A method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel.

By performing channel measurements under multiple frequency bands, locations, and propagation conditions, and estimating multipath parameters of overlapping local subarrays, the near-field propagation characteristics problem of channel modeling for centimeter-wave ultra-large-scale antenna arrays was solved, enabling accurate characterization of channel parameters and evaluation of system performance.

CN123122707APending Publication Date: 2026-11-17SOUTHEAST UNIV +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202611332528.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2026-08-31
Publication Date
2026-11-17

AI Technical Summary

Technical Problem

Existing technologies cannot effectively describe the near-field propagation characteristics of centimeter-wave ultra-large-scale antenna array channels, including spherical wavefront characteristics and spatial non-stationarity, leading to inaccurate channel modeling.

Method used

By measuring the channel under multiple frequency bands, locations, and propagation conditions, and using overlapping local subarrays to estimate multipath parameters, the spatial variation of channel parameters along the array aperture is obtained, characterizing the near-field spherical wavefront effect, path power variation, multipath component generation and extinction, and spatial decorrelation characteristics.

Benefits of technology

It provides a consistent foundation of channel parameters, offering a basis for channel modeling, array configuration, and communication system performance evaluation of centimeter-wave ultra-large-scale antenna arrays, thereby improving the accuracy of channel modeling and system performance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN123122707A_ABST
    Figure CN123122707A_ABST
Patent Text Reader

Abstract

This invention discloses a method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna (UML) wireless channel, relating to the field of wireless channel research. It involves transmitting a broadband sounding signal and acquiring received signals from array elements under multi-band, multi-location, and multi-propagation conditions. The channel impulse response is obtained through back-to-back calibration and inverse Fourier transform. A sliding window is constructed along the array aperture, and a multipath extraction threshold is determined based on the noise floor and local peaks. Spatial alternation generalized expectation-maximization (SAGE) parameter estimation is performed within each subarray. Delay, angle, power, etc., are calculated, and the spatial variation of these parameters along the array aperture is analyzed. This invention combines the sliding window, SAGE estimation, and location index output into a unified processing flow to obtain parameters such as delay power spectral density, angular power spectral density, delay spread, angle spread, and spatial coherence distance, providing verifiable spatial parameters for near-field wireless channel modeling, array partitioning, and system evaluation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] This invention relates to the field of wireless channel research, and in particular to a method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel. Background Technology

[0002] To meet the comprehensive demands of 6G mobile communication in terms of high speed, large capacity, and wide coverage, future wireless communication systems will further develop towards multi-band collaboration. Compared with lower frequency bands, the centimeter wave band has richer available bandwidth and higher spatial resolution; compared with millimeter wave and terahertz bands, its propagation loss and blockage sensitivity are relatively lower, making it more suitable for achieving a trade-off between system capacity and coverage. As an important candidate spectrum for 6G, the centimeter wave band is usually located between the traditional cellular low-frequency band and the millimeter wave band, and has received widespread attention from academia and industry in recent years.

[0003] In traditional mobile communication systems, the physical aperture of base station antenna arrays is relatively limited, and communication terminals and major scatterers are typically located in the far-field region of the array. Therefore, channel modeling mostly adopts the plane wave assumption. Under far-field conditions, the angles at which the same propagation path reaches different array elements can be considered approximately the same, and the array response is mainly determined by the angle of arrival or departure. Thus, channel modeling in the array dimension can be relatively simplified. However, with the continuous increase in antenna array aperture in 6G systems, the boundary between the far-field and near-field regions has changed. Under near-field conditions, the traditional plane wave assumption is no longer entirely applicable. Signals exhibit spherical wavefront characteristics on the array aperture, and the propagation distance, phase, and power from different array elements to the same scatterer may also differ significantly. Therefore, near-field propagation characteristics have become an important issue that must be considered in 6G large-aperture array channel modeling.

[0004] In centimeter-wave communication systems, further increasing the size of the antenna array is a crucial means to improve system capacity and link reliability. The shorter wavelength of the centimeter-wave band compared to traditional low-frequency cellular bands makes it easier to integrate more antenna elements within a limited physical size. Simultaneously, very large-scale antenna arrays (VMAs) can utilize large-scale element collaboration to form high-gain narrow beams, thereby compensating for propagation losses at higher frequencies and improving the system's spatial resolution and spectral efficiency. However, with the continuous increase in the array's physical aperture, VMA channels exhibit new characteristics different from traditional small-scale array channels, including spherical wavefront features and spatial non-stationarity. Therefore, in centimeter-wave VMA communication systems, the channel can no longer be simply described using uniform stationary statistical parameters across the entire array; it is necessary to further consider near-field propagation, visible region variations, cluster formation and extinction, and the evolution of channel parameters along the array.

[0005] Therefore, it is necessary to conduct measurement and characteristic analysis of the near-field channel of centimeter-wave ultra-large-scale antennas. Summary of the Invention

[0006] The technical problem to be solved by this invention is to overcome the shortcomings of the prior art and provide a method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel. By performing multi-band, multi-location, and multi-propagation condition measurements on the channel of a centimeter-wave ultra-large-scale antenna array, and by using overlapping local subarrays on the array aperture to estimate multipath parameters, the spatial variation of channel parameters along the array aperture is obtained. This characterizes the propagation characteristics such as near-field spherical wavefront effect, path power variation, multipath component generation and extinction, and spatial decorrelation, providing channel parameters for near-field channel modeling, array configuration, and communication system performance evaluation.

[0007] To solve the above-mentioned technical problems, the present invention adopts the following technical solution:

[0008] A method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel, proposed according to the present invention, includes:

[0009] Step S1: Configure the transmitter and receiver. The transmitter transmits a broadband channel sounding signal, and the receiver collects the received signal according to the array element index under different frequency bands, different measurement locations, and different propagation conditions.

[0010] Step S2: Perform data preprocessing on the received signal to obtain the channel impulse response;

[0011] Step S3: Perform local multipath parameter estimation on the channel impulse response to obtain the multipath parameter set;

[0012] Step S4: Calculate the channel characteristic parameters based on the channel impulse response and multipath parameter set;

[0013] Step S5: Analyze the obtained channel characteristic parameters one by one along the antenna array at the receiving end to obtain the spatial variation results of the channel characteristic parameters along the array aperture;

[0014] Step S6: Based on the spatial variation results, analyze the array domain spatial non-stationarity, spherical wavefront effect, and spatial decorrelation characteristics to form a characterization result of the spatial propagation characteristics of the near-field channel of the ultra-large scale antenna.

