CN110611626B

CN110611626B - Channel estimation method, device and equipment

Info

Publication number: CN110611626B
Application number: CN201810617693.7A
Authority: CN
Inventors: 黄杰; 秦一平; 李言召
Original assignee: Shanghai Huawei Technologies Co Ltd
Current assignee: Shanghai Huawei Technologies Co Ltd
Priority date: 2018-06-15
Filing date: 2018-06-15
Publication date: 2022-07-29
Anticipated expiration: 2038-06-15
Also published as: CN110611626A

Abstract

The embodiment of the application discloses a channel estimation method, a device and equipment. The method comprises the following steps: generating a signal data matrix according to the parameters without noise reduction; performing Singular Value Decomposition (SVD) on the signal data matrix to obtain a signal intensity matrix; determining each eigenvalue in the signal strength matrix as a signal domain eigenvalue or a noise domain eigenvalue; generating a filter matrix according to the signal domain eigenvalue and the noise domain eigenvalue; and performing filtering and noise reduction on corresponding channel estimation parameters in a channel estimation matrix to be filtered by using each filter coefficient in the filter matrix to obtain a target channel estimation matrix. According to the technical scheme of the embodiment of the application, the accuracy of channel estimation can be improved, the performance of the channel estimation is improved, and the performance of a wireless communication system is further improved.

Description

Channel estimation method, device and equipment

Technical Field

The embodiment of the application relates to the technical field of communication, in particular to a channel estimation method, a channel estimation device and channel estimation equipment.

Background

The performance of a wireless communication system is mainly affected by a wireless channel, which has great randomness and is affected by various factors. Based on this, the receiving end, for example, a base station, in order to correctly demodulate the signal sent by the transmitting end and accurately predict the downlink channel, channel estimation becomes an essential operation process, wherein the channel estimation is an operation of estimating channel parameters according to the received data. It can be seen that channel estimation is a key technology of wireless communication, and the accuracy of channel estimation directly affects the performance of wireless communication.

Currently, the mainstream wireless communication systems include Long Term Evolution (LTE) and 5th-Generation (5G) systems, and LTE and 5G New air interface (NR) support Massive Multiple Input and Multiple Output (Massive MIMO) communication technologies. The Massive MIMO communication technology is a communication technology in which a base station transmits signals on the same time-frequency resource using multiple antennas, and User Equipment (UE) receives the signals on the multiple antennas. In view of the fact that antennas at the base station side have certain sparsity, in a Massive MIMO communication system, the base station performs channel estimation on uplink received signals.

The Massive MIMO communication system has higher carrier frequency than other wireless communication systems, so that the path loss is larger in the transmission process of uplink receiving signals, the energy of the uplink receiving signals reaching a base station is smaller, and the signal-to-noise ratio is lower. On the other hand, due to the array gain of the base station side antennas in the Massive MIMO communication system, the signal on each antenna contains more noise, and therefore, the signal-to-noise ratio of the received signal on each antenna is also low. However, the lower the snr of the uplink received signal is, the larger the error of the obtained channel estimation is, so that in a Massive MIMO communication system, the performance of the channel estimation is poor, and the accuracy is low, thereby causing the performance of the Massive MIMO communication system to be poor.

Disclosure of Invention

The embodiment of the application provides a channel estimation method, a device and equipment, which are used for solving the problem of poor channel estimation performance caused by low signal-to-noise ratio of signals in a Massive MIMO communication system.

In a first aspect, an embodiment of the present application provides a channel estimation method, which includes,

generating a signal data matrix according to the parameters without noise reduction;

performing Singular Value Decomposition (SVD) on the signal data matrix to obtain a signal intensity matrix;

determining each eigenvalue in the signal strength matrix as a signal domain eigenvalue or a noise domain eigenvalue, wherein the signal domain eigenvalue is an eigenvalue indicating that the signal strength is greater than the noise strength, and the noise domain eigenvalue is an eigenvalue indicating that the signal strength is less than the noise strength;

generating a filter matrix according to the signal domain eigenvalue and the noise domain eigenvalue;

and performing filtering and noise reduction on corresponding channel estimation parameters in a channel estimation matrix to be filtered by using each filter coefficient in the filter matrix to obtain a target channel estimation matrix, wherein the channel estimation parameters in the channel estimation matrix to be filtered and the filter coefficients in the filter matrix are parameters of the same spatial domain.

The technical scheme of the embodiment of the application aims at performing noise reduction on signals on each antenna, so that parameters which are not subjected to noise reduction on each antenna are used as original data, distribution characteristics of noise under a Massive MIMO scene are combined, data carrying noise are decomposed from each dimension such as an incident direction and signal intensity of each incident direction through Singular Value Decomposition (SVD), parameters such as the incident direction, the signal intensity and the noise intensity of each incident direction can be quantized, corresponding relations are established, and then noise reduction is performed based on information according to quantization.

By adopting the implementation mode, the signal data matrix generated by the parameters without noise reduction is utilized to obtain the signal intensity and the noise intensity in each incidence direction, and then the filter coefficient is generated corresponding to each incidence direction, and the channel estimation parameters in the corresponding incidence directions are filtered and subjected to noise reduction, so that the accuracy of channel estimation can be improved, the performance of channel estimation is improved, and the performance of a wireless communication system is improved.

In an alternative design, the determining each eigenvalue in the signal strength matrix as a signal domain eigenvalue or a noise domain eigenvalue includes:

Sequentially calculating differences between every two n eigenvalues in the signal intensity matrix according to the sequence from large to small to obtain n-1 differences, wherein n is an integer larger than 1;

dividing the weighted average of the first difference value to the kth difference value by the weighted average of the kth difference value to the (n-1) th difference value in sequence to obtain k change rates, wherein k is greater than 1 and less than or equal to n-x, and x is an integer less than n and greater than or equal to 1;

determining a k value corresponding to the maximum change rate;

and determining k eigenvalues in the signal intensity matrix as the signal domain eigenvalues and determining the (k + 1) th to the nth eigenvalues as the noise domain eigenvalues in sequence from big to small.

Since the noise is not directional, the intensity of the noise is substantially the same for each incident direction. The signal transmission has a certain directivity, so the signal energy in each characteristic direction is different, and further, in the signal intensity matrix, a larger characteristic value can be understood as a larger signal intensity, and a smaller characteristic value can be understood as a smaller signal intensity, even no signal intensity.

Based on this, with the implementation, the eigenvalues in the signal intensity matrix can be classified, and then the different classes of eigenvalues are used as reference data for generating the filter matrix, thereby providing parameter basis for generating the filter matrix.

reading the maximum eigenvalue in the signal intensity matrix;

determining the value of the preset proportion of the maximum characteristic value as a critical value;

and determining the characteristic value which is greater than or equal to the critical value in the signal intensity matrix as the signal domain characteristic value, and determining the characteristic value which is smaller than the critical value in the signal intensity matrix as the noise domain characteristic value.

