Vehicle-mounted communication optimization system based on active antenna

By using an active antenna-based vehicle communication optimization system, beamforming and resource allocation are adaptively adjusted, solving the problems of multipath effect and rapid channel state changes in vehicle communication, improving communication quality and system adaptability, and optimizing resource utilization and spectrum efficiency.

CN119010975AActive Publication Date: 2024-11-22SHENZHEN FENGYUN TECH CO LTD

Patent Information

Application Number
CN202411487330.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-24
Publication Date
2024-11-22
Estimated Expiration
2044-10-24

AI Technical Summary

Technical Problem

Multipath effects in vehicle communication cause signal distortion and interference, which are difficult to overcome with traditional methods. When the channel state changes rapidly, the system configuration cannot be optimized in real time. The dense presence of vehicle terminals leads to severe interference. Traditional methods cannot make full use of spatial degrees of freedom for beamforming and resource allocation. The system performance and flexibility are insufficient, and it cannot meet the bandwidth and capacity requirements of vehicle networking and autonomous driving.

Method used

An on-board communication optimization system based on active antennas is adopted, including a multi-beam active antenna array module, a clustering network module, an adaptive beamforming module, and a multi-user multiple access module. By adaptively adjusting the weighting coefficients of the active antenna elements, an optimization algorithm is used to select the best beam for data transmission, and resources are dynamically allocated according to channel state information.

Benefits of technology

It significantly improves communication quality and reliability, enables dynamic response to signal clusters, enhances the system's adaptability and spectrum utilization in complex environments, improves the accuracy of signal analysis and noise immunity, optimizes resource utilization, and enhances user experience.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119010975A_ABST
    Figure CN119010975A_ABST
Patent Text Reader

Abstract

The invention belongs to the technical field of vehicle-mounted communication, and discloses a vehicle-mounted communication optimization system based on an active antenna. Comprising a multi-beam active antenna array module composed of a plurality of active antenna units; the clustering formation network module is connected with the multi-beam active antenna array module and is used for receiving the multipath signals and preprocessing the multipath signals to obtain multipath standard signals; clustering the multipath standard signals to obtain n signal clusters; the self-adaptive beam forming module is used for adjusting the complex weight coefficient of each active antenna unit according to the n signal clusters so as to form a plurality of beams; the beam screening module is used for acquiring channel state information of each beam and selecting an excellent beam by using an optimization algorithm; and the multi-user multiple access module is used for carrying out data transmission by using the excellent wave beams and allocating resources, and remarkable effects are achieved in the aspects of improving the communication quality, enhancing the system adaptability, optimizing the resource utilization and improving the user experience.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle-mounted communication, and more specifically, to a vehicle-mounted communication optimization system based on an active antenna. Background Art

[0002] The patent with application publication number CN108562876A discloses a broadband low sidelobe analog multi-beam array reconnaissance system, including a broadband active antenna array, a Rotman lens network, a beam selection switch, a broadband digital receiver, and a signal processing unit. The broadband active antenna array is connected to the analog multi-beam network. The Rotman lens network is used to form multiple beams simultaneously within a large instantaneous bandwidth. At the same time, the corresponding amplitude and phase weights are assigned to each amplitude and phase weighting module of the antenna array. The beam selection switch performs radio frequency selection in multiple beams according to a control command, and sends the selected signal to a broadband digital receiver. The digital receiver processes data from multiple beams and frequency bands in parallel, measures parameters of the data, forms a radar pulse description word, and then sends it to a signal processing unit. The signal processing unit processes the signal to form a radar description word, thereby achieving high system sensitivity, wide frequency domain coverage, wide spatial domain coverage, flexible beam scheduling, and good adaptability to complex electromagnetic environments.

[0003] In terms of in-vehicle communications, due to the serious multipath effect in the mobile vehicle environment, signal distortion and interference occur, which seriously reduces the communication quality and reliability. Due to the high-speed movement of vehicles and the effects of urban buildings, the multipath effect is more severe and difficult to predict, and it is difficult for traditional methods to effectively overcome this influence. Secondly, in the in-vehicle communication scenario, the channel state changes rapidly, and traditional methods lack the ability to quickly track and adapt to these changes, and cannot optimize the system configuration in real time, resulting in performance degradation. Furthermore, the in-vehicle terminals are densely distributed and the interference is serious. Traditional methods cannot fully utilize the spatial degrees of freedom for beamforming and resource allocation, and it is difficult to maximize the system capacity. With the rise of applications such as the Internet of Vehicles and autonomous driving, in-vehicle terminals will be more dense, and the demand for system bandwidth and capacity will increase significantly. In addition, traditional methods lack special optimization designs for in-vehicle communication scenarios, and their performance and flexibility need to be improved. In addition, in-vehicle communications have special requirements for reliability, real-time and adaptability, and existing general technologies are difficult to fully adapt to them.

[0004] In view of this, the present invention proposes an in-vehicle communication optimization system based on active antenna to solve the above problems. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art and to achieve the above-mentioned purpose, the present invention provides the following technical solutions: an active antenna-based vehicle communication optimization system, comprising: a multi-beam active antenna array module, composed of a plurality of active antenna units; The clustering network module is connected to the multi-beam active antenna array module to receive multipath signals and pre-process the multipath signals to obtain multipath standard signals; cluster the multipath standard signals to obtain n signal clusters; An adaptive beamforming module, used to adaptively adjust the complex weight coefficient of each active antenna unit according to n signal clusters, so as to form a number of beams; A beam screening module is used to obtain the channel state information of each beam and select one or m beams as good beams using an optimization algorithm according to the channel state information; The multi-user multiple access module is used to utilize excellent beams for data transmission and allocate resources to different vehicle terminals based on the channel state information of the excellent beams; each module is connected by wire and / or wireless means.