[0015] As a further optimization scheme for the centimeter-wave ultra-large-scale antenna wireless channel measurement and characteristic analysis method described in this invention, step S1 includes:

[0016] S101. Configure at least one center frequency in the 7 GHz to 24 GHz frequency band, and set the bandwidth, sampling rate and waveform period of the broadband channel probe signal;

[0017] S102. Transmit the broadband channel detection signal using the transmitter;

[0018] S103. The control receiver uses four parallel 1-to-32 RF switch boxes to collect the received signals of each array element according to the array element number.

[0019] S104. Repeatedly perform signal transmission and acquisition at different measurement locations and under different propagation conditions. At each measurement location, continuously acquire multiple channel snapshots and associate and store the original received signal with the center frequency, measurement location, propagation conditions, array element number and snapshot number.

[0020] As a further optimization scheme for the centimeter-wave ultra-large scale antenna wireless channel measurement and characteristic analysis method described in this invention, the measurement position refers to determining the Rayleigh distance based on the physical aperture and measurement frequency band of the ultra-large scale receiving antenna array, and setting multiple measurement positions within the Rayleigh distance range; the propagation conditions include line-of-sight, non-line-of-sight, and obstructed line-of-sight propagation conditions. Under the obstructed line-of-sight propagation condition, a portion of the receiving antenna elements of the ultra-large scale receiving antenna array is obstructed by local obstacles, while the remaining receiving antenna elements remain visible at line of sight, thereby enabling the ultra-large scale receiving antenna array to cross the boundary of the visible area.

[0021] As a further optimization scheme for the centimeter-wave ultra-large-scale antenna wireless channel measurement and characteristic analysis method described in this invention, in step S2, data preprocessing includes using back-to-back calibration response to perform frequency domain calibration on the received signal, and transforming the calibrated channel frequency domain response into channel impulse response.

[0022] As a further optimization scheme for the centimeter-wave ultra-large-scale antenna wireless channel measurement and characteristic analysis method described in this invention, step S2 includes:

[0023] S201. Obtain the calibration signal when the transmitter and receiver are connected back to back, and calculate the calibration frequency domain response;

[0024] S202. Divide the frequency domain response of the received signal of each array element by the corresponding calibration frequency domain response to eliminate the hardware frequency response of the transmit channel, receive channel and RF switch branch, and obtain the calibrated channel frequency domain response.

[0025] S203. Perform an inverse Fourier transform on the calibrated channel frequency domain response to obtain the channel impulse response of each array element.

[0026] As a further optimization scheme for the centimeter-wave ultra-large-scale antenna wireless channel measurement and characteristic analysis method described in this invention, step S3 includes:

[0027] S301. Arrange the channel impulse response according to the spatial order of the array elements on the array aperture;

[0028] S302. Select continuous array elements with a preset window length to form a local subarray, and move the local subarray along the array aperture with a preset step size to form multiple overlapping local subarrays.

[0029] S303. Calculate the noise floor power and local maximum power for each local subarray, and take the larger value between the noise floor power level plus the preset noise margin and the local maximum power level minus the preset dynamic range limit as the multipath extraction threshold.

[0030] S304. Select candidate multipath components that are not lower than the multipath extraction threshold as the initial estimation objects, and perform spatial alternating generalized expectation maximization estimation based on the channel impulse response of each receiving antenna element in the corresponding local subarray. Alternately perform expectation step and maximization step between each candidate multipath component to update the complex gain, arrival delay and arrival azimuth of each multipath component respectively. Stop iterating when the difference between the likelihood function values ​​of two adjacent iterations is lower than the convergence threshold, the iterative changes of the complex gain, arrival delay and arrival azimuth of each multipath component are lower than the corresponding preset parameter threshold or the maximum number of iterations is reached.

[0031] S305. Obtain the complex gain, arrival delay and arrival azimuth of each multipath component, remove multipath components below the multipath extraction threshold, and associate them with the element index range and window center position covered by the corresponding local subarray.

[0032] As a further optimization scheme for the centimeter-wave ultra-large-scale antenna wireless channel measurement and characteristic analysis method described in this invention, in step S4, the channel characteristic parameters include path power, time delay power spectral density, angular power spectral density, root mean square time delay spread, root mean square azimuth spread, and coherence distance.

[0033] As a further optimization scheme of the centimeter-wave ultra-large-scale antenna wireless channel measurement and characteristic analysis method described in this invention, step S4 includes: calculating the path power of the multipath components based on the complex gain of each multipath component; calculating the delay power spectral density based on the channel impulse response of each receiving antenna element; calculating the root mean square delay spread based on the multipath complex gain and arrival delay; calculating the angular power spectral density and root mean square arrival azimuth spread based on the multipath complex gain and arrival azimuth; calculating the normalized spatial correlation coefficient based on the channel impulse response between receiving antenna elements with different spatial spacing, and determining the coherence distance based on the normalized spatial correlation coefficient.

[0034] As a further optimization scheme for the centimeter-wave ultra-large-scale antenna wireless channel measurement and characteristic analysis method described in this invention, step S5 includes:

[0035] S501. Compare the time delay power spectral density corresponding to different receiving antenna array elements and the angular power spectral density corresponding to different local subarrays to obtain the variation results of time delay power spectral density and angular power spectral density along the array aperture.

[0036] S502. Based on the root mean square time delay spread and root mean square azimuth spread corresponding to different local subarrays, calculate the mean and standard deviation of the root mean square time delay spread and root mean square azimuth spread between the sliding windows.

[0037] S503. Determine the coherent distance based on the normalized spatial correlation coefficient corresponding to different array element intervals, and statistically analyze the distribution of coherent distance under different center frequencies and propagation conditions to obtain the spatial correlation variation results of the channel.

[0038] As a further optimization scheme for the centimeter-wave ultra-large-scale antenna wireless channel measurement and characteristic analysis method described in this invention, step S6 includes:

[0039] S601. Analyze the path power variation of the line-of-sight path based on the variation of the time delay power spectral density along the array aperture.

[0040] S602. Based on the changes in time delay power spectral density and angular power spectral density along the array aperture, analyze the angle of arrival drift and the generation and extinction of multipath components.

[0041] S603. Based on the spatial mean and spatial standard deviation of the root mean square delay spread and root mean square arrival azimuth spread corresponding to different local subarrays, analyze the channel dispersion level and spatial fluctuation degree under different propagation conditions.

[0042] S604. The arrival angle drift is used as a characterization of the near-field spherical wavefront effect. The path power change of the line-of-sight path, the birth and death of multipath components, and the channel dispersion level and spatial fluctuation degree of the root mean square time delay spread and root mean square arrival azimuth spread are used as characterizations of channel spatial non-stationarity. The spatial correlation change results of the channel are used as a characterization of channel spatial decorrelation, thereby forming the near-field channel spatial propagation characteristic characterization results.