In an alternative design, the generating a filter matrix according to the signal domain eigenvalue and the noise domain eigenvalue includes:

calculating the corresponding signal power of the corresponding signal domain characteristic value according to each signal domain characteristic value;

calculating the average noise power according to all the noise domain characteristic values;

and calculating a filter coefficient corresponding to the corresponding signal power according to each signal power and the noise average power to obtain the filter matrix.

In an alternative design, the filter matrix P satisfies:

wherein,

refers to the signal power, i-1 … … k, and σ refers to the noise average power.

In the embodiment of the present application, each signal domain feature value includes a certain amount of noise, and a part or all of the noise domain feature values also include a small amount of signals. Based on this, adopt this implementation, make relatively accurate estimation with noise and signal to in improving the filter matrix, the accuracy nature of each filter coefficient, and then, reach better noise reduction effect.

In an alternative design, the performing filtering denoising on the corresponding channel estimation parameters in the channel estimation matrix to be filtered by using each filter coefficient in the filter matrix includes:

generating a filtering weight matrix W according to the filtering matrix P, wherein the filtering weight matrix W satisfies the following conditions:

wherein, U _k Is an incidence direction matrix corresponding to k signal domain eigenvalues,

refers to the matrix U _k The conjugate transpose of (1);

filtering the channel estimation parameters in the channel estimation matrix H to be filtered by using each weight in the filtering weight matrix W to obtain the target channel estimation matrix

The target channel estimation matrix

Satisfies the following conditions:

by adopting the implementation mode, the part with almost no signal distribution is removed, the energy value of the noise is estimated according to the noise domain characteristic value, the filter coefficient corresponding to the corresponding incident direction is calculated according to the noise energy value and the signal energy value of each incident direction, and the parameters corresponding to the incident direction with larger signal intensity can also be filtered respectively, so that the noise reduction effect is better.

In an alternative design, the generating a signal data matrix according to the non-noise reduction parameters includes:

acquiring an antenna domain signal parameter matrix or a beam domain signal parameter matrix or a channel estimation matrix without noise reduction;

performing autocorrelation on the obtained matrix to obtain an autocorrelation matrix;

and calculating the average value of each parameter in the autocorrelation matrix to obtain the signal data matrix.

Since the signal data matrix mainly needs to reflect information such as signal values, antennas corresponding to signals, and REs corresponding to signals, and the non-noise-reduction parameter matrix contains more information, in order to simplify data information, the non-noise-reduction parameter matrix may be further processed to obtain the signal data matrix.

Based on this, with the implementation mode, not only can unnecessary information in the parameter matrix without noise reduction be eliminated, but also the dimension of the parameter matrix without noise reduction can be reduced, so that the calculation amount can be reduced on the basis of keeping the necessary parameter information.

In an alternative design, the generating a signal data matrix according to the non-noise reduction parameter includes:

and acquiring an estimation matrix of the channel without noise reduction as the signal data matrix.

By adopting the implementation mode, the accuracy of the generated filter coefficient can be ensured to the maximum extent, so that the accuracy of the filter coefficient can be improved to the maximum extent, and the noise reduction effect is optimized.

In an alternative embodiment, the signal power is

Satisfies the following conditions:

λ _i referring to the signal domain characteristic value, the noise average power σ satisfies:

wherein, sigma _n-k Refers to a matrix formed by the eigenvalues of the noise domain.

In an alternative embodiment, the signal power is

Satisfies the following conditions:

Deleting the noise domain eigenvalue from the signal intensity matrix to obtain a matrix formed by the signal domain eigenvalue;

and determining a matrix formed by the signal domain characteristic values as the filtering matrix.

After the eigenvalues in the signal strength matrix are summarized into the signal domain eigenvalue and the noise domain eigenvalue, the signal domain eigenvalue can be considered to contain all signal energy, and the noise domain eigenvalue can be similar to not carrying signals. Therefore, the implementation mode is simple in operation and small in calculation amount.

In an optional design, the target channel estimation matrix

Satisfies the following conditions:

wherein λ is _i Is the signal domain characteristic value, i 1 … … k, U _k Is an incidence direction matrix corresponding to k signal domain eigenvalues,

the method refers to the conjugate transpose of a sampling point matrix corresponding to the k signal domain eigenvalues.

By adopting the implementation mode, the part which has almost no signal distribution can be completely discarded, and only the part with stronger signal intensity is reserved, so that most signals are reserved, and most noises are filtered.

In an optional design, before performing filtering denoising on a corresponding channel estimation parameter in a channel estimation matrix to be filtered by using each filter coefficient in the filter matrix, the method further includes:

Receiving an antenna domain signal to obtain an antenna domain signal matrix;

performing channel estimation on each signal parameter in the antenna domain signal matrix to obtain the channel estimation matrix to be filtered, which contains corresponding channel estimation parameters; or,

receiving an antenna domain signal to obtain an antenna domain signal matrix;

performing Discrete Fourier Transform (DFT) on each signal parameter in the antenna domain signal matrix to obtain the beam domain signal matrix;

and performing channel estimation on each signal parameter in the beam domain signal to obtain the channel estimation matrix to be filtered, which contains the corresponding channel estimation parameter.

In a second aspect, an embodiment of the present application further provides a channel estimation apparatus, which includes a module configured to perform the method steps in the implementations of the first aspect and the first aspect.

In a third aspect, an embodiment of the present application provides a channel estimation device, which includes a transceiver, a processor, and a memory. The transceiver, the processor and the memory can be connected through a bus system. The memory is for storing a program, instructions or code, and the processor is for executing the program, instructions or code in the memory to perform the method of the first aspect, or any one of the possible designs of the first aspect.

In a fourth aspect, the present application provides a computer-readable storage medium having stored therein instructions which, when run on a computer, cause the computer to perform the method of the first aspect or any possible design of the first aspect.

In order to solve the problem of poor channel estimation performance, the embodiment of the application adds noise reduction processing on the basis of the existing channel estimation processing process. Specifically, a spatial distribution matrix of signal intensity is obtained by performing Singular Value Decomposition (SVD) on a data matrix of the non-noise-reduced signal. Because the signal has a certain directivity and the distribution of the noise on the antenna array is relatively uniform, based on this, the eigenvalue in the signal intensity matrix is respectively summarized into a signal domain eigenvalue and a noise domain eigenvalue, and then a filter matrix is generated according to the signal domain eigenvalue and the noise domain eigenvalue, and the filter coefficient in the filter matrix is adopted to perform filtering and noise reduction on the channel estimation parameter to be denoised. Therefore, according to the technical scheme of the embodiment of the application, the distribution characteristic of noise under a Massive MIMO scene is combined, the filtering matrix is generated based on the signal matrix without noise reduction, and then filtering noise reduction is performed on the channel estimation parameters to be subjected to noise reduction, so that the accuracy of channel estimation can be improved, the performance of channel estimation is improved, and the performance of a wireless communication system is improved.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly described below, and it is obvious for those skilled in the art to obtain other drawings without creative efforts.

