CN108370283A

CN108370283A - A kind of channel statistical information acquisition methods and receiver

Info

Publication number: CN108370283A
Application number: CN201580085395.2A
Authority: CN
Inventors: 王悦
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2015-12-18
Filing date: 2015-12-18
Publication date: 2018-08-03
Also published as: WO2017101097A1

Abstract

The embodiment of the invention discloses a kind of channel statistical information acquisition methods and receiver, method may include：Receiver receives the reception pilot signal for the pilot signal for including transmitter transmission in F time slot, wherein the F is the integer more than or equal to 1；The receiver receives the covariance matrix that pilot signal calculates reception signal of the receiver on the channel for transmitting the pilot signal using receiving F；The receiver carries out the signal reconstruction based on compression covariance perception using the covariance matrix, to obtain the reconstruction signal of the sparse statistical information of the channel；The receiver obtains the final statistical information of the channel using the reconstruction signal.The embodiment of the present invention can improve the accuracy of channel statistical information.

Description

Channel statistical information acquisition method and receiver

Technical Field

The present invention relates to the field of communications, and in particular, to a channel statistics information obtaining method and a receiver.

Background

With the development of communication technology, a scheme for optimizing a precoding design by using channel statistical information has been proposed in the industry, and the scheme can optimize the precoding design by using the channel statistical information to achieve a maximum transmission rate or achieve a minimum symbol error rate. However, at present, the final channel statistical information is obtained by performing channel estimation and then obtaining channel statistical information. However, the accuracy of the channel statistics estimation in this scheme may not be high.

Disclosure of Invention

The embodiment of the invention provides a channel statistical information acquisition method and a receiver, which can improve the accuracy of channel statistical information.

In a first aspect, an embodiment of the present invention provides a method for acquiring channel statistics information, including:

a receiver receives a receiving pilot signal comprising a pilot signal sent by a transmitter in F time slots, wherein F is an integer greater than or equal to 1;

the receiver calculates a covariance matrix of a received signal by using the received F received pilot signals, wherein the received signal refers to the covariance matrix of the received signal of the receiver on a channel for transmitting the pilot signals;

the receiver uses the covariance matrix to carry out signal reconstruction based on compressed covariance perception so as to obtain a reconstructed signal of sparse statistical information of the channel;

the receiver uses the reconstructed signal to obtain the final statistical information of the channel.

In the technical scheme, the statistical information of the signals is obtained by using the received pilot signals including the pilot signals sent by the transmitter, so that compared with the prior art of obtaining the channel statistical information, the accuracy of the channel statistical information can be improved.

In a first possible implementation manner of the first aspect, the calculating manner that the receiver calculates the covariance matrix of the received signal by using the received F received pilot signals may include:

the receiver performs column vectorization on the received F received pilot signals to obtain column vectorized vectors of the F received pilot signals, and calculates a covariance matrix of the received signals using the column vectorized vectors of the F received pilot signals.

In this embodiment, since the covariance matrix of the received signal is calculated using the column-vectorized vector, and the column-vectorized vector is less complicated in calculation than the matrix calculation, the calculation complexity can be reduced in calculating the covariance matrix of the received signal.

With reference to the first aspect or the first possible implementation manner of the first aspect, in a second possible implementation manner of the first aspect, the performing, by a receiver, signal reconstruction based on compressed covariance perception by using the covariance matrix to obtain a reconstructed signal of sparse statistical information of the channel may include:

the receiver takes a column vectorization vector of the covariance matrix, an observation matrix and L as inputs of an Orthogonal Matching Pursuit (OMP) algorithm, and executes the OMP algorithm to obtain a reconstructed signal of sparse statistical information of the channel, wherein the observation matrix is a matrix acquired in advance for signal reconstruction, and the L is sparsity.

With reference to the second possible implementation manner of the first aspect, in a third possible implementation manner of the first aspect, the tracking an OMP algorithm with the column vectorization vector of the covariance matrix, the observation matrix, and L as inputs of the orthogonal matching and executing the OMP algorithm to obtain a reconstructed signal of sparse statistical information of the channel may include:

the receiver calculates an inner product of each diagonal column vector of the observation matrix and a residual vector of the i minus 1 iteration, and takes an index of the diagonal column vector with the maximum absolute value of the inner product as an index of the current iteration, wherein the diagonal column vector is a column vector corresponding to an index identifier of a diagonal element in a matrix corresponding to the reconstruction signal in the observation matrix, i is a label of the current iteration, and the residual vector of the i minus 1 iteration is a residual of a column vectorization vector of the covariance matrix after i minus 1 iteration;

the receiver judges whether the i is 1, if so, the current iteration index is updated to a set to be updated, if not, the current iteration index and an extrapolation element are updated to the set to be updated in the index of the reconstruction signal, wherein when the current iteration index is larger than a historical iteration index, the extrapolation element comprises an element of which the index of the reconstruction signal is equal to the difference value of the current iteration index minus a row index difference and also comprises an element of which the index of the reconstruction signal is equal to the sum value of the historical iteration index plus the row index difference, when the current iteration index is smaller than the historical iteration index, the extrapolation element comprises an element of which the index of the reconstruction signal is equal to the sum value of the current iteration index plus the row index difference and also comprises an element of which the index of the reconstruction signal is equal to the difference value of the historical iteration index minus the row index difference, the row index difference is an absolute value of a difference between a row vector index corresponding to the current iteration index in a matrix corresponding to the reconstruction signal and a row vector index corresponding to the historical iteration index in a matrix corresponding to the reconstruction signal, and the historical iteration index is an iteration index of any iteration before the current iteration;

the receiver carries out reconstruction vector estimation based on least square on a partial matrix formed by column vectors in the observation matrix corresponding to indexes in the set to be updated and a column vectorization vector of the covariance matrix so as to obtain an ith reconstruction signal;

when the i is equal to the L, the receiver determines the reconstructed signal of the ith time as a reconstructed signal of sparse statistical information of the channel;

and when the i is smaller than the L, the receiver takes the result of subtracting the product of the partial matrix and the ith reconstruction signal from the column vectorization vector of the covariance matrix as the residual vector of the ith iteration, adds 1 to the i, and triggers the step of calculating the inner product of each diagonal column vector of the observation matrix and the residual vector of the ith minus 1 iteration.

In this implementation, through the above steps, only the elements on the diagonal of the matrix corresponding to the reconstructed signal may be searched and the extrapolated elements may be calculated when the OMP algorithm is executed, so that, compared with the conventional OMP algorithm, in this embodiment, the number of iterations is much smaller than that in the conventional OMP algorithm, so as to reduce the amount of computation.

With reference to any one of the foregoing possible implementation manners of the first aspect, in a fourth possible implementation manner of the first aspect, the obtaining, by the receiver, final statistical information of the channel using the reconstructed signal may include:

and the receiver acquires the final statistical information corresponding to the reconstructed signal as the final statistical information of the channel according to the relationship information between the pre-acquired sparse statistical information and the final statistical information.

In a second aspect, an embodiment of the present invention provides a receiver configured to implement the functions of the foregoing method, and implemented by hardware/software, where the hardware/software includes units corresponding to the foregoing functions.