[0043] Compared with the prior art, the present invention, employing the above technical solution, has the following technical effects:

[0044] (1) This invention addresses the near-field propagation conditions of centimeter-wave ultra-large-scale antenna arrays and forms a complete processing method including multi-band signal acquisition, data preprocessing, local multipath parameter estimation, channel characteristic parameter calculation, and spatial propagation characteristic characterization.

[0045] (2) First, this invention eliminates the hardware frequency response generated by the transceiver equipment, RF channel, and connecting cables through back-to-back calibration, and improves the comparability between the received signals of different array elements and their channel impulse responses through frequency and time synchronization between the transmitter and receiver. Second, this invention constructs overlapping local subarrays along the array aperture, and performs spatial alternating generalized expectation-maximization estimation within each local subarray, enabling multipath parameter estimation to reflect the propagation state at different local locations of the VMI array; by combining the noise floor power and the local maximum power to determine the multipath extraction threshold, it can adapt to the noise level and power dynamic range of different local subarrays. Third, this invention jointly calculates path power, time delay power spectral density, angular power spectral density, root mean square time delay spread, root mean square azimuth spread, spatial correlation coefficient, and coherence distance, and analyzes the spatial variation of these parameters along the receiving antenna array, thereby characterizing the path power variation caused by obstruction, the angle of arrival drift caused by the near-field spherical wavefront, the local birth and death of multipath components, the spatial non-stationarity of the array domain, and the spatial decorrelation characteristics of the channel. Finally, this invention enables the use of a unified measurement and parameter processing method across different frequency bands, measurement locations, and propagation conditions including line-of-sight, non-line-of-sight, and obstructed line-of-sight, providing a consistent data foundation for comparing channel characteristics under different frequency bands and propagation conditions.

[0046] (3) In summary, the present invention can obtain near-field channel parameters and spatial propagation characteristics associated with the local location of the array, providing channel-level parameter basis for centimeter-wave ultra-large-scale antenna array channel modeling, array partitioning, beamforming, link simulation and communication system performance evaluation. Attached Figure Description

[0047] Figure 1 This is a flowchart of the centimeter-wave near-field ultra-large-scale antenna channel measurement and channel characteristic characterization method in Embodiment 1 of the present invention;

[0048] Figure 2 This is a schematic diagram of the near-field channel measurement system for a centimeter-wave ultra-large-scale antenna in Embodiment 1 of the present invention;

[0049] Figure 3 This is a schematic diagram of the measurement layout in Embodiment 1 of the present invention;

[0050] Figure 4 This is a spatial variation diagram of the time delay power spectral density along the receiving antenna array under the line-of-sight obstruction condition in the 7 GHz band in Embodiment 1 of the present invention.

[0051] Figure 5 This is a spatial variation diagram of the angular power spectral density along the receiving antenna array in the 7 GHz band in Embodiment 1 of the present invention;

[0052] Figure 6This is a cumulative distribution diagram of the root mean square delay spread of the 7 GHz and 10 GHz frequency bands under different propagation conditions in Embodiment 1 of the present invention;

[0053] Figure 7 This is a cumulative distribution diagram of the root mean square arrival azimuth angle spread of the 7 GHz and 10 GHz frequency bands under different propagation conditions in Embodiment 1 of the present invention;

[0054] Figure 8 This is a graph showing the spatial mean and spatial standard deviation of the root mean square delay spread of the 7 GHz band under different propagation conditions in Embodiment 1 of the present invention.

[0055] Figure 9 This is a graph showing the spatial mean and spatial standard deviation of the root mean square arrival azimuth angle spread of the 7 GHz band under different propagation conditions in Embodiment 1 of the present invention.

[0056] Figure 10 This is a cumulative distribution diagram of the coherence distance in the 7 GHz and 10 GHz frequency bands under different propagation conditions in Embodiment 1 of the present invention. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0058] This invention belongs to the field of wireless channel research, and specifically relates to a method for near-field channel measurement, multipath parameter extraction, and channel characteristic characterization of a very large-scale antenna array for centimeter-wave communication in the sixth generation (6G). See also... Figure 1 The main process of this invention includes the following steps:

[0059] S1. Configure the transmitting and receiving equipment, wherein the transmitting end transmits a broadband channel detection signal, and the receiving end collects and receives the signal according to the array element index at different frequency bands, different measurement locations, and different propagation conditions;

[0060] S2. Perform data preprocessing on the received signal to obtain the channel impulse response;

[0061] S3. Perform local multipath parameter estimation on the channel impulse response to obtain the multipath parameter set;

[0062] S4. Calculate the channel characteristic parameters based on the channel impulse response and multipath parameter set;

[0063] S5. Analyze the channel characteristic parameters obtained by one along the receiving antenna array to obtain the spatial variation results of the channel characteristic parameters along the array aperture;

[0064] S6. Based on the spatial variation results, analyze the array domain spatial non-stationarity, spherical wavefront effect, and spatial decorrelation characteristics to form a characterization result of the spatial propagation characteristics of the near-field channel of the ultra-large-scale antenna.

[0065] Furthermore, S1 includes the following steps:

[0066] S101, see also Figure 2 The transmitter is configured to include a vector signal transceiver, an upconverter, a power amplifier, and a vertically polarized biconical omnidirectional antenna connected in sequence. The vector signal transceiver generates a broadband intermediate frequency (IF) detection signal, which is converted to the target radio frequency band by the upconverter. After being amplified by the power amplifier, the signal is transmitted by the vertically polarized biconical omnidirectional antenna.

[0067] S102, see also Figure 2 The receiver is configured to include a very large-scale receiving antenna array, an RF switch group, a downconverter group, and a multi-channel vector signal transceiver. The very large-scale receiving antenna array includes multiple vertically polarized array elements arranged according to a preset array element spacing. The RF switch group switches the received signals of different array elements to the corresponding receiving channels according to the array element index. The switched received signals are then converted by the downconverter group and input to the vector signal transceiver.

[0068] S103. Determine the channel measurement parameters, wherein the channel measurement parameters include:

[0069] Center frequency: Multiple centimeter wave center frequencies are used to obtain channel measurement results in different frequency bands and analyze the changes in channel characteristics with operating frequency;

[0070] Signal bandwidth: Preset signal bandwidth, used to provide the time delay resolution capability required for broadband channel measurement;

[0071] Sampling rate: The preset sampling rate is used for digital acquisition of broadband channel probe signals;

[0072] Waveform period: Preset waveform period, used to determine the duration of a single channel probe waveform;

[0073] Array size: A very large-scale array consisting of multiple array elements arranged in a vertical polarization manner to obtain the received signals at different spatial locations of the very large-scale antenna array;

[0074] Element spacing: set according to half wavelength at the corresponding center frequency. For example, the element spacing for a 7 GHz center frequency is 21.4 mm, and the element spacing for a 10 GHz center frequency is 15.0 mm.

