A hybrid wavefront signal model construction method and channel estimation method

Abstract

The present invention relates to a method for constructing a hybrid wavefront signal model for 5G signals and a channel estimation method based on passive measurement of the hybrid wavefront signal model. In the 5G receiving signal model, according to the distance #imgabs0# from the receiving end antenna to the previous scattering point and the Rayleigh distance d Rayleigh Different wavefront signal models #imgabs1# are used to fit the propagation of 5G signals in space based on the size comparison. The hybrid wavefront signal model is used to iteratively update the signal state information (CSI) of the receiving end using the spatial alternating generalized expectation maximization algorithm (SAGE) until the channel characteristic parameters #imgabs2# converge, thereby realizing channel parameter estimation. The present invention uses a hybrid of the plane wavefront model and the spherical wavefront model for estimation, which can better adapt to the indoor environment under the 5G environment, improve the accuracy of channel modeling in the indoor environment, and better describe the propagation of electromagnetic waves in the indoor environment.

Description

Technical Field

[0001] The present invention relates to the technical field of signal modeling and channel estimation, and in particular to a method for constructing a hybrid wavefront signal model for 5G signals and a channel estimation method based on passive measurement of the hybrid wavefront signal model. Background Art

[0002] With the rapid development of mobile communications technology, mobile internet traffic has exploded. Emerging application scenarios such as industrial control, telemedicine, and autonomous driving are placing increasingly stringent demands on latency and reliability. 4G LTE can no longer meet the market's demand for high throughput, low latency, and high reliability, leading to the emergence of 5G technology. In indoor environments, multipath effects, due to extensive reflection, transmission, and scattering, significantly impact 5G communication performance. Extensive channel measurements are essential for characterizing and verifying 5G indoor channel quality.

[0003] Channel measurement uses instruments and equipment to measure transmitted signals, thereby understanding the transmission characteristics of electromagnetic signals in space. Based on the measurement method, it can be divided into active and passive measurements. Active channel measurement offers advantages such as mature technology, high channel resolution, and high accuracy in channel characteristic estimation. However, it also has problems such as inconsistency between the measured channel and the channel experienced by users, the inability to build large-scale MIMO, and poor traversability due to limited measurement in specific areas. Compared to active channel measurement, passive channel detection methods have the advantages of measuring channels that are completely consistent with the channels experienced by users, allowing for widespread measurement, and achieving better traversability. However, it also has problems such as the inability to synchronize the transmitter and receiver, and limited signal bandwidth.

[0004] Compared with 4G LTE passive measurements, 5G's large bandwidth ensures higher channel feature resolution, so 5G's passive channel measurement can be applied to small-scale parameter estimation and channel measurement in outdoor and indoor environments with significant multipath. In the paper "Measurement-Based Channel Characterization of 5G Downlink Based on Passive Probing in 5G Commercial Networks" published in the IEEE Transactions on Wireless Communications and the paper "5G Channel Random Cluster Model Based on Passive Probing in Service Networks" published at the 2020 IEEE Sixth International Conference on Computers and Communications, Tianqi Wu et al. and Xuyan Hou et al. conducted research on 5G passive channel sounding measurements, conducting passive measurements in multiple outdoor and indoor scenarios, analyzing broadband channel characteristics, and comparing them with the channel characteristics of existing active measurement channel models (WINNER II and 3GPP). These research works demonstrate the feasibility and advantages of using passive probing to estimate small-scale channel parameters for commercial 5G networks in complex multipath indoor environments.

[0005] Channel modeling based on extensive measurements is crucial for describing electromagnetic wave propagation in real-world environments. Traditional general multipath models are derived from plane wavefronts and are not applicable to indoor environments. Regarding indoor environments, in "A Parameterized UWB Propagation Channel Estimation and Its Performance Verification in an Anechoic Chamber," published in the IEEE Transactions on Microwave Theory and Techniques, and "Scatterer Localization Using a Large Antenna Array Based on a Spherical Wavefront Parametric Model," published in the IEEE Transactions on Wireless Communications, K. Haneda et al. and X. Yin et al. simply categorize indoor environments as near-field environments. They propose different spherical wavefront signal models to describe indoor signal propagation and demonstrate that, in near-field environments, spherical wavefront models simulate measured data better than plane wavefront models. However, with the advancement of mobile communication technology, communication signal frequencies are increasing, resulting in a shrinking near-field range. Therefore, indoor environments are no longer fully applicable to near-field spherical wavefront models. Summary of the Invention

[0006] In response to the problems existing in the prior art, the purpose of the present invention is to provide a method for constructing a hybrid wavefront signal model for 5G signals and a channel estimation method based on passive measurement of the hybrid wavefront signal model, which can better adapt to the indoor environment under the 5G environment, improve the accuracy of channel modeling in the indoor environment, and better describe the propagation of electromagnetic waves in the indoor environment.