In a third aspect, an embodiment of the present invention provides a receiver, including: the system comprises a processor, a network interface, a memory and a communication bus, wherein the communication bus is used for realizing the connection and communication among the processor, the network interface and the memory, and the processor executes a program stored in the memory for realizing the method.

In addition, in the embodiment of the present invention, the receiver may directly use the reconstructed signal of the sparse statistical information as the final statistical information of the channel. This allows to quickly obtain statistical information of the signal.

In addition, embodiments of the present invention also provide a computer readable storage medium storing one or more programs, the one or more programs including instructions, which when executed by a receiver including a screen and a plurality of application programs, cause the receiver to perform the method according to any one of the implementation manners provided by the first aspect.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.

Fig. 1 is a system architecture diagram applicable to a channel statistic information obtaining method provided by an embodiment of the present invention;

fig. 2 is a schematic flowchart of a channel statistic information obtaining method according to an embodiment of the present invention;

fig. 3 is a schematic flow chart of another channel statistic information obtaining method according to an embodiment of the present invention;

fig. 4 is a schematic flow chart of signal reconstruction according to an embodiment of the present invention;

fig. 5 to 8 are schematic views of effects provided by the embodiment of the present invention;

fig. 9 is a schematic structural diagram of a receiver according to an embodiment of the present invention;

fig. 10 is a schematic structural diagram of another receiver provided in the embodiment of the present invention;

fig. 11 is a schematic structural diagram of another receiver according to an embodiment of the present invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

Referring to fig. 1, fig. 1 is a system architecture diagram applicable to a channel statistic information obtaining method according to an embodiment of the present invention, as shown in fig. 1, including a transmitter 11 and a receiver 12, where: the transmitter 12 may be understood as a device that transmits signals in a communication system, for example: a base station, an Access Point (AP), or the like may communicate with a user equipment, or the transmitter 11 may also be a device that transmits signals in other communication scenarios, such as a device that transmits signals in Machine to Machine (M2M) communication. The receiver 12 may be understood as a device in a communication system that receives signals, for example: a user device or a device receiving signals in M2M communication, wherein the user device may include a cell phone, a tablet, a computer, a wearable device or a vehicle-mounted device, etc.

In addition, the system to which the above system architecture is applicable may include a millimeter wave communication system or a Multiple-Input Multiple-Output (MIMO) communication system, or a communication system that is applicable to a combination of millimeter wave communication technology and MIMO technology.

In the above system architecture, the transmitter 11 transmits pilot signals to the receiver 12 in F time slots, so that the receiver 12 receives F received pilot signals including pilot signals in the F time slots, so that the receiver 12 can use the received F received pilot signals to calculate a covariance matrix of received signals of the receiver on a channel transmitting the pilot signals, and use the covariance matrix to perform signal reconstruction based on compressed covariance perception to obtain a reconstructed signal of sparse statistical information of the channel, so that the reconstructed signal can be used to obtain final statistical information of the channel.

In addition, in the embodiment of the present invention, the statistical information of the channel may include a covariance matrix of the channel, and the covariance matrix of the channel may be referred to as a channel covariance matrix.

Referring to fig. 2, fig. 2 is a schematic flow chart of a channel statistic information obtaining method according to an embodiment of the present invention, as shown in fig. 2, including the following steps:

201. a receiver receives a received pilot signal including a pilot signal transmitted by a transmitter in F slots, wherein F is an integer greater than or equal to 1.

In this embodiment, the transmitter may transmit the pilot signal to the receiver in F allocated time slots, where the transmission may be to transmit one pilot signal in each of the F time slots, so that step 201 may receive F received pilot signals. It should be noted that, when receiving, the receiver needs to pass through a wireless channel and a noise signal exists in the channel, so that the receiver receives the received pilot signal not equal to the pilot signal. For example: the pilot signal is a compressed training sequence, which may be a random signal generated according to a certain random distribution, and the random signal may be subjected to a random distribution such as bernoulli and gaussian. Wherein, the pilot signal of the compressed training sequence can be represented by the following formula:

wherein X is the above-mentioned pilot signal, U_tIs N_t×N_tDiscrete Fourier Transform (DFT) matrix of (d) ()^H()^-1Respectively representing the operations of conjugate transpose on matrix and inversion on matrix, Z being N_tThe random signal matrix of the x T can be distributed randomly according to Bernoulli, Gaussian and the like, N_tIs the number of transmitting antennas, and T represents the length of time resource occupied by the pilot signal in a training duration, or can be understood as the duration of the pilot signal.

Thus, the received pilot signal received by the receiver can be represented by the following formula:

Y＝HX+W

wherein Y is N_tX T received pilot signal, X being N_tPilot signal of x T, H being N_r×N_tChannel matrix of, N_rIs the number of receiving antennas, W is N_rAdditive noise of x T.

202. The receiver uses the received F received pilot signals to calculate a covariance matrix of the received signals of the receiver on the channel on which the pilot signals are transmitted.

After all the F received pilot signals are received, the F received pilot signals can be used to calculate the covariance matrix of the received signal of the receiver on the channel on which the pilot signal is transmitted. The received signal here may be a received signal understood as any signal transmitted in the transmitter received in the above-described channel by the receiver.

Similarly, the received signal can be represented by the following formula:

Y＝HX+W

where Y is the received signal, X is the signal transmitted by the transmitter, and H is N_r×N_tChannel matrix of, N_rIs the number of receiving antennas, W is N_rAdditive noise of x T.

203. And the receiver uses the covariance matrix to carry out signal reconstruction based on compressed covariance perception so as to obtain a reconstructed signal of sparse statistical information of the channel.

Compressed covariance sensing is a signal reconstruction technique for reconstructing covariance information of an observation object in a non-compressed domain according to the covariance information of the observation object in a compressed domain.

204. The receiver uses the reconstructed signal to obtain the final statistical information of the channel.

After the reconstructed signal is obtained, the final statistical information of the channel can be obtained by using the reconstructed signal. For example: the relationship between the sparse statistical information and the final statistical information of the channel is preset, and then the final statistical information can be obtained through the relationship. Of course, in some scenarios, the reconstructed signal of the sparse statistical information may be directly used as the final statistical information of the channel.

In this embodiment, a receiver receives a received pilot signal including a pilot signal transmitted by a transmitter in F slots, where F is an integer greater than or equal to 1; the receiver using the received F received pilot signals to calculate a covariance matrix of the received signals of the receiver on the channel on which the pilot signals are transmitted; the receiver uses the covariance matrix to perform signal reconstruction based on compressed covariance perception so as to obtain a reconstructed signal of sparse statistical information of the channel; the receiver uses the reconstructed signal to obtain the final statistical information of the channel. Since the statistical information of the signal is obtained by using the received pilot signal including the pilot signal transmitted by the transmitter, the accuracy of the channel statistical information can be improved compared with the prior art of obtaining the channel statistical information.

Referring to fig. 3, fig. 3 is a schematic flow chart of another channel statistics information obtaining method according to an embodiment of the present invention, as shown in fig. 3, including the following steps:

301. a receiver receives a received pilot signal including a pilot signal transmitted by a transmitter in F slots, wherein F is an integer greater than or equal to 1.