[0075] Radio frequency switches: Four parallel 1-to-32 radio frequency switches are used to switch the received signals of different array elements according to the array element index and to collect the data through four receiving channels;

[0076] Synchronization parameters: Rubidium atomic clocks are set at both the transmitting and receiving ends. The rubidium atomic clocks provide a 10 MHz frequency reference signal and a 1 pulse trigger signal per second to achieve frequency and time synchronization between the transmitting and receiving ends.

[0077] Channel snapshot count: Multiple channel snapshots are continuously collected at each center frequency, each measurement location, and each propagation condition to improve the signal-to-noise ratio of the received signal and the statistical stability of subsequent channel parameters;

[0078] S104, see also Figure 3 The Rayleigh distance is determined based on the physical aperture and measurement frequency band of the ultra-large scale receiving antenna array, and multiple measurement locations are set within the Rayleigh distance range; Figure 3 The measurement locations cover line-of-sight, non-line-of-sight, and obstructed line-of-sight propagation conditions. Among them, under the obstructed line-of-sight propagation condition, some elements of the ultra-large-scale receiving antenna array are obstructed, while the remaining elements remain line-of-sight visible.

[0079] S105. The transmitter transmits broadband channel detection signals at each center frequency. The receiver collects the received signals of each array element sequentially or in parallel according to the array element index. The received signals are then associated with the center frequency, measurement position, propagation conditions, array element index, receiving channel number, and channel snapshot number to form measurement data to be preprocessed.

[0080] Furthermore, S2 specifically includes the following steps:

[0081] S201. Take multiple channel snapshots continuously collected under the same center frequency, measurement position, propagation conditions and receiving antenna array element as a group, and perform average processing on the received signals at the same frequency point in each snapshot to obtain the average received signal frequency domain response corresponding to the receiving antenna array element.

[0082] S202. Perform back-to-back calibration between the transmitter and receiver, directly connecting the transmitter and receiver channels, and acquiring the calibration signal without passing through a wireless propagation channel to obtain the calibration frequency domain response of the measurement system. The calibration frequency domain response is associated with the corresponding center frequency, transmit channel, receive channel, and RF switch branch.

[0083] S203, Receive signal frequency domain response Divide by the corresponding calibration frequency domain response Obtain the calibrated channel frequency domain response Its expression is:

[0084]

[0085] The frequency domain calibration is used to eliminate hardware frequency responses generated by vector signal transceivers, upconverters, power amplifiers, downconverters, RF switches, and connecting cables;

[0086] S204. Frequency domain response of the calibrated channel Performing the inverse fast Fourier transform, we obtain the channel impulse response h(τ), which is expressed as:

[0087]

[0088] S205. Based on the center frequency, measurement area, measurement route, measurement location, propagation conditions, array element index, and receiving channel number, the channel impulse response obtained after averaging is associated and stored to form channel impulse response data for local multipath parameter estimation.

[0089] Furthermore, S3 specifically includes the following steps:

[0090] S301. Organize the channel impulse response of each element according to the arrangement order of the receiving antenna array elements on the array aperture, select a preset number of adjacent array elements from the continuously arranged receiving antenna array elements to form a local subarray, and move the local subarray along the array aperture direction to form multiple overlapping local subarrays.

[0091] S302. Determine the processing parameters used for local multipath parameter estimation, wherein the processing parameters include:

[0092] Sliding window length: A preset window length used to achieve a balance between local stationarity, parameter estimation accuracy, and computational complexity;

[0093] Movement step size: The preset movement step size ensures that adjacent local subarrays overlap and obtains multipath parameters corresponding to different local array aperture ranges;

[0094] Maximum number of multipath components: The preset maximum number of multipath components is used to limit the number of multipath components extracted in each local subarray, so as to balance the multipath extraction accuracy and computational complexity.

[0095] Noise margin: Set the noise margin according to the preset noise suppression requirements to eliminate noise components close to the noise floor;

[0096] Dynamic range: Sets a dynamic range limit based on the preset multipath extraction dynamic range to limit the multipath extraction range relative to the local strongest multipath component;

[0097] S303. Calculate the noise floor power and local maximum power for each local subarray, and determine the multipath extraction threshold based on the two.

[0098] S304. Perform spatial alternation generalized expectation-maximization estimation for each local subarray, and calculate the channel impulse response corresponding to the local subarray. Represented as the superposition of multiple multipath components, its signal model is:

[0099]

[0100] in, For the first Complex gain of a multipath component For the first Arrival delay of each multipath component For the first The azimuth angle of arrival of each multipath component. For receiving antenna array to the azimuth angle of arrival The array response vector, Noise term;

[0101] S305. The complex gain, arrival delay, and arrival azimuth of each multipath component are iteratively estimated using the spatial alternation generalized expectation-maximization algorithm. Iteration stops when the difference in likelihood function values ​​between two adjacent iterations is lower than the convergence threshold, the iterative changes in the complex gain, arrival delay, and arrival azimuth of each multipath component are all lower than the corresponding preset parameter thresholds, or the maximum number of iterations is reached. Multipath components with power not lower than the multipath extraction threshold are retained, and the complex gain, arrival delay, and arrival azimuth of the multipath components are associated with the element index range and window center position covered by the corresponding local subarray to form a multipath parameter set corresponding to each local subarray.

[0102] Furthermore, S4 specifically includes the following steps:

[0103] S401. Based on the corresponding channel impulse response and the multipath complex gain, arrival delay, and arrival azimuth angle obtained in S3, calculate the following channel characteristic parameters:

[0104] Path power: the first The complex gain of the multipath components is denoted as The path power of the multipath components The calculation is as follows:

[0105]

[0106] Delay power spectral density: For the l-th receiving antenna element covered, based on the channel impulse response of that receiving antenna element. Calculate the time delay power spectral density Its expression is:

[0107]

[0108] The time delay power spectral density is used to characterize the distribution of the received signal power at different time delay positions;

[0109] Angular power spectral density: Calculate the angular power spectral density based on the complex gain and azimuth angle of each multipath component. Its expression is:

[0110]

[0111] Where K represents the number of multipath components extracted within the local subarray. Indicates the first The azimuth angle of arrival of each multipath component. Indicates the location at the arrival azimuth angle The Dirac function at that location.

[0112] Root mean square delay spread: Calculate the root mean square delay spread based on the path power and arrival delay of the multipath components. :

[0113]

[0114] in, Indicates the first The arrival delay of each multipath component, and the root mean square delay spread are used to characterize the degree of multipath dispersion of the channel in the delay domain.