[0007] To achieve the above object, the technical solution adopted by the present invention is:

[0008] A method for constructing a hybrid wavefront signal model for 5G signals, the construction method comprising:

[0009] 5G signal reception model In the calculation, the distance from the receiving antenna to the previous scattering point is and Rayleigh distance d Rayleigh The size comparison of different wavefront signal models To fit the propagation of 5G signals in space: When the plane wavefront model is used, A spherical wavefront model is used when ; where Indicates the The signal transmitted by the Rx antenna is W(t), which represents the Gaussian white noise received by the Rx antenna.

[0010] use Indicates the signal observed by the nth antenna element in the ith period. Signals transmitted along the path;

[0011] when The plane wavefront model is used when , and the plane wavefront model is expressed as:

[0012]

[0013] Among them, the channel characteristic parameters and They represent the first The path delay, Doppler frequency, arrival direction unit vector and the complex attenuation coefficient experienced by the signal received with p polarization, p∈[1,2] represents two mutually orthogonal polarization directions, The azimuth of the multipath arrival angle and pitch angle The only certainty, is the plane wavefront model observed by the nth antenna in the i-th snapshot cycle. u(t) is the input signal of the system, i.e. the CSI-RS transmission signal of the base station; c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with a unit vector of Ω in the direction of arrival; t i，n w is the starting time of the nth element of the receiving array in the i-th snapshot period; i，n (t) is the Gaussian white noise received on the nth Rx antenna array element in the i-th snapshot cycle;

[0014] when The spherical wavefront model is used when , which means:

[0015]

[0016] Among them, the channel characteristic parameters and They represent the first The time delay of each path to the reference position of the receiving antenna, the Doppler frequency and the complex attenuation coefficient experienced by the signal received with p polarization, is the reference position of the receiving antenna array, r n is the specific position of the nth antenna in the receiving antenna array, is the direction of arrival unit vector of the receiving antenna reference position; The azimuth of the arrival angle of the multipath at the antenna reference position and pitch angle The only certainty,

[0017] A channel estimation method based on passive measurement of a hybrid wavefront signal model, the method comprising the following steps:

[0018] Step S10: Receive and process 5G downlink information to obtain a channel state information reference signal;

[0019] Step S20: Calculate SAGE estimates of channel parameters using a spherical wavefront model;

[0020] Step S30: judging whether the distance to the scattering point is greater than the Rayleigh distance according to the estimation result of step S20, if so, proceeding to step S40; if not, proceeding to step S50;

[0021] Step S40: Calculating SAGE estimates of channel parameters using a plane wavefront model;

[0022] Step S50: Determine the path Is it equal to the set path number pathnum? Then order Then return to step S20; if The calculation ends.

[0023] Before step 20, the initial value of the channel characteristic parameter θ to be estimated is set. in, Represents the estimated value of the channel characteristic parameter of the previous iteration; represents the initial value of the channel parameter to be estimated;

[0024] The step 20 is specifically as follows:

[0025] Step S21: Using the spherical wavefront model, according to the received incomplete data Y(t)=y(t) and To estimate the complete data And by calculating Y(t)=y(t) and assuming Conditional expectation of get Natural estimate of

[0026] in, Represents the first The signal transmitted by the path Indicates the first Gaussian white noise of the paths; Y(t) represents the total received signal, y(t) represents the observation of the received signal, is a real coefficient, Acceptable data, N0 is a positive constant;

[0027] Step S22: Use Natural estimate of replace Find the given observation value Under these conditions, the channel characteristics The log-likelihood function of

[0028] in, is the population of unobservable complete data, for Observation, for Natural estimate of For the Channel characteristic parameters to be estimated of the stripe diameter;

[0029] Step S23: Channel characteristics are analyzed by performing the following two steps: and To perform a SAGE estimate:

[0030]

[0031]

[0032] in, is the complex attenuation coefficient of this round of iteration Other channel characteristic parameters Estimates, For the The complex attenuation coefficient of the strip diameter Other channel characteristic parameters Estimates, Contains the complex attenuation coefficients for both p=1 and p=2 polarization directions, is the first spherical wavefront received by each antenna element The arrival direction unit vector corresponding to the strip diameter; f(θ) and D(Ω) are intermediate variables used in the derivation of the formula;