302. And performing column vectorization on the received F received pilot signals to obtain column vectorization vectors of the F received pilot signals, and calculating a covariance matrix of the received signals of the receiver on a channel for transmitting the pilot signals by using the column vectorization vectors of the F received pilot signals.

In this embodiment, step 302 may be to obtain the F column vectorization vectors, then calculate the product of each column vectorization vector of the F column vectorization vectors and the conjugate transpose vector thereof, and then use the average of the obtained F products as the covariance matrix.

Alternatively, in this embodiment, step 302 may calculate the covariance matrix by the following disclosure:

wherein R is_yThe covariance matrix is represented, f represents the f-th observation time slot, vec () represents the matrix is subjected to the column vectorization, Y is the received signal vector after the column vectorization, that is, Y ═ vec (Y), and Y represents the received pilot signal.

Alternatively, in this embodiment, in some scenarios, one column-vectored vector may be directly selected from the column-vectored vectors of the F received pilot signals as the column-vectored vector of the covariance matrix to obtain the covariance matrix.

In step 302, since the covariance matrix of the received signal is calculated using the column-vectorized vector, and the column-vectorized vector is less complicated in calculation than the matrix, in this embodiment, the calculation complexity can be reduced in calculating the covariance matrix of the received signal.

303. And the receiver uses the covariance matrix to carry out signal reconstruction based on compressed covariance perception so as to obtain a reconstructed signal of sparse statistical information of the channel.

In this embodiment, step 303 may include:

and the receiver takes the column vectorization vector of the covariance matrix, an observation matrix and L as the input of an Orthogonal Matching Pursuit (OMP) algorithm, and executes the OMP algorithm to obtain a reconstructed signal of sparse statistical information of the channel, wherein the observation matrix is a matrix which is acquired in advance and used for signal reconstruction, and L is sparsity.

There are three input parameters in the OMP algorithm, namely, the observation matrix, the acquisition vector and the sparsity, and when these three input parameters are input into the OMP algorithm, a reconstructed signal can be obtained, where the OMP algorithm is a characteristic of the existing algorithm and will not be described in detail here. Then, in this embodiment, the column vectorization vector of the covariance matrix, the observation matrix, and L are used as the inputs of the OMP algorithm, so as to obtain the reconstructed signal of the sparse statistical information of the channel.

In addition, the observation matrix may be composed of Z and U_rWherein Z is N_tRandom signal matrix of x T, N_tIs the number of transmitting antennas, U_rIs N_r×N_rDFT matrix of, N_rIs the number of receiving antennas. For example: the above observation matrix can be expressed by the following disclosure:

wherein Φ is the observation matrix, and represents a Kronecker product operation ()^*Indicating a conjugate operation.

It should be noted that, in this embodiment, a conventional OMP algorithm may be used to obtain a reconstructed signal of the sparse statistical information of the above channels. However, considering that the conventional OMP algorithm searches the matrix by searching the elements of each position in the matrix, the amount of calculation is large. While in some scenarios the statistical information of the channel has sparse characteristics, such as: in a propagation environment of a system combining millimeter wave technology and MIMO technology, joint sparsity exists in statistical information of channels. In this example embodiment, the reconstructed signal of the sparse statistical information of the channel may be obtained through the following steps:

a) and the receiver calculates the inner product of each diagonal column vector of the observation matrix and the residual vector of the i minus 1 iteration, and takes the index of the diagonal column vector with the maximum absolute value of the inner product as the index of the current iteration, wherein the diagonal column vector is the column vector corresponding to the index identifier of the diagonal element in the matrix corresponding to the reconstruction signal in the observation matrix, i is the index of the current iteration, and the residual vector of the i minus 1 iteration is the residue of the column vectorization vector of the covariance matrix after i minus 1 iteration.

It should be noted that, since the signal is reconstructed from the sparse statistical information, and the dimension of the matrix corresponding to the sparse statistical information is related to the number of the transmitting antennas and the number of the receiving antennas, the dimension of the matrix corresponding to the reconstructed signal may be determined before reconstruction, that is, the number of rows and columns of the matrix corresponding to the reconstructed signal may be determined. In addition, in this embodiment, the reconstructed signal may be a column-vectorized vector, and then a matrix corresponding to the reconstructed signal is a matrix obtained by matrixing the column vector of the reconstructed signal. The diagonal elements are elements on the diagonal of the matrix corresponding to the reconstructed signal. In addition, the diagonal column vector is a column vector corresponding to an index mark of a diagonal element in a matrix corresponding to the reconstructed signal in the observation matrix, and it is understood that a column vector index of the diagonal column vector is the same as a row vector index of the diagonal element in the reconstructed signal, and the index mark of the diagonal element is a row vector index of the diagonal element in the reconstructed signal. Here, the matrix corresponding to the reconstructed signal is a matrix of 9 by 9, then the reconstructed signal is a column vectorization vector of 81 by 1, and the observation matrix may be a matrix including 81 columns. Thus, the diagonal elements of the matrix corresponding to the reconstructed signal have row vector indices of 1, 11, 21, 31, 41, 51, 61, 71, and 81 in the reconstructed signal, and thus the diagonal column vectors in the observation matrix include column vectors having column vector indices of 1, 11, 21, 31, 41, 51, 61, 71, and 81.

In addition, the above steps are described by taking a first iteration as an example, and the residual vector in the first iteration is a column-vectorized vector of the covariance matrix. In addition, since the residual vector is continuously updated, a different column vector index can be obtained for each iteration.

It should be noted that, here, the current iteration index is a column vector index when in the observation matrix, and is a row vector index when in the reconstructed signal, because the index here is only a sequence number or a numerical value.

b) The receiver judges whether the i is 1, if so, the current iteration index is updated to a set to be updated, if not, the current iteration index and an extrapolation element are updated to the set to be updated in the index of the reconstruction signal, wherein when the current iteration index is larger than a historical iteration index, the extrapolation element comprises an element of which the index of the reconstruction signal is equal to the difference value of the current iteration index minus a row index difference and an element of which the index of the reconstruction signal is equal to the sum value of the historical iteration index plus the row index difference, when the current iteration index is smaller than the historical iteration index, the extrapolation element comprises an element of which the index of the reconstruction signal is equal to the sum value of the current iteration index plus the row index difference and an element of which the index of the reconstruction signal is equal to the difference value of the historical iteration index minus the row index difference, the row index difference is an absolute value of a difference between a row vector index corresponding to the current iteration index in a matrix corresponding to the reconstruction signal and a row vector index corresponding to the historical iteration index in the matrix corresponding to the reconstruction signal, and the historical iteration index is an iteration index of any iteration before the current iteration.