[0006] Furthermore, the method of preprocessing the multipath signal includes: The noise component in the multipath signal is eliminated to obtain a denoised multipath signal. The complex fading coefficient of each path in the denoised multipath signal is calculated using a preset pilot sequence. According to the complex fading coefficient of each path, the denoised multipath signal is equalized to obtain a multipath standard signal.

[0007] Furthermore, the noise component removal method includes: The spectrum coefficient of the multipath signal is calculated by using the improved short-time Fourier transform. The calculation formula of the spectrum coefficient is: SF(t,k)= ∑ [ (x(p) ∙ w(p - t) ∙ g(p) ∙ exp ( - j ∙ 2 π k(p+ t ) N )) ∙ exp (j ∙θ ) ] ;in, is a multipath signal, is the moving window function; is the windowing function; is an imaginary unit, is the time index, To improve the frequency index of the short-time Fourier transform, is an adjustable delay parameter, is the phase correction term; is the number of sampling points in the frequency domain; For time and frequency Spectral coefficients at ; is the signal point index of the multipath signal; Perform wavelet transform on the multipath signal to obtain a wavelet coefficient matrix, find the local maximum point in the wavelet coefficient matrix as the starting point of the ridge line; start from each starting point, extend along the direction of the local maximum to form a signal ridge line, and obtain A signal ridge lines in total; based on the A signal ridge lines, obtain the effective area R; The effective area R is represented by a binary function Indicates that within the effective region R, , otherwise ; Based on the spectrum coefficient and the binary function, the effective signal energy is calculated; the calculation formula of the effective signal energy is: ;in, is the effective signal energy, is a hard criterion function; ;in, for The phase concentration, for The higher-order cumulants of is a unit step function; is the amplitude judgment threshold, is the concentration judgment threshold, is the cumulative amount judgment threshold; Defining energy thresholds ;in, is the empirical coefficient, is the weighted average noise power; if , then the time-frequency component is judged to be valid; if , then the time-frequency component is judged to be the noise component, and the noise component is removed to obtain the denoised multipath signal.

[0008] Furthermore, the method of obtaining the effective area R includes: A preset length threshold is set. For each signal ridgeline, its length is calculated. If the length is less than the preset length threshold, the corresponding signal ridgeline is removed. For the remaining signal ridge lines, calculate their comprehensive curvature. The calculation formula for the comprehensive curvature is: ; Among them, the signal ridge line is at time and scale The coordinates on , for The first derivative of for The second-order derivative of and For the general The weighted parameters, is the regularization parameter; Define the curvature threshold, record the comprehensive curvature of the remaining signal ridge lines, and remove the signal ridge lines whose comprehensive curvature is greater than the curvature threshold; Each signal ridgeline that is finally retained is recorded as the best ridgeline; the best ridgeline corresponds to a time-frequency component of the multipath signal; Get the duration of each optimal ridgeline , get the bandwidth of each optimal ridgeline ; Construct area judgment function ;in, and is the corresponding weight parameter; Preset judgment threshold , calculate the value of the regional judgment function for each optimal ridge line ,like ; then the corresponding best ridge line is determined to be the effective signal component; For the best ridgeline determined as a valid signal component, the time-frequency region it covers is recorded, and the final valid region R is defined as the union of the time-frequency regions corresponding to all the best ridgelines.

[0009] Furthermore, the method for obtaining the weighted average noise power includes: Construct a minimum mean square adaptive filter, the input is a multipath signal, then the filter coefficient vector is The error signal is ; According to the least mean square criterion, use LMS or RLS algorithm to continuously adjust , when convergence, calculate Power , as the integrated noise power; Set a smoothing factor ,and ; Calculate the weighted average noise power ;in, for The weighted average noise power at the previous time point.

[0010] Furthermore, the calculation method of the complex fading coefficient includes: In the experimental environment, the transmitter sends a known pilot sequence Insert the historical multipath signal sent, and use the historical multipath signal sent by the receiving end Extract the corresponding transmission pilot sequence ; Using the known pilot sequence ,right Perform correlation operations to obtain the channel impulse response estimate ; The channel impulse response estimate Perform peak detection to obtain the paths Delay ; Extract at the delay The value of each path The complex fading coefficient ; The method of performing equalization processing includes: Define an adaptive equalizer, perform Fourier transform on the complex fading coefficients, and obtain the preliminary frequency response of the channel , based on the preliminary frequency response Calculate the frequency response of an adaptive equalizer ; The frequency response is calculated as: ;in, for The complex conjugate of is the power spectral density of Gaussian white noise, is the energy of sending the symbol, is the power spectral density of inter-symbol interference; right Perform inverse Fourier transform to obtain the impulse response of the adaptive equalizer, where the coefficients of the adaptive equalizer are the values ​​of the impulse response at the sampling point; only the first M1 coefficients are retained as the coefficients of the adaptive equalizer; Convolve the denoised multipath signal with the coefficients of the adaptive equalizer to obtain the multipath standard signal ; The formula for convolution is: ;in, is the unit impulse function, The adaptive equalizer coefficients.

[0011] Furthermore, the method of clustering the multipath standard signals includes: The multipath standard signal is represented as N2 D-dimensional data points, where N2 is the number of multipath components and D is the dimension of the feature. If D is equal to 2, the feature includes the arrival angle and the delay. The D-dimensional data space to which the multipath standard signal belongs is divided into a number of identical grid cells. For each grid cell, the number of data points falling within the grid cell is counted as the density of the grid cell; a density threshold is preset, and if the density of any grid cell is greater than the density threshold, the corresponding grid cell is marked as a dense cell; The density thresholds can be preset in the following ways: For the grid cells , define its neighborhood as a set of grid cells that differ by at most 1 in any dimension, and calculate the mean density of all grid cells in its neighborhood and density standard deviation , then set the grid unit Density threshold ;in, is a density parameter greater than 0; are the relevant control parameters, is the correlation coefficient between the density of a grid cell and the mean density of all grid cells in its neighborhood; Taking dense units as nodes, if two dense units are adjacent, an edge is connected between them to form an undirected graph. In this undirected graph, any set of connected nodes corresponds to a dense area; for data points falling into the same dense area, they are clustered into a cluster; finally, n signal clusters are obtained.