[0033]

[0034]

[0035]

[0036] Among them, c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with a unit vector of Ω in the arrival direction, where p∈[1,2] represents two mutually orthogonal polarization directions. and They represent the first and second positions of the receiving antenna reference position of the spherical wavefront model. The time delay and Doppler frequency of each path, the distance from the receiving antenna to the previous scattering point, and the complex attenuation coefficient experienced by the signal received with p polarization, is the reference position of the receiving antenna array, r n is the specific position of the nth antenna in the receiving antenna array, For the The unit vector of the direction of arrival of the receiving antenna reference position of the sliver path, is f(θ) of the nth element in the antenna array;

[0037] The spherical wavefront model is received by the nth element of the receiving antenna. The arrival direction unit vector of the stripe, D n (Ω) is the D(Ω) of the nth element in the antenna array;

[0038] c is the speed of light, t i，n is the starting time of the nth element of the receiving array in the i-th snapshot period, u(t) is the input signal of the system, that is, the CSI-RS transmission signal of the base station, T sc is the scanning interval of the receiving end; is an intermediate variable;

[0039] Step S24: By calculating Determine whether the channel parameters converge; if not, let Return to step S21; if converged, execute step S25; the convergence is The difference is less than the threshold; is the estimated value of the channel characteristic parameter of this round of iteration, is the estimated value of the channel characteristic parameter in the previous iteration;

[0040] The step 30 is specifically as follows:

[0041] judge in Is it greater than the Rayleigh distance d? Rayleigh If the estimated Less than or equal to the Rayleigh distance d Rayleigh , then choose to use the spherical wavefront signal model and use the estimated Reconstruct this multipath Execute step S50; if the estimated Greater than the Rayleigh distance d Rayleigh , choose to use the plane wavefront signal model instead and execute step S40.

[0042] The spherical wavefront model is expressed as:

[0043]

[0044] Among them, the channel characteristic parameters and They represent the first The time delay of each path to the reference position of the receiving antenna, the Doppler frequency and the complex attenuation coefficient experienced by the signal received with p polarization, is the reference position of the receiving antenna array, r n is the specific position of the nth antenna in the receiving antenna array, is the direction of arrival unit vector of the receiving antenna reference position; The azimuth of the arrival angle of the multipath at the antenna reference position and pitch angle The only certainty,

[0045] is the wavefront observed by the nth antenna element in the i-th snapshot cycle in the spherical wavefront model. u(t) is the input signal of the base station CSI-RS, c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with the arrival direction unit vector Ω, where p∈[1,2] represents two mutually orthogonal polarization directions, and t i，n is the starting time of the nth element of the receiving array in the i-th snapshot period, w i，n (t) is the Gaussian white noise received on the nth Rx antenna.

[0046] The step S40 is specifically as follows:

[0047] Step S41: Using the plane wavefront model, according to the received incomplete data Y(t)=y(t) and To estimate the complete data By calculating Y(t)=y(t) and assuming Conditional expectation of get Natural estimate of

[0048] in, Represents the first The signal transmitted by the path Indicates the first Gaussian white noise of the paths; Y(t) represents the total received signal, y(t) represents the observation of the received signal, is a real coefficient, Acceptable data, N0 is a positive constant;

[0049] Step S42: Due to is unobservable, using Natural estimate of replace Find the given observation value conditions, The log-likelihood function of

[0050] in, is the population of unobservable complete data, for Observation, for Natural estimate of For the Channel characteristic parameters to be estimated of the stripe diameter;

[0051] Step S43: Channel characteristics are analyzed by performing the following two steps: and Perform SAGE estimation, where and They are respectively the first The time delay, Doppler frequency, direction of arrival unit vector and complex attenuation coefficient of the stripe path; Contains complex attenuation coefficients for two polarization directions: p = 1 and p = 2;

[0052]

[0053] in, is the complex attenuation coefficient of this round of iteration Other channel characteristic parameters Estimates, For the The complex attenuation coefficient of the strip diameter Other channel characteristic parameters Estimates, f(θ) and D(Ω) are intermediate variables for formula derivation and arrangement;

[0054]

[0055] in, is the complex attenuation coefficient of this iteration Estimates, For the Complex attenuation coefficient of strip diameter Estimation, I is the number of snapshots, P is the power of the input signal u(t), T sc is the scanning interval of the receiving end;

[0056] in,

[0057]

