Millimeter wave MIMO-OFDM system channel estimation enhancement method

Abstract

The invention relates to a millimeter wave MIMO-OFDM (Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing) system channel estimation enhancement method, which mainly solves the problems of unnecessary calculation overhead and precision loss in an estimation process due to the fact that the prior information of a stable path is not effectively utilized in the existing method. The method comprises the following implementation steps: 1) for each subcarrier, a base station (BS) is equipped with different beamforming vectors in continuous time frames, and a user (UE) independently detects a transmission signal in each time subframe; 2) separating known path information from the received signal to obtain a tensor slice combined with prior and unknown information, and constructing the tensor slice into a third-order tensor based on a parallel factor (PARAFAC) model; and 3) estimating three factor matrixes by using a prior information optimization alternative least square (ALS) algorithm, and estimating unknown channel parameters by using maximum likelihood (ML). According to the channel estimation enhancement method provided by the invention, the ALS algorithm can be optimized by using the prior information of a stable path, the estimation error can be reduced, and the system performance can be improved.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a method for enhancing channel estimation of a millimeter wave MIMO-OFDM system. Background Art

[0002] Millimeter wave communication is an advanced communication technology that uses the 30GHz to 300GHz frequency band for wireless transmission. Due to the short wavelength of millimeter waves, multiple-input multiple-output (MIMO) antenna arrays can be deployed in a limited space, and the signal quality can be significantly enhanced through beamforming technology. This frequency band provides abundant spectrum resources, but also faces many challenges, including high propagation path loss, complex multipath effects, and rapid channel changes in dynamic environments. Therefore, accurate channel state information (CSI) estimation is the key to its successful application.

[0003] In millimeter wave communications, channels may change rapidly due to factors such as mobility and environmental changes, causing traditional channel estimation methods to fail. In scenarios with multiple paths, the reflection path information from fixed objects (such as ceilings and walls) is usually relatively stable, so it can be used as prior information to provide a stable reference framework for channel estimation. In a rapidly changing environment, prior information of known paths can help the system better track channel changes. By jointly estimating known and unknown paths, channel estimation errors can be reduced and overall system performance can be improved.

[0004] At present, there are many papers on CSI estimation of millimeter wave channels. In the paper by Z. Gao, C. Hu, L. Dai and Z. Wang (Channel Estimation for Millimeter-Wave Massive MIMO With Hybrid Precoding Over Frequency-Selective Fading Channels [J]. IEEE Commun. Lett., vol. 20, no. 6, pp. 1259-1262, June 2016.), a distributed compressed sensing method is proposed for millimeter wave systems running on frequency-selective wideband channels, taking advantage of the angular domain structure sparsity of millimeter wave wideband frequency-selective fading channels. In the paper (Low-Rank Tensor Decomposition-Aided Channel Estimation for Millimeter Wave MIMO-OFDM Systems [J]. IEEE J. Sel. Areas Commun. , vol. 35, no. 7, pp. 1524-1538, July 2017.) by Z. Zhou, J. Fang, L. Yang, H. Li, Z. Chen, and RS Blum, a channel parameter estimation method based on parallel factor (PARAFAC) decomposition was proposed by taking advantage of the sparse scattering characteristics of the millimeter wave channel. However, these studies did not fully consider the differences in stability between paths. They usually regarded all paths as unknown paths for estimation, and failed to effectively utilize the prior information of stable paths, resulting in unnecessary computational overhead and accuracy loss in the estimation process. Summary of the invention

[0005] Purpose of the invention: In view of the defects of the prior art, the present invention proposes a channel estimation enhancement method for a millimeter wave MIMO orthogonal frequency division multiplexing (OFDM) system to obtain a CSI with higher accuracy.