[0012] Furthermore, the grid unit division method includes: For dimensions, set the number of grid cells to , then The width of the grid cell in dimensions ;in, For the The adaptive coefficient of the dimension; For the The maximum value of the feature of the dimension, No. The minimum value of the feature in each dimension, To uniformly control the parameters, For the The standard deviation of the features in the dimension, For the The mean of the features in the dimensions, For the The density correction factor of each dimension; Density correction factor ;in, For the The density of data points in each dimension; is the preset expected density; is a dense control parameter; According to the calculated width of the grid unit, the D-dimensional data space is divided into the same grid units.

[0013] Furthermore, the beam forming method includes: For n signal clusters, each signal cluster corresponds to a main arrival angle. The main arrival angle of the signal cluster , construct its guiding vector ; a_Y= [1,exp( e ∙ Vsin i _Y),exp( e ∙ 2Vsin i _Y),...,exp( e ∙ (d '- 1)Vsin i _Y)] T ;in, is an imaginary unit, is the comprehensive wave number; is the number of active antenna units; [ ] T is the transposition operation; comprehensive wave number ;in, is the signal wavelength, is the spacing between active antenna elements; The steering vectors of all n signal clusters form an array response matrix A1=[a_Y] , the constrained minimum variance algorithm is applied to solve the array response matrix and obtain the preliminary complex weight coefficients; For active antenna unit, measure the deviation between its actual position and ideal position; and measure its phase error , that is, the difference between the actual phase and the ideal phase; For active antenna unit, calculate its compensation factor ; The calculation formula of the compensation factor is: ;in, It is The distance from the active antenna element to the beam reference point, yes The square root of the sum of the squares of the offset; The initial complex weight coefficient is multiplied by the corresponding compensation factor to obtain the compensated complex weight coefficient, and the compensated complex weight coefficient is normalized and distributed to each active antenna unit to form the final beam; each beam corresponds to a signal cluster.

[0014] Further, the channel state information of the beam includes signal-to-noise ratio, bit error rate, Doppler shift, delay spread and angle spread; The selection method of the excellent beam includes: A molecular model is constructed, where several beams are represented as several molecules, and the state of each molecule is described by its position, velocity, and acceleration. Based on the channel state information of each beam, an initial position and initial velocity are assigned to the corresponding molecule. The total potential energy of the molecular model is defined as the sum of the single-molecule potential energy of each molecule. The calculation formula of the single-molecule potential energy is: ;in, For the The signal-to-noise ratio of a molecule, , , , and is the corresponding channel state weight coefficient; For the The bit error rate of each molecule, For the The Doppler shift of a molecule, For the The effective delay spread of the molecules is For the The effective angular spread of each molecule; For the The single-molecule potential energy of a molecule; is the preliminary adjustment coefficient; Based on The single-molecule potential energy of the molecule is calculated as Forces on molecules ; Using the classical Newtonian equations of motion, calculate Molecules under force The new position, new velocity and new acceleration under the action; and calculate the total energy of the molecular model. If the total energy remains stable within a fixed time, the molecular model reaches equilibrium; In the equilibrium state, the beam corresponding to the molecule with the smallest single-molecule potential energy or the beam corresponding to the first m molecules with the smallest single-molecule potential energies is selected as the excellent beam.

[0015] The technical effects and advantages of the vehicle-mounted communication optimization system based on active antenna of the present invention are as follows: The present invention can effectively pre-process, cluster and adaptively beamform multipath signals, can quickly track and adapt to changes in channel states, significantly improve communication quality and reliability, realize dynamic response to signal clusters, and greatly improve the adaptability of the system in complex environments. The introduced active antenna unit error compensation mechanism greatly improves the accuracy of beamforming; the optimization algorithm based on the molecular model realizes the comprehensive utilization of channel state information, making beam selection more intelligent and efficient; then, resources are dynamically allocated according to the real-time channel state, which significantly improves the spectrum utilization; in addition, the accuracy of signal analysis and the ability to resist noise are improved through the improved signal processing algorithm; remarkable results have been achieved in improving communication quality, enhancing system adaptability, optimizing resource utilization and improving user experience, providing a comprehensive solution for improving the performance of vehicle-mounted communications. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic diagram of an active antenna-based vehicle communication optimization system of the present invention; Figure 2 is a schematic diagram of an active antenna unit of the present invention; Figure 3 It is a schematic diagram of a vehicle-mounted communication optimization method based on active antenna according to the present invention. DETAILED DESCRIPTION

[0017] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0018] Example 1 See also Figure 1 As shown, the present embodiment is an active antenna-based vehicle communication optimization system, comprising: a multi-beam active antenna array module, which is composed of a plurality of active antenna units; The clustering network module is connected to the multi-beam active antenna array module to receive multipath signals and pre-process the multipath signals to obtain multipath standard signals; cluster the multipath standard signals to obtain n signal clusters; An adaptive beamforming module, used to adaptively adjust the complex weight coefficient of each active antenna unit according to n signal clusters, so as to form a number of beams; A beam screening module is used to obtain the channel state information of each beam and select one or m beams as good beams using an optimization algorithm according to the channel state information; The multi-user multiple access module is used to utilize excellent beams for data transmission and allocate resources to different vehicle terminals based on the channel status information of the excellent beams; each module is connected by wire and / or wireless means to achieve data transmission between modules.

