簇对，首跳簇和末跳簇一一对应，在单簇模型中对应着Nqp(t)个簇；微观上，对第n条路径上的簇和进行分析，能见簇内存在Mn(t)个散射体，表示中的第m个散射体，表示中的第m个散射体；从路径的角度来看，理解为到的第m个子径连接的散射体,理解为至的第m个子径连接的散射体；因此，和是时刻t时从至的第m个子径对应的方位离开角和俯仰离开角，和是时刻t时从至的第m个子径对应的方位到达角和俯仰到达角，此外，模型对发送端、接收端以及簇的运动情况分别建模，支持收发端及簇的任意速度和轨迹的三维运动，其中分别表示发送端、接收端，首跳簇以及末跳簇的运动速度，分别表示发送端，接收端，首跳簇以及末跳簇的运动方向的方位角，分别表示发送端，接收端，首跳簇以及末跳簇的运动方向的俯仰角。
6G普适几何随机信道模型的信道矩阵可表示为：
H＝[PL·SH·BL·WE·AL]1/2·Hs
其中，PL,SH,BL,WE,AL为大尺度衰落，PL为路径损耗，SH为阴影衰落，BL为阻挡效应，AL为大气吸收损耗，例如毫米波频段的氧气吸收损耗和太赫兹频段的分子吸收损耗等，WE为天气影响损耗，例如卫星通信场景中的雨水衰落。本发明主要聚焦于小尺度衰落Hs的计算，方法如下：
其中，MT(MR)为发送端(接收端)天线阵列中天线阵元数量，为发送端天线阵元与接收天线阵元之间的信道冲激响应，可以表示为视距(LoS)分量与非视距(NLoS)分量的叠加：
其中，KR(t)为莱斯因子。和分别计算如下：