FIG. 1 is a schematic diagram of a scenario for Massive MIMO;

fig. 2 is a method flow diagram of a prior art channel estimation method;

FIG. 3 is a flow chart of a second embodiment of a prior art channel estimation method;

FIG. 4 is a flow chart of a third embodiment of a prior art channel estimation method;

FIG. 5 is a diagram illustrating the channel energy distribution in the ideal beam domain;

FIG. 6 is a diagram illustrating the channel energy distribution in the beam domain under noise conditions;

fig. 7 is a flowchart of a method of a channel estimation method according to an embodiment of the present application;

fig. 8 is a flowchart of a first example of a channel estimation method provided in an embodiment of the present application;

fig. 9 is a flowchart of a second example of a channel estimation method provided in an embodiment of the present application;

fig. 10 is a flowchart of a third example of a channel estimation method provided in an embodiment of the present application;

fig. 11 is a schematic structural diagram of a channel estimation apparatus according to an embodiment of the present application;

Fig. 12 is a schematic structural diagram of a channel estimation device according to an embodiment of the present application.

Detailed Description

Referring to fig. 1, fig. 1 is a schematic view of a scenario of Massive MIMO, wherein fig. 1 shows a base station including at least one antenna array consisting of p transmit antennas (pT) and p receive antennas (pR). The UE under the coverage of the base station comprises q receiving antennas and q transmitting antennas. Wherein both p and q are greater than 1. Based on this, the MIMO technology is a wireless communication technology in which a transmitting end transmits signals to a plurality of receiving antennas of a receiving end through a plurality of transmitting antennas, and in a Massive MIMO scenario, p is generally greater than or equal to 16.

Based on the scene diagram shown in fig. 1, the number of transmitting antennas in the coverage area of the base station is usually much smaller than the number of receiving antennas at the base station side, so that the uplink receiving channel at the base station side has a certain sparsity in the spatial domain and has a condition for performing channel estimation, and therefore, in a Massive MIMO system, the base station performs channel estimation based on the uplink receiving signal. Generally, the method for the base station to perform channel estimation includes the following steps:

referring to fig. 2, fig. 2 is a flowchart of a method of a conventional channel estimation method, and method 01 includes the following steps:

Step S10, an uplink signal is received.

And step S11, executing channel estimation according to the uplink signal to obtain channel estimation parameters.

On the one hand, compared with a non-Massive MIMO system, the Massive MIMO system has a higher carrier frequency, which causes a larger path loss of the Massive MIMO system, so that the energy of the uplink received signal is smaller, the noise is larger, and the signal-to-noise ratio of the uplink received signal is lower. On the other hand, due to the gain of the base station antenna array, the signal on each antenna contains more noise, which also results in a lower signal-to-noise ratio of the received signal on each antenna. Based on this, the uplink signal received by the base station contains relatively more noise, and in the method 01, the base station directly performs channel estimation based on the uplink received signal, and the obtained channel estimation parameter has poor accuracy, thereby causing poor channel estimation performance.

In order to solve this problem and improve the performance of channel estimation, method 02 is proposed on the basis of method 01. Referring to fig. 3, fig. 3 is a method flowchart of a second embodiment of a conventional channel estimation method, and a method 02 shown in fig. 3 includes the following steps:

in step S20, an antenna domain signal is received.

Step S21, performing Discrete Fourier Transform (DFT) on the antenna domain signal to obtain a beam domain signal.

And step S22, performing channel estimation according to the beam domain signal to obtain channel estimation parameters.

After the uplink received signal is transmitted to the base station, the signal energy is uniformly distributed on each antenna in the antenna array of the base station, so the uplink received signal is also referred to as an antenna domain signal. After the antenna domain signal is subjected to DFT, signal energy is concentrated on a small number of beams, and therefore, for the sake of convenience of distinguishing from the antenna domain signal, the signal after DFT is referred to as a beam domain signal.

Since the signal-to-noise ratio of the signal is low when the signal energy is uniformly distributed on each antenna, the accuracy of the channel estimation parameter obtained by the method 01 is poor. Based on this, the method 02 is adopted, the signal energy which is uniformly distributed is converged on a small number of beams through DFT, and then channel estimation is performed, so that the signal-to-noise ratio of the signal can be improved, and further, the performance of channel estimation is optimized.

Further, since the signal energy of the beam domain signal is focused on a small number of beams, the signal-to-noise ratio of the beams focused with signal energy only is relatively high, and based on this, an alternative embodiment of the method 02 is described below. Referring to fig. 4, fig. 4 is a flowchart of a third embodiment of a conventional channel estimation method, and a method 021 shown in fig. 4 includes the following steps:

In step S20, an antenna domain signal is received.

Step S22, performing channel estimation according to the beam domain signal to obtain a channel estimation parameter corresponding to each beam.

In step S23, a Reference Signal Receiving Power (RSRP) of each beam is calculated.

And step S24, sequentially selecting the channel estimation parameters corresponding to a plurality of beams as final channel estimation parameters according to the RSRP from large to small.

Wherein, the larger RSRP indicates the larger signal energy on the corresponding beam, therefore, the beams are selected in the order from the larger RSRP to the smaller RSRP, and the beam with the larger signal energy can be determined. It should be understood that the number of selected beams can be flexibly set according to the RSRP size, for example, 64 beams in an alternative embodiment, and 16 beams can be selected according to the RSRP size. And method 02 may be understood as selecting all beams.

In addition, when the uplink received Signal is a Sounding Reference Signal (SRS), each channel may also be accompanied by a demodulation Reference Signal (DMRS). Based on this, in this embodiment, channel estimation is performed on the SRS and the DMRS respectively according to the method 02, and further, since the channels corresponding to the SRS and the DMRS are the same, the RSRP of each beam corresponding to the SRS or the DMRS is determined, and then, according to step S24 in the method 021, the channel estimation parameter of the corresponding beam is selected as the final channel estimation parameter.

Based on this, when performing DFT, it is necessary to determine the direction of a beam and the direction of a channel, and each beam is usually directed to a fixed direction, while one channel may include incident paths in different directions, and the base station cannot know the incident paths of the channels, so when performing DFT, it is usually default that the incident path direction of the channel coincides with the beam direction. Based on this feature, after performing DFT, the signal energy that should be converged on one beam is spread to the adjacent beams of the corresponding beam. In an ideal environment without noise, the resulting signal energy distribution is shown in fig. 5, while in an actual noisy environment, the resulting signal energy distribution is shown in fig. 6.

With reference to fig. 6, a beam with large signal energy is extracted according to the method 021, although noise on other beams can be filtered, a part of noise still exists on the extracted beam, and a part of signal also exists on the filtered beam, so that the extracted beam is affected by the part of noise, and a part of signal is lost, and the channel estimation performance is still not good. Channel estimation is performed according to method 02, which preserves the entire signal, but also preserves much noise, making the resulting channel estimation less accurate. In view of this, the technical solution of the embodiment of the present application is provided.

Embodiments of the present application will be described below with reference to the accompanying drawings.

Referring to fig. 7, fig. 7 is a flowchart of a method of channel estimation provided in an embodiment of the present application, and the method 100 shown in fig. 7 adds noise reduction processing on the basis of an existing channel estimation processing procedure, so as to improve accuracy of channel estimation and performance of channel estimation. The method 100 comprises the steps of:

and step S101, generating a signal data matrix according to the non-noise reduction parameters.