[0019] Further, such as Figure 2 As shown, the active antenna unit integrates a bandpass LC filter, a limiter, a low noise amplifier and two transceiver switches (transceiver switch 2 and transceiver switch 3); the two transceiver switches are controlled in a unified manner, the bandpass LC filter is used to suppress signals outside the working frequency band to prevent the low noise amplifier from saturating; the limiter is used to prevent high-power signal input from burning the subsequent circuit, for example, it can withstand a peak power of 1000W, a pulse width of 50ns, a period of 1ms, and a limit of ≤18dBm; it should be noted that, during implementation, there will be an RF feeder, and the active antenna unit needs to receive and transmit RF signals, so an RF feeder is required to introduce or lead the RF signal into or out of the active antenna unit.

[0020] Furthermore, the method of preprocessing the multipath signal includes: Multipath signal refers to the signal on different paths generated by the multipath effect. The noise component in the multipath signal is removed to obtain the denoised multipath signal. The complex fading coefficient of each path in the denoised multipath signal is calculated using the preset pilot sequence. According to the complex fading coefficient of each path, the denoised multipath signal is equalized to obtain the multipath standard signal.

[0021] The noise component removal methods include: The spectrum coefficient of the multipath signal is calculated by using the improved short-time Fourier transform. The calculation formula of the spectrum coefficient is: SF(t,k)= ∑ [ (x(p) ∙ w(p - t) ∙ g(p) ∙ exp ( - j ∙ 2 π k(p+ t ) N )) ∙ exp (j ∙θ ) ] ;in, is the multipath signal (discrete time series), is a moving window function (using a Gaussian window function or a rectangular window function) used to extract time Local signal at is a windowing function (using Kaiser window function) to further suppress spectral leakage; is an imaginary unit, is the time index, To improve the frequency index of the short-time Fourier transform, is an adjustable delay parameter, is a phase correction term used to compensate for the phase distortion caused by the introduction of time delay; is the number of sampling points in the frequency domain, which is a positive integer; For time and frequency Spectral coefficients at ; is the signal point index of the multipath signal.

[0022] The multipath signal is subjected to wavelet transform to obtain a wavelet coefficient matrix, and the local maximum point is found in the wavelet coefficient matrix as the starting point of the ridge line; starting from each starting point, it is extended along the direction of the local maximum to form a signal ridge line, and a total of A signal ridge lines are obtained; based on the A signal ridge lines, the effective area R is obtained.

[0023] A preset length threshold is set, and for each signal ridge line, its length is calculated. If the length is less than the preset length threshold, the corresponding signal ridge line is considered to be too short and is removed. Too short signal ridge lines usually correspond to noise or short transient components, which are insufficient to describe the main time-frequency behavior of the signal.

[0024] For the remaining signal ridge lines, calculate their comprehensive curvature. The calculation formula for the comprehensive curvature is: ; Among them, the signal ridge line is at time and scale The coordinates on , for The first derivative of for The second-order derivative of and For the general The weighted parameter is used to adaptively weight the curvature according to time t. is a regularization parameter used to avoid numerical instability caused by a too small denominator; the parameters of the formula are adaptively determined based on the data of the actual signal ridgeline using machine learning or optimization algorithms.

[0025] Define the curvature threshold, record the comprehensive curvature of the remaining signal ridge lines, and remove the signal ridge lines whose comprehensive curvature is greater than the curvature threshold. It is considered that the curvature of the signal ridge line is too large at this location. Excessive curvature means that the signal ridge line is discontinuous or not smooth, which is caused by noise or interference.

[0026] Each signal ridgeline that is finally retained is recorded as the best ridgeline; the best ridgeline corresponds to a time-frequency component of the multipath signal.

[0027] Get the duration of each optimal ridgeline (defined as the length of the best ridgeline, or the time range in which the best ridgeline exists), obtain the bandwidth of each best ridgeline (defined as the width of the optimal ridgeline in the scale / frequency direction).

[0028] Constructing a region judgment function ;in, and is the corresponding weight parameter; preset judgment threshold , calculate the value of the regional judgment function for each optimal ridge line ,like ; then the corresponding optimal ridge line is determined to be the effective signal component.

[0029] For the best ridgeline determined as a valid signal component, the time-frequency region it covers is recorded, and the final valid region R is defined as the union of the time-frequency regions corresponding to all the best ridgelines.

[0030] The effective area R is represented by a binary function (function of time and frequency) means that within the effective region R, , otherwise .

[0031] Based on the spectrum coefficient and the binary function, the effective signal energy is calculated; the calculation formula of the effective signal energy is: ;in, is the effective signal energy, is a hard criterion function, which is used to further eliminate suspicious time-frequency components.

[0032] ;in, for The phase concentration, for Higher-order cumulants (such as fourth-order cumulants); is a unit step function. When the value in the brackets is greater than or equal to 0, the unit step function is equal to 1, otherwise it is 0; is the amplitude judgment threshold, is the concentration judgment threshold, is the cumulative amount judgment threshold; that is, only when all three criterion items are 1, the hard criterion function is equal to 1, and the time-frequency point is judged as a valid signal component, otherwise it is eliminated; it needs to be explained that the variance or standard deviation of the angle is calculated using the modulus and angle of the spectral coefficient, and the phase concentration is defined as the inverse of the variance / standard deviation.

[0033] Defining energy thresholds ;in, is the empirical coefficient, is the weighted average noise power; the calculated effective signal energy Energy threshold For comparison, if , then the time-frequency component is judged to be valid; if , then the time-frequency component is judged to be the noise component, and the noise component is removed (directly set to zero) to obtain the denoised multipath signal.

[0034] The weighted average noise power is obtained by: Construct a minimum mean square adaptive filter, the input is a multipath signal, then the filter coefficient vector is The error signal is ; According to the least mean square criterion, use LMS or RLS algorithm to continuously adjust , when converged, , calculate this time Power , as the integrated noise power.

[0035] Set a smoothing factor ,and ; Calculate the weighted average noise power ;in, for The weighted average noise power at the previous time point; By adaptively adjusting the threshold value, the change of the statistical characteristics of the noise can be tracked and the threshold value can be dynamically updated, thereby improving the accuracy and robustness of detecting the existence of a valid signal.