其中，{*}T表示转置，fc表示载波频率，和表示在不同频段上天线单元对应垂直极化和水平极化的方向图，为交叉极化功率比，μ表征联合极化不均衡，和是时刻t时从至的LoS路径对应的方位离开角以及俯仰离开角,和是时刻t时从至的LoS路径对应的方位到达角以及俯仰到达角。和是服从(0,2π]均匀分布的随机相位，ψl,m＝108/fc 2为法拉第旋转角，此处计算法拉第旋转角时fc的单位是GHz，是NLoS条件下到的第n个路径的第m个子径的功率，时刻t时LoS路径的时延是在时刻t时发送端天线阵元与接收天线阵元间的矢量距离，c为光速。是在时刻t时与之间的第n个路径的第m个子径的时延。以上所有参数均为时变参数。
值得注意的是，在海洋通信信道场景中，本模型分别将LoS径、粗糙的海洋表面和海面上空蒸发波导的多径传播这三个部分建模为和并使用功率系数S1和S2控制对应部分随两船距离变化的消失和出现，即将信道冲激响应矩阵计算中的NLoS部分分为两部分，其中S1+S2＝1。在工业物联网信道中将镜面反射分量和密集多径分量分别建模为和其中和的建模方法与相同，只是参数值和簇的分布不同。在可重构智能表面场景中，将信道分为发送端到可重构智能表面的子信道HTI，可重构智能表面到接收端的子信道HIR和发送端到接收端的子信道HTR，对三个子信道分别建模并引入相移对角矩阵Φ实现对信道环境的智能调控。HIR,HTI和HTR与的计算方法与Hs相同，只是参数值和簇的分布不同。
在可见光频段时，一方面光信号波长极短，接收端尺寸通常为几百万个波长，不会发生几个波长上的信号快速衰落；另一方面由于可见光通信***中LED灯发出的是非相干光，光信号没有相位信息，在接收端的实数多径信号叠加后并不会引起快衰落，而表现为缓慢变化的阴影衰落。因此，虽然现在的可见光模型表示形式是多径叠加的信道冲激响应形式，其本质是建模了PL和SH的大尺度模型，所以Hs＝0，pH,pV为LED面阵的行和列数。
考虑多用户场景时，假设基站数目为NBS，用户数为NMS，多链路信道模型的信道传输矩阵如下公式所示：
每条链路对应的i＝1,2…NBS,j＝1,2…NMS，即为前述的单链路信道模型H。
信道系数生成详细分为以下步骤：
S1、设置传播场景和传播条件，确定载波频率和天线类型、收发端布局以及收发端运动轨迹等；
S2、生成路径损耗、阴影衰落、氧气吸收以及阻挡效应等大尺度衰落值；本发明主要聚焦于小尺度衰落的建模，此部分可参考标准化信道模型或业界经典模型。
S3、根据收发端位置和运动情况生成具有空间一致性的时延扩展、4个角度扩展等大尺度参数；
除SH外，其余相关大尺度参数有时延扩展(delay spread，DS)、方位到达角扩展(azimuth spread of arrival，ASA)、方位离开角扩展(azimuth spread of departure，ASD)、俯仰到达角扩展(elevation spread of arrival，ESA)、俯仰离开角扩展(elevation spread of departure，ESD)、莱斯因子KR和交叉极化比(cross-polarization ratio，XPR)。大尺度参数的生成方法相同，此处以生成时延扩展DS为例，如下公式所示：
其中，P＝(PT,PR)由收发端位置矢量组成，PT(t)＝(xT(t),yT(t),zT(t))和PR(t)＝(xR(t),yR(t),zR(t))分别表示t时刻发送端的坐标矢量和接收端的坐标矢量，其初始值根据仿真环境及要求生成。XDS(P)是用正弦波叠加法生成的服从均值为0、方差为1的空间一致性的正态分布变量，表示DS在fc频段的均值，表示DS在fc频段的方差。根据终端的高度hUT，的配置值可分为三种类型。对于陆地移动通信场景(1.5m≤hUT≤22.5m)，可以参考3GPPTR38.901标准化文档中的表7.5-6中的值；对于无人机场景(22.5m≤hUT≤300m)，取值参考3GPPTR36.777标准化文档中的表B1.2中的值；对于低轨卫星通信场景，取值参考3GPPTR38.811标准化文档中的表6.7-2中的值。城市宏小区(UMa)场景NLoS条件下，载波频率介于2-4GHz情况下的计算如下
8个大尺度参数均采用这种方法独立生成，乘以大尺度参数间的互相关矩阵可以得到具有空间一致性的全部大尺度参数在对数域的值，然后，需要将对数域的值转化到线性域。至此，信道的大尺度参数都可得到。
S4、生成服从椭球高斯散射分布的散射体，并根据收发端和散射体的地理位置信息计算簇的时延、角度和功率，生成信道系数；
步骤S401、使用椭球高斯散射分布，获取散射体的位置，即以为中心的第n个簇内的散射体服从标准差在三个坐标轴上分别为和的高斯分布，在获得散射体的位置后可将它们转换成球坐标，相对于第一根发送天线的位置和第一根接收天线的位置的第n个簇内的散射体的位置和可以表示为和其中，和分别表示第n个簇第m个子径到X(X∈{T,R}，分别表示发送端和接收端)天线阵列的距离、方位角和俯仰角；
步骤S402、初始时刻簇内子径时延计算，在多跳信道模型中，簇内子径的时延可通过计算，其中，表示和之间虚拟链路的时延，为t0时刻到之间的距离，为t0时刻到之间的距离。为首跳簇和末跳簇间的距离，τlink为服从指数分布的非负随机变量。
步骤S403、在大规模天线阵中，簇内子径的功率会沿着时间轴和阵列轴变化，通常将其建模为随时间变化的对数正态过程和沿阵列变化的对数正态过程，非归一化的簇内子径功率为：
其中，DS为均方根时延扩展，Zn是以dB为单位的每簇阴影项，rτ为时延分布比例因子，ξn(p,q)是一个二维空间对数正态过程，用于模拟天线阵列上平滑的功率变化。
在大带宽场景下，需要考虑频域非平稳特性，所以我们在频域上将功率值乘以项，其中，是依赖于频率的常数因子。最后，通过将所有簇的功率归一化即可得到最终簇内子径的功率如果簇是新产生的，可以通过将替换成得到和之间第n簇内第m个子径的初始功率。
步骤S404、对于幸存的簇，需要在不同的时刻更新簇内子径的功率、时延等小尺度参数。对于时刻t1处的轨迹段，即在簇生成后的下一个时刻,第p根发送天线的坐标为其中，初始时刻第p根发送天线的坐标可以通过计算，可以通过 计算。在t1时刻，到首跳簇的距离可以通过计算得到，同理可得到的距离t1时刻的簇内子径的时延利用前一时刻发送端、接收端和散射体的地理位置可以得到和
S5、根据收发端运动和簇的生灭过程，进行大、小尺度参数的更新，生成新的信道系数。
模型的空-时-频非平稳性主要体现在两个方面，一方面是空-时-频变化的参数，另一个方面是簇在空-时-频域的生灭过程。t时刻簇数量计算如下：
Nqp(t)＝Nsurv(t)+Nnew(t)
其中，Nqp(t)为簇的数量，Nsurv(t)为幸存簇的数目，由簇的生存概率Psurv(Δt,Δr,Δf)决定，Nnew(t)为新生簇的数目，服从均值为E[Nnew(t)]的泊松分布。定义λG为簇的出生率、λR为簇的结合率(灭亡率)。为了更精确地建模簇的空-时-频演进过程，我们引入两种采样间隔，一种是时域采样间隔Δt，频域采样间隔Δf和空间域(阵列域)采样间隔Δr，信道参数连续更新；另一种是的Δt，Δf和Δr的整数倍，分别是ΔtBD，ΔfBD和ΔrBD，在这些采样点发生簇的生灭演进过程。发送端和接收端簇沿阵列轴和时间轴的生存概率：