[0058]

[0059]

[0060] Among them, c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with a unit vector of Ω in the arrival direction, where p∈[1,2] represents two mutually orthogonal polarization directions. is f(θ) of the nth element in the antenna array; D n (Ω) is the D(Ω) of the nth element in the antenna array; c is the speed of light, t i，n is the starting time of the nth element of the receiving array in the i-th snapshot period, u(t) is the input signal of the system, that is, the CSI-RS transmission signal of the base station, is an intermediate variable;

[0061] Step S44: By calculating Determine whether the channel parameters converge; if not, let Return to step S41; if converged, use the estimated Reconstruct this multipath Execute step S50; the convergence is The difference is less than the threshold.

[0062] The plane wavefront model is expressed as:

[0063]

[0064] Among them, the channel characteristic parameters and They represent the first The time delay of each path, the Doppler frequency, the direction of arrival unit vector, and the complex attenuation coefficient experienced by the signal received with p polarization, The azimuth of the multipath arrival angle and pitch angle The only certainty,

[0065] is the wavefront observed by the nth antenna element in the i-th snapshot cycle in the plane wavefront model. u(t) is the input signal of the system, i.e. the CSI-RS transmission signal of the base station; c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with a unit vector of Ω in the direction of arrival; t i，nw is the starting time of the nth element of the receiving array in the i-th snapshot period; i，n (t) is the Gaussian white noise received by the nth Rx antenna element in the i-th snapshot cycle.

[0066] After adopting the above scheme, the present invention uses a mixture of plane wavefront models and spherical wavefront models for estimation, which can better adapt to the indoor environment under the 5G environment, improve the accuracy of channel modeling in the indoor environment, and better describe the propagation of electromagnetic waves in the indoor environment. It has important theoretical significance and application value for the optimization of 5G technical solutions for indoor environments and multipath signal-assisted precise indoor positioning based on electromagnetic wave propagation in the environment. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 This is a flow chart of the channel estimation method of the present invention;

[0068] Figure 2 Schematic diagram of the channel estimation example flow based on the hybrid model. DETAILED DESCRIPTION

[0069] The present invention discloses a method for constructing a hybrid wavefront signal model for 5G signals. In the calculation, the distance from the receiving antenna to the previous scattering point is and Rayleigh distance d Rayleigh Different wavefront signal models are used for the size comparison To fit the propagation of 5G signals in space: When the plane wavefront model is used, The spherical wavefront model is used when . Indicates the The signal transmitted by the Rx antenna is W(t), which represents the Gaussian white noise received by the Rx antenna.

[0070] Specifically, using It represents the signal observed by the nth antenna in the ith period. The signal is transmitted along the path.

[0071] when The plane wavefront model is used when , and the plane wavefront model is expressed as:

[0072]

[0073] Among them, the channel characteristic parameters and They represent the first The path delay, Doppler frequency, arrival direction unit vector and the complex attenuation coefficient experienced by the signal received with p polarization, p∈[1,2] represents two mutually orthogonal polarization directions, The azimuth of the multipath arrival angle and pitch angle The only certainty, is the plane wavefront model observed by the nth antenna in the i-th snapshot cycle. u(t) is the input signal of the system, i.e. the CSI-RS transmission signal of the base station. n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with a unit vector of Ω in the direction of arrival. i，n w is the starting time of the nth element of the receiving array in the i-th snapshot period. i，n (t) is the Gaussian white noise received by the nth Rx antenna element in the i-th snapshot cycle.

[0074] when The spherical wavefront model is used when , which means:

[0075]

[0076] Among them, the channel characteristic parameters and They represent the first The time delay of each path to the reference position of the receiving antenna, the Doppler frequency and the complex attenuation coefficient experienced by the signal received with p polarization, is the reference position of the receiving antenna array, r n is the specific position of the nth antenna in the receiving antenna array, is the direction of arrival unit vector of the receiving antenna reference position; The azimuth of the arrival angle of the multipath at the antenna reference position and pitch angle The only certainty,

[0077] In two In the different expressions of u(t), both are the input signals of the base station CSI-RS, c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with a unit vector of Ω in the direction of arrival, t i，n are the starting time of the nth element of the receiving array in the i-th snapshot period, w i，n (t) are all Gaussian white noises received on the nth Rx antenna.