[0006] Technical solution: A millimeter wave MIMO-OFDM system channel estimation enhancement method described in the present invention includes:

[0007] For each subcarrier, the base station (BS) is equipped with different beamforming vectors in consecutive time frames, and the user (UE) detects the transmission signal independently in each time subframe;

[0008] Separate the known path information from the received signal to obtain a tensor slice of the joint prior and unknown information, and construct it into a third-order tensor based on the PARAFAC model;

[0009] The alternating least squares (ALS) algorithm is optimized using prior information to estimate three factor matrices, and then the unknown channel parameters are estimated using maximum likelihood (ML).

[0010] Further, for each subcarrier, the BS configures different beamforming vectors in consecutive time frames, and the UE independently detects the transmission signal in each time subframe, specifically including:

[0011] Assume that the BS is equipped with Antenna and RF chains, UE equipped Antenna and RF chains, both BS and UE use uniform linear arrays, and , Since channel estimation is performed independently for each UE in the downlink scenario, only one UE is considered and one RF chain is configured for each UE. The total number of OFDM subcarriers is set to , assuming that the previous subcarriers are used for training. To simplify the problem, the interference between adjacent cells caused by frequency reuse is ignored. Subcarrier The beamforming vector expression for a time frame is:

[0012]

[0013] in, For the The pilot symbol vector on subcarriers, is the digital precoding matrix, is a general RF precoding matrix. Considering the geometric broadband millimeter wave channel model, that is, there is scattering points, each scattering point corresponds to a path with a corresponding departure angle , Angle of Arrival and delay , and each path has a complex path gain , so the channel matrix in the delay domain can be written as:

[0014]

[0015] in, is the delta function, and are the steering vectors at UE and BS respectively, Indicates transpose.

[0016] For a uniform linear array, the steering vector is:

[0017]

[0018] in, is the angle of arrival or departure, is the number of antennas, is the signal wavelength, is the distance between adjacent antenna units. Assume that the departure angle, arrival angle and delay corresponding to different scattering points are different, that is, when hour, , , .

[0019] Based on the delay domain channel model, we can get The frequency domain channel matrix on the subcarriers for:

[0020]

[0021] in, is the sampling frequency.

[0022] Assume that Among the paths, there are The path of the transmission signal through fixed scattering points (such as buildings, etc.) to reach the UE, and its channel parameter information is known; the rest The path is the path where the transmission signal reaches the UE through unstable scattering points (such as vehicles, etc.). Its channel parameter information is unknown and channel estimation is required. Since the order of the paths does not affect the channel estimation, it is assumed that the paths 1 to 2 are The paths are known paths, then:

[0023]

[0024]

[0025]

[0026] in, is the known part of the channel matrix, is the index of the known path channel parameter, is the unknown part of the channel matrix, The channel parameter index of the unknown path.

[0027] In each time frame, the UE uses The received signal is detected by using a combination of RF vectors, which are common to all subcarriers. In each time subframe, the received signal is combined in the RF domain, and after removing the cyclic prefix, it is converted back to the frequency domain through discrete Fourier transform. Subcarrier The received signal of a time subframe can be expressed as:

[0028]

[0029] in Indicates The combination vector of time subframes, represents additive Gaussian noise.

[0030] Furthermore, the known path information is separated from the received signal to obtain a tensor slice of the prior and unknown information, which is constructed into a third-order tensor based on the PARAFAC model, including:

[0031] In each time frame The received signals are arranged to obtain:

[0032]

[0033] in , , ,and The items in , is uniformly distributed. and , the signal must be received from The unknown part of the estimated channel matrix Assume that the digital precoding matrix and pilot matrix on each subcarrier are consistent, that is, ,have , , which facilitates channel parameter estimation based on tensor decomposition with minimal measurement cost. , , , , then:

[0034]

[0035] The UE can receive signals from multiple subcarriers, so the received signal can be expressed as a third-order tensor Represents, and the first Item is .make , , , we can get the tensor Slices:

[0036]

[0037] Will By stacking the subcarriers, we get the third-order tensor form of the received signal:

[0038]

[0039] Among them, the symbol represents the outer product, , It is the third-order tensor form of Gaussian noise.