[0115] Root mean square (RMS) azimuth spread: Calculate the RMS azimuth spread based on the path power and azimuth of arrival of the multipath components. :

[0116]

[0117] The root mean square azimuth extension is used to characterize the degree of multipath dispersion of the channel in the angular domain.

[0118] Coherence distance: based on spatial spacing The channel impulse response of the two receiving antenna elements is used to calculate the normalized spatial correlation coefficient. The coherence distance is determined according to the following formula. :

[0119]

[0120] in, The spatial correlation threshold is set to 0.5; the coherence distance represents the maximum spatial range in which the channel impulse response maintains a preset degree of correlation across the array aperture.

[0121] S402. Associate the time delay power spectral density with the corresponding receiving antenna element index, associate the path power, angular power spectral density, root mean square time delay spread, and root mean square arrival azimuth spread with the corresponding local subarray positions, and associate the spatial correlation coefficient and coherence distance with the corresponding center frequency, measurement position, and propagation conditions to form a set of channel characteristic parameters.

[0122] Furthermore, S5 specifically includes the following steps:

[0123] S501. For the same center frequency, measurement location and propagation conditions, obtain the corresponding set of channel characteristic parameters according to the arrangement order of the receiving antenna array elements and local subarrays along the array aperture.

[0124] S502. Determine the parameters for spatial variation analysis along the array aperture, the parameters including:

[0125] Delay power spectral density: By comparing the delay power spectral density corresponding to different receiving antenna element indices, the variation of the received signal power with element position in the time delay domain is obtained, such as... Figure 4 As shown;

[0126] Angular power spectral density: By comparing the angular power spectral density corresponding to different local subarrays, the variation of multipath signal power in the angular domain with the position of the local subarray is obtained, such as... Figure 5 As shown;

[0127] Root mean square time delay spread: Calculate the root mean square time delay spread for different sliding windows at the same measurement location, and obtain the spatial mean and spatial standard deviation of the root mean square time delay spread between sliding windows. Its cumulative distribution is shown below. Figure 6 As shown, the distribution of its spatial mean and spatial standard deviation is as follows: Figure 8 As shown;

[0128] Root mean square (RMS) azimuth spread: Calculate the RMS azimuth spread for different sliding windows at the same measurement location, and obtain the spatial mean and spatial standard deviation of the RMS azimuth spread between sliding windows. Its cumulative distribution is shown below. Figure 7 As shown, the distribution of its spatial mean and spatial standard deviation is as follows: Figure 9 As shown;

[0129] Coherence distance: Compare the spatial correlation coefficients corresponding to different array element spacings, and statistically analyze the distribution, mean, standard deviation, and quantiles of coherence distance under different center frequencies and propagation conditions to obtain the spatial correlation variation results of the channel along the array aperture, such as... Figure 10 As shown;

[0130] S503. The spatial variation results of time delay power spectral density, angular power spectral density, root mean square time delay spread, root mean square arrival azimuth spread, and coherence distance are correlated with the corresponding center frequency, measurement location, and propagation conditions to form parameter variation results for spatial propagation characteristic analysis.

[0131] Furthermore, S6 specifically includes the following steps:

[0132] S601. For the same center frequency, measurement location, and propagation conditions, obtain the spatial variation results of the channel characteristic parameters formed in S5 along the array aperture;

[0133] S602, according to Figures 4 to 10 The analysis content included in the characterization of the spatial propagation characteristics of the near-field channel is determined. The analysis content includes: path power variation, angle of arrival drift, multipath component generation and extinction, time delay and angular dispersion parameter variation, spatial decorrelation and other characteristics.

[0134] S603. The arrival angle drift is used as a characterization of the near-field spherical wavefront effect. The path power change of the line-of-sight path, the birth and death of multipath components, and the channel dispersion level and spatial fluctuation degree of the root mean square time delay spread and root mean square arrival azimuth spread are used as characterizations of channel spatial non-stationarity. The channel spatial correlation change result characterized by coherent distance is used as a characterization of channel spatial decorrelation. This forms a characterization result of near-field channel spatial propagation characteristics associated with center frequency, measurement position, propagation conditions, and receiving antenna element index range.

[0135] Example 1:

[0136] See Figure 1 This embodiment provides a method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel. The specific steps are as follows:

[0137] S1. Configure the transmitting and receiving equipment, wherein the transmitting end transmits a broadband channel detection signal, and the receiving end collects and receives the signal according to the array element index at different frequency bands, different measurement locations, and different propagation conditions.

[0138] The specific steps are as follows:

[0139] See Figure 2 This embodiment employs a time-domain channel measurement system based on a vector signal transceiver. The transmitting end includes a vector signal transceiver, an upconverter, a power amplifier, and a vertically polarized biconical omnidirectional antenna connected in sequence. The vector signal transceiver generates a broadband intermediate frequency (IF) probe signal with a bandwidth of 200 MHz. The upconverter converts the broadband IF probe signal to the target radio frequency (RF) band. The power amplifier amplifies the frequency-converted RF signal, which is then transmitted by the vertically polarized biconical omnidirectional antenna. The height of the transmitting antenna is set to 1.4 m.

[0140] The receiving end includes a 128×8 vertically polarized VML receiving antenna array (its eight vertically polarized physical antenna elements are connected to the same RF output port, and their received signals are combined into one equivalent received signal in the RF domain), four parallel 1-to-32 RF switches, four downconverters, and a four-channel vector signal transceiver. The signal received by the VML receiving antenna array is input to the RF switch group according to the array element index, switched to four receiving channels by the four parallel RF switches, converted to intermediate frequency signals by the corresponding downconverters, and then acquired by the four-channel vector signal transceiver. The height of the receiving antenna array is set to 1.6 m, and both the transmitting and receiving antenna arrays adopt vertical polarization.

[0141] Both the transmitter and receiver are equipped with rubidium atomic clocks. The rubidium atomic clock on the transmitter side provides a 10 MHz frequency reference signal and a trigger signal of 1 pulse per second to the transmitter vector signal transceiver, while the rubidium atomic clock on the receiver side provides a 10 MHz frequency reference signal and a trigger signal of 1 pulse per second to the receiver vector signal transceiver, thereby achieving frequency and time synchronization between the transmitter and receiver.

[0142] This embodiment performs channel measurements at two center frequencies, 7 GHz and 10 GHz. Both frequency bands use the same 200 MHz signal bandwidth, 250 MS / s sampling rate, and 40 μs waveform period. Adjacent receiving antenna elements are arranged according to half the wavelength at their corresponding center frequencies. The element spacing is set to 21.4 mm at the 7 GHz center frequency and 15.0 mm at the 10 GHz center frequency. By using consistent probe signal bandwidth, sampling method, and measurement procedure at both center frequencies, channel measurement data suitable for frequency difference comparison is obtained.