Each row of data of the signal data matrix is signal data of one antenna, that is, one incident direction, and each column of data is signal data on one Resource Element (RE). The non-noise reduction parameters include antenna domain signal parameters, beam domain signal parameters, and non-noise reduction channel estimation parameters.

And if the signal parameters for executing the channel estimation are received for the first time, taking the antenna domain signal parameters, the beam domain signal parameters or the non-noise reduction channel estimation parameters in the current data transmission period as the non-noise reduction parameters. If the signal parameters for performing channel estimation are received for the second time and subsequently, the antenna domain signal parameters, the beam domain signal parameters, or the non-noise reduction channel estimation parameters received in the current data transmission period may be used as the non-noise reduction parameters. In order to avoid using the parameters in the current data transmission period to cause delay, the antenna domain signal parameters, the beam domain signal parameters, or the non-noise reduction channel estimation parameters received in the previous data transmission period may also be used as non-noise reduction parameters.

It should be noted that, in the embodiment of the present application, signal data is distributed on each antenna and each RE, and therefore, for convenience of expressing and processing each parameter, the parameter of each processing procedure is represented in a matrix form, and based on this, each type of parameter is described as a matrix of related parameters in the following embodiment of the present application.

Furthermore, since the signal data matrix mainly needs to represent information such as signal values, antennas corresponding to signals, and REs corresponding to signals, and the non-noise-reduction parameter matrix contains more information, in order to simplify data information, the non-noise-reduction parameter matrix may be further processed to obtain the signal data matrix.

Specifically, because the antenna domain signal parameter matrix and the beam domain signal parameter matrix both include modulated pilot sequences, and the dimensions of the antenna domain signal parameter matrix and the beam domain signal parameter matrix are both large, when the non-noise-reduction parameter matrix is the antenna domain signal parameter matrix or the beam domain signal parameter matrix, on one hand, in order to cancel the pilot sequences, and on the other hand, in order to reduce the dimensions, the non-noise-reduction parameter matrix is subjected to autocorrelation to obtain an autocorrelation matrix, and then, an average value of each parameter in the autocorrelation matrix is calculated to obtain the signal data matrix.

Since the pilot sequence in the signal parameter matrix is already cancelled when the channel estimation operation is performed on the signal parameter matrix, when the non-noise-reduced parameter matrix is the non-noise-reduced channel estimation matrix, the non-noise-reduced channel estimation matrix may be directly used as the signal data matrix, or the non-noise-reduced channel estimation matrix may be subjected to autocorrelation and averaging to obtain the signal data matrix, which is not limited in the embodiment of the present application.

And step S102, executing SVD on the signal data matrix to obtain a signal intensity matrix.

In combination with the description of the signal data matrix in step S101, based on the physical meaning of SVD, after performing SVD on the signal data matrix, the obtained left singular matrix is the incident direction matrix of the signal, and the middle singular matrix is the signal intensity matrix, that is, each feature value in the middle singular matrix represents the intensity of the signal in one incident direction.

Specifically, the signal strength matrix Σ satisfies:

the signal intensity matrix Σ is a matrix of m rows and n columns, each of the n columns corresponds to an incident direction, and the eigenvalue λ in each column represents a signal intensity value in the incident direction.

Step S103, determining each eigenvalue in the signal intensity matrix as a signal domain eigenvalue or a noise domain eigenvalue.

The signal domain characteristic value is a characteristic value indicating that the signal intensity is greater than the noise intensity, and the noise domain characteristic value is a characteristic value indicating that the signal intensity is less than the noise intensity.

Note that since noise is not directional, after SVD, the energy distributed in each characteristic direction is equivalent, that is, the noise intensity corresponding to each incident direction in the signal intensity matrix is substantially the same. Because the signal transmission has a certain directivity, the signal energy in each characteristic direction is different, and further, in the signal intensity matrix, a larger characteristic value can be understood as a larger signal intensity, and a smaller characteristic value can be understood as a smaller signal intensity, even no signal intensity. Based on this, the embodiment of the application can estimate the noise intensity and the signal intensity in each incident direction, and further can perform filtering and denoising on the channel estimation to be filtered according to the estimated noise intensity and the signal intensity in each incident direction. Specifically, the embodiments of the present application provide two ways to determine the signal domain eigenvalue and the noise domain eigenvalue.

The first method is as follows: according to the embodiment of the application, the signal domain characteristic value and the noise domain characteristic value critical characteristic value can be determined according to the turning point of the characteristic value change degree from large to small. Since the difference can directly reflect the change between the two eigenvalues, in an optional embodiment of the present application, the difference between every two of the n eigenvalues can be sequentially calculated in the descending order to obtain n-1 differences. Then, the weighted average value of the first difference value to the kth difference value is divided by the weighted average value of the kth difference value to the (n-1) th difference value in sequence to obtain k change rates, further, k values corresponding to the maximum change rates are determined, k eigenvalues in the signal intensity matrix are determined as signal domain eigenvalues in sequence from large to small, and the (k + 1) th eigenvalue to the nth eigenvalue are determined as noise domain eigenvalues. Wherein n is an integer greater than 1, k is greater than 1 and less than or equal to n-x, and x is an integer less than n and greater than or equal to 1.

Namely, k satisfies:

wherein k is more than 1 and less than or equal to n-x.

The change degree of the characteristic value is smaller in the incident direction in which the signal energy is not distributed or the distributed signal energy is small, so that the closer the k value is to n-1 when calculating the change rate, the larger the obtained change rate is, and the signal domain characteristic value and the noise domain characteristic value are determined according to the k obtained at the moment, so that the signal intensity represented by the signal domain characteristic value is inaccurate. Based on this, k is set to be greater than 1 and smaller than or equal to n-x, and the value of x can be flexibly set according to the size and distribution of the characteristic value, which is not limited in the embodiment of the present application.

The second method comprises the following steps: since a larger eigenvalue indicates a larger signal intensity and a smaller eigenvalue indicates a smaller signal intensity, based on this, after determining the maximum eigenvalue, a value of a preset ratio of the maximum eigenvalue may be used as a threshold value, for example, a value corresponding to 10% of the maximum eigenvalue may be used as a threshold value. And determining the characteristic value which is greater than or equal to the critical value in the signal intensity matrix as a signal domain characteristic value, and determining the characteristic value which is smaller than the critical value in the signal intensity matrix as a noise domain characteristic value.

It should be understood that the above two manners are optional embodiments provided in the embodiments of the present application, and the embodiments of the present application may further include other determining manners based on the attributes of the signal domain characteristic value and the noise domain characteristic value, and the distribution characteristics of the signal and the noise, and are not described in detail herein.

And step S104, generating a filter matrix according to the signal domain eigenvalue and the noise domain eigenvalue.

The eigenvalues in the filter matrix are all filter coefficients, and the filter coefficients correspond to the eigenvalues in the signal domain one to one.