其中，和分别表示发送(接收)天线单元在阵列轴上以及时间轴上的位置差。和分别表示在阵列轴和时间轴依赖于场景的相关因子。发送端和接收端簇的联合生存概率为：
新生簇的平均数量为：
当研究大带宽时，在频率轴上也会存在簇的生灭过程。频率轴上簇的生存概率为：
其中，F(ΔfBD)和可以通过信道测量决定，表示在频率轴依赖于场景的相关因子。
综上所述，同时考虑空-时-频域簇的生灭过程时，簇的生存概率为：
新生簇的平均数量为：
超高速列车场景中，考虑真空管道超高速列车场景的波导效应和管道壁粗糙度对信道的影响，新生簇的平均数量为：

其中，Dqp(t)为t时刻收发端的直线距离，D为收发端的初始距离，ρs为管道壁散射系数，为粗糙度σh＝0时的散射系数。
基于前述方法和利用发送端、接收端和散射体之间的几何关系，可以得到不同天线对的小尺度参数，至此可以得到信道矩阵中所有参数值，模型的建模方法和对应参数总结如下表所示：

表1.模型参数与建模方法
3.5模型简化
通过调整参数，可将6G普适信道模型简化为多个专用信道模型。如表2所示。
表2. 6G普适信道模型简化总结表


下面结合附图和具体实施例对本发明进行详细说明。本实施例以本发明技术方案为前提进行实施，给出了详细的实施方式和具体的操作过程，但本发明的保护范围不限于下述的实施例。
以毫米波段超大规模天线阵列场景为例，P6GCM的信道矩阵可表示为：
H＝[PL·SH·BL·AL]1/2·Hs
其中，PL为路径损耗，SH为阴影衰落，BL为阻挡效应，AL为大气吸收损耗。
其中，MT(MR)为发送端(接收端)天线阵列中天线阵元数量，为与之间的信道冲激响应，可以表示为LoS分量与非视距(NLoS)分量的叠加：
其中，KR(t)为莱斯因子。和分别计算如下：