Taking as an example when the current iteration index is smaller than the historical iteration index, the extrapolation element includes an element whose index in the reconstructed signal is equal to a difference between the current iteration index and a line index difference subtracted by the current iteration index, and it can be understood that the included extrapolation element has an index in the reconstructed signal equal to a difference between the current iteration index and a line index difference subtracted by the current iteration index. Similarly, the above-mentioned element included in the index of the reconstructed signal equal to the sum of the historical iteration index plus the row index difference may be understood as including an extrapolated element in the index of the reconstructed signal equal to the sum of the current iteration index plus the row index difference. Here again by way of example a matrix corresponding to the reconstructed signal is a 9 by 9 matrix,

if the column vector index of the observation matrix selected by the diagonal search in the first iteration is 21, adding the index 21 into the set to be updated, and performing no diagonal extrapolation in the first iteration; the column vector index of the observation matrix searched out along the diagonal in the second iteration is 51, the index 51 is taken as the second iteration index, and the indexes 21 and 51 correspond to the element (3,3) and the element (6,6) respectively in the matrix corresponding to the reconstructed signal, and the row index difference of the element (3,3) and the element (6,6) is 3, then the indexes 51, minus 3 and equal to 48 and 21, plus 3 and equal to 24 can be extrapolated, wherein the indexes 48 and 24 correspond to the element (3,6) and the element (6,3) in the matrix corresponding to the reconstructed signal; the column vector index of the observation matrix searched out along the diagonal in the third iteration is 81, then the index 81 is taken as the third iteration index, and the indexes 21, 51 and 81 respectively correspond to the element (3,3), the element (6,6) and the element (9,9) in the corresponding matrix of the reconstructed signal, wherein the indexes 21 and 51 are history iteration indexes, and the row index difference between the element (3,3) and the element (9,9) is 6, and the row index difference between the element (6,6) and the element (9,9) is 3, then the third iteration may extrapolate an index of 81 minus 3 equal to 78, an index of 51 plus 3 equal to 54, an index of 81 minus 6 equal to 75, and an index of 21 plus 6 equal to 27, wherein the indexes 78, 54, 75 and 27 correspond to the element (6,9) element (9,6), element (3,9) and element (9, 3).

c) And the receiver carries out reconstruction vector estimation based on least square on a partial matrix formed by the column vectors in the observation matrix corresponding to the indexes in the set to be updated and the column vectorization vector of the covariance matrix so as to obtain the ith reconstruction signal.

Through the steps a) and b), a set to be updated including the index can be obtained, so that a partial matrix can be formed by using the column vector in the observation matrix corresponding to the index in the set to be updated, and then the partial matrix and the column vectorization vector of the covariance matrix are subjected to reconstruction vector estimation based on least square to obtain the ith reconstructed signal. The partial matrix may be a column vector index using the index in the set to be updated as a partial matrix, that is, a partial matrix including a column vector with an index being the index in the set to be updated is constructed.

d) When the i is equal to the L, the receiver determines the reconstructed signal of the ith time as a reconstructed signal of sparse statistical information of the channel.

e) And when the i is smaller than the L, the receiver takes the result of subtracting the product of the partial matrix and the ith reconstruction signal from the column vectorization vector of the covariance matrix as the residual vector of the ith iteration, adds 1 to the i, and triggers the step of calculating the inner product of each diagonal column vector of the observation matrix and the residual vector of the ith minus 1 iteration.

Through the steps a) to e), only the elements on the diagonal line of the matrix corresponding to the reconstructed signal can be searched and the extrapolated elements can be calculated when the OMP algorithm is executed, so that compared with the conventional OMP algorithm, in the embodiment, the number of iterations is far smaller than that in the conventional OMP algorithm, and the calculation amount is reduced. In addition, the above-described OMP algorithm may be defined as a DS-OMP (diagonalsearch Orthogonal Matching Pursuit) algorithm based on Diagonal Search.

304. The receiver uses the reconstructed signal to obtain the final statistical information of the channel.

In this embodiment, the relationship information between the sparse statistical information and the final statistical information may be obtained in advance, and step 304 may include:

For example: the sparse statistical information and the final statistical information have the following relationship:

wherein the above is the final statistical information, is the reconstructed signal of the sparse statistical information, and satisfies the condition that the U can be obtained by matrixing the vector_tIs N_t×N_tDFT matrix of, U_rIs N_r×N_rRepresents a Kronecker product operation ()^*Indicating a conjugate operation.

The above formula can be used for the receiver to obtain the statistical information of the channel according to the signal reconstruction result after the receiver performs the signal reconstruction algorithm based on the compressed covariance perception, and then to matrixing the vector into the final statistical information of the channel

Wherein, the above description is as follows:

column vector r of a covariance matrix of a received signal received by a receiver in an ideal scene or in a noise-free scene_yCan be transformed into the following formula:

wherein, U_rIs N_r×N_rRepresents the Kronecker product operation, ()^*Denotes a conjugation operation, R_hCovariance matrix representing the channel: and R is_hAnd can be represented by the following formula:

R_h＝E{hh^H}

where E { } denotes the mathematical expectation, H ═ vec (H), H is the channel matrix, and H can be further expressed by channel decomposition as:

wherein H_vIs a sparse matrix because in some propagation environments, for example: in a propagation environment combining the millimeter wave technology and the MIMO technology, the number of effective scatterers is limited, so that the number L of effective multipath channels is much smaller than the dimension of the channel, that is, the channel sparsity is L.

In addition, the channel covariance matrix R is based on an expansion of the product of the decomposed representation of the channel matrix H and Kronecker (Kronecker)_hCan be further expressed as:

wherein h is_v＝vec(H_v)。

It can be seen that the channel covariance matrix comprises a plurality of sparse channels h in time slot_v＝vec(H_v) The independent sparsity of the channels within each time slot is represented by the joint sparsity between the multi-slot channels.

The above formula R_hBringing in

Further comprising the following steps:

in addition, the mathematical symbols are further simplified, and the above formula can be expressed as

It should be noted that the above derivation is only a derivation between formulas, and the receiver does not need to perform the above derivation in practical applications, and the following disclosure may be obtained and used:

in addition, in the present embodiment, when calculating the final statistical information of the channel, the calculation is not limited to using the above formula, and a relationship between the reconstructed signal and the final statistical information may be obtained by a large amount of experimental data, and the final statistical information of the channel may be obtained by using the relationship.

The following will exemplify the steps of the receiver performing signal reconstruction based on compressed covariance sensing by using the covariance matrix to obtain a reconstructed signal of sparse statistical information of the channel by using fig. 4 as an example:

and the signal reconstruction based on compressed covariance perception at the receiver in this example translates to the following mathematical problem:

wherein argmin represents that solving makes the objective function minimum as a reconstruction result, | | | | | calculation₁|| ||₂Respectively representing a 1-norm and a 2-norm of the vector, wherein the 1-norm is a partial manifestation to joint sparsity, the 2-norm is a constraint to additive noise, and λ represents a lagrange parameter.

Specifically, in this example, the step shown in fig. 4 is used to obtain the reconstructed signal of the sparse statistical information, and as shown in fig. 4, the method includes:

1. and (4) inputting. Wherein the step ofMay be the input r_yΦ, L, where L is sparsity, or is understood as iteration number, or is understood as the number of nonzero elements on the diagonal line in the corresponding matrix.

2. And (5) initializing. This step may be an iteration index initialization i-0 and a reconstruction vector initialization r⁽⁰⁾0 (where bold zero represents a zero vector with all zero values of the elements), and initializing the residual vector with u⁽⁰⁾＝r_yAnd set initialization, wherein Γ, Λ represent the set to be updated and the diagonal search space, respectively.