After the eigenvalues in the signal strength matrix are summarized into the signal domain eigenvalue and the noise domain eigenvalue, the signal domain eigenvalue may be considered to contain all signal energy, and the noise domain eigenvalue may be similar to a signal not carried, so the noise domain eigenvalue may be deleted from the signal strength matrix to obtain a matrix formed by the signal domain eigenvalue, and the matrix formed by the signal domain eigenvalue is determined as a filter matrix in step S103. In this embodiment, each signal domain feature value may be considered as a filter coefficient.

Specifically, based on the SVD characteristics, the eigenvalues in the signal strength matrix are arranged from large to small, i.e., λ in Σ ₁ ≥λ ₂ …≥λ _m . Therefore, after determining the signal domain eigenvalues and the noise domain eigenvalues, the signal strength matrixΣ can be expressed as:

wherein, sigma _k Is a signal domain eigenvalue matrix, sigma _n-k Is referred to as a noise domain eigenvalue matrix. Then, in the present embodiment, the filter matrix is Σ _k Of a filter matrix sigma _k Characteristic value λ of ₁ To lambda _k I.e. the filter coefficients.

The filtering matrix is a key for performing filtering and noise reduction, and the accuracy of the filtering coefficient in the filtering matrix directly influences the noise reduction effect. Based on this, in this embodiment, the signal domain eigenvalue matrix is directly used as a filter matrix, and in order to ensure the noise reduction effect, the calculation process of data should be reduced as much as possible, so that in this embodiment, the channel estimation matrix without noise reduction should be directly used as a signal data matrix, so as to improve the accuracy of the filter coefficient to the greatest extent.

It will be appreciated that in fact each signal domain feature value contains a certain amount of noise, and that some or all of the noise domain feature values likewise contain a small amount of signal. Based on this, in order to achieve a better noise reduction effect, in another embodiment, the base station may calculate a signal power corresponding to a corresponding signal domain eigenvalue according to each signal domain eigenvalue, calculate a noise average power according to all the noise domain eigenvalues, and further calculate a filter coefficient corresponding to the corresponding signal power according to each signal power and the noise average power to obtain a filter matrix.

Specifically, in this embodiment, the filter matrix P satisfies:

Wherein,

i is 1 … … k, and σ is the noise average power.

Based on the description of step S101, it can be seen that the signal data matrix is generated in two ways in the embodiment of the present application, and the calculation process of the non-noise reduction parameter matrix is performed in the two waysIn contrast, the eigenvalue meaning of the signal data matrix obtained in one generation mode is different from the eigenvalue meaning of the signal data matrix obtained in another generation mode, and further, the eigenvalue meaning is different in the matrices obtained in two different signal data matrices SVD. While filtering the signal power in the matrix P

And the noise average power sigma is obtained by calculation according to the eigenvalue in each matrix after SVD, and the signal power is obtained by two modes of correspondingly generating a signal data matrix based on the eigenvalue

And the noise power σ is expressed as follows:

if the signal data matrix is obtained after the autocorrelation and the average of the non-noise reduction parameter matrix, the signal power

Satisfies the following conditions:

λ _i referring to the signal domain characteristic value, i is 1 … … k, and the noise average power σ satisfies:

wherein, sigma _n-k Is a matrix formed by noise domain eigenvalues.

If the signal data matrix is an un-denoised channel estimation matrix, the signal power

Satisfies the following conditions:

the noise average power σ satisfies:

Step S105, using each filter coefficient in the filter matrix to perform filtering and noise reduction on the corresponding channel estimation parameter in the channel estimation matrix to be filtered, so as to obtain a target channel estimation matrix.

The channel estimation matrix to be filtered is obtained by performing channel estimation on the antenna domain signal parameters received in the current data transmission period, or is obtained by performing channel estimation on the antenna domain signal parameters received in the current data transmission period after the antenna domain signal parameters are subjected to DFT conversion to obtain beam domain signal parameters. Specifically, before step S105, the channel estimation matrix to be filtered is obtained by referring to method 01, method 02 or method 021, which is not described herein again in this embodiment of the present application.

Based on the description of step S104, the filter matrix obtained in the embodiments of the present application includes two forms, and corresponding to the filter matrices of the two forms, the embodiments of the present application respectively provide a filtering method, and respectively obtain a target channel estimation matrix of one form.

In particular, when the filter matrix is sigma _k Time of day, sigma _k ＝diag(λ ₁ …λ _k ) Meridian sigma _k Target channel estimation matrix obtained by filtering and denoising

Satisfies the following conditions:

wherein, U _k Refers to an incidence direction matrix corresponding to k signal domain eigenvalues,

The method refers to the conjugate transpose of sampling point matrixes corresponding to k signal domain eigenvalues.

In the embodiment, the filtering and noise reduction is performed on the channel estimation matrix to be filtered in this way, all parts with almost no signal distribution are discarded, and only the part with strong signal strength is reserved, so that most signals are reserved, and most noise is filtered.

When the filter matrix is P, the embodiment of the present application generates a filter weight matrix W according to the filter matrix P, where the filter weight matrix W satisfies:

then, each weight in the filtering weight matrix W is used for filtering the channel estimation parameters in the channel estimation matrix H to be filtered, and the target channel estimation matrix is obtained

Target channel estimation matrix

Satisfies the following conditions:

wherein,

refers to the matrix U _k The conjugate transpose of (c).

It should be understood that DFT is the mapping of signal parameters of one spatial domain to another spatial domain, and thus, antenna domain signal parameters and beam domain signal parameters are parameters of two spatial domains, respectively. Based on this, the channel estimation parameters obtained by the antenna domain signal parameters executing the channel estimation belong to the antenna domain, and the channel estimation parameters obtained by the beam domain signal parameters executing the channel estimation belong to the beam domain. Since the parameters of different spatial domains are different from the antenna, the beam, the incident direction, and the like, the parameters of one spatial domain cannot participate in the calculation process of the parameters of another spatial domain, and therefore, in the embodiment of the present application, the spatial domain corresponding to the filter matrix is the same as the spatial domain corresponding to the channel estimation matrix to be denoised.

As can be seen from

methods

01 and 02, the spatial domain corresponding to the channel estimation matrix to be filtered depends on the spatial domain of the signal parameters for which channel estimation is performed, while the spatial domain corresponding to the filtering matrix depends on the spatial domain corresponding to the non-noise-reduced parameter matrix. Based on this, if the spatial domain of the filtering matrix is the same as that of the channel estimation matrix to be filtered, it is necessary to ensure that the spatial domain of the parameter matrix without noise reduction is the same as that of the signal parameter matrix for generating the channel estimation matrix to be filtered.

For example, the non-noise reduction parameter matrix is the antenna domain signal parameter matrix received in the last data transmission period, and the obtained filtering matrix also belongs to the antenna domain, so the channel estimation matrix for performing filtering and noise reduction using the filtering matrix is generated from the antenna domain signal parameter matrix received in the current data transmission period. For another example, the non-noise reduction parameter matrix is a non-noise reduction channel estimation matrix generated by the beam domain signal parameter matrix in the previous data transmission period, and since the non-noise reduction channel estimation matrix belongs to the beam domain, the obtained filter matrix also belongs to the beam domain, and then the channel estimation matrix for performing filtering and noise reduction using the filter matrix is obtained by performing channel estimation after DFT on the antenna domain signal parameter matrix received in the current data transmission period. When the non-noise reduction parameter matrix is another matrix, the correspondence relationship refers to the above description, and the embodiment of the present application is not described in detail here.