[0143] See Figure 3 In this embodiment, measurements are conducted in a mixed indoor environment, encompassing four regions with different spatial structures and scatterer distributions. All measurement locations cover three propagation conditions: line-of-sight, non-line-of-sight, and obstructed line-of-sight. The first and second regions simultaneously include line-of-sight, non-line-of-sight, and obstructed line-of-sight measurement locations; the third region includes line-of-sight measurement locations; and the fourth region includes non-line-of-sight and obstructed line-of-sight measurement locations. Under obstructed line-of-sight conditions, a portion of the elements of the ultra-large-scale receiving antenna array are obstructed by metal pillars, wall corners, or other local obstacles, while the remaining elements remain visible at line of sight, thereby allowing the receiving antenna array to cross the boundary of the visible area.

[0144] During the measurement process, the receiver's position and the direction of the receiving antenna array remain fixed, while the transmitter moves sequentially to discrete measurement positions along each measurement route. At each center frequency and each measurement position, the transmitter continuously transmits broadband channel sounding signals, and the receiver switches and acquires the received signals of each array element according to the element index. At each measurement position, 100 channel snapshots are continuously acquired, and the center frequency, measurement area, measurement route, measurement position, propagation conditions, array element index, receiving channel number, and channel snapshot number are recorded, forming multi-band near-field channel measurement data for subsequent data preprocessing.

[0145] S2. Perform data preprocessing on the received signal to obtain the channel impulse response.

[0146] The specific steps are as follows:

[0147] S201. Perform snapshot averaging on the acquired received signals. For each center frequency, measurement area, measurement route, measurement location, and receiving antenna element, while maintaining the transmit signal configuration and transceiver operating status, average 100 consecutively acquired channel snapshots according to their corresponding frequency points to obtain the frequency domain response of the received signal corresponding to that receiving antenna element. Snapshot averaging reduces the impact of random noise and single measurement fluctuations on subsequent parameter estimation. The averaged frequency domain response of the received signal is denoted as... .

[0148] S202. Perform back-to-back calibration on the frequency domain response of the received signal. Before the wireless channel measurement, connect the transmitter and receiver measurement links back-to-back with the same center frequency, signal bandwidth, and sampling parameters as the actual measurement. Collect the system frequency domain response formed by the transmitting equipment, frequency converter, amplifier, RF switch, receiving equipment, and connecting cables, and use it as the back-to-back calibration response. For the two center frequencies of 7 GHz and 10 GHz, the frequency domain response of the received signal is calibrated using the back-to-back calibration response of the corresponding frequency band, so as to eliminate the influence of the measurement system's own frequency response and the amplitude and phase characteristics of the transceiver link on the wireless channel measurement results.

[0149] S203, Receive signal frequency domain response With the corresponding back-to-back calibration response Divide the two values ​​and perform an inverse fast Fourier transform on the ratio to obtain the channel impulse response. .

[0150] S204. The channel impulse response obtained after averaging is associated and stored according to the center frequency, measurement area, measurement route, measurement location, propagation conditions, receiving antenna array element index and receiving channel number.

[0151] S3. Perform local multipath parameter estimation on the channel impulse response to obtain the multipath parameter set.

[0152] The specific steps are as follows:

[0153] S301. Construct overlapping sliding subarrays along the element index direction of the receiving antenna array. For the channel impulse response of 128 receiving antenna elements obtained under the same center frequency, measurement area, measurement route, measurement position, and propagation conditions, select 20 receiving antenna elements consecutively according to their arrangement order on the array aperture to form a local subarray. Move the window sequentially along the array aperture with a sliding step size of one element. Retain 19 identical elements between adjacent local subarrays, thus constructing 109 local subarrays on the 128 receiving antenna elements. Record the starting element index, ending element index, and window center position of each local subarray to establish the correspondence between the multipath parameter estimation results and the array aperture position.

[0154] S302. Determine a multipath extraction threshold for each local subarray. Jointly process the channel impulse responses of the 20 receiving antenna elements within the local subarray to obtain the noise floor and local maximum power level corresponding to that local subarray. Use the noise floor plus 6 dB as the first candidate threshold, and the local maximum power level minus 30 dB as the second candidate threshold. Select the larger of the two candidate thresholds as the multipath extraction threshold for that local subarray. Use the threshold to exclude candidate components with excessively low noise and power. Simultaneously, set the upper limit of the number of multipath components extracted within each local subarray to 100 to balance parameter estimation accuracy and computational complexity.

[0155] S303. Perform spatial alternating generalized expectation-maximization estimation within each local subarray. Represent the channel impulse response measured within the local subarray as a superposition of K multipath components and complex Gaussian white noise.

[0156] Candidate components not lower than the multipath extraction threshold are used as initial multipath components. A desired step and a maximization step are sequentially performed on each multipath component. In the desired step, the corresponding signal is reconstructed based on the current estimates of the remaining multipath components, and the current multipath component to be estimated is separated from the measurement signal. In the maximization step, the arrival delay, arrival azimuth, and complex gain of the current multipath component are updated. The above estimation process is repeated in the order of each multipath component. Iteration stops when the difference in likelihood function values ​​between two adjacent iterations is lower than the convergence threshold, the iterative changes in the complex gain, arrival delay, and arrival azimuth of each multipath component are all lower than the corresponding preset parameter thresholds, or the preset maximum number of iterations is reached.

[0157] After the estimation is completed, record the complex gain of each multipath component within each local subarray. Arrival delay and arrival azimuth The obtained multipath parameters are associated with the element index range and window center position of the local subarray to obtain the multipath parameter set corresponding to each local subarray.

[0158] S4. Calculate the channel characteristic parameters based on the channel impulse response and multipath parameter set.

[0159] S401. Calculate the time delay power spectral density based on the channel impulse response of each receiving antenna element to characterize the distribution of received power on arrival delay and identify the arrival time and power variation of different multipath components; calculate the angular power spectral density based on the multipath complex gain and arrival azimuth angle obtained in S3 to characterize the distribution of received power in different arrival directions.

[0160] S402. Calculate the root mean square delay spread (RMS) based on the power and arrival delay of each multipath component to characterize the multipath dispersion of the signal within the local subarray in the delay domain; calculate the root mean square azimuth spread (RMS) based on the power and arrival azimuth of each multipath component to characterize the dispersion of the received signal within the local subarray in the angular domain. Each local subarray containing 20 consecutive receiving antenna elements obtains its corresponding RMS delay spread and RMS azimuth spread.

[0161] S403. Calculate the coherence distance based on the channel impulse response correlation between different receiving antenna elements, with the correlation threshold set to 0.5. The coherence distance characterizes the spatial range in which the channel maintains correlation across the array aperture and reflects the speed at which the channel undergoes spatial decorrelation with the element position.