[0040] Furthermore, the ALS algorithm is optimized using prior information to estimate three factor matrices, and then ML is used to estimate the unknown channel parameters, including:

[0041] Due to the sparse scattering characteristics of the millimeter wave channel, the number of paths is usually small relative to the dimension of a tensor. Therefore, the received signal tensor has an inherent low-rank structure, which ensures that the scale ambiguity and permutation ambiguity of the received tensor PARAFAC decomposition are unique. Perform PARAFAC decomposition to obtain unknown channel parameters The estimated value of .

[0042] The factor matrices of the known part of the noise-free received signal in the three modes are:

[0043]

[0044]

[0045]

[0046] The factor matrices of the unknown part of the noise-free received signal in the three modes are:

[0047]

[0048]

[0049]

[0050] Assume the factor matrix to be estimated , and their corresponding estimated values ​​are set as .

[0051] In order to achieve PARAFAC decomposition of the unknown path part in the received signal tensor, the following problems must be solved:

[0052]

[0053] in, is the true value of the unknown path part of the received signal tensor.

[0054] The ALS algorithm is used to solve the above problem. Specifically, two of the factor matrices are fixed to minimize the fitting error of the third factor matrix, and the three factor matrices are alternately rotated to achieve effective optimization. The formula is as follows:

[0055]

[0056]

[0057]

[0058] in, , , They are Expanded form in three modes, corner mark Indicates the number of ALS algorithm iterations.

[0059] There is a fuzzy relationship between the estimated factor matrix and the true factor matrix:

[0060]

[0061]

[0062]

[0063] Among them, the unknown matrix is satisfied A non-singular diagonal matrix of Represents the unit matrix, matrix The meaning of is that the size of each column in the factor matrix estimate is a scale of its true value; the permutation matrix is also unknown, meaning that the order of the columns in the factor matrix estimate is different from the order of the columns in its true value; and , and is the estimation error of the three factor matrices.

[0064] Since the factor matrix Each column of Therefore, the angle of arrival can be estimated by ML. ,Right now:

[0065]

[0066] in, express No. Similar to the solution of the arrival angle, the departure angle and delay estimates are:

[0067]

[0068]

[0069] in, express No. List, No. List for .

[0070] Using the estimated and Regenerate the factor matrix and , and then use the estimated structure .because , and is based on the factor matrix estimate , and So , and and , and have the same column vector arrangement order. It is also known that the permutation matrix in the fuzzy relation is universal and can be ignored, so a non-singular diagonal matrix can be obtained and , combined with the conditions It can be calculated . Using the structure and factor matrix estimates , the channel gain can be estimated by inverse operation .

[0071] Finally, the unknown part of the channel matrix According to the estimated Recovered and combined with the known path prior information, the channel matrix can be obtained The estimated value of .

[0072] Beneficial effects: Compared with the prior art, the main advantages of the present invention are: the present invention can effectively utilize the prior information of the stable path, reduce the computational overhead and precision loss in the estimation process; the proposed method can enhance the CSI estimation and improve the system performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1A millimeter wave MIMO-OFDM system channel estimation enhancement method of the present invention;

[0074] Figure 2 A schematic diagram of the structure of a millimeter wave MIMO-OFDM system of the present invention;

[0075] Figure 3 The present invention is based on different known path numbers and the number of unknown paths Next, channel parameters Performance plots of mean square error (MSE) and corresponding Cramer-Rao bound (CRB) as signal-to-noise ratio varies;

[0076] Figure 4 The present invention is based on different known path numbers and the number of unknown paths Next, channel Performance graph of the normalized mean square error (NMSE) as the signal-to-noise ratio changes;

[0077] Figure 5 The present invention is based on different known path numbers and the number of unknown paths Next, channel parameters Performance diagram of the MSE and corresponding CRB as the number of subcarriers used for training changes;

[0078] Figure 6 The present invention is based on different known path numbers and the number of unknown paths Next, channel Performance plot of NMSE as the number of subcarriers used for training varies. DETAILED DESCRIPTION

[0079] In order to make the features and advantages of the present invention more obvious and understandable, the present invention is described in detail below with reference to the accompanying drawings.