The technical solution of the embodiment of the present application aims to perform noise reduction on signals on each antenna, and what the signals on each antenna are, and the implementation of the embodiment of the present application is not limited. Based on this, if the signal received by each antenna includes the SRS and the DMRS, the filter matrix generated according to the SRS-related parameters may also be used to reduce noise of the DMRS channel estimation parameters, and similarly, the filter matrix generated according to the DMRS-related parameters may also be used to reduce noise of the SRS channel estimation parameters.

Therefore, in the embodiment of the application, the signal data matrix generated by the parameters without noise reduction is used for obtaining the signal intensity and the noise intensity in each incidence direction, and then the filter coefficient is generated corresponding to each incidence direction, and the channel estimation parameters in the corresponding incidence directions are filtered and subjected to noise reduction, so that the accuracy of channel estimation can be improved, the performance of channel estimation is improved, and the performance of a wireless communication system is improved.

The embodiments of the present application will be described below with reference to specific examples.

In an optional embodiment of the present application, the non-noise reduction parameter matrix is an antenna domain signal parameter matrix Y received in a previous data transmission period, each row of data of the antenna domain signal parameter matrix Y is data of one antenna, and each column of data is data of one time domain sampling point. Referring to fig. 8, the method 200 shown in fig. 8 is an implementation method of the present embodiment, and includes the following steps:

step S201, performing autocorrelation and averaging on the antenna domain signal parameter matrix Y to obtain a signal data matrix a.

In this embodiment, Y may be directly read from the buffer memory.

Autocorrelation array of antenna domain signal parameter matrix Y

Satisfies the following conditions:

the signal data matrix A satisfies:

where n refers to the total number of antennas.

In step S202, SVD is performed on the signal data matrix a.

Executing SVD to obtain A ═ U ∑ U ^H Where the left singular matrix U is the eigenvector of the spatial signal incident direction, each row of U represents one incident direction, and the middle singular matrix Σ is the eigenvector of the signal intensity of the spatial signal in each incident direction, where, for example

Is a moment of m rows and n columnsArray, each column in n rows corresponds to an incident direction, the eigenvalue lambda in each column represents the signal intensity value in the incident direction, each row in m rows corresponds to an antenna, and a right singular matrix U ^H Is the conjugate transpose of the left singular matrix U.

Step S203, dividing the eigenvalue λ in the intermediate singular matrix into a signal domain eigenvalue and a noise domain eigenvalue.

Calculating a value of k, said k being indicative of the number of λ that are signal domain characteristic values. Wherein,

wherein k is more than 1 and less than or equal to n/3. Further, the eigenvalue λ in the intermediate singular matrix Σ is set ₁ To lambda _k As the signal domain eigenvalue, the eigenvalue λ is set _k+1 To lambda _n As a signal domain characteristic value. In the present step, the first step is carried out,

wherein, sigma _k Is a signal domain eigenvalue matrix, is an eigenvalue lambda ₁ To lambda _k Diagonal matrix of (e), sigma _n-k Is a noise domain eigenvalue matrix, is an eigenvalue lambda _k+1 To lambda _n Diagonal matrix of (2).

For example, in this embodiment, if n is 40 and k is 16, λ in the intermediate singular matrix Σ is set ₁ To lambda ₁₆ As a signal domain eigenvalue, will ₁₇ To lambda ₄₀ As a noise domain characteristic value, obtaining

Wherein,

step S204, generating a filtering weight matrix.

According to the signal domain eigenvalue matrix sigma _k Of each eigenvalue λ _i I 1 … … k, calculating the corresponding incident sideUpward signal power

In the present embodiment, the first and second electrodes are,

according to the noise domain eigenvalue matrix sigma _n-k The noise domain feature value in (c), the noise average power sigma is calculated,

then, according to the signal power

And the noise mean power sigma generates a filter matrix P,

further, a filter weight matrix W is generated,

wherein, U _k The method refers to an incidence direction matrix corresponding to k signal domain eigenvalues.

In step S205, filtering and denoising are performed on the channel estimation matrix to be filtered by using the filtering weight matrix.

It should be understood that, in this embodiment, the parameters of the non-noise reduction parameter matrix belong to the antenna domain, and therefore, the obtained filtering weight matrix W also belongs to the antenna domain. Based on this, if the antenna domain signal is received in the current data transmission period and the channel estimation is directly performed on the antenna domain signal parameter matrix, the obtained channel estimation matrix H to be filtered is the channel estimation matrix H to be filtered, the filtering weight matrix W of the embodiment may be used to perform filtering and noise reduction to obtain the filtering weight matrix W

Is a target channel estimation matrix of an embodiment of the present application.

It should be understood that the method 200 is only an exemplary method description of the present application, and the method 200 is equally applicable if the non-noise reduction parameter matrix is a beam domain signal parameter matrix obtained in a previous data transmission period or a non-noise reduction channel estimation matrix, or an antenna domain signal parameter matrix, a beam domain signal parameter matrix, or even a channel estimation matrix to be filtered in a current data transmission period.

Of course, when the non-noise-reduction parameter matrix is another parameter matrix, some of the calculated parameters need to be adaptively adjusted, for example, when the non-noise-reduction parameter matrix is a non-noise-reduction channel estimation matrix, the matrix for performing the autocorrelation in step S201 is

And if the channel estimation matrix without noise reduction belongs to the beam domain, the channel estimation matrix H to be filtered is obtained by performing channel estimation by the beam domain signal parameter matrix. The embodiments of the present application will not be described in detail herein.

The channel estimation method described in this embodiment not only removes a portion where almost no signal is distributed, but also can perform filtering respectively for the parameter of the incident direction in which the signal intensity is large, so that the noise reduction effect is better.

In a second optional embodiment of the present application, the non-noise reduction parameter matrix is a non-noise reduction channel estimation matrix h of a previous data transmission period, the non-noise reduction channel estimation matrix h is obtained according to an antenna domain signal matrix, and is a matrix of m × l, where m rows respectively correspond to m antennas, and l is a data stream number. Referring to fig. 9, the method 300 shown in fig. 9 is an implementation method of the present embodiment, and includes the following steps:

Step S301, obtaining the non-denoised channel estimation matrix h on each RE, and arranging the channel estimation matrices according to the columns to obtain a signal data matrix A.

Wherein A ═ h ₀ h ₁ … h _n ]。

In step S302, SVD is performed on the signal data matrix a.

Performing SVD to obtain A ═ U ∑ V ^H The left singular matrix U and the middle singular matrix Σ are as described in the method 200, and are not described herein again. Right singular matrix V ^H Is the conjugate transpose of the feature vector of the signal sample point in the time domain.

Step S303, divide the eigenvalue λ in the intermediate singular matrix Σ into a signal domain eigenvalue and a noise domain eigenvalue.

Step S304, generating a filtering weight matrix W.