[0036] The calculation method of complex fading coefficient includes: In the experimental environment, the transmitter sends a known pilot sequence Insert the sent historical multipath signal (the multipath signal in the experimental environment) and use the historical multipath signal sent by the receiving end Extract the corresponding transmission pilot sequence ; Using the known pilot sequence ,right Perform correlation operations to obtain the channel impulse response estimate .

[0037] The channel impulse response estimate Perform peak detection to obtain the paths Delay ; Extract at the delay The value of each path The complex fading coefficient .

[0038] Specifically, a cyclic correlation operation is performed on the received signal and the conjugate complex sequence of the known pilot sequence to obtain a correlation function. The correlation function will have peaks at certain delays, and these peak points correspond to the delays of different multipath components. Specifically, if there is a peak point, it means that there is a multipath component, and its delay relative to the main path is k0 sampling periods. The peak value of the correlation function gives the delay and complex fading coefficient estimate of each multipath component, thereby reconstructing the estimate of the entire channel impulse response.

[0039] The methods for performing equalization include: Define an adaptive equalizer, perform Fourier transform on the complex fading coefficients, and obtain the preliminary frequency response of the channel , based on the preliminary frequency response Calculate the frequency response of an adaptive equalizer .

[0040] The frequency response is calculated as: ;in, for The complex conjugate of is the power spectral density of Gaussian white noise, is the energy of sending the symbol, is the power spectral density of inter-symbol interference. Inter-symbol interference is caused by the multipath effect and limited bandwidth of the channel. It will cause interference between adjacent symbols, thus affecting signal detection and judgment.

[0041] right Perform inverse Fourier transform to obtain the impulse response of the adaptive equalizer. The coefficients of the adaptive equalizer are the values ​​of the impulse response at the sampling point. Since the length of the impulse response is usually infinite, it is truncated and only the main first M1 coefficients are retained as the coefficients of the adaptive equalizer. Convolve the denoised multipath signal with the coefficients of the adaptive equalizer to obtain the multipath standard signal .

[0042] The formula for convolution is: ;in, is the unit impact function (its function is to convert the first path Delay , to compensate for the relative delays between different multipath components, such as the Kronecker function), The adaptive equalizer coefficients.

[0043] Furthermore, the method of clustering the multipath standard signals includes: The multipath standard signal is represented as N2 D-dimensional data points, where N2 is the number of multipath components, D is the dimension of the feature, and when D is equal to 2, the features include arrival angle and time delay; the D-dimensional data space to which the multipath standard signal belongs is divided into a number of identical grid cells.

[0044] The grid cell division methods include: For dimensions, set the number of grid cells to , then The width of the grid cell in dimensions ;in, For the The adaptive coefficient of a dimension is used to automatically adjust the grid cell width according to the data distribution of that dimension; For the The maximum value of the feature of the dimension, No. The minimum value of the feature in each dimension, is a uniform control parameter, when When it is larger, it responds more to the uneven distribution of data. For the The standard deviation of the features in the dimension, For the The mean of the features in the dimensions, For the The density correction factor of each dimension.

[0045] Density correction factor ;in, For the The density of data points in each dimension (number of data points / data range); is the preset expected density; is a dense control parameter, when When larger, the response to density deviations is higher.

[0046] According to the calculated width of the grid cell, the D-dimensional data space is divided into the same grid cells; for each grid cell, the number of data points falling within the grid cell is counted as the density of the grid cell; a density threshold is preset, and if the density of any grid cell is greater than the density threshold, the corresponding grid cell is marked as a dense cell.

[0047] For the grid cells , define its neighborhood as a set of grid cells that differ by at most 1 in any dimension, and calculate the mean density of all grid cells in its neighborhood and density standard deviation , then set the grid unit Density threshold ;in, is a density parameter greater than 0, which controls the size of the density threshold. When is large, only grid cells with a density significantly greater than the neighborhood mean will be judged as dense cells, resulting in fewer dense areas; When it is smaller, the opposite is true; are the relevant control parameters, is the correlation coefficient between the density of a grid cell and the mean density of all grid cells in its neighborhood.

[0048] Taking dense units as nodes, if two dense units are adjacent, an edge is connected between them to form an undirected graph. In this undirected graph, any set of connected nodes corresponds to a dense area; for data points falling into the same dense area, they are clustered into a cluster; finally, n signal clusters are obtained.

[0049] Beam forming methods include: Each active antenna unit has an adjustable complex weight coefficient. For n signal clusters, each signal cluster corresponds to a main arrival angle. The main arrival angle of the signal cluster , construct its guiding vector a_Y= [1,exp( e ∙ Vsin i _Y),exp( e ∙ 2Vsin i _Y),...,exp( e ∙ (d '- 1)Vsin i _Y)] T ;in, is an imaginary unit, is the comprehensive wave number; is the number of active antenna units; [ ] T is the transposition operation; comprehensive wave number ;in, is the signal wavelength, is the spacing between active antenna elements.

[0050] The steering vectors of all n signal clusters form an array response matrix A1=[a_Y] , the constrained minimum variance algorithm is applied to solve the array response matrix and obtain the preliminary complex weight coefficients.

[0051] For active antenna unit, measure the deviation between its actual position and the ideal position (preset); and measure its phase error , which is the difference between the actual phase and the ideal phase (preset).

[0052] For active antenna unit, calculate its compensation factor ; The calculation formula of the compensation factor is: ;in, Ideally, The distance from each active antenna element to the beam reference point (a virtual point in the desired beam direction, usually a point far enough away from the array antenna), is the actual distance after taking into account the position offset, that is The square root of the sum of the squares of the offset.

[0053] The initial complex weight coefficient is multiplied by the corresponding compensation factor to obtain the compensated complex weight coefficient, and the compensated complex weight coefficient is normalized and distributed to each active antenna unit to form the final beam; each beam corresponds to a signal cluster.

