Search

CN-116248210-B - Method, system and medium for predicting large-scale multiple-input multiple-output system channel

CN116248210BCN 116248210 BCN116248210 BCN 116248210BCN-116248210-B

Abstract

The invention provides a base extension extrapolation large-scale multi-transmission system channel prediction method, a system and a medium, wherein the method comprises the steps of modeling an uplink channel by using a base extension model, representing the uplink channel as a small number of unknown BEM coefficients, and reconstructing the uplink channel by estimating the small number of unknown BEM coefficients; the estimated uplink channel is expressed as a linear combination of the discrete ellipsoid BEM and the corresponding coefficient to reduce the complexity of the downlink channel prediction, the DPS-BEM coefficient of the downlink time-varying is predicted, and the DPS-BEM coefficient obtained by prediction is used for the downlink channel recovery. The invention relieves the problem of channel aging in a large-scale MIMO TDD system.

Inventors

  • ZHU XU
  • ZHANG YANFENG
  • JIANG YUFEI

Assignees

  • 哈尔滨工业大学（深圳）

Dates

Publication Date
20260505
Application Date
20221220

Claims (7)

  1. 1. A method for base-extended extrapolation of large-scale multiple-input multiple-output system channel prediction, the method comprising the steps of: modeling an uplink channel by using a base extension model, representing the uplink channel as a small number of unknown BEM coefficients, and reconstructing the uplink channel by estimating the small number of unknown BEM coefficients; Representing the estimated uplink channel as a linear combination of the discrete ellipsoid BEM and the corresponding coefficients to reduce the complexity of the downlink channel prediction; predicting the DPS-BEM coefficient of the downlink time-varying, and using the DPS-BEM coefficient obtained by prediction for the recovery of a downlink channel; the step of modeling the uplink channel using a base extension model, representing the uplink channel as a small number of unknown BEM coefficients, reconstructing the uplink channel by estimating the BEM coefficients, considers a multi-user massive MIMO TDD system, one equipped with Base station service for root antenna Different time resources are allocated to uplink and downlink, the uplink transmission block comprises reference signals and data, the downlink only transmits data, and the mobile terminal has speed And the base station keep relatively moving; The method uses a base expansion model to model an uplink channel, represents the uplink channel as a small number of unknown BEM coefficients, and considers the adoption of an orthogonal frequency division multiplexing modulation mode in the step of reconstructing the uplink channel by estimating the BEM coefficients, wherein the frequency domain transmitting signals are as follows Wherein Representing the number of sub-carriers, obtaining time domain transmitting signals after inverse discrete Fourier transform Wherein The method comprises the steps of obtaining a time domain transmitting signal, obtaining a discrete Fourier transform matrix, avoiding inter-symbol interference by adding a cyclic prefix which is long enough to the time domain transmitting signal, and at a receiving end, after the cyclic prefix is removed, representing the received frequency domain signal as: (1) Wherein the method comprises the steps of Represent the first The received signal from the root receive antenna is, Represent the first The mean value corresponding to the root receiving antenna is 0 and the variance is Is added to the complex gaussian white noise of (c), Represent the first The time domain channel corresponding to the root receiving antenna is a pseudo cyclic matrix, expressed as: (2) in a high mobility scenario, the channel is time-varying, channel matrix The number of unknown parameters is Wherein Representing the number of paths of the channel; the step of modeling the upstream channel using the base extension model, representing the upstream channel as a small number of unknown BEM coefficients, and reconstructing the upstream channel by estimating the BEM coefficients comprises: Modeling a time-varying channel using a base expansion model: set the first The corresponding first antenna Time-varying channel of a path The complex exponential base expansion model is used to represent: (3) Wherein the method comprises the steps of Is that A base expansion matrix of the order complex exponential base expansion model, As the base expansion coefficient, To model errors, each column of the base expansion matrix is ; That is to say the channel By only Substituting the formula (3) into the formula (1), and obtaining the following receiving signal model by the base station end: (4) Wherein the method comprises the steps of Representing a permutation matrix if By applying a matrix to a unit Left cyclic shift of columns Secondary get, if Conversely, the method can be used for controlling the temperature of the liquid crystal display device, Is a matrix formed by discrete Fourier transform A sub-matrix of columns is formed, 。
  2. 2. The method of base-extended extrapolated large-scale multiple-input multiple-output (mimo) system channel prediction as claimed in claim 1, wherein said step of modeling an uplink channel using a base-extended model, representing the uplink channel as a small number of unknown BEM coefficients, and reconstructing the uplink channel by estimating the BEM coefficients further comprises: using pilot-assisted channel estimation in transmitting signals Is inserted into Each effective pilot frequency is arranged in front and behind The guard pilots are used to protect the received signal from inter-carrier interference: assume that the index set of active pilots is The received signal is expressed as: (5) Wherein the method comprises the steps of Aggregation of ; (5) The formula is rewritten as a merged form: (6) The received signals of all the receiving antennas at the base station end are expressed as: (7) Wherein the method comprises the steps of In order to receive the signal(s), For the complex exponential base extension coefficient, Is a noise matrix.