The execution process of step S303 is the same as step S203, and the details refer to the description of step S203. The step S304 is performed in the same manner as the step S204, however, in the present embodiment, the signal power is

Satisfies the following conditions:

the noise average power σ satisfies:

in step S305, filtering and denoising is performed on the channel estimation matrix H to be filtered by using the filtering weight matrix W.

The step is performed in the same way as step S205, and the details are described in step S205 and will not be described in detail here.

It should be understood that the method 300 is equally applicable to the channel estimation matrix H to be filtered, i.e., the non-denoised channel estimation matrix for the current data transmission period. In addition, the parameters of the channel estimation matrix without noise reduction in this embodiment may be in an antenna domain or a beam domain. When in the beam domain, the right singular matrix V ^H The conjugate transpose of the eigenvector of the signal sampling point in the frequency domain is obtained, and the obtained filtering weight matrix W is suitable for filtering and denoising the channel estimation matrix to be filtered in the beam domain.

In a third optional embodiment of the present application, the non-noise reduction parameter matrix is a channel estimation matrix H to be filtered, that is, a non-noise reduction channel estimation matrix of a current data transmission period. The channel estimation matrix H to be filtered is a matrix of m × l, where m rows correspond to m antennas respectively, and l is the number of data streams. Referring to fig. 10, the method 400 shown in fig. 10 is an implementation method of the present embodiment, and includes the following steps:

step S401, obtaining a channel estimation matrix H to be filtered on each RE, and arranging the channel estimation matrices H according to columns to obtain a signal data matrix A.

In step S402, SVD is performed on the signal data matrix a.

Step S403, dividing the eigenvalue λ in the intermediate singular matrix Σ into a signal domain eigenvalue and a noise domain eigenvalue.

The execution processes of step S401, step S402 and step S403 refer to the descriptions of step S301, step S302 and step S303, and are not described in detail here.

And S404, taking the signal domain characteristic value matrix as a filtering matrix, and recovering a channel estimation matrix H to be filtered.

Specifically, in this embodiment, the target channel estimation matrix

It can be seen that the method 400 essentially selects the portion with the larger signal strength as the target channel estimation parameter, but does not perform any noise reduction processing on the portion with the larger signal strength. Based on this, in order to ensure the accuracy of the target channel estimation parameters and better noise reduction effect, when the method 400 is used to perform channel estimation, the non-noise reduction channel estimation matrix of the current data transmission period is selected as the signal data matrix.

It should be understood that the

methods

200, 300 and 400 are only exemplary descriptions for supporting the present solution, and the technical solution of the embodiments of the present application is not limited. As can be known to those skilled in the art, with the evolution of SVD and the appearance of new service scenarios, the technical solution provided in the embodiment of the present application is also applicable to similar technical problems.

In summary, the channel estimation method according to the embodiment of the present application combines the noise distribution characteristics in a Massive MIMO scene, generates a filter matrix based on a signal matrix without noise reduction, and then performs filtering noise reduction on a channel estimation parameter to be noise reduced, so that the accuracy of channel estimation can be improved, the performance of channel estimation can be improved, and the performance of a wireless communication system can be improved.

Fig. 11 is a schematic structural diagram of a channel estimation apparatus 1100 according to an embodiment of the present application. The channel estimation apparatus 1100 may be configured to perform the methods corresponding to fig. 7 to fig. 10. As shown in fig. 11, the channel estimation apparatus 1100 includes a generation module 1101, a calculation module 1102, a determination module 1103, and a noise reduction module 1104. The generating module 1101, the calculating module 1102 and the determining module 1103 can be specifically configured to perform the processing of the various parameters described in the method 100, the method 200, the method 300 and the method 400; the denoising module 1104 is specifically configured to perform filtering denoising of parameters in the channel estimation matrix to be denoised in the

methods

100, 200, 300 and 400.

For example, the generating module 1101 may be configured to generate a signal data matrix according to the non-noise reduction parameters, and may be further configured to generate a filter matrix according to the signal domain eigenvalue and the noise domain eigenvalue. The calculation module 1102 may be configured to perform Singular Value Decomposition (SVD) on the signal data matrix to obtain a signal strength matrix. The determining module 1103 may be configured to determine each eigenvalue in the signal strength matrix as a signal domain eigenvalue or a noise domain eigenvalue, where the signal domain eigenvalue is an eigenvalue indicating that the signal strength is greater than the noise strength, and the noise domain eigenvalue is an eigenvalue indicating that the signal strength is less than the noise strength. The denoising module 1104 may be configured to perform filtering denoising on corresponding channel estimation parameters in a channel estimation matrix to be filtered by using each filter coefficient in the filter matrix, to obtain a target channel estimation matrix, where the channel estimation parameters in the channel estimation matrix to be filtered and the filter coefficients in the filter matrix are parameters of the same spatial domain.

For specific content, reference may be made to descriptions of relevant parts in the

methods

100, 200, 300, and 400, which are not described herein again.

It should be understood that the above division of the modules is only a division of logical functions, and the actual implementation may be wholly or partially integrated into one physical entity or may be physically separated. In this embodiment, the generating module 1101, the calculating module 1102, the determining module 1103 and the noise reducing module 1104 may be implemented by a processor. As shown in fig. 12, the channel estimation device 1200 may include a processor 1201, a transceiver 1202, and a memory 1203. The memory 1203 may be used to store a program/code preinstalled in the channel estimation device 1200 at the time of factory shipment, or may store a code or the like used when the processor 1201 executes.

It is to be understood that the channel estimation apparatus 1200 according to the embodiment of the present application may correspond to a base station in the

methods

100, 200, 300 and 400 according to the embodiment of the present application, wherein the transceiver 1202 is configured to perform transceiving of various parameters, e.g., receiving of non-noise reduction parameters, described in the

methods

100, 200, 300 and 400, and the processor 1201 is configured to perform processing of various parameters in the base station described in the

methods

100, 200, 300 and 400. And will not be described in detail herein.

In a specific implementation, corresponding to the channel estimation device 1200, an embodiment of the present application further provides a computer storage medium, where the computer storage medium disposed in the base station may store a program, and when the program is executed, part or all of the steps of the channel estimation methods provided in fig. 7 to 10 may be implemented. The storage medium in the base station may be a magnetic disk, an optical disk, a read-only memory (ROM) or a Random Access Memory (RAM).

In the embodiment of the present application, the transceiver may be a wired transceiver, a wireless transceiver, or a combination thereof. The wired transceiver may be, for example, an ethernet interface. The ethernet interface may be an optical interface, an electrical interface, or a combination thereof. The wireless transceiver may be, for example, a wireless local area network transceiver, a cellular network transceiver, or a combination thereof. The processor may be a Central Processing Unit (CPU), a Network Processor (NP), or a combination of a CPU and an NP. The processor may further include a hardware chip. The hardware chip may be an application-specific integrated circuit (ASIC), a Programmable Logic Device (PLD), or a combination thereof. The PLD may be a Complex Programmable Logic Device (CPLD), a field-programmable gate array (FPGA), a General Array Logic (GAL), or any combination thereof. The memory may include a volatile memory (RAM), such as a random-access memory (RAM); the memory may also include a non-volatile memory (ROM), such as a read-only memory (ROM), a flash memory (flash memory), a hard disk (HDD) or a solid-state drive (SSD); the memory may also comprise a combination of memories of the kind described above.