其中，{*}T表示转置，fc表示载波频率，和表示在不同频段上天线单元对应垂直极化和水平极化的方向图，为交叉极化功率比，μ表征联合极化不均衡，和是时刻t时从至的第m个子径对应的方位离开角和俯仰离开角，和是时刻t时从至的第m个子径对应的方位到达角和俯仰到达角,和是时刻t时从至的LoS路径对应的方位离开角以及俯仰离开角,和是时刻t时从至的LoS路径对应的方位到达角以及俯仰到达角，和是服从(0,2π]均匀分布的随机相位，是NLoS条件下到的第n个路径的第m个子径的功率，是在时刻t时LoS路径的时延， 是在时刻t时发送端天线阵元与接收天线阵元间的矢量距离，c为光速。是在时刻t时与之间的第n个路径的第m个子径的时延。以上所有参数均为时变参数。
信道系数生成详细分为以下步骤：
S1、设置传播场景和传播条件，确定载波频率和天线类型、收发端布局以及收发端运动轨迹等；
S2、生成路径损耗、阴影衰落、氧气吸收以及阻挡效应等大尺度衰落值；本发明主要聚焦于小尺度衰落的建模，此部分可参考标准化信道模型或业界经典模型。
S3、根据收发端位置和运动情况生成具有空间一致性的时延扩展、4个角度扩展等大尺度参数；
除SH外，其余相关大尺度参数有时延扩展(delay spread，DS)、方位到达角扩展(azimuth spread of arrival，ASA)、方位离开角扩展(azimuth spread of departure，ASD)、俯仰到达角扩展(elevation spread of arrival，ESA)、俯仰离开角扩展(elevation spread of departure，ESD)、莱斯因子KR和交叉极化比(cross-polarization ratio，XPR)。大尺度参数的生成方法相同，此处以生成时延扩展DS为例，如下公式所示：
其中，P＝(PT,PR)由收发端位置矢量组成，PT(t)＝(xT(t),yT(t),zT(t))和PR(t)＝(xR(t),yR(t),zR(t))分别表示t时刻发送端的坐标矢量和接收端的坐标矢量，其初始值根据仿真环境及要求生成。XDS(P)是用正弦波叠加法生成的服从均值为0、方差为1的空间一致性的正态分布变量，表示DS在fc频段的均值，表示DS在fc频段的方差。根据终端的高度hUT，的配置值可分为三种类型。本实施例中的值可以参考3GPP TR 38.901标准化文档中的表7.5-6。8个大尺度参数均采用这种方法独立生成，乘以大尺度参数间的互相关矩阵可以得到具有空间一致性的全部大尺度参数在对数域的值，然后，需要将对数域的值转化到线性域。至此，信道的大尺度参数都可得到。
S4、生成服从椭球高斯散射分布的散射体，并根据收发端和散射体的地理位置信息计算簇的时延、角度和功率，生成信道系数；
步骤S401、使用椭球高斯散射分布，获取散射体的位置，即以为中心的第n个簇内的散射体服从标准差在三个坐标轴上分别为和的高斯分布，在获得散射体的位置后可将它们转换成球坐标，相对于第一根发送天线的位置和第一根接收天线的位置的第n个簇内的散射体的位置和可以表示为和其中，和分别表示第n个簇第m个子径到X(X∈{T,R}，分别表示发送端和接收端)天线阵列的距离、方位角和俯仰角；
步骤S402、初始时刻簇内子径时延计算，在多跳信道模型中，簇内子径的时延可通过计算，其中，表示和之间虚拟链路的时延，为t0时刻到之间的距离，为t0时刻到之间的距离。为首跳簇和末跳簇间的距离，τlink为服从指数分布的非负随机变量。
步骤S403、在大带宽大规模天线阵中，簇内子径的功率会沿着时间轴、频率轴和阵列轴变化，通常将其建模为随时间变化的对数正态过程和沿阵列变化的对数正态过程，非归一化的簇内子径功率为：
其中，DS为均方根时延扩展，Zn是以dB为单位的每簇阴影项，rτ为时延分布比例因子，ξn(p,q)是一个二维空间对数正态过程，用于模拟天线阵列上平滑的功率变化。
在毫米波段，大带宽场景下，需要考虑频域非平稳特性，所以我们在频域上将功率值乘以项，其中，是依赖于频率的常数因子。通过将所有簇的功率归一化即可得到最终簇内子径的功率如果簇是新产生的，可以通过将替换成得到和之间第n簇内第m个子径的初始功率。
步骤S404、对于幸存的簇，需在不同的时刻更新簇内子径的功率、时延等小尺度参数。对于时刻t1处的轨迹段，即在簇生成后的下一个时刻,第p根发送天线的坐标为其中，初始时刻第p根发送天线 的坐标可以通过计算，可以通过 计算。在t1时刻，到首跳簇的距离可以通过计算得到，同理可得到的距离t1时刻的簇内子径的时延利用前一时刻发送端、接收端和散射体的地理位置可以得到和
S5、根据收发端运动和簇的生灭过程，进行大、小尺度参数的更新，生成新的信道系数。
模型的空-时-频非平稳性主要体现在两个方面，一方面是空-时-频变化的参数，另一个方面是簇在空-时-频域的生灭过程。t时刻簇数量计算如下：
Nqp(t)＝Nsurv(t)+Nnew(t)
其中，Nqp(t)为簇的数量，Nsurv(t)为幸存簇的数目，由簇的生存概率Psurv(Δt,Δr,Δf)决定，Nnew(t)为新生簇的数目，服从均值为E[Nnew(t)]的泊松分布。定义λG为簇的出生率、λR为簇的结合率(灭亡率)。为了更精确地建模簇的空-时-频演进过程，我们引入两种采样间隔，一种是时域采样间隔Δt，频域采样间隔Δf和空间域(阵列域)采样间隔Δr，信道参数连续更新；另一种是的Δt，Δf和Δr的整数倍，分别是ΔtBD，ΔfBD和ΔrBD，在这些采样点发生簇的生灭演进过程。发送端和接收端簇沿阵列轴和时间轴的生存概率：

其中，和分别表示发送(接收)天线单元在阵列轴上以及时间轴上的位置差。和分别表示在阵列轴和时间轴依赖于场景的相关因子。发送端和接收端簇的联合生存概率为：
新生簇的平均数量为：
当研究大带宽时，在频率轴上也会存在簇的生灭过程。频率轴上簇的生存概率为：
其中，F(ΔfBD)和可以通过信道测量决定，表示在频率轴依赖于场景的相关因子。
综上所述，同时考虑空-时-频域簇的生灭过程时，簇的生存概率为：
新生簇的平均数量为：
基于前述方法和利用发送端、接收端和散射体之间的几何关系，可以得到不同天线对的小尺度参数，至此可以得到信道矩阵中所有参数值。

Claims (9)