[0078] Combine Figure 1 and Figure 2 As shown, based on the same inventive concept, the present invention also discloses a channel estimation method based on passive measurement of the hybrid wavefront signal model, which specifically includes the following steps:

[0079] Step S10: Receive and process 5G downlink information to obtain a channel state information reference signal;

[0080] Specifically, step S11: using a receiving antenna array to receive a 5G downlink signal, and using a signal acquisition device to acquire the signal received by the receiving antenna array.

[0081] Step S12: Use the 5G terminal to extract the radio resource control (RRC) parameters of the 5G downlink signal.

[0082] Step S13: The signal collected by the signal collection device is processed using the RRC parameters, and the CSI-RS signal is extracted as the received signal y(t).

[0083] Step S20: Calculate SAGE estimates of channel parameters using the spherical wavefront model.

[0084] Before step S20 begins, the initial value of the channel characteristic parameter θ to be estimated is set. in Represents the estimated value of the channel characteristic parameter of the previous iteration; represents the initial value of the channel parameter to be estimated. Step S20 specifically includes the following:

[0085] Step S21: Using the spherical wavefront model, according to the received incomplete data Y(t)=y(t) and To estimate the complete data because is unobservable, by calculating given Y(t)=y(t) and assuming Conditional expectation of get A natural estimate of . Represents the first The signal transmitted by the path Indicates the first Gaussian white noise of the path. Y(t) represents the total received signal, y(t) represents the observation of the received signal, is a real coefficient, Acceptable data, N0 is a positive constant.

[0086] Step S22: Use Natural estimate of replace Find the given observation value conditions, The log-likelihood function of in is the population of unobservable complete data, for Observation of (not available), for The natural estimate of is the estimated value of the channel characteristic parameter in the previous iteration. For the The channel characteristic parameters to be estimated of the stripe diameter, is the signal observed by the receiving antenna array in the spherical wavefront signal model. The signal is transmitted along the path.

[0087] Step S23: Channel characteristics are analyzed by performing the following two steps: and Perform SAGE estimation: and They are respectively the first The time delay of the stripe path, the Doppler frequency, the distance from the receiving antenna to the previous scattering point, the arrival direction unit vector of the receiving antenna reference position and the complex attenuation coefficient. Contains the complex attenuation coefficients for two polarization directions: p=1 and p=2.

[0088]

[0089] in is the complex attenuation coefficient of this round of iteration Other channel characteristic parameters Estimates, For the The complex attenuation coefficient of the strip diameter Other channel characteristic parameters Estimates, for The natural estimate of For the The complex attenuation coefficient of the strip diameter Other channel characteristic parameters besides is the first spherical wavefront received by each antenna array element The arrival direction unit vector corresponding to the strip diameter. f(θ) and D(Ω) are intermediate variables used in the derivation and organization of the formula.

[0090]

[0091] in is the complex attenuation coefficient of this iteration Estimates, For the Complex attenuation coefficient of strip diameter Estimates, for The natural estimate of I is the number of snapshots, P is the power of the input signal u(t), T sc is the scanning interval of the receiver, is the first spherical wavefront received by each antenna element The arrival direction unit vector corresponding to the strip diameter. f(θ) and D(Ω) are intermediate variables used in the derivation and organization of the formula.

[0092] in, and They represent the first The time delay of each path to the reference position of the receiving antenna, the Doppler frequency and the complex attenuation coefficient experienced by the signal received with p polarization;

[0093]

[0094] where c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with a unit vector of Ω in the arrival direction, where p∈[1,2] represents two mutually orthogonal polarization directions. and They represent the first and second positions of the receiving antenna reference position of the spherical wavefront model. The time delay and Doppler frequency of each path, the distance from the receiving antenna to the previous scattering point, and the complex attenuation coefficient experienced by the signal received with p polarization, is the reference position of the receiving antenna array, r n is the specific position of the nth antenna in the receiving antenna array, For the The unit vector of the direction of arrival of the receiving antenna reference position of the sliver path, is f(θ) of the nth element in the antenna array.

[0095]

[0096] in, The spherical wavefront model is received by the nth element of the receiving antenna. The arrival direction unit vector of the stripe, c n，p(Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with the arrival direction unit vector Ω, where p∈[1,2] represents two mutually orthogonal polarization directions. n (Ω) is the D(Ω) of the nth element in the antenna array.

[0097]

[0098] in, and They represent the first The time delay, Doppler frequency and distance from the receiving antenna to the previous scattering point of each path are is the reference position of the receiving antenna array, r n is the specific position of the nth antenna in the receiving antenna array, For the The unit vector of the direction of arrival of the receiving antenna reference position of the sliver path, is the observation of complete data, c is the speed of light, t i，n is the starting time of the nth element of the receiving array in the i-th snapshot period, u(t) is the input signal of the system, that is, the CSI-RS transmission signal of the base station, T sc is the scanning interval of the receiver. is an intermediate variable.