3. A diagonal search space is defined. This step may be to define the subsequent search space at individual element positions on the diagonal of the channel covariance matrix, rather than at all element positions of the entire channel covariance matrix. I.e. only the non-zero elements on the diagonal of the channel covariance matrix are searched. The channel covariance matrix is the corresponding matrix or the matrix after matrixing

4. Searching along a diagonal line. In each iteration, the iteration index is firstly updated, i is i +1, and then the column vector index of the matrix phi is searched in the defined diagonal search space, wherein the index number t is the index number t⁽ⁱ⁾(wherein, the upper foot mark⁽ⁱ⁾Reference numeral denoting the current iteration) is

Wherein arg max-<u^(i-1),Φ_.,j>I represents a column vector index for which the search maximizes the absolute value of the inner product of the residual vector u updated in the last iteration and the column vector of the observation matrix Φ, and j ∈ Λ represents that the search range is limited only to the element position of the diagonal.

The inner product of each diagonal column vector of the matrix phi and the residual vector of the i minus 1 iteration can be calculated through the steps, and the index of the diagonal column vector with the maximum absolute value of the inner product is used as the index of the iteration.

5. And (4) judging whether the current iteration is the first iteration, if so, directly performing the step 7, and otherwise, performing the step 6.

6. Diagonal extrapolation is performed. The off-diagonal estimation means that the off-diagonal element position is obtained according to the previously searched diagonal element position and the currently searched diagonal element position, the operation is to utilize the hermitian structure characteristic of the channel covariance matrix, the non-zero elements on the diagonal reflect the autocorrelation characteristic of the sparse channel, and the non-zero elements outside the corresponding diagonal reflect the cross-correlation characteristic of the sparse channel. Taking a simple example of off-diagonal estimation as an example, in a matrix with a channel covariance matrix of 9 by 9, if a column vector index 21 in the matrix Φ is obtained in the first iteration and the index 21 corresponds to an element (3,3) in a covariance matrix (the matrix after the matrixing), that is, the diagonal element position of the covariance matrix selected in the diagonal search in the first iteration is (3,3), that is, an element (3,3) in the covariance matrix is a nonzero element, the off-diagonal estimation is not performed in the first iteration; the second iteration yields a column vector index 51 in the matrix Φ, and the index 51 corresponds to the element (6,6) in the covariance matrix, i.e., the diagonal element position of the covariance matrix selected by the diagonal search in the second iteration is (6,6), the off-diagonal element position of the covariance matrix derived by the second diagonal extrapolation is (3,6) (6,3), and the column vector indices of the elements (3,6) and (6,3) in the matrix Φ are 48 and 24; the third iteration yields a column vector index 81 in the matrix Φ, and the index 81 corresponds to the element (9,9) in the covariance matrix, i.e., the diagonal element position of the covariance matrix selected by the diagonal search in the third iteration is (9,9), the off-diagonal element position of the covariance matrix derived by the third diagonal extrapolation is (3,9) (9,3) (6,9) (9,6), and the column vector indices of the elements (3,9) (9,3) (6,9) (9,6) in the above are 75, 27, 78, and 54, and so on if there is a subsequent iteration.

7. And (4) updating the set. The step may be to update the index obtained by diagonal search and diagonal extrapolation in the current iteration into the set Γ to be updated.

8. And (6) least square estimation. This step may be a partial matrix formed from the column vector indices corresponding to the updated set Γ and r_yTo perform a least-squares based reconstruction vector estimation, specifically a least-squares reconstruction vector estimation by the following formula

Wherein, the lower foot mark is T-shaped⁽ⁱ⁾And, gamma⁽ⁱ⁾Respectively representing the corresponding element position of the vector and the corresponding column vector position of the matrix, and representing the pseudo-inverse operation of the matrix. It should be noted that, since the above is a column vectorization vector, the index in the above is the position in the above.

9. And (5) updating the residue. This step may be the last step of the current iteration, i.e. on the residual vector u⁽ⁱ⁾Perform a residual update, such as: expressed by the following formula:

10. and (6) outputting. The step is to output the least square reconstruction vector in the last (L-th) iteration as a final reconstruction result.

The statistical information of the channel obtained by the receiver according to the signal reconstruction result can be obtained through the steps, namely, the final estimation of the channel covariance matrix is obtained according to the following formula after the vector matrixing is carried out

In this embodiment, various optional embodiments are added on the basis of the embodiment shown in fig. 2, and the accuracy of the channel statistical information can be improved.

Referring now to fig. 5-8, fig. 5-8 are experimental data charts comparing an embodiment of the present invention with a prior art method of estimating channel statistics.

In fig. 5 to 8, Pr represents the probability of correct channel statistic estimation, and SNR represents the signal-to-noise ratio, and compared with the prior art (dashed square), the performance of the technical solution (solid triangle) provided by the embodiment of the present invention is better. For example: when the SNR is 10, the Pr in the prior art is about 0.4, and when the SNR is 10, the Pr in the technical scheme provided by the embodiment of the invention can reach 0.7.

In addition, L in fig. 6 represents sparsity, and as can be seen from fig. 6, along with the deterioration of sparsity of the channel (that is, "L" increases, the channel changes from extra sparsity to general sparsity), the performance variation of the technical scheme provided by the embodiment of the present invention is relatively gradual, and the performance variation is much smaller than the performance attenuation degree of the prior art, in other words, the technical scheme provided by the embodiment of the present invention has better robustness of the sparse channel environment.

In fig. 7 and 8, T represents the length of the time resource occupied by the pilot signal in each time slot, F represents the number of observation time slots of the pilot signal, or both can also be understood as the observation overhead used for estimating the statistical information of the channel. As can be seen from fig. 7 and 8, on the premise of obtaining the same estimation accuracy, the technical solution provided by the embodiment of the present invention uses less observation overhead than the prior art solution, and the observation overhead, regardless of the length "T" of the time resource in each time slot or the number "F" of the observation time slots, is less than the overhead used by the prior art solution.

Moreover, because the technical scheme provided by the embodiment of the invention only needs to carry out signal reconstruction operation once, the complexity of calculation in the embodiment of the invention is lower, and the embodiment of the invention also can adopt a DS-OMP algorithm which only searches on a diagonal line, thereby further reducing the complexity of calculation, and the advantages brought by the invention are that the hardware complexity and the power consumption of the system can be greatly reduced.

For convenience of description, only the relevant parts of the embodiments of the present invention are shown, and details of the specific technology are not disclosed.

Referring to fig. 9, fig. 9 is a schematic structural diagram of a receiver according to an embodiment of the present invention, as shown in fig. 9, including: a receiving unit 91, a calculating unit 92, a reconstructing unit 93 and an obtaining unit 94, wherein:

a receiving unit 91, configured to receive a received pilot signal including a pilot signal transmitted by a transmitter in F time slots, where F is an integer greater than or equal to 1.

The calculating unit 92 is configured to calculate a covariance matrix of received signals of the receiver on a channel on which the pilot signals are transmitted, using the received F received pilot signals.

The reconstructing unit 93 is configured to perform signal reconstruction based on compressed covariance sensing by using the covariance matrix to obtain a reconstructed signal of sparse statistical information of the channel.