[0054] Furthermore, the channel state information of the beam includes signal-to-noise ratio, bit error rate, Doppler frequency shift, delay spread and angle spread; the channel response is estimated using a known pilot sequence, and then the signal-to-noise ratio and bit error rate are calculated; the delay spread is estimated by the power delay profile, the angle spread is estimated by the angle delay profile, and the Doppler frequency shift is estimated by the power Doppler profile.

[0055] The selection methods of good beams include: A molecular model is constructed, and several beams are represented as several molecules. The state of each molecule is described by its position, velocity and acceleration. According to the channel state information of each beam, an initial position and initial velocity are assigned to the corresponding molecule.

[0056] The total potential energy of the molecular model is defined as the sum of the single-molecule potential energy of each molecule. The calculation formula of the single-molecule potential energy is: ;in, For the The signal-to-noise ratio of a molecule (beam), , , , and is the corresponding channel state weight coefficient, which is used to balance the importance of each channel state information; For the The bit error rate of each molecule (beam), For the The Doppler shift of a molecule (beam), For the The effective delay spread of each molecule (beam) (based on the delay spread acquisition), For the The effective angular spread of each molecule (beam) (acquired based on angular spread); For the The single-molecule potential energy of a molecule; This is the initial adjustment coefficient to ensure that the denominator is not 0; Based on The single-molecule potential energy of the molecule is calculated as Forces on molecules (The single molecule potential energy negative gradient with respect to its position).

[0057] The effective delay spread refers to the part of the delay spread where the energy is concentrated within a certain delay range. Similarly, the effective angular spread refers to the part of the angular spread where the energy is concentrated within a certain delay range.

[0058] Using the classical Newtonian equations of motion, calculate Molecules under force The new position, new velocity and new acceleration under the action; and calculate the total energy of the molecular model (the sum of total potential energy and total kinetic energy, the total kinetic energy is calculated from mass and velocity). If the total energy remains stable within a fixed time, the molecular model reaches equilibrium. Specifically, an energy convergence threshold is set, which is a very small positive number. Every time step, the current total energy is calculated, and the change in the total energy within the time step is calculated. If the change is less than the energy convergence threshold, the total energy remains basically stable within the time step.

[0059] At the same time, a stable time threshold is set, which indicates the shortest time required to maintain continuous stability; if within the stable time threshold, the change in the total energy of each time step is less than the energy convergence threshold, the total energy is judged to remain stable within the stable time threshold; otherwise, if within the stable time threshold, the change in the total energy of any time step is greater than the energy convergence threshold, the total energy has not reached stability within the stable time threshold.

[0060] In the equilibrium state, the beam corresponding to the molecule with the smallest single-molecule potential energy or the beam corresponding to the first m molecules with the smallest single-molecule potential energies is selected as the excellent beam.

[0061] Methods of using good beams for data transmission and allocating resources to different vehicle terminals based on the channel state information of the good beams include: According to the channel state information, an appropriate modulation and coding method is assigned to each excellent beam, that is, beams with high signal-to-noise ratio and low bit error rate can be assigned higher-order modulation and coding, such as 64QAM, and beams with low signal-to-noise ratio and high bit error rate can be assigned lower-order modulation and coding, such as QPSK.

[0062] According to the delay spread estimate, the required time domain spreading factor is allocated, and sufficient OFDM symbol resources are allocated. According to the Doppler shift estimate, the required frequency domain spreading factor is allocated, and sufficient subcarrier resources are allocated. According to the angle spread, appropriate beam bandwidth resources are allocated to the beam.

[0063] By measuring the power or signal-to-noise ratio from each vehicle-mounted terminal to each excellent beam, each vehicle-mounted terminal is associated with its best one or several excellent beams; according to the modulation and coding method, time-frequency resources and beam bandwidth resources of the associated excellent beams, a corresponding resource mapping table is constructed to allocate available resource blocks to each vehicle-mounted terminal.

[0064] Data is sent to the respective associated vehicle terminals through the beam bandwidth of the excellent beam, transmitted on the allocated resource block, encoded using the allocated modulation and coding method, and the characteristics of the beam bandwidth are utilized to radiate the signal mainly in the direction of the associated vehicle terminal.

[0065] This embodiment can effectively pre-process, cluster and adaptively beamform multipath signals, can quickly track and adapt to changes in channel states, significantly improve communication quality and reliability, realize dynamic response to signal clusters, and greatly improve the adaptability of the system in complex environments. The introduced active antenna unit error compensation mechanism greatly improves the accuracy of beamforming; the optimization algorithm based on the molecular model realizes the comprehensive utilization of channel state information, making beam selection more intelligent and efficient; then, resources are dynamically allocated according to the real-time channel state, which significantly improves the spectrum utilization rate; in addition, the accuracy of signal analysis and the ability to resist noise are improved through improved signal processing algorithms; remarkable results have been achieved in improving communication quality, enhancing system adaptability, optimizing resource utilization and improving user experience, providing a comprehensive solution for improving the performance of vehicle-mounted communications.

[0066] Example 2 See also Figure 3 As shown, the part not described in detail in this embodiment is described in Example 1, and a vehicle communication optimization method based on an active antenna is provided, including: S1. Receive multipath signals and preprocess the multipath signals to obtain multipath standard signals; cluster the multipath standard signals to obtain n signal clusters; S2. Adaptively adjust the complex weight coefficient of each active antenna unit according to the n signal clusters, so as to form several beams; S3, obtaining channel state information of each beam, and selecting one or m beams as good beams by using an optimization algorithm according to the channel state information; S4. Use the excellent beams for data transmission and allocate resources to different vehicle terminals based on the channel state information of the excellent beams.

[0067] Example 3 This embodiment discloses an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the operation mode of the above-mentioned vehicle communication optimization method based on active antenna is implemented.