[0099] Step S24: By calculating Determine whether the channel parameters converge. If not, let Return to step S21; if converged, proceed to step S25. The difference is less than a certain threshold. is the estimated value of the channel characteristic parameter of this round of iteration, is the estimated value of the channel characteristic parameter in the previous iteration.

[0100] Step S30: judging whether the distance to the scattering point is greater than the Rayleigh distance according to the estimation result of step S20, if so, proceeding to step S40; if not, proceeding to step S50;

[0101] Specifically, judge in Is it greater than the Rayleigh distance d? Rayleigh If the estimated Less than or equal to the Rayleigh distance d Rayleigh , then choose to use the spherical wavefront signal model and use the estimated Reconstruct this multipath Execute step S50. Greater than the Rayleigh distance d Rayleigh, select and use the plane wavefront signal model instead, and execute step S40. is the estimated value of the channel characteristic parameter of this round of iteration, This is the iteration number The estimated distance from the receiving antenna of the stripe to the previous scattering point, Indicates the Estimation of the signal transmitted along each path, is reconstructed by the first The signal is transmitted along the path.

[0102] Step S40: Calculating SAGE estimates of channel parameters using a plane wavefront model;

[0103] The step S40 specifically includes the following:

[0104] Step S41: Using the plane wavefront model, according to the received incomplete data Y(t)=y(t) and To estimate the complete data By calculating Y(t)=y(t) and assuming Conditional expectation of get A natural estimate of . Represents the first The signal transmitted by the path Indicates the first Gaussian white noise of the path, Y(t) represents the total received signal, y(t) represents the observation of the received signal, is a real coefficient, Acceptable data, N0 is a positive constant.

[0105] Step S42: Due to is unobservable, using Natural estimate of replace Find the given observation value conditions, The log-likelihood function of in is the population of unobservable complete data, for Observation of (not available), for The natural estimate of is the estimated value of the channel characteristic parameter in the previous iteration. For the The channel characteristic parameters to be estimated of the stripe diameter, is the signal observed by the receiving antenna array in the plane wavefront signal model. The signal is transmitted along the path.

[0106] Step S43: Channel characteristics are analyzed by performing the following two steps: and Perform SAGE estimation: and They are respectively the first The time delay, Doppler frequency, direction of arrival unit vector and complex attenuation coefficient of the sliver path. Contains the complex attenuation coefficients for two polarization directions: p=1 and p=2.

[0107]

[0108] in is the complex attenuation coefficient of this round of iteration Other channel characteristic parameters Estimates, For the The complex attenuation coefficient of the strip diameter Other channel characteristic parameters Estimates, for The natural estimate of For the The complex attenuation coefficient of the strip diameter Other channel characteristic parameters besides For the plane wavefront The arrival direction unit vector corresponding to the strip diameter. f(θ) and D(Ω) are intermediate variables used in the derivation and organization of the formula.

[0109]

[0110] in is the complex attenuation coefficient of this iteration Estimates, For the Complex attenuation coefficient of strip diameter Estimates, for The natural estimate of I is the number of snapshots, P is the power of the input signal u(t), T sc is the scanning interval of the receiver, For the plane wavefront The arrival direction unit vector corresponding to the strip diameter. f(θ) and D(Ω) are intermediate variables used in the derivation and organization of the formula.

[0111] in,

[0112]

[0113] Among them, c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with a unit vector of Ω in the arrival direction, where p∈[1,2] represents two mutually orthogonal polarization directions. and They represent the first and second positions of the receiving antenna reference position of the plane wavefront model. The time delay of each path, the Doppler frequency, the direction of arrival unit vector, and the complex attenuation coefficient experienced by the signal received with p polarization, is f(θ) of the nth element in the antenna array.

[0114]

[0115] in, For the plane wavefront model The arrival direction unit vector of the stripe, c n，p (Ω) is the p-polarized antenna response of the nth Rx antenna to the incident wave with the arrival direction unit vector Ω, where p∈[1,2] represents two mutually orthogonal polarization directions. n (Ω) is the D(Ω) of the nth element in the antenna array.