  1. 一种适用于全频段全场景的6G普适信道建模方法，其特征在于，所述的建模方法中，6G普适几何随机信道模型中发送端和接收端均采用大规模均匀线阵，模型为多跳传播，其中，为发送端天线阵列的第p个阵元，为接收端天线阵列的第q个阵元，发送端和接收端天线阵元间的距离为δT(δR)；为xy平面内发送端和接收端天线阵列的方位角，为发送端和接收端天线阵列的俯仰角；到的第n条传播路径，n＝1,2,3,...,Nqp(t)，其中为第n条路径上靠近发送端的首跳簇，为靠近接收端的末跳簇，将两簇之间的传播路径建模为虚拟的链路；当首跳簇和末跳簇间的虚拟链路时延为零时，模型退化为单跳模型，此外，Nqp(t)为时刻t时到的路径数量，在双簇模型中对应着Nqp(t)个簇对，首跳簇和末跳簇一一对应，在单簇模型中对应着Nqp(t)个簇；微观上，对第n条路径上的簇和进行分析，簇内存在Mn(t)个散射体，表示中的第m个散射体，表示中的第m个散射体；从路径的角度来看，理解为到的第m个子径连接的散射体,理解为至的第m个子径连接的散射体；此外，和是时刻t时从至的第m个子径对应的方位离开角和俯仰离开角，和是时刻t时从至的第m个子径对应的方位到达角和俯仰到达角，此外，模型对发送端、接收端以及簇的运动情况分别建模，支持收发端及簇的任意速度和轨迹的三维运动，其中vT(t),vR(t),分别表示发送端、接收端，首跳簇以及末跳簇的运动速度，分别表示发送端，接收端，首跳簇以及末跳簇的运动方向的方位角，分别表示发送端，接收端，首跳簇以及末跳簇的运动方向的俯仰角；
    6G普适几何随机信道模型的信道矩阵表示为：
    H＝[PL·SH·BL·WE·AL]1/2·Hs
    其中，PL,SH,BL,WE,AL为大尺度衰落，PL为路径损耗，SH为阴影衰落，BL为阻挡效应，AL为大气吸收损耗，WE为天气影响损耗，Hs为小尺度衰落。
  2. 根据权利要求1所述的适用于全频段全场景的6G普适信道建模方法，其特征在于，所述小尺度衰落Hs如下：
    其中，MT为发送端天线阵列中天线阵元数量，MR为接收端天线阵列中天线阵元数量，为时刻t时发送端天线阵元与接收天线阵元之间的信道冲激响应，表示为视距LoS分量与非视距NLoS分量的叠加：
    其中，KR(t)为莱斯因子，和分别如下：