[0068] Since the electronic device introduced in this embodiment is an electronic device used to implement a vehicle-mounted communication optimization method based on an active antenna in the embodiment of the present application, based on the vehicle-mounted communication optimization method based on an active antenna introduced in the embodiment of the present application, a person skilled in the art can understand the specific implementation of the electronic device of the present embodiment and its various variations, so how the electronic device implements the method in the embodiment of the present application is not described in detail here. As long as a person skilled in the art implements an electronic device used in a vehicle-mounted communication optimization method based on an active antenna in the embodiment of the present application, it belongs to the scope of protection of this application.

[0069] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters and thresholds in the formula are set by technicians in this field according to actual conditions.

[0070] The above is only a preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technical users in this technical field, some improvements and modifications without departing from the principle of the present invention should also be regarded as the protection scope of the present invention.

Claims

1. A vehicle communication optimization system based on active antenna, characterized in that: include: Multi-beam active antenna array module, consisting of several active antenna units; The clustering network module is connected to the multi-beam active antenna array module to receive multipath signals and pre-process the multipath signals to obtain multipath standard signals; Cluster the multipath standard signals to obtain n signal clusters; An adaptive beamforming module, used to adaptively adjust the complex weight coefficient of each active antenna unit according to n signal clusters, so as to form a number of beams; A beam screening module is used to obtain the channel state information of each beam and select one or m beams as good beams using an optimization algorithm according to the channel state information; The multi-user multiple access module is used to utilize excellent beams for data transmission and allocate resources to different vehicle terminals based on the channel state information of the excellent beams; each module is connected by wire and / or wireless means.

2. The vehicle-mounted communication optimization system based on active antenna according to claim 1, characterized in that: The method of preprocessing the multipath signal includes: The noise component in the multipath signal is eliminated to obtain a denoised multipath signal. The complex fading coefficient of each path in the denoised multipath signal is calculated using a preset pilot sequence. According to the complex fading coefficient of each path, the denoised multipath signal is equalized to obtain a multipath standard signal.

3. The vehicle-mounted communication optimization system based on active antenna according to claim 2, characterized in that: The noise component removal method includes: The spectrum coefficient of the multipath signal is calculated by using the improved short-time Fourier transform. The calculation formula of the spectrum coefficient is: ;in, is a multipath signal, is the moving window function; is the windowing function; is an imaginary unit, is the time index, To improve the frequency index of the short-time Fourier transform, is an adjustable delay parameter, is the phase correction term; is the number of sampling points in the frequency domain; For time and frequency Spectral coefficients at ; is the signal point index of the multipath signal; Perform wavelet transform on the multipath signal to obtain a wavelet coefficient matrix, find the local maximum point in the wavelet coefficient matrix as the starting point of the ridge line; start from each starting point, extend along the direction of the local maximum to form a signal ridge line, and obtain A signal ridge lines in total; based on the A signal ridge lines, obtain the effective area R; The effective area R is represented by a binary function Indicates that within the effective region R, , otherwise ; Based on the spectrum coefficient and the binary function, the effective signal energy is calculated; the calculation formula of the effective signal energy is: ;in, is the effective signal energy, is a hard criterion function; ;in, for The phase concentration, for The higher-order cumulants of is a unit step function; is the amplitude judgment threshold, is the concentration judgment threshold, is the cumulative amount judgment threshold; Defining energy thresholds ;in, is the empirical coefficient, is the weighted average noise power; if , then the time-frequency component is judged to be valid; if , then the time-frequency component is judged to be the noise component, and the noise component is removed to obtain the denoised multipath signal.

4. The vehicle-mounted communication optimization system based on active antenna according to claim 3 is characterized in that: The method of obtaining the effective area R includes: A preset length threshold is set. For each signal ridgeline, its length is calculated. If the length is less than the preset length threshold, the corresponding signal ridgeline is removed. For the remaining signal ridge lines, calculate their comprehensive curvature. The calculation formula for the comprehensive curvature is: ; Among them, the signal ridge line is at time and scale The coordinates on , for The first derivative of for The second-order derivative of and For the general The weighted parameters, is the regularization parameter; Define the curvature threshold, record the comprehensive curvature of the remaining signal ridge lines, and remove the signal ridge lines whose comprehensive curvature is greater than the curvature threshold; Each signal ridgeline that is finally retained is recorded as the best ridgeline; the best ridgeline corresponds to a time-frequency component of the multipath signal; Get the duration of each optimal ridgeline , get the bandwidth of each optimal ridgeline ; Construct area judgment function ;in, and is the corresponding weight parameter; Preset judgment threshold , calculate the value of the regional judgment function for each optimal ridge line ,like ; then the corresponding best ridge line is determined to be the effective signal component; For the best ridgeline determined as a valid signal component, the time-frequency region it covers is recorded, and the final valid region R is defined as the union of the time-frequency regions corresponding to all the best ridgelines.

5. The vehicle-mounted communication optimization system based on active antenna according to claim 4 is characterized in that: The method for obtaining the weighted average noise power includes: Construct a minimum mean square adaptive filter, the input is a multipath signal, then the filter coefficient vector is The error signal is ; According to the least mean square criterion, use LMS or RLS algorithm to continuously adjust , when convergence, calculate Power , as the integrated noise power; Set a smoothing factor ,and ; Calculate the weighted average noise power ;in, for The weighted average noise power at the previous time point.