  3. 3. The method of base-extended extrapolated large-scale multiple-input multiple-output (mimo) system channel prediction as claimed in claim 2, wherein said step of modeling the uplink channel using the base-extended model, representing the uplink channel as a small number of unknown BEM coefficients, and reconstructing the uplink channel by estimating the BEM coefficients further comprises: Modeling a massive MIMO channel using a parameterized model: Wherein, the first The channels corresponding to the channel paths and all the receiving antennas are expressed as: (8) Wherein the method comprises the steps of , In order to activate the number of clusters, Is the first The number of rays activating a cluster of rays, Represent the first The complex gain of the individual rays is used, Is the first The maximum doppler shift of the individual rays, Representing sampling period, assuming that the base station end adopts uniform linear array and guiding vector Expressed as: (9) Wherein the method comprises the steps of The angle of arrival of the kth active cluster for the ith ray, As a function of the wavelength of the signal, For the spacing of any two antennas, the angle of arrival of each active cluster is expressed as Here, where Representing the center angle of arrival of the kth active cluster, Represents an angular offset relative to the center angle and satisfies Wherein An angular extension for the kth active cluster.
  4. 4. The method for base-extended extrapolated large-scale multiple-input multiple-output (mimo) system channel prediction as claimed in claim 3, wherein said step of modeling an uplink channel using a base-extended model, representing the uplink channel as a small number of unknown BEM coefficients, and reconstructing the uplink channel by estimating the BEM coefficients further comprises: Modeling spatial channels using spatial basis expansion models Expressed as: (10) Wherein the method comprises the steps of The spatial basis expansion model is represented as, Is the spatial domain base expansion coefficient and is used for generating a spatial domain base expansion coefficient, Selecting generalized complex index base expansion model to construct airspace base expansion model The generalized complex exponential base expansion model is defined as Wherein , For the modeling frequency of the complex exponential basis extension model, For the order of the complex exponential base expansion model, For the resolution parameter of the modeling frequency, ; Substituting the formula (10) into the formulas (3) and (7) to obtain a new received signal model: (11) Wherein the method comprises the steps of In order to measure the matrix of the device, For the GCE-BEM coefficient to be estimated, a linear minimum mean square error estimator is adopted to obtain the solution of the uplink GCE-BEM coefficient: (12) Wherein the method comprises the steps of For the estimated upstream GCE-BEM coefficients, Is a matrix Noise variance per column, upstream channel pass Is restored to 。
  5. 5. The method for base extension extrapolation large scale Multiple Input Multiple Output (MIMO) system channel prediction as claimed in claim 4, wherein the step of predicting the coefficients of the downstream time varying DPS-BEM and using the predicted DPS-BEM coefficients for downstream channel recovery comprises: suppose that one frame up contains The uplink GCE-BEM coefficient estimated by the j-th symbol of OFDM symbol is Expanding into a vector form: Then The uplink GCE-BEM coefficients corresponding to the OFDM symbols are expressed as In order to predict the downlink GCE-BEM coefficient according to the uplink GCE-BEM coefficient, firstly, the uplink GCE-BEM coefficient is approximately simulated by utilizing a Legend polynomial to obtain a group of fitting parameters, and then the downlink GCE-BEM coefficient is predicted by iteratively extrapolating the fitting parameters; The legendre polynomial is obtained by the recursive formula: (13) Wherein the method comprises the steps of , Fitting the upstream GCE-BEM coefficients with Legendre polynomials is expressed as: (14) Wherein the method comprises the steps of Is that In the (j) th row and (m) th column, The order of the discrete legendre polynomial, Is a discrete legendre polynomial coefficient, Is the fitting error of the discrete legendre polynomial, Is the additive Gaussian noise of the downlink, and (14) is written in a matrix vector form as follows: (15) Wherein the method comprises the steps of Is a discrete Legendre polynomial base matrix and , Is a coefficient matrix of discrete Legendre polynomials, and , Is the total error term; the fitting coefficients for the kth iteration are calculated as follows: (16) Wherein the method comprises the steps of For weighting the error matrix, in the absence of a priori error information Assume that the step size of each extrapolation is Then The discrete legendre polynomial extrapolation model for the steps is expressed as: (17) In the k+1st iteration, the predicted downstream GCE-BEM coefficients are added to the existing GCE-BEM coefficient samples, updated as follows: (18) The discrete legendre polynomial matrix also needs to be updated: (19) Predicted first The step-down GCE-BEM coefficients are expressed as: (20) Thus predicted (th) The step down channel is expressed as: (21)。
  6. 6. A base-extended extrapolated large-scale multiple-input multiple-output system channel prediction system, characterized in that the system comprises a memory, a processor and a base-extended extrapolated large-scale multiple-input system channel prediction program stored on the processor, which when run by the processor performs the steps of the method according to any one of claims 1 to 5.
  7. 7. A computer readable storage medium, characterized in that the computer readable storage medium stores a base extended extrapolated large-scale multiple-input multiple-output system channel prediction program, which when run by a processor performs the steps of the method according to any of claims 1 to 5.