[0080] Figure 2 Schematic diagram of the millimeter wave MIMO-OFDM system structure of the present invention, as shown in Figure 2 In the downlink communication system shown in the figure, both the BS and UE use uniform linear arrays, and the BS is equipped with Antenna, UE equipped Due to unfavorable propagation conditions, it is assumed that the direct path between the BS and the UE is highly attenuated. Therefore, for the proposed millimeter wave downlink communication system, the direct path is not taken into account, and the transmission signal is reflected by the BS through several scattering points and transmitted to the UE via known and unknown paths for reception.

[0081] Implementation Example 1

[0082] See also Figure 3 and Figure 4 These two figures show the present invention under different known path numbers. and the number of unknown paths Next, channel parameters MSE and corresponding CRB and channel Performance diagram of NMSE as the signal-to-noise ratio changes. The system parameters are: , , , It can be found that when the total number of paths remains unchanged, the number of known paths The more, the channel parameters MSE performance and channel The better the NMSE performance is. And the channel parameters The CRB decreases exponentially with the increase of SNR, except in the case of low SNR, where the channel parameter The MSE of is very close to the corresponding CRB. In addition, when the total number of paths is the same, as the signal-to-noise ratio increases, the number of known paths increases. The curves are always located on the known path number This also means that no matter which parameter configuration is adopted, the channel estimation accuracy is higher with more prior information. The present invention effectively utilizes the prior information to enhance the CSI estimation.

[0083] Implementation Example 2

[0084] See also Figure 5 and Figure 6 These two figures show the present invention under different known path numbers. and the number of unknown paths Next, channel parameters MSE and corresponding CRB and channel Performance diagram of NMSE of varies with the number of subcarriers used for training. The system parameters are: , , , , the signal-to-noise ratio is set to 15dB. It can be found that the number of known paths The more unknown paths there are, the The less, the channel parameter MSE performance and channel The better the NMSE performance is. Except for a small number of training subcarriers, the channel parameters The MSE of is very close to the corresponding CRB. In addition, when the total number of paths is the same, as the number of subcarriers used for training increases, the information that can be used for estimation increases, and the number of known paths increases. The curves are always located on the known path number The curve with less a priori information indicates that the channel estimation accuracy is higher with more a priori information. The present invention effectively utilizes the a priori information to enhance the CSI estimation.

[0085] In summary, the present invention considers a channel estimation enhancement method for a millimeter wave MIMO-OFDM system, in which a pilot is sent by the BS, which is received by the UE after being reflected by several scattering points, and the CSI estimation is enhanced by effectively utilizing the prior information with the help of the optimized ALS fitting algorithm.

[0086] The above embodiments are only for helping to understand the method and main idea of ​​the present invention. The contents of this specification cannot be used to limit the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the attached claims.