    其中，{*}T表示转置，fc表示载波频率，和表示在不同频段上天线单元对应垂直极化和水平极化的方向图，为交叉极化功率比，μ表征联合极化不均衡，和是时刻t时从至的LoS路径对应的方位离开角以及俯仰离开角,和是时刻t时从至的LoS路径对应的方位到达角以及俯仰到达角，和是服从(0,2π]均匀分布的随机相位，为法拉第旋转角，此处计算法拉第旋转角时fc的单位是GHz，是NLoS条件下到的第n条路径中的第m个子径的功率，是在时刻t时LoS路径的时延，是在时刻t时发送端天线阵元与接收天线阵元间的矢量距离，c为光速。是在时刻t时与之间的第n条路经的第m个子径的时延，是在时刻t时与之间的第n条路经的第m个子径的功率，以上所有参数均为时变参数。
  3. 根据权利要求1所述的适用于全频段全场景的6G普适信道建模方法，其特征在于，所述6G普适信道建模方法，在海洋通信信道场景中，模型分别将LoS径、粗糙的海洋表面和海面上空蒸发波导的多径传播这三个部分建模为和并使用功率系数，S1和S2控制对应部分随两船距离变化的消失和出现，即将计算公式中的NLoS部分分为和两部分，其中S1+S2＝1，在工业物联网信道中将镜面反射分量和密集多径分量分别建模为和其中和的建模方法与相同，只是参数值和簇的分布不同。
  4. 根据权利要求1所述的适用于全频段全场景的6G普适信道建模方法，其特征在于，所述6G普适信道建模方法，在可重构智能表面场景中，将信道分为发送端到可重构智能表面的子信道HTI，可重构智能表面到接收端的子信道HIR和发送端到接收端的子信道HTR，对三个子信道分别建模并引入相移对角矩阵Φ实现对信道环境的智能调控，HIR,HTI和HTR与的计算方法与Hs相同，只是参数值和簇的分布不同。
  5. 根据权利要求1所述的适用于全频段全场景的6G普适信道建模方法，其特征在于，所述6G普适信道建模方法，针对可见光频段信道建模时，一方面光信号波长极短，接收端尺寸通常为几百万个波长，不会发生几个波长上的信号快速衰落；另一方面由于可见光通信***中LED灯发出的是非相干光，光信号没有相位信息，在接收端的实数多径信号叠加后并不会引起快衰落，而表现为缓慢变化的阴影衰落，因此，虽然现在的可见光模型表示形式是多径叠加的信道冲激响应形式，其本质是建模了PL和SH的大尺度模型，所以Hs＝0， pH,pV为LED面阵的行和列数。
  6. 根据权利要求1所述的适用于全频段全场景的6G普适信道建模方法，其特征在于，所述6G普适信道建模方法，在多链路 场景中，假设基站数目为NBS，用户数为NMS，多链路信道模型的信道传输矩阵如下公式所示：
    每条链路对应的即为前述的单链路信道模型H。
  7. 根据权利要求1所述的适用于全频段全场景的6G普适信道建模方法，其特征在于，所述6G普适信道建模方法，信道矩阵H生成详细分为以下步骤：
    S1、设置传播场景和传播条件，确定载波频率和天线类型、收发端布局以及收发端运动轨迹等；
    S2、生成路径损耗、阴影衰落、氧气吸收以及阻挡效应的大尺度衰落；本方法主要聚焦于小尺度衰落的建模，此部分计算可参考标准化信道模型或业界经典模型；
    S3、根据收发端位置和运动情况生成具有空间一致性的时延扩展、4个角度扩展的大尺度参数；
    除SH外，其余相关大尺度参数有时延扩展DS、方位到达角扩展ASA、方位离开角扩展ASD、俯仰到达角扩展ESA、俯仰离开角扩展ESD、莱斯因子KR和交叉极化比XPR，时延扩展DS的生成如下公式所示：
    其中，P＝(PT,PR)由收发端位置矢量组成，PT(t)＝(xT(t),yT(t),zT(t))和PR(t)＝(xR(t),yR(t),zR(t))分别表示t时刻发送端的坐标矢量和接收端的坐标矢量，其初始值根据仿真环境及要求生成；XDS(P)是用正弦波叠加法生成的服从均值为0、方差为1的空间一致性的正态分布变量，表示DS在fc频段的均值，表示DS在fc频段的方差。根据终端的高度hUT，的配置值可分为三种类型；对于陆地移动通信场景1.5m≤hUT≤22.5m，可以参考3GPP TR 38.901标准化文档中的表7.5-6中的值；对于无人机场景22.5m≤hUT≤300m，取值参考3GPP TR 36.777标准化文档中的表B1.2中的值；对于低轨卫星通信场景，取值参考3GPP TR 38.811标准化文档中的表6.7-2中的值，城市宏小区UMa场景NLoS条件下，载波频率介于2-4 GHz情况下的计算如下
    其他的大尺度参数的生成与时延扩展DS的生成过程相同，8个大尺度参数都生成后，乘以大尺度参数间的互相关矩阵可以得到具有空间一致性的全部大尺度参数在对数域的值，然后，需要将对数域的值转化到线性域，至此，信道的大尺度参数都可得到；
    S4、生成服从椭球高斯散射分布的散射体，并根据收发端和散射体的地理位置信息计算簇的时延、角度和功率，生成信道系数；
    S5、根据收发端运动和簇的生灭过程，进行大、小尺度参数的更新，生成新的信道系数，
    模型的空-时-频非平稳性主要体现在两个方面，一方面是空-时-频变化的参数，另一个方面是簇在空-时-频域的生灭过程，t时刻簇数量计算如下：Nqp(t)＝Nsurv(t)+Nnew(t)
    其中，Nqp(t)为簇的数量，Nsurv(t)为幸存簇的数目，由簇的生存概率Psurv(Δt,Δr,Δf)决定，Nnew(t)为新生簇的数目，服从均值为E[Nnew(t)]的泊松分布，定义λG为簇的出生率、λR为簇的结合率即灭亡率。
  8. 根据权利要求7所述的适用于全频段全场景的6G普适信道建模方法，其特征在于，所述步骤S4具体为：
    步骤S401、使用椭球高斯散射分布，获取散射体的位置，即以为中心的第n个簇内的散射体服从标准差在三个坐标轴上分别为和的高斯分布，在获得散射体的位置后可将它们转换成球坐标，相对于第一根发送天线的位置和第一根接收天线的位置的第n个簇内的散射体的位置和可以表示为和其中，和分别表示第n个簇第m个子径到X(X∈{T,R}，分别表示发送端和接收端)天线阵列的距离、方位角和俯仰角；
    步骤S402、初始时刻簇内子径时延计算，在多跳信道模型中，簇内子径的时延可通过计算，其中，表示和之间虚拟链路的时延，为t0时刻到之间的距离，为t0时刻到之间的距离。为首跳簇和末跳簇间的距离，τlink为服从指数分布的非负随机变量；
    步骤S403、在大规模天线阵中，簇内子径的功率会沿着时间轴和阵列轴变化，通常将其建模为随时间变化的对数正态过程和沿阵列变化的对数正态过程，非归一化的簇内子径功率为：
    其中，DS为均方根时延扩展，Zn是以dB为单位的每簇阴影项，rτ为时延分布比例因子，ξn(p,q)是一个二维空间对数正态过程，用于模拟天线阵列上平滑的功率变化；
    在大带宽场景下，需要考虑频域非平稳特性，所以我们在频域上将功率值乘以项，其中，是依赖于频率的常数因子，最后，通过将所有簇的功率归一化即可得到最终簇内子径的功率如果簇是新产生的，通过将替换成得到和之间第n簇内第m个子径的初始功率；
    步骤S404、对于幸存的簇，需要在不同的时刻更新簇内子径的功率、时延等小尺度参数，对于时刻t1处的轨迹段，即在簇生成后的下一个时刻,第p根发送天线的坐标为其中，初始时刻第p根发送天线的坐标通过计算，在t1时刻，第n个首跳簇中第m个散射体的坐标通过计算。在t1时刻，到的距离可以通过计算得到，同理可得到的距离t1时刻的簇内子径的时延利用前一时刻发送端、接收端和散射体的地理位置可以得到和 (t＝t2,t3,...)。
  9. 根据权利要求7所述的适用于全频段全场景的6G普适信道建模方法，其特征在于，所述步骤S5中：为了更精确地建模簇的空-时-频演进过程，引入两种采样间隔，一种是时域采样间隔Δt，频域采样间隔Δf和空间域(阵列域)采样间隔Δr，信道参数连续更新；另一种是的Δt，Δf和Δr的整数倍，分别是ΔtBD，ΔfBD和ΔrBD，在这些采样点发生簇的生灭演进过程，发送端和接收端簇沿阵列轴和时间轴的生存概率：