[0162] After completing the above calculations, path power, time delay power spectral density, angular power spectral density, root mean square time delay spread, root mean square arrival azimuth spread, and coherence distance are obtained, and the corresponding channel characteristic parameters can be obtained.

[0163] S5. Analyze the channel characteristic parameters obtained by analyzing each antenna array at the receiving end to obtain the spatial variation results of the channel characteristic parameters along the array aperture.

[0164] The specific steps are as follows:

[0165] S501. For the same center frequency and measurement location, the array aperture position of the time delay power spectral density is taken as the receiving antenna element index, and the array aperture position of the window center of the local subarray is taken as the angular power spectral density. Arrival delay and arrival azimuth angle are used as parameter dimensions, and the power magnitude at each parameter position is represented by the corresponding received power. This yields the spatial variation results of the time delay power spectral density and angular power spectral density along the receiving antenna array, such as... Figure 4 , Figure 5 As shown.

[0166] S502. For each measurement location, analyze the root mean square delay spread and root mean square azimuth spread of each of the 109 local subarrays along the receiving antenna array, and calculate the mean and standard deviation of the above parameters among the local subarrays. The mean is used to characterize the overall level of the channel dispersion parameter at that measurement location, and the standard deviation is used to characterize the degree of fluctuation of the parameter along the array aperture.

[0167] S503, according to the center frequencies of 7 GHz and 10 GHz and the line-of-sight, obstructed line-of-sight, and non-line-of-sight propagation conditions, the root mean square delay spread of each local subarray is calculated. Figure 6 ), Root Mean Square Azimuth Extension ( Figure 7 ), and the coherence distance calculated based on different array element intervals ( Figure 10 We perform classification statistics to obtain the cumulative distribution results of the corresponding parameters, and fit the root mean square time delay spread and root mean square arrival azimuth spread with a log-normal distribution.

[0168] Through the above processing, the path power, time delay power spectral density, angular power spectral density, root mean square time delay spread, root mean square arrival azimuth spread, and coherence distance along the array aperture are obtained as well as statistical results, which are used by S6 to perform near-field channel spatial propagation characteristic analysis.

[0169] S6. Based on the spatial variation results, analyze the array domain spatial non-stationarity, spherical wavefront effect, and spatial decorrelation characteristics to form a characterization result of the spatial propagation characteristics of the near-field channel of the ultra-large scale antenna.

[0170] The specific steps are as follows:

[0171] S601. Analyze the multipath propagation characteristics based on the variations of time-delay power spectral density and angular power spectral density along the receiving antenna array. (See also...) Figure 4 In the medium-delay power spectral density, local occlusion will cause significant changes in the line-of-sight path power and cause some multipath components to appear or disappear along the array aperture. Figure 5 The mid-angle power spectral density and the significant drift in the azimuth angle of arrival along the line-of-sight path reflect the spherical wavefront effect in near-field propagation. The changes in path power, the drift in the angle of arrival, and the emergence and demise of multipath components collectively indicate that the channel parameters change with the position of the array elements.

[0172] S602. Combining the statistical results of time delay power spectral density, angular power spectral density, root-mean-square time delay spread, and root-mean-square arrival azimuth spread, analyze the spatial nonstationarity under different propagation conditions. See also Figure 6 , Figure 7The root mean square (RMS) azimuth spread of both frequency bands showed the largest value under non-line-of-sight conditions, followed by the line-of-sight condition with obstruction, and the smallest value under line-of-sight conditions. The RMS delay spread, however, exhibited larger values ​​under line-of-sight conditions in the 7 GHz band and under non-line-of-sight conditions in the 10 GHz band, respectively. See also... Figure 8 In the 7 GHz band, for the root mean square delay spread (RMSD), the spatial standard deviation is relatively small at most measurement locations under line-of-sight conditions, indicating that the delay dispersion of the line-of-sight channel is relatively stable along the array aperture. Under obstructed line-of-sight conditions, multiple measurement locations exhibit larger spatial standard deviations, indicating that local obstruction and the visible area boundary enhance the fluctuation of the RMS delay spread along the array aperture. Under non-line-of-sight conditions, the spatial mean of the RMS delay spread is generally large, but its spatial fluctuation varies depending on the specific measurement location. See also... Figure 9 In the 7 GHz band, for the root mean square angle of arrival spread, the spatial mean and spatial standard deviation under line-of-sight conditions are generally small, indicating that the dominant line-of-sight path makes the angular dispersion of the received signal weak and relatively stable along the array. The spatial mean under non-line-of-sight conditions is generally large, indicating that the scattering path has a wider angle of arrival distribution. The spatial standard deviation under obstructed line-of-sight conditions increases significantly, indicating that when the array crosses the boundary of the visible area, the angular dispersion will fluctuate strongly along the array aperture. Figure 8 and Figure 9 Both studies demonstrate that the spatial mean can characterize the overall level of the dispersion parameter, while the spatial standard deviation can characterize the spatial non-stationarity of the dispersion parameter along the array aperture.

[0173] S603. Obtain the spatial correlation variation results of the channel based on the statistical results of the coherence distance, and analyze the spatial decorrelation characteristics of the channel accordingly. See also Figure 10 Under line-of-sight conditions, the dominant line-of-sight path can maintain spatial correlation over a large array area; under non-line-of-sight conditions, abundant scattering causes the channel to decorrelate rapidly with increasing element spacing; under obstructed line-of-sight conditions, the array crosses the boundary of the visible area, causing the spatial correlation to exhibit the most significant fluctuations. Compared with the 7 GHz band, the average coherence distance of the 10 GHz band is reduced under all propagation conditions, with the extreme spatial fluctuations under obstructed line-of-sight conditions being more pronounced.

[0174] This results in a comprehensive characterization of near-field spherical wavefronts, path generation and extinction, local occlusion effects, dispersion parameter variations, and spatial decorrelation characteristics.

[0175] In summary, this embodiment conducts ultra-large-scale antenna channel measurements in indoor line-of-sight, obstructed line-of-sight, and non-line-of-sight scenarios at 7 GHz and 10 GHz. Channel impulse response is obtained through back-to-back calibration, and local multipath parameters are extracted using a sliding window and the SAGE algorithm. Based on this, path power, delay power spectral density, angular power spectral density, root mean square delay spread, root mean square azimuth spread, and coherence distance are calculated. The variations of each parameter along the array aperture are analyzed, achieving a comprehensive characterization of the near-field spherical wavefront, array domain spatial non-stationarity, and spatial decorrelation characteristics. This provides a measurement basis for centimeter-wave ultra-large-scale antenna channel modeling and system performance evaluation.