[0116]

[0117] in, and They represent the first and second positions of the receiving antenna reference position of the plane wavefront model. The time delay of each path, the Doppler frequency, the direction of arrival unit vector, and the complex attenuation coefficient experienced by the signal received with p polarization, is the observation of complete data, c is the speed of light, t i，n is the starting time of the nth element of the receiving array in the i-th snapshot period, u(t) is the input signal of the system, that is, the CSI-RS transmission signal of the base station, Ts c为 Scan interval of the receiver. is an intermediate variable.

[0118] Step S44: By calculating Determine whether the channel parameters converge. If not, let Return to step S41; if converged, use the estimated Reconstruct this multipath Execute step S50. The convergence is The difference is less than a certain threshold. is the estimated value of the channel characteristic parameter of this round of iteration, is the estimated value of the channel characteristic parameter in the previous iteration. Indicates the Estimation of the signal transmitted along each path, is reconstructed by the first The signal is transmitted along the path.

[0119] Step S50: Determine the path Is it equal to the set path number pathnum? Then order Then return to step S20; if Then the calculation ends, Number the path.

[0120] When this method is applied to passive channel detection and estimation in indoor environments, the plane wavefront and spherical wavefront signal models are switched according to the Rayleigh distance between the far and near fields of the signal. This solves the problem that traditional plane wavefront or spherical wavefront signal models are not applicable in indoor environments, improves the accuracy of channel modeling in indoor environments, and better describes the propagation of electromagnetic waves in indoor environments.

[0121] The above description is merely an embodiment of the present invention and does not limit the technical scope of the present invention. Therefore, any minor modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention are still within the scope of the technical solution of the present invention.

Claims

1. A channel estimation method based on passive measurement of a hybrid wavefront signal model, characterized by: The method comprises the following steps: Step S10: Receive and process 5G downlink information to obtain a channel state information reference signal; Step S20: Calculate SAGE estimates of channel parameters using a spherical wavefront model; Before step 20, the channel characteristic parameters to be estimated are first set Initial value of ,in, Represents the estimated value of the channel characteristic parameter of the previous iteration; represents the initial value of the channel parameter to be estimated; The step 20 is specifically as follows: Step S21: Using the spherical wavefront model, according to the received incomplete data and To estimate the complete data , and by calculating the given And assume Conditional expectation of get Natural estimate of in, Represents the first The signal transmitted by the path Indicates the first Gaussian white noise of the paths; represents the total number of received signals, represents the observation of the received signal, is a real coefficient, Acceptable data, is a positive constant; Step S22: Use Natural estimate of replace , find the given observation value Under these conditions, the channel characteristics The log-likelihood function of ; in is the population of unobservable complete data, for Observation, for Natural estimate of For the Channel characteristic parameters to be estimated of the stripe diameter; Step S23: Channel characteristics are analyzed by performing the following two steps: and Conduct SAGE estimation; Step S24: By calculating Determine whether the channel parameters converge; if not, let , return to step S21; if converged, execute step S30; the convergence refers to The difference is less than the threshold; is the estimated value of the channel characteristic parameter of this round of iteration, is the estimated value of the channel characteristic parameter in the previous iteration; Step S30: judging whether the distance to the scattering point is greater than the Rayleigh distance according to the estimation result of step S20, if so, proceeding to step S40; if not, proceeding to step S50; The step 30 is specifically as follows: judge in Is it greater than the Rayleigh distance? If the estimated Less than or equal to the Rayleigh distance , then choose to use the spherical wavefront signal model and use the estimated Reconstruct this multipath , execute step S50; if the estimated Greater than the Rayleigh distance , select and use the plane wavefront signal model as a replacement, and execute step S40; Step S40: Calculating SAGE estimates of channel parameters using a plane wavefront model; Step S50: Determine the path Is it equal to the set number of paths? ;like , then let , then return to step S20; if , then the calculation ends.

2. The channel estimation method based on passive measurement of a hybrid wavefront signal model according to claim 1, characterized in that: The step S23 is specifically as follows: The channel characteristics are analyzed by performing the following two steps and To perform a SAGE estimate: in, is the complex attenuation coefficient of this round of iteration Other channel characteristic parameters Estimates, , For the The complex attenuation coefficient of the strip diameter Other channel characteristic parameters Estimates, Include and The complex attenuation coefficients for the two polarization directions, is the first spherical wavefront received by each antenna array element The unit vector of the arrival direction corresponding to the strip diameter; and It is an intermediate variable in formula derivation and arrangement; in, It is The unit vector of the arrival direction of the Rx antenna pair is The incident wave Polarized antenna response, represents two mutually orthogonal polarization directions, 、 , and They represent the first and second positions of the receiving antenna reference position of the spherical wavefront model. The time delay and Doppler frequency of each path, the distance from the receiving antenna to the previous scattering point, and the The complex attenuation coefficient experienced by the polarization received signal, is the reference position of the receiving antenna array, The receiving antenna array The specific location of the antenna, For the The unit vector of the direction of arrival of the receiving antenna reference position of the sliver path, The first array element ; The receiving antenna for the spherical wavefront model The array element receives the The arrival direction unit vector of the stripe, ; The first array element ; is the speed of light, For the The receiving array in the snapshot cycle The starting time of each array element, is the input signal of the system, i.e. the CSI-RS transmission signal of the base station. is the scanning interval of the receiving end; is an intermediate variable.