    其中，和分别表示发送和接收天线单元在阵列轴上以及时间轴上的位置差，和分别表示在阵列轴和时间轴依赖于场景的相关因子，发送端和接收端簇的联合生存概率为：
    新生簇的平均数量为：
    当研究大带宽时，在频率轴上也会存在簇的生灭过程，频率轴上簇的生存概率为：
    其中，F(ΔfBD)和可以通过信道测量决定，表示在频率轴依赖于场景的相关因子，
    综上所述，同时考虑空-时-频域簇的生灭过程时，簇的生存概率为：
    新生簇的平均数量为：
    超高速列车场景中，考虑真空管道超高速列车场景的波导效应和管道壁粗糙度对信道的影响，新生簇的平均数量为：

    其中，Dqp(t)为t时刻收发端的直线距离，D为收发端的初始距离，ρs为管道壁散射系数，为粗糙度σh＝0时的散射系数。
PCT/CN2023/082380 2022-03-10 2023-03-19 一种适用于全频段全场景的6g普适信道建模方法 WO2023169590A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210235058.9 2022-03-10
CN202210235058.9A CN114598408B (zh) 2022-03-10 2022-03-10 一种适用于全频段全场景的6g普适信道建模方法

Publications (1)

Publication Number Publication Date
WO2023169590A1 true WO2023169590A1 (zh) 2023-09-14

Family

ID=81817474

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/082380 WO2023169590A1 (zh) 2022-03-10 2023-03-19 一种适用于全频段全场景的6g普适信道建模方法

Country Status (2)

Country Link
CN (1) CN114598408B (zh)
WO (1) WO2023169590A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118101000A (zh) * 2024-04-25 2024-05-28 南京邮电大学 一种辅助6g车与车通信的混合智能反射面功率分配优化方法

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114598408B (zh) * 2022-03-10 2024-02-27 东南大学 一种适用于全频段全场景的6g普适信道建模方法
CN115150861A (zh) * 2022-06-28 2022-10-04 南京航空航天大学 机舱内部5g宽带信号覆盖性能评估方法
CN114844584B (zh) * 2022-07-04 2022-10-21 北京交通大学 一种基于随机孪生簇的波束信道的仿真方法
CN115589268B (zh) * 2022-09-08 2023-06-13 江南大学 面向工业物联网的无线信道建模方法及***
CN116633423B (zh) * 2023-07-24 2023-10-13 成都本原星通科技有限公司 一种基于可重构智能表面的低轨卫星辅助通信方法
CN117395669A (zh) * 2023-10-17 2024-01-12 山东大学 一种面向6g无人机空对地通信场景的信道聚簇及建模方法
CN118282545A (zh) * 2024-06-03 2024-07-02 南京信息工程大学 一种空地通信信道建模方法、装置、预测方法及介质

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105553584A (zh) * 2015-12-10 2016-05-04 国网山东省电力公司烟台供电公司 一种3d mimo信道建模的方法
CN107425895A (zh) * 2017-06-21 2017-12-01 西安电子科技大学 一种基于实测的3d mimo统计信道建模方法
CN114598408A (zh) * 2022-03-10 2022-06-07 东南大学 一种适用于全频段全场景的6g普适信道建模方法

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4457222B2 (ja) * 2006-11-14 2010-04-28 独立行政法人情報通信研究機構 チャネル特性解析装置及び方法
CN113922900B (zh) * 2021-10-22 2024-01-26 东南大学 一种考虑天线尺寸与天线互耦的无线电信道建模方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105553584A (zh) * 2015-12-10 2016-05-04 国网山东省电力公司烟台供电公司 一种3d mimo信道建模的方法
CN107425895A (zh) * 2017-06-21 2017-12-01 西安电子科技大学 一种基于实测的3d mimo统计信道建模方法
CN114598408A (zh) * 2022-03-10 2022-06-07 东南大学 一种适用于全频段全场景的6g普适信道建模方法