The obtaining unit 94 is configured to obtain final statistical information of the channel by using the reconstructed signal.

Referring to fig. 10, fig. 10 is a schematic structural diagram of another receiver according to an embodiment of the present invention, as shown in fig. 10, including: a receiving unit 101, a calculating unit 102, a reconstructing unit 103 and an obtaining unit 104, wherein:

a receiving unit 101, configured to receive a received pilot signal including a pilot signal transmitted by a transmitter in F time slots, where F is an integer greater than or equal to 1.

A calculating unit 102, configured to perform column vectorization on the received F received pilot signals by the calculating unit to obtain column vectorized vectors of the F received pilot signals, and calculate a covariance matrix of received signals of the receiver on a channel on which the pilot signals are transmitted by using the column vectorized vectors of the F received pilot signals.

A reconstruction unit 103, configured to perform signal reconstruction based on compressed covariance perception using the covariance matrix to obtain a reconstructed signal of sparse statistical information of the channel.

In this embodiment, the reconstruction unit 103 may be configured to use a column vectorization vector of the covariance matrix, an observation matrix, and L as inputs of an orthogonal matching pursuit OMP algorithm, and execute the OMP algorithm to obtain a reconstructed signal of sparse statistical information of the channel, where the observation matrix is a matrix that is obtained in advance and used for signal reconstruction, and L is sparsity.

In this embodiment, the reconstruction unit 103 may include:

a first calculating subunit 1031, configured to calculate an inner product of each diagonal column vector of the observation matrix and a residual vector of the i-th minus 1-time iteration, and use an index of the diagonal column vector with a largest absolute value of the inner product as an index of the current iteration, where the diagonal column vector is a column vector corresponding to an index identifier of a diagonal element in a matrix corresponding to the reconstruction signal in the observation matrix, i is a label of the current iteration, and the residual vector of the i-th minus 1-time iteration is a residual of a column vectorization vector of the covariance matrix after i-minus 1-time iteration;

a determining unit 1032, configured to determine whether i is 1, if yes, update the current iteration index into a set to be updated, if no, update the current iteration index and an extrapolated element into the set to be updated in the index of the reconstructed signal, where, when the current iteration index is greater than a historical iteration index, the extrapolated element includes an element whose index is equal to a difference value obtained by subtracting a row index difference from the current iteration index and also includes an element whose index is equal to a sum value obtained by adding the row index difference to the historical iteration index when the current iteration index is less than the historical iteration index, and when the current iteration index is less than the historical iteration index, the extrapolated element includes an element whose index is equal to a sum value obtained by adding the row index difference from the current iteration index and also includes an element whose index is equal to a difference value obtained by subtracting the row index difference from the historical iteration index when the reconstructed signal is less than the historical iteration index, the row index difference is an absolute value of a difference between a row vector index corresponding to the current iteration index in a matrix corresponding to the reconstruction signal and a row vector index corresponding to the historical iteration index in a matrix corresponding to the reconstruction signal, and the historical iteration index is an iteration index of any iteration before the current iteration;

a least square unit 1033, configured to perform a least square-based reconstruction vector estimation on a partial matrix formed by column vectors in the observation matrix corresponding to the indexes in the set to be updated and a column vectorization vector of the covariance matrix to obtain an i-th reconstructed signal;

a determining unit 1034, configured to determine the ith reconstructed signal as a reconstructed signal of sparse statistical information of the channel when i is equal to L;

a second calculating subunit 1035, configured to, when i is smaller than L, subtract a result of subtracting a product of the partial matrix and the i-th reconstructed signal from a column-vectorization vector of the covariance matrix as a residual vector of an i-th iteration, add i to 1, and trigger the operation of calculating an inner product of each diagonal column vector of the observation matrix and the residual vector of the i-th iteration minus 1.

An obtaining unit 104, configured to obtain final statistical information of the channel by using the reconstructed signal.

In this embodiment, the obtaining unit 104 may be configured to obtain final statistical information corresponding to the reconstructed signal as final statistical information of the channel according to relationship information between pre-obtained sparse statistical information and final statistical information.

In this embodiment, various optional embodiments are added on the basis of the embodiment shown in fig. 9, and the accuracy of the channel statistical information can be improved.

Referring to fig. 11, fig. 11 is a schematic structural diagram of another receiver according to an embodiment of the present invention, as shown in fig. 11, including: a processor 111, a network interface 112, a memory 113 and a communication bus 114, wherein the communication bus 114 is used for realizing the connection communication among the processor 111, the network interface 112 and the memory 113, and the processor 111 executes the program stored in the memory 113 for realizing the following method:

receiving a received pilot signal including a pilot signal transmitted by a transmitter in F time slots, wherein F is an integer greater than or equal to 1;

calculating a covariance matrix of received signals of the receiver on a channel on which the pilot signals are transmitted using the received F received pilot signals;

performing signal reconstruction based on compressed covariance perception by using the covariance matrix to obtain a reconstructed signal of sparse statistical information of the channel;

and acquiring final statistical information of the channel by using the reconstructed signal.

In this embodiment, the process executed by the processor 111 for calculating the covariance matrix of the received signals of the receiver on the channel for transmitting the pilot signals by using the received F received pilot signals may include:

and performing column vectorization on the received F received pilot signals to obtain column vectorization vectors of the F received pilot signals, and calculating a covariance matrix of the received signals of the receiver on a channel for transmitting the pilot signals by using the column vectorization vectors of the F received pilot signals.

In this embodiment, the process executed by the processor 111 for performing signal reconstruction based on compressed covariance perception by using the covariance matrix to obtain a reconstructed signal of sparse statistical information of the channel may include:

and taking a column vectorization vector, an observation matrix and L of the covariance matrix as the input of an orthogonal matching pursuit OMP algorithm, and executing the OMP algorithm to obtain a reconstructed signal of sparse statistical information of the channel, wherein the observation matrix is a matrix which is acquired in advance and used for signal reconstruction, and the L is sparsity.

In this embodiment, the program executed by the processor 111 to take the column vectorization vector of the covariance matrix, the observation matrix, and L as the inputs of the orthogonal matching pursuit OMP algorithm, and execute the OMP algorithm to obtain the reconstructed signal of the sparse statistical information of the channel may include:

calculating an inner product of each diagonal column vector of the observation matrix and a residual vector of the i minus 1 iteration, and taking an index of the diagonal column vector with the maximum absolute value of the inner product as an index of the current iteration, wherein the diagonal column vector is a column vector corresponding to an index identifier of a diagonal element in a matrix corresponding to the reconstruction signal in the observation matrix, i is a label of the current iteration, and the residual vector of the i minus 1 iteration is a residual of a column vectorization vector of the covariance matrix after i minus 1 iteration;

judging whether the i is 1, if so, updating the current iteration index into a set to be updated, otherwise, updating the current iteration index and an extrapolation element into the set to be updated, wherein when the current iteration index is larger than a historical iteration index, the extrapolation element comprises an element of which the index of the reconstructed signal is equal to the difference value of the current iteration index minus a row index difference and also comprises an element of which the index of the reconstructed signal is equal to the sum value of the historical iteration index plus the row index difference, when the current iteration index is smaller than the historical iteration index, the extrapolation element comprises an element of which the index of the reconstructed signal is equal to the sum value of the current iteration index plus the row index difference and also comprises an element of which the index of the reconstructed signal is equal to the difference value of the historical iteration index minus the row index difference, the row index difference is an absolute value of a difference between a row vector index corresponding to the current iteration index in a matrix corresponding to the reconstruction signal and a row vector index corresponding to the historical iteration index in a matrix corresponding to the reconstruction signal, and the historical iteration index is an iteration index of any iteration before the current iteration;