3. The channel estimation method based on passive measurement of a hybrid wavefront signal model according to claim 2, characterized in that: The spherical wavefront model is expressed as: Among them, the channel characteristic parameters , 、 and They represent the first The time delay, Doppler frequency and The complex attenuation coefficient experienced by the polarization received signal, is the reference position of the receiving antenna array, The receiving antenna array The specific location of the antenna, is the direction of arrival unit vector of the receiving antenna reference position; The azimuth of the arrival angle of the multipath at the antenna reference position and pitch angle The only certainty, ; is the first The first in the snapshot cycle The observations of the antenna array element are Signals transmitted along the path; is the input signal of the base station CSI-RS, It is The unit vector of the arrival direction of the Rx antenna pair is The incident wave Polarization antenna response, represents two mutually orthogonal polarization directions, It is The receiving array in the snapshot cycle The starting time of each array element, It is Gaussian white noise received on each Rx antenna.

4. The channel estimation method based on passive measurement of a hybrid wavefront signal model according to claim 1, characterized in that: The step S40 is specifically as follows: Step S41: Using the plane wavefront model, according to the received incomplete data and To estimate the complete data , by calculating the given And assume Conditional expectation of get Natural estimate of in, Represents the first The signal transmitted by the path Indicates the first Gaussian white noise of the paths; represents the total amount of received signals, represents the observation of the received signal, is a real coefficient, Acceptable data, is a positive constant; Step S42: Due to is unobservable, using Natural estimate of replace , find the given observation value conditions, The log-likelihood function of ; in is the population of unobservable complete data, for Observation, for Natural estimate of For the Channel characteristic parameters to be estimated of the stripe diameter; Step S43: Channel characteristics are analyzed by performing the following two steps: and Perform SAGE estimation, where 、 、 and They are respectively the first The time delay, Doppler frequency, direction of arrival unit vector and complex attenuation coefficient of the stripe; Include and Complex attenuation coefficients for both polarization directions; in, is the complex attenuation coefficient of this round of iteration Other channel characteristic parameters Estimates, , For the The complex attenuation coefficient of the strip diameter Other channel characteristic parameters Estimates, and It is an intermediate variable in formula derivation and arrangement; in, is the complex attenuation coefficient of this iteration Estimates, For the Complex attenuation coefficient of strip diameter Estimates, For the snapshot number, For input signal The power, is the scanning interval of the receiving end; in, in, It is The unit vector of the arrival direction of the Rx antenna pair is The incident wave Polarization antenna response, represents two mutually orthogonal polarization directions, The first array element ; The first array element ; is the speed of light, For the The receiving array in the snapshot cycle The starting time of each array element, is the input signal of the system, i.e. the CSI-RS transmission signal of the base station. is an intermediate variable; Step S44: By calculating Determine whether the channel parameters converge; if not, let , return to step S41; if converged, use the estimated Reconstruct this multipath , execute step S50; the convergence refers to The difference is less than the threshold.

5. The channel estimation method based on passive measurement of a hybrid wavefront signal model according to claim 4, characterized in that: The plane wavefront model is expressed as: Among them, the channel characteristic parameters , 、 、 and They represent the first The path delay, Doppler frequency, arrival direction unit vector and The complex attenuation coefficient experienced by the polarization received signal, The azimuth of the multipath arrival angle and pitch angle The only certainty, ; The plane wavefront model The first in the snapshot cycle The observations of the antenna array element are Signals transmitted along the path; is the input signal of the system, i.e., the CSI-RS transmission signal of the base station; It is The unit vector of the arrival direction of the Rx antenna pair is The incident wave Polarization antenna response; For the The receiving array in the snapshot cycle The starting time of each array element; It is The first in the snapshot cycle Gaussian white noise received on the Rx antenna array element.

Images