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
BIAN JI; WANG CHENG-XIANG; GAO XIQI; YOU XIAOHU; ZHANG MINGGAO: "A General 3D Non-Stationary Wireless Channel Model for 5G and Beyond", IEEE TRANSACTIONS ON WIRELESS COMMUNICATIONS, IEEE SERVICE CENTER, PISCATAWAY, NJ., US, vol. 20, no. 5, 8 January 2021 (2021-01-08), US , pages 3211 - 3224, XP011853719, ISSN: 1536-1276, DOI: 10.1109/TWC.2020.3047973 *
WANG CHENG-XIANG; LV ZHEN; GAO XIQI; YOU XIAOHU; HAO YANG; HAAS HARALD: "Pervasive Wireless Channel Modeling Theory and Applications to 6G GBSMs for All Frequency Bands and All Scenarios", IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY, IEEE, USA, vol. 71, no. 9, 1 June 2022 (2022-06-01), USA, pages 9159 - 9173, XP011920074, ISSN: 0018-9545, DOI: 10.1109/TVT.2022.3179695 *
WANG, CHENGXIANG ET AL.: "6G Oriented Wireless Communication Channel Characteristics Analysis and Modeling", CHINESE JOURNAL ON INTERNET OF THINGS, vol. 4, no. 1, 31 March 2020 (2020-03-31), XP081728179 *
ZHENG YI; YU LONG; YANG RUNRUO; WANG CHENG-XIANG: "A General 3D Non-Stationary Massive MIMO GBSM for 6G Communication Systems", 2021 IEEE WIRELESS COMMUNICATIONS AND NETWORKING CONFERENCE (WCNC), IEEE, 29 March 2021 (2021-03-29), pages 1 - 6, XP033908498, DOI: 10.1109/WCNC49053.2021.9417601 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118101000A (zh) * 2024-04-25 2024-05-28 南京邮电大学 一种辅助6g车与车通信的混合智能反射面功率分配优化方法

Also Published As

Publication number Publication date
CN114598408B (zh) 2024-02-27
CN114598408A (zh) 2022-06-07

Similar Documents

Publication Publication Date Title
WO2023169590A1 (zh) 一种适用于全频段全场景的6g普适信道建模方法
Wang et al. Pervasive wireless channel modeling theory and applications to 6G GBSMs for all frequency bands and all scenarios
Wang et al. 6G wireless channel measurements and models: Trends and challenges
Ma et al. Impact of UAV rotation on MIMO channel characterization for air-to-ground communication systems
Lv et al. 6G-enabled network in box for internet of connected vehicles
Cheng et al. Channel nonstationarity and consistency for beyond 5G and 6G: A survey
CN109639378B (zh) 矩形隧道无线传播信道建模方法
CN104994517B (zh) 一种高空平台mimo通信***三维信道建模方法
Cheng et al. Modeling and simulation for UAV air-to-ground mmWave channels
CN103747456A (zh) 基于三维空间域多天线mimo统计信道的建模方法
CN114124266B (zh) 一种基于irs辅助无人机与无人船通信的信道建模方法
CN109951213B (zh) 高空平台mimo三维几何随机模型建立方法及通信方法
Pang et al. Investigation and comparison of 5G channel models: From QuaDRiGa, NYUSIM, and MG5G perspectives
Liu et al. 6G enabled advanced transportation systems
CN109450575B (zh) 非平稳的三维宽带高空平台mimo几何随机模型建立方法
CN114499724B (zh) 一种面向低轨卫星通信的空时频非平稳传输特性分析方法
Zhang et al. A 3-D semi-deterministic MIMO beam channel model for cellular-assisted millimeter-wave vehicular communications
Xiong et al. Channel characteristics of high‐speed railway station based on ray‐tracing simulation at 5G mmWave band
Yuan et al. A 3D geometry-based reconfigurable intelligent surfaces-assisted MmWave channel model for high-speed train communications
CN117040670A (zh) 一种面向卫星信道的几何随机信道建模方法
Ranjith et al. A multi-fusion integrated end-to-end deep kernel CNN based channel estimation for hybrid range UM-MIMO 6G communication systems
CN115664567A (zh) 一种太赫兹mimo无人机非平稳性信道建模方法
Gulfam et al. Angle and time of arrival characteristics of 3D air-to-ground radio propagation environments
Huang et al. A 3D wideband GBSM for THz communications in indoor scenarios
Zhang et al. A new model for estimating troposcatter loss and delays based on ray-tracing and beam splitting with ERA5

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23766167

Country of ref document: EP

Kind code of ref document: A1

WWE Wipo information: entry into national phase

Ref document number: 18696233

Country of ref document: US