6. The vehicle-mounted communication optimization system based on active antenna according to claim 5, characterized in that: The calculation method of the complex fading coefficient includes: In the experimental environment, the transmitter sends a known pilot sequence Insert the historical multipath signal sent, and use the historical multipath signal sent by the receiving end Extract the corresponding transmission pilot sequence ; Using the known pilot sequence ,right Perform correlation operations to obtain the channel impulse response estimate ; The channel impulse response estimate Perform peak detection to obtain the paths Delay ; Extract at the delay The value of each path The complex fading coefficient ; The method of performing equalization processing includes: Define an adaptive equalizer, perform Fourier transform on the complex fading coefficients, and obtain the preliminary frequency response of the channel , based on the preliminary frequency response Calculate the frequency response of an adaptive equalizer ; The frequency response is calculated as: ;in, for The complex conjugate of is the power spectral density of Gaussian white noise, is the energy of sending the symbol, is the power spectral density of inter-symbol interference; right Perform inverse Fourier transform to obtain the impulse response of the adaptive equalizer, where the coefficients of the adaptive equalizer are the values ​​of the impulse response at the sampling point; only the first M1 coefficients are retained as the coefficients of the adaptive equalizer; Convolve the denoised multipath signal with the coefficients of the adaptive equalizer to obtain the multipath standard signal ; The formula for convolution is: ;in, is the unit impulse function, The adaptive equalizer coefficients.

7. The vehicle-mounted communication optimization system based on active antenna according to claim 6, characterized in that: The method of clustering the multipath standard signals includes: The multipath standard signal is represented as N2 D-dimensional data points, where N2 is the number of multipath components and D is the dimension of the feature. If D is equal to 2, the feature includes the arrival angle and the delay. The D-dimensional data space to which the multipath standard signal belongs is divided into a number of identical grid cells. For each grid cell, the number of data points falling within the grid cell is counted as the density of the grid cell; a density threshold is preset, and if the density of any grid cell is greater than the density threshold, the corresponding grid cell is marked as a dense cell; The density thresholds can be preset in the following ways: For the grid cells , define its neighborhood as a set of grid cells that differ by at most 1 in any dimension, and calculate the mean density of all grid cells in its neighborhood and density standard deviation , then set the grid unit Density threshold ;in, is a density parameter greater than 0; are the relevant control parameters, is the correlation coefficient between the density of a grid cell and the mean density of all grid cells in its neighborhood; Taking dense units as nodes, if two dense units are adjacent, an edge is connected between them to form an undirected graph. In this undirected graph, any set of connected nodes corresponds to a dense area; for data points falling into the same dense area, they are clustered into a cluster; finally, n signal clusters are obtained.

8. The vehicle-mounted communication optimization system based on active antenna according to claim 7, characterized in that: The grid unit division method includes: For dimensions, set the number of grid cells to , then The width of the grid cell in dimensions ;in, For the The adaptive coefficient of the dimension; For the The maximum value of the feature of the dimension, No. The minimum value of the feature in each dimension, To uniformly control the parameters, For the The standard deviation of the features in the dimension, For the The mean of the features in the dimensions, For the The density correction factor of each dimension; Density correction factor ;in, For the The density of data points in each dimension; is the preset expected density; is a dense control parameter; According to the calculated width of the grid unit, the D-dimensional data space is divided into the same grid units.

9. The vehicle-mounted communication optimization system based on active antenna according to claim 8, characterized in that: The beam forming method includes: For n signal clusters, each signal cluster corresponds to a main arrival angle. The main arrival angle of the signal cluster , construct its guiding vector ; ;in, is an imaginary unit, is the comprehensive wave number; is the number of active antenna units; is the transposition operation; comprehensive wave number ;in, is the signal wavelength, is the spacing between active antenna elements; The steering vectors of all n signal clusters form an array response matrix , the constrained minimum variance algorithm is applied to solve the array response matrix and obtain the preliminary complex weight coefficients; For active antenna unit, measure the deviation between its actual position and ideal position; and measure its phase error , that is, the difference between the actual phase and the ideal phase; For active antenna unit, calculate its compensation factor ; The calculation formula of the compensation factor is: ;in, It is The distance from the active antenna element to the beam reference point, yes The square root of the sum of the squares of the offset; The initial complex weight coefficient is multiplied by the corresponding compensation factor to obtain the compensated complex weight coefficient, and the compensated complex weight coefficient is normalized and distributed to each active antenna unit to form the final beam; each beam corresponds to a signal cluster.

10. The vehicle-mounted communication optimization system based on active antenna according to claim 9, characterized in that: The channel state information of the beam includes signal-to-noise ratio, bit error rate, Doppler shift, delay spread and angle spread; The selection method of the excellent beam includes: A molecular model is constructed, where several beams are represented as several molecules, and the state of each molecule is described by its position, velocity, and acceleration. Based on the channel state information of each beam, an initial position and initial velocity are assigned to the corresponding molecule. The total potential energy of the molecular model is defined as the sum of the single-molecule potential energy of each molecule. The calculation formula of the single-molecule potential energy is: ;in, For the The signal-to-noise ratio of a molecule, , , , and is the corresponding channel state weight coefficient; For the The bit error rate of each molecule, For the The Doppler shift of a molecule, For the The effective delay spread of the molecules is For the The effective angular spread of each molecule; For the The single-molecule potential energy of a molecule; is the preliminary adjustment coefficient; Based on The single-molecule potential energy of the molecule is calculated as Forces on molecules ; Using the classical Newtonian equations of motion, calculate Molecules under force The new position, new velocity and new acceleration under the action; and calculate the total energy of the molecular model. If the total energy remains stable within a fixed time, the molecular model reaches equilibrium; In the equilibrium state, the beam corresponding to the molecule with the smallest single-molecule potential energy or the beam corresponding to the first m molecules with the smallest single-molecule potential energies is selected as the excellent beam.

Citation Information

Patent Citations

  • Method and apparatus for implementing reception of diversity with intelligent antenna

    CN101013916A

  • OFDM (Orthogonal Frequency Division Multiplexing) soft information extraction method based on channel weighting

    CN115021847A

  • Mobile phone signal enhancement method based on multi-channel adaptive beam forming

    CN118264297A

Cited By

  • Adaptive RS channel coding optimization method based on fuzzy reasoning

    CN120110606A

  • An adaptive RS channel coding optimization method based on fuzzy reasoning

    CN120110606B