Description

Method, system and medium for predicting large-scale multiple-input multiple-output system channel Technical Field The invention relates to the technical field of wireless communication, in particular to a channel prediction method, a system and a medium for a base extension extrapolation large-scale multiple-input multiple-output system. Background Massive multiple-input multiple-output (MIMO) has been widely recognized as one of the key technologies for 5G and 6G, because massive MIMO technology can well utilize spatial multiplexing gain, and greatly improve the spectral efficiency of channels. Time division duplexing (time division duplex, TDD) is a widely used duplexing scheme because it has the advantages of higher spectral efficiency, flexible modulation of upstream and downstream traffic, low operating cost, channel reciprocity, etc. However, in a high mobility scenario, the channel exhibits a fast time-varying characteristic under the influence of doppler spread, and the downlink may suffer from channel aging phenomenon, so that channel reciprocity in the TDD system is not available. While estimating the downlink channel requires a large number of reference signals, resulting in reduced system spectral efficiency, the mobile end is often inadequate to support significant channel estimation computation overhead. Disclosure of Invention The invention mainly aims to provide a channel prediction method, a system and a medium for a base extension extrapolation large-scale MIMO multi-receiving system, aiming at alleviating the problem of channel aging in a large-scale MIMO TDD system. In order to achieve the above object, the present invention provides a method for predicting a channel of a base extension extrapolation massive multiple-input multiple-output system, comprising the steps of: modeling an uplink channel by using a base extension model, representing the uplink channel as a small number of unknown BEM coefficients, and reconstructing the uplink channel by estimating the small number of unknown BEM coefficients; Representing the estimated uplink channel as a linear combination of the discrete ellipsoid BEM and the corresponding coefficients to reduce the complexity of the downlink channel prediction; and predicting the coefficients of the DPS-BEM of the downlink time variation, and using the DPS-BEM coefficients obtained by prediction for the recovery of the downlink channel. The invention further provides a method for constructing an uplink channel by modeling the uplink channel using a base extension model, representing the uplink channel as a small number of unknown BEM coefficients, and reconstructing the uplink channel by estimating the BEM coefficients, considering a multi-user massive MIMO TDD system, and providing the system with the systemBase station service for root antennaDifferent time resources are allocated to uplink and downlink, the uplink transmission block comprises reference signals and data, the downlink only transmits data, and the mobile terminal has speedAnd the base station remain relatively mobile. The invention further adopts the technical proposal that the uplink channel is modeled by utilizing a base expansion model, the uplink channel is expressed as a small quantity of unknown BEM coefficients, and in the step of reconstructing the uplink channel by estimating the BEM coefficients, the orthogonal frequency division multiplexing modulation mode is considered to be adopted, and the frequency domain transmitting signal isWhereinRepresenting the number of sub-carriers, obtaining time domain transmitting signals after inverse discrete Fourier transformWhereinThe method comprises the steps of providing a discrete Fourier transform matrix, avoiding intersymbol interference by adding a cyclic prefix which is long enough to a time domain transmitting signal, and at a receiving end, removing the cyclic prefix, and then expressing a received frequency domain signal as follows: (1) Wherein the method comprises the steps of Represent the firstThe received signal from the root receive antenna is,Represent the firstThe mean value corresponding to the root receiving antenna is 0 and the variance isIs added to the complex gaussian white noise of (c),Represent the firstThe time domain channel corresponding to the root receiving antenna is a pseudo cyclic matrix, which can be expressed as: (2) in a high mobility scenario, the channel is time-varying, channel matrix The number of unknown parameters isWhereinIndicating the number of paths of the channel. The invention further provides a method for reconstructing an uplink channel by estimating a BEM coefficient, wherein the method comprises the steps of modeling the uplink channel by using a base extension model, expressing the uplink channel as a small number of unknown BEM coefficients, and reconstructing the uplink channel by estimating the BEM coefficients, wherein the method comprises the following steps: Modeling a time-varying c