[0176] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for centimeter wave ultra-massive antenna wireless channel measurement and characteristic analysis, characterized in that, include: Step S1: Configure the transmitter and receiver. The transmitter transmits a broadband channel sounding signal, and the receiver collects the received signal according to the array element index under different frequency bands, different measurement locations, and different propagation conditions. Step S2: Perform data preprocessing on the received signal to obtain the channel impulse response; Step S3: Perform local multipath parameter estimation on the channel impulse response to obtain the multipath parameter set; Step S4: Calculate the channel characteristic parameters based on the channel impulse response and multipath parameter set; Step S5: Analyze the obtained channel characteristic parameters one by one along the antenna array at the receiving end to obtain the spatial variation results of the channel characteristic parameters along the array aperture; Step S6: Based on the spatial variation results, analyze the array domain spatial non-stationarity, spherical wavefront effect, and spatial decorrelation characteristics to form a characterization result of the spatial propagation characteristics of the near-field channel of the ultra-large scale antenna.

2. The method of claim 1, wherein, Step S1 includes: S101. Configure at least one center frequency in the 7 GHz to 24 GHz frequency band, and set the bandwidth, sampling rate and waveform period of the broadband channel probe signal; S102. Transmit the broadband channel detection signal using the transmitter; S103. The control receiver uses four parallel 1-to-32 RF switch boxes to collect the received signals of each array element according to the array element number. S104. Repeatedly perform signal transmission and acquisition at different measurement locations and under different propagation conditions. At each measurement location, continuously acquire multiple channel snapshots and associate and store the original received signal with the center frequency, measurement location, propagation conditions, array element number and snapshot number.

3. The method of claim 2, wherein the method further comprises: Measurement location refers to determining the Rayleigh distance based on the physical aperture and measurement frequency band of the ultra-large scale receiving antenna array, and setting up multiple measurement locations within the Rayleigh distance range; propagation conditions include line-of-sight, non-line-of-sight, and obstructed line-of-sight propagation conditions. Under obstructed line-of-sight propagation conditions, a portion of the receiving antenna elements of the ultra-large scale receiving antenna array is obstructed by local obstacles, while the remaining receiving antenna elements remain visible at line of sight, thereby enabling the ultra-large scale receiving antenna array to cross the boundary of the visible area.

4. The method of claim 3, wherein the method further comprises: In step S2, data preprocessing includes using back-to-back calibration response to perform frequency domain calibration on the received signal and transforming the calibrated channel frequency domain response into a channel impulse response.

5. The method of claim 3, wherein the method further comprises: Step S2 includes: S201. Obtain the calibration signal when the transmitter and receiver are connected back to back, and calculate the calibration frequency domain response; S202. Divide the frequency domain response of the received signal of each array element by the corresponding calibration frequency domain response to eliminate the hardware frequency response of the transmit channel, receive channel and RF switch branch, and obtain the calibrated channel frequency domain response. S203. Perform an inverse Fourier transform on the calibrated channel frequency domain response to obtain the channel impulse response of each array element.

6. The method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel according to claim 1, characterized in that, Step S3 includes: S301. Arrange the channel impulse response according to the spatial order of the array elements on the array aperture; S302. Select continuous array elements with a preset window length to form a local subarray, and move the local subarray along the array aperture with a preset step size to form multiple overlapping local subarrays. S303. Calculate the noise floor power and local maximum power for each local subarray, and take the larger value between the noise floor power level plus the preset noise margin and the local maximum power level minus the preset dynamic range limit as the multipath extraction threshold. S304. Select candidate multipath components that are not lower than the multipath extraction threshold as the initial estimation objects, and perform spatial alternating generalized expectation maximization estimation based on the channel impulse response of each receiving antenna element in the corresponding local subarray. Alternately perform expectation step and maximization step between each candidate multipath component to update the complex gain, arrival delay and arrival azimuth of each multipath component respectively. Stop iterating when the difference between the likelihood function values ​​of two adjacent iterations is lower than the convergence threshold, the iterative changes of the complex gain, arrival delay and arrival azimuth of each multipath component are lower than the corresponding preset parameter threshold or the maximum number of iterations is reached. S305. Obtain the complex gain, arrival delay and arrival azimuth of each multipath component, remove multipath components below the multipath extraction threshold, and associate them with the element index range and window center position covered by the corresponding local subarray.

7. The method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel according to claim 1, characterized in that, In step S4, the channel characteristic parameters include path power, delay power spectral density, angular power spectral density, root mean square delay spread, root mean square azimuth spread, and coherence distance.

8. The method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel according to claim 7, characterized in that, Step S4 includes: calculating the path power of the multipath components based on the complex gain of each multipath component; calculating the delay power spectral density based on the channel impulse response of each receiving antenna element; calculating the root mean square delay spread based on the multipath complex gain and arrival delay; calculating the angular power spectral density and the root mean square arrival azimuth spread based on the multipath complex gain and arrival azimuth; calculating the normalized spatial correlation coefficient based on the channel impulse response between receiving antenna elements with different spatial spacings, and determining the coherence distance based on the normalized spatial correlation coefficient.

9. The method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel according to claim 1, characterized in that, Step S5 includes: S501. Compare the time delay power spectral density corresponding to different receiving antenna array elements and the angular power spectral density corresponding to different local subarrays to obtain the variation results of time delay power spectral density and angular power spectral density along the array aperture. S502. Based on the root mean square time delay spread and root mean square azimuth spread corresponding to different local subarrays, calculate the mean and standard deviation of the root mean square time delay spread and root mean square azimuth spread between the sliding windows. S503. Determine the coherent distance based on the normalized spatial correlation coefficient corresponding to different array element intervals, and statistically analyze the distribution of coherent distance under different center frequencies and propagation conditions to obtain the spatial correlation variation results of the channel.

10. The method for measuring and analyzing the characteristics of a centimeter-wave ultra-large-scale antenna wireless channel according to claim 1, characterized in that, Step S6 includes: S601. Analyze the path power variation of the line-of-sight path based on the variation of the time delay power spectral density along the array aperture. S602. Based on the changes in time delay power spectral density and angular power spectral density along the array aperture, analyze the angle of arrival drift and the generation and extinction of multipath components. S603. Based on the spatial mean and spatial standard deviation of the root mean square delay spread and root mean square arrival azimuth spread corresponding to different local subarrays, analyze the channel dispersion level and spatial fluctuation degree under different propagation conditions. S604. The arrival angle drift is used as a characterization of the near-field spherical wavefront effect. The path power change of the line-of-sight path, the birth and death of multipath components, and the channel dispersion level and spatial fluctuation degree of the root mean square time delay spread and root mean square arrival azimuth spread are used as characterizations of channel spatial non-stationarity. The spatial correlation change results of the channel are used as a characterization of channel spatial decorrelation, thereby forming the near-field channel spatial propagation characteristic characterization results.