Also included in fig. 12 is a bus interface, which may include any number of interconnected buses and bridges, with various circuits of one or more processors, represented by a processor, and memory, represented by a memory, linked together. The bus interface may also link together various other circuits such as peripherals, voltage regulators, power management circuits, and the like, which are well known in the art, and therefore, will not be described any further herein. The bus interface provides an interface. The transceiver provides a means for communicating with various other apparatus over a transmission medium. The processor is responsible for managing the bus architecture and the usual processing, and the memory may store data used by the processor in performing operations.

Those of skill in the art will further appreciate that the various illustrative logical blocks and steps (step) set forth in the embodiments of the present application may be implemented in electronic hardware, computer software, or combinations of both. Whether such functionality is implemented as hardware or software depends upon the particular application and design requirements of the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the embodiments of the present application.

The various illustrative logical units and circuits described in this application may be implemented or operated upon by design of a general purpose processor, a digital signal processor, an Application Specific Integrated Circuit (ASIC), a Field Programmable Gate Array (FPGA) or other programmable logic device, discrete gate or transistor logic, discrete hardware components, or any combination thereof. A general-purpose processor may be a microprocessor, but in the alternative, the processor may be any conventional processor, controller, microcontroller, or state machine. A processor may also be implemented as a combination of computing devices, e.g., a digital signal processor and a microprocessor, a plurality of microprocessors, one or more microprocessors in conjunction with a digital signal processor core, or any other similar configuration.

The steps of a method or algorithm described in the embodiments herein may be embodied directly in hardware, in a software element executed by a processor, or in a combination of the two. The software cells may be stored in RAM memory, flash memory, ROM memory, EPROM memory, EEPROM memory, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. For example, a storage medium may be coupled to the processor such the processor can read information from, and write information to, the storage medium. In the alternative, the storage medium may be integral to the processor. The processor and the storage medium may reside in an ASIC, which may be disposed in a UE. In the alternative, the processor and the storage medium may reside in different components in the UE.

It should be understood that, in the various embodiments of the present application, the size of the serial number of each process does not mean the execution sequence, and the execution sequence of each process should be determined by its function and inherent logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the application to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, from one website site, computer, server, or data center to another website site, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.

All parts of the specification are described in a progressive mode, the same and similar parts of all embodiments can be referred to each other, and each embodiment is mainly introduced to be different from other embodiments. In particular, for the apparatus and device embodiments, since they are substantially similar to the method embodiments, the description is relatively simple and reference may be made to the description of the method embodiments in relevant places.

While the preferred embodiments of the present application have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all alterations and modifications as fall within the scope of the application.

It will be apparent to those skilled in the art that various changes and modifications may be made in the present application without departing from the spirit and scope of the application. Thus, if such modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is intended to include such modifications and variations as well.

Claims

1. A method of channel estimation, the method comprising:

Generating a signal data matrix according to the non-noise reduction parameters, wherein each row of data of the signal data matrix is signal data in one incident direction;

performing Singular Value Decomposition (SVD) on the signal data matrix to obtain a signal intensity matrix, wherein each eigenvalue in the signal intensity matrix represents the intensity of a signal in an incident direction;

determining each eigenvalue in the signal strength matrix as a signal domain eigenvalue or a noise domain eigenvalue, wherein the signal domain eigenvalue is an eigenvalue indicating that the signal strength in one incident direction is greater than the noise strength, and the noise domain eigenvalue is an eigenvalue indicating that the signal strength in one incident direction is less than the noise strength;

generating a filter matrix according to the signal domain eigenvalue and the noise domain eigenvalue, wherein the eigenvalues in the filter matrix are all filter coefficients, and the filter coefficients are in one-to-one correspondence with the signal domain eigenvalue;

2. The channel estimation method of claim 1, wherein the determining each eigenvalue in the signal strength matrix as a signal domain eigenvalue or a noise domain eigenvalue comprises:

determining a k value corresponding to the maximum change rate;

3. The channel estimation method of claim 1, wherein the determining each eigenvalue in the signal strength matrix as a signal domain eigenvalue or a noise domain eigenvalue comprises:

reading the maximum eigenvalue in the signal strength matrix;

And determining the characteristic value which is greater than or equal to the critical value in the signal strength matrix as the signal domain characteristic value, and determining the characteristic value which is less than the critical value in the signal strength matrix as the noise domain characteristic value.

4. The channel estimation method of any of claims 1 to 3, wherein the generating a filter matrix from the signal domain eigenvalues and the noise domain eigenvalues comprises:

5. The channel estimation method of claim 4, wherein the filtering matrix P satisfies:

wherein,

6. The channel estimation method of claim 5, wherein the performing filtering noise reduction on the corresponding channel estimation parameters in the channel estimation matrix to be filtered using each filter coefficient in the filter matrix comprises:

refers to the matrix U _k The conjugation transpose of (1);

The target channel estimation matrix

Satisfies the following conditions:

7. the channel estimation method of claim 1, wherein the generating a signal data matrix from the non-noise reduction parameters comprises:

8. The channel estimation method of claim 1, wherein the generating a signal data matrix from the non-noise reduction parameters comprises:

and acquiring an un-denoised channel estimation matrix as the signal data matrix.

9. The channel estimation method of claim 5 or 7, characterized in that the signal power

Satisfies the following conditions:

10. The channel estimation method of claim 5 or 8, characterized in that the signal power

Satisfies the following conditions:

11. The channel estimation method of any of claims 1 to 3, 8, wherein the generating a filter matrix according to the signal domain eigenvalue and the noise domain eigenvalue comprises:

12. The channel estimation method of claim 11, wherein the target channel estimation matrix

Satisfies the following conditions:

wherein λ is _i Means the signal domain characteristic value, i is 1 … … k, U _k Is an incident direction matrix V corresponding to k eigenvalues of the signal domain _k ^H The method refers to the conjugate transpose of a sampling point matrix corresponding to the k signal domain eigenvalues.

13. The channel estimation method of claim 1, wherein before performing filtering denoising on the corresponding channel estimation parameters in the channel estimation matrix to be filtered using each filter coefficient in the filter matrix, further comprising:

receiving an antenna domain signal to obtain an antenna domain signal matrix;

performing Discrete Fourier Transform (DFT) on each signal parameter in the antenna domain signal matrix to obtain a beam domain signal matrix;

14. A channel estimation device, characterized in that it comprises means for performing the channel estimation method of any one of claims 1 to 13.

15. A channel estimation device comprising a processor and a memory, wherein:

the memory to store program instructions;

the processor, configured to invoke and execute program instructions stored in the memory to cause the channel estimation device to perform the channel estimation method of any of claims 1 to 13.

16. A computer-readable storage medium comprising instructions that, when executed on a computer, cause the computer to perform the channel estimation method of any of claims 1 to 13.