performing reconstruction vector estimation based on least square on a partial matrix formed by column vectors in the observation matrix corresponding to the indexes in the set to be updated and a column vectorization vector of the covariance matrix to obtain an ith reconstruction signal;

when the i is equal to the L, determining the reconstructed signal of the ith time as a reconstructed signal of sparse statistical information of the channel;

and when the i is smaller than the L, taking a result obtained by subtracting the product of the partial matrix and the ith reconstruction signal from the column vectorization vector of the covariance matrix as a residual vector of the ith iteration, adding 1 to the i, and triggering the step of calculating the inner product of each diagonal column vector of the observation matrix and the residual vector of the ith iteration minus 1.

In this embodiment, the procedure executed by the processor 111 for obtaining the final statistical information of the channel by using the reconstructed signal may include:

and acquiring final statistical information corresponding to the reconstructed signal as final statistical information of the channel according to relationship information between the pre-acquired sparse statistical information and the final statistical information.

In addition, embodiments of the present invention also provide a computer readable storage medium storing one or more programs, the one or more programs including instructions, which when executed by a receiver including a screen and a plurality of application programs, cause the receiver to perform the method according to any one of the implementations provided by the embodiments of the present invention.

It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.

The above disclosure is only for the purpose of illustrating the preferred embodiments of the present invention, and it is therefore to be understood that the invention is not limited by the scope of the appended claims.

Claims

A method for acquiring channel statistical information is characterized by comprising the following steps:

a receiver receives a receiving pilot signal comprising a pilot signal sent by a transmitter in F time slots, wherein F is an integer greater than or equal to 1;

the receiver using the received F received pilot signals to calculate a covariance matrix of the received signals of the receiver on the channel on which the pilot signals are transmitted;

the receiver uses the covariance matrix to perform signal reconstruction based on compressed covariance perception so as to obtain a reconstructed signal of sparse statistical information of the channel;

the receiver uses the reconstructed signal to obtain the final statistical information of the channel.
The method of claim 1, wherein the receiver uses the received F received pilot signals to compute a covariance matrix of the receiver's received signals on a channel over which the pilot signals are transmitted, comprising:

the receiver performs column vectorization on the received F received pilot signals to obtain column vectorized vectors of the F received pilot signals, and calculates a covariance matrix of the received signals of the receiver on a channel on which the pilot signals are transmitted using the column vectorized vectors of the F received pilot signals.
The method of claim 1 or 2, wherein the receiver performs compressed covariance perception-based signal reconstruction using the covariance matrix to obtain a reconstructed signal of sparse statistics of the channel, comprising:

and the receiver takes the column vectorization vector of the covariance matrix, an observation matrix and L as the input of an orthogonal matching pursuit OMP algorithm, and executes the OMP algorithm to obtain a reconstructed signal of sparse statistical information of the channel, wherein the observation matrix is a matrix which is acquired in advance and used for signal reconstruction, and the L is sparsity.
The method of claim 3, wherein the receiver tracks an OMP algorithm with column vectorization vectors of the covariance matrix, an observation matrix, and L as inputs to the OMP algorithm and executes the OMP algorithm to obtain a reconstructed signal of sparse statistics of the channel, comprising:

the receiver calculates an inner product of each diagonal column vector of the observation matrix and a residual vector of the i minus 1 iteration, and takes an index of the diagonal column vector with the maximum absolute value of the inner product as an index of the current iteration, wherein the diagonal column vector is a column vector corresponding to an index identifier of a diagonal element in a matrix corresponding to the reconstruction signal in the observation matrix, i is a label of the current iteration, and the residual vector of the i minus 1 iteration is a residual of a column vectorization vector of the covariance matrix after i minus 1 iteration;

the receiver judges whether the i is 1, if so, the current iteration index is updated to a set to be updated, if not, the current iteration index and an extrapolation element are updated to the set to be updated in the index of the reconstruction signal, wherein when the current iteration index is larger than a historical iteration index, the extrapolation element comprises an element of which the index of the reconstruction signal is equal to the difference value of the current iteration index minus a row index difference and also comprises an element of which the index of the reconstruction signal is equal to the sum value of the historical iteration index plus the row index difference, when the iteration current index is smaller than the historical iteration index, the extrapolation element comprises an element of which the index of the reconstruction signal is equal to the sum value of the current iteration index plus the row index difference and also comprises an element of which the index of the reconstruction signal is equal to the difference value of the historical iteration index minus the row index difference, the row index difference is an absolute value of a difference between a row vector index corresponding to the current iteration index in a matrix corresponding to the reconstruction signal and a row vector index corresponding to the historical iteration index in a matrix corresponding to the reconstruction signal, and the historical iteration index is an iteration index of any iteration before the current iteration;

the receiver carries out reconstruction vector estimation based on least square on a partial matrix formed by column vectors in the observation matrix corresponding to indexes in the set to be updated and a column vectorization vector of the covariance matrix so as to obtain an ith reconstruction signal;

when the i is equal to the L, the receiver determines the reconstructed signal of the ith time as a reconstructed signal of sparse statistical information of the channel;

and when the i is smaller than the L, the receiver takes the result of subtracting the product of the partial matrix and the ith reconstruction signal from the column vectorization vector of the covariance matrix as the residual vector of the ith iteration, adds 1 to the i, and triggers the step of calculating the inner product of each diagonal column vector of the observation matrix and the residual vector of the ith minus 1 iteration.
The method of any of claims 1-4, wherein the receiver uses the reconstructed signal to obtain final statistics for the channel, comprising:

and the receiver acquires final statistical information corresponding to the reconstructed signal as final statistical information of the channel according to the relationship information between the pre-acquired sparse statistical information and the final statistical information.
A receiver, comprising: receiving unit, computational element, rebuild unit and acquisition unit, wherein:

the receiving unit is configured to receive a received pilot signal including a pilot signal transmitted by a transmitter in F slots, where F is an integer greater than or equal to 1;

the calculation unit is used for calculating a covariance matrix of a received signal of the receiver on a channel for transmitting the pilot signal by using the received F received pilot signals;

the reconstruction unit is used for performing signal reconstruction based on compressed covariance perception by using the covariance matrix so as to obtain a reconstructed signal of sparse statistical information of the channel;

the obtaining unit is configured to obtain final statistical information of the channel by using the reconstructed signal.
The receiver of claim 6, wherein the calculation unit is configured to column-vector the received F received pilot signals to obtain column-vectored vectors for the F received pilot signals, and to calculate a covariance matrix of received signals of the receiver on a channel on which the pilot signals are transmitted using the column-vectored vectors for the F received pilot signals.
The receiver of claim 6 or 7, wherein the reconstruction unit is configured to track an OMP algorithm with a column vectorization vector of the covariance matrix, an observation matrix and L as inputs of the OMP algorithm, and execute the OMP algorithm to obtain a reconstructed signal of sparse statistical information of the channel, wherein the observation matrix is a pre-acquired matrix for signal reconstruction, and the L is a sparsity.
The receiver of claim 8, wherein the reconstruction unit comprises:

a first calculating subunit, configured to calculate an inner product of each diagonal column vector of the observation matrix and a residual vector of the i minus 1 iteration, and use an index of the diagonal column vector with a largest absolute value of the inner product as an index of the current iteration, where the diagonal column vector is a column vector corresponding to an index identifier of a diagonal element in a matrix corresponding to the reconstruction signal in the observation matrix, i is a label of the current iteration, and the residual vector of the i minus 1 iteration is a residue of a column vectorized vector of the covariance matrix after i minus 1 iteration;

a determining unit, configured to determine whether i is 1, if so, update the current iteration index into a set to be updated, otherwise, update the current iteration index and an extrapolated element into the set to be updated in the index of the reconstructed signal, where, when the current iteration index is greater than a historical iteration index, the extrapolated element includes an element whose index is equal to a difference value obtained by subtracting a row index difference from the current iteration index and also includes an element whose index is equal to a sum value obtained by adding the row index difference to the historical iteration index when the current iteration index is less than the historical iteration index, and when the current iteration index is less than the historical iteration index, the extrapolated element includes an element whose index is equal to a sum value obtained by adding the row index difference to the current iteration index and also includes an element whose index is equal to a difference value obtained by subtracting the row index difference from the historical iteration index when the reconstructed signal is less than the historical iteration index, the row index difference is an absolute value of a difference between a row vector index corresponding to the current iteration index in a matrix corresponding to the reconstruction signal and a row vector index corresponding to the historical iteration index in a matrix corresponding to the reconstruction signal, and the historical iteration index is an iteration index of any iteration before the current iteration;

a least square unit, configured to perform least square-based reconstruction vector estimation on a partial matrix formed by column vectors in the observation matrix corresponding to the indexes in the set to be updated and a column-vectorized vector of the covariance matrix to obtain an i-th reconstructed signal;

a determining unit, configured to determine the ith reconstructed signal as a reconstructed signal of sparse statistical information of the channel when i is equal to L;

and a second calculating subunit, configured to, when i is smaller than L, take a result obtained by subtracting a product of the partial matrix and the i-th reconstructed signal from a column-vectorized vector of the covariance matrix as a residual vector of an i-th iteration, add i to 1, and trigger the operation of calculating an inner product of each diagonal column vector of the observation matrix and the residual vector of the i-th iteration minus 1.
The receiver according to any of claims 6-9, wherein the obtaining unit is configured to obtain final statistical information corresponding to the reconstructed signal as final statistical information of the channel according to relationship information between pre-obtained sparse statistical information and final statistical information.
A receiver, comprising: the system comprises a processor, a network interface, a memory and a communication bus, wherein the communication bus is used for realizing connection communication among the processor, the network interface and the memory, and the processor executes a program stored in the memory to realize the following method:

receiving a received pilot signal including a pilot signal transmitted by a transmitter in F time slots, wherein F is an integer greater than or equal to 1;

calculating a covariance matrix of received signals of the receiver on a channel on which the pilot signals are transmitted using the received F received pilot signals;

performing signal reconstruction based on compressed covariance perception by using the covariance matrix to obtain a reconstructed signal of sparse statistical information of the channel;

and acquiring final statistical information of the channel by using the reconstructed signal.
The receiver of claim 11, wherein the processor executes a program for using received F received pilot signals to compute a covariance matrix for a received signal of the receiver on a channel over which the pilot signals are transmitted, comprising:

and performing column vectorization on the received F received pilot signals to obtain column vectorization vectors of the F received pilot signals, and calculating a covariance matrix of the received signals of the receiver on a channel for transmitting the pilot signals by using the column vectorization vectors of the F received pilot signals.
The receiver of claim 11 or 12, wherein the processor performs a compressed covariance perception-based signal reconstruction using the covariance matrix to obtain a reconstructed signal of sparse statistical information for the channel, comprising:

and taking a column vectorization vector, an observation matrix and L of the covariance matrix as the input of an orthogonal matching pursuit OMP algorithm, and executing the OMP algorithm to obtain a reconstructed signal of sparse statistical information of the channel, wherein the observation matrix is a matrix which is acquired in advance and used for signal reconstruction, and the L is sparsity.
The receiver of claim 13, wherein the processor executes a program that tracks an OMP algorithm with a column vectorization vector of the covariance matrix, an observation matrix, and L as inputs to the OMP algorithm and executes the OMP algorithm to obtain a reconstructed signal of sparse statistics of the channel, comprising:

calculating an inner product of each diagonal column vector of the observation matrix and a residual vector of the i minus 1 iteration, and taking an index of the diagonal column vector with the maximum absolute value of the inner product as an index of the current iteration, wherein the diagonal column vector is a column vector corresponding to an index identifier of a diagonal element in a matrix corresponding to the reconstruction signal in the observation matrix, i is a label of the current iteration, and the residual vector of the i minus 1 iteration is a residual of a column vectorization vector of the covariance matrix after i minus 1 iteration;

judging whether the i is 1, if so, updating the current iteration index into a set to be updated, otherwise, updating the current iteration index and an extrapolation element into the set to be updated, wherein when the current iteration index is larger than a historical iteration index, the extrapolation element comprises an element of which the index of the reconstructed signal is equal to the difference value of the current iteration index minus a row index difference and also comprises an element of which the index of the reconstructed signal is equal to the sum value of the historical iteration index plus the row index difference, when the current iteration index is smaller than the historical iteration index, the extrapolation element comprises an element of which the index of the reconstructed signal is equal to the sum value of the current iteration index plus the row index difference and also comprises an element of which the index of the reconstructed signal is equal to the difference value of the historical iteration index minus the row index difference, the row index difference is an absolute value of a difference between a row vector index corresponding to the current iteration index in a matrix corresponding to the reconstruction signal and a row vector index corresponding to the historical iteration index in a matrix corresponding to the reconstruction signal, and the historical iteration index is an iteration index of any iteration before the current iteration;

performing reconstruction vector estimation based on least square on a partial matrix formed by column vectors in the observation matrix corresponding to the indexes in the set to be updated and a column vectorization vector of the covariance matrix to obtain an ith reconstruction signal;

when the i is equal to the L, determining the reconstructed signal of the ith time as a reconstructed signal of sparse statistical information of the channel;

and when the i is smaller than the L, taking a result obtained by subtracting the product of the partial matrix and the ith reconstruction signal from the column vectorization vector of the covariance matrix as a residual vector of the ith iteration, adding 1 to the i, and triggering the step of calculating the inner product of each diagonal column vector of the observation matrix and the residual vector of the ith iteration minus 1.
The receiver of any of claims 11-14, wherein the processor executes a procedure for obtaining final statistics for the channel using the reconstructed signal, comprising:

and acquiring final statistical information corresponding to the reconstructed signal as final statistical information of the channel according to relationship information between the pre-acquired sparse statistical information and the final statistical information.