Claims

1. A millimeter wave MIMO-OFDM system channel estimation enhancement method, characterized in that The method includes: For each subcarrier, the base station (BS) is equipped with different beamforming vectors in consecutive time frames, and the user (UE) independently detects the transmission signal in each time subframe. Specifically, it is assumed that the BS is equipped with Antenna and RF chains, UE equipped Antenna and RF chains, both BS and UE use uniform linear arrays, and , ,Since the channel estimation is performed independently by each UE in the downlink scenario, only one UE is considered, and one RF chain is equipped for each UE. The total number of orthogonal frequency division multiplexing (OFDM) subcarriers is set to , assuming that the previous subcarriers are used for training, Subcarrier The beamforming vector expression for a time frame is ， in, For the The pilot symbol vector on subcarriers, is the digital precoding matrix, is a general RF precoding matrix, considering the geometric broadband millimeter wave channel model, that is, there is scattering points, each scattering point corresponds to a path with a corresponding departure angle , Angle of Arrival and delay , and each path has a complex path gain , so the channel matrix in the delay domain can be written as ， in, is the delta function, and are the steering vectors at UE and BS respectively, represents transposition. Assuming that different scattering points correspond to different departure angles, arrival angles, and delays, based on the delay domain channel model, we can get The frequency domain channel matrix on the subcarriers for ， in, is the sampling frequency, assuming that Among the paths, there are The path of the transmission signal through fixed scattering points (such as buildings, etc.) to reach the UE, and its channel parameter information is known. The path is the path where the transmission signal reaches the UE through unstable scattering points (such as vehicles, etc.). Its channel parameter information is unknown and channel estimation is required. Since the path order does not affect the channel estimation, it is assumed that the paths 1 to 2 are If a path is a known path, then ， ， ， in, is the known part of the channel matrix, is the unknown part of the channel matrix. In each time frame, the UE uses RF combination vectors are used to detect the transmitted signal, and these RF combination vectors are common to all subcarriers. Subcarrier The received signal of a time subframe can be expressed as ， in Indicates The combination vector of time subframes, represents additive Gaussian noise; Separate the known path information from the received signal, obtain the tensor slice of the prior and unknown information, and construct it into a third-order tensor based on the parallel factor (PARAFAC) model, including: The received signals are arranged to obtain ， in , , ,and The items in , is uniformly distributed, for and , the signal must be received from The unknown part of the estimated channel matrix is , assuming that the digital precoding matrix and pilot matrix on each subcarrier are consistent, that is, ,have , , which facilitates channel parameter estimation based on tensor decomposition with minimal measurement cost, and defines , , , , then ， The UE can receive signals from multiple subcarriers, so the received signal can be expressed as a third-order tensor Represents, and the first Item is ,make , , , we can get the tensor Slices ， Will By stacking the subcarriers, the third-order tensor form of the received signal is obtained ， Among them, the symbol represents the outer product, , is the third-order tensor form of Gaussian noise; The alternating least squares (ALS) algorithm is optimized using prior information to estimate three factor matrices, and then the maximum likelihood (ML) is used to estimate the unknown channel parameters, including: due to the sparse scattering characteristics of the millimeter wave channel, the number of paths The dimension of a tensor is usually small, so the received signal tensor It has an inherent low-rank structure, which ensures that the scale ambiguity and permutation ambiguity of the received tensor PARAFAC decomposition are unique. By performing PARAFAC decomposition, we can obtain the unknown channel parameters The estimated value of the noise-free received signal is the factor matrix of the known part in the three modes. ， ， ， The factor matrices of the unknown part of the noise-free received signal in the three modes are: ， ， ， Assume the factor matrix to be estimated , and their corresponding estimated values ​​are set as In order to realize the PARAFAC decomposition of the unknown path part in the received signal tensor, the problem must be solved ， in, In order to receive the true value of the unknown path part of the signal tensor, the ALS algorithm is used to solve the above problem. The formula is as follows ， ， ， in, , , They are Expanded form in three modes, corner mark Represents the number of ALS algorithm iterations. There is a fuzzy relationship between the estimated factor matrix and the real factor matrix as follows ， ， ， Among them, the unknown matrix is satisfied A non-singular diagonal matrix of Represents the unit matrix, permutation matrix is also unknown, and , and is the estimated error of the three factor matrices. Each column of Therefore, the angle of arrival can be estimated by ML. ,Right now ， in, express No. Column, similar to the solution of arrival angle, the departure angle and delay estimation values ​​are ， ， in, express No. List, No. List for , using the estimated and Regenerate the factor matrix and , and then use the estimated structure ,because , and is based on the factor matrix estimate , and So , and and , and Have the same column vector arrangement order, and the permutation matrix in the fuzzy relation is known is universal and can be ignored, so a non-singular diagonal matrix can be obtained and , combined with the conditions It can be calculated , using the constructed and factor matrix estimates , the channel gain can be estimated by inverse operation , and finally, the unknown part of the channel matrix According to the estimated Recovered and combined with the known path prior information, the channel matrix can be obtained The estimated value of .

Images