CN104702543B

CN104702543B - Method for precoding and device

Info

Publication number: CN104702543B
Application number: CN201310648796.7A
Authority: CN
Inventors: 陈庆春; 曾韬; 张琴琴; 孙德福
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2013-12-04
Filing date: 2013-12-04
Publication date: 2018-05-29
Anticipated expiration: 2033-12-04
Also published as: CN104702543A

Abstract

An embodiment of the present invention provides a kind of method for precoding, the pilot signal sent by receiving transmitting terminal；Channel state information estimation is carried out according to the pilot signal, to obtain channel average and channel covariancc；According to the channel average and channel covariancc, pre-coding matrix is calculated；Transmitting terminal carries out the data-signal that needs are sent precoding according to the pre-coding matrix, and the receiving terminal separation pre-coding matrix carries out solution precoding to the precoded signal of reception.It can realize that the MIMO OFDM precodings based on average Signal to Interference plus Noise Ratio criterion can obtain better MIMO precodings unfailing performance and mimo system capacity under the conditions of statistical channel status information, be obtained under low relevant environment better than unfailing performance and mimo system capacity based on codebook precoding.

Description

Precoding method and device

Technical Field

The invention relates to the field of mobile communication, in particular to a precoding method and a precoding device.

Background

Multiple Input Multiple Output (MIMO) technology can improve the system spatial throughput without increasing the system bandwidth, and can also significantly improve the signal transmission quality by effectively utilizing spatial diversity. In the MIMO system, the precoding technique is a signal processing technique that uses Channel State Information (CSI) to preprocess a transmission signal at a transmitting end to achieve the purpose of eliminating inter-symbol interference (ISI) or inter-user interference and improving system capacity. Common precoding types include mainly linear precoding and nonlinear precoding. The linear precoding mainly includes two linear precoding based on a codebook and a non-codebook. The precoding based on the codebook is that a known codebook set is shared at a transmitting end and a receiving end, the codebook set comprises a plurality of precoding matrixes, the receiving end selects the precoding matrix which can optimize the system performance in the codebook set according to a certain performance index of a channel matrix estimated by a channel, then the codebook sequence number is fed back to the transmitting end, and the transmitting end selects the precoding matrix according to the related sequence number for precoding. Problems with codebook-based precoding include that the codebook design should match the channel characteristics, and in practical applications, the system may work in different fading environments, the inter-antenna spacing is also different, and the antenna pattern and polarization are also different, which makes the codebook design very complicated; in addition, the design of codebooks is often required to meet constant modulus properties, finite character and nesting properties, and the like. These factors greatly influence and restrict the codebook-based precoding scheme to more flexibly meet different requirements of the actual system. Compared with codebook-based MIMO precoding, non-codebook MIMO precoding can flexibly select a precoding matrix matched with channel characteristics according to dynamic changes of channel conditions, and can better meet the MIMO precoding requirement under a complex time-varying environment.

The existing MIMO precoding is difficult to simultaneously consider two performance indexes of precoding reliability and MIMO system capacity realized by a system under the condition of medium and low signal to noise ratio.

Disclosure of Invention

The invention aims to provide a precoding method to realize a precoding method based on an average signal-to-interference-and-noise ratio criterion, thereby obtaining the reliability and the MIMO system capacity which are superior to those based on codebook precoding in a low-correlation environment.

In order to achieve the above object, an aspect of the embodiments of the present invention provides a precoding method, where the method includes:

receiving an uplink/downlink pilot signal sent by a sending end;

estimating channel state information according to the uplink/downlink pilot signals to obtain a channel mean value and a channel covariance;

calculating a precoding matrix according to the channel mean value and the channel covariance;

when the uplink/downlink channel does not meet the symmetry, the precoding matrix is fed back to the sending end, so that the sending end can carry out precoding on the data signals needing to be sent according to the precoding matrix;

and according to the precoding matrix, performing de-precoding on the received data signal.

Based on the first aspect, in a first possible implementation manner, the performing channel state information estimation according to the uplink/downlink pilot signals to obtain a channel mean and a channel covariance further includes:

performing channel state estimation of a pilot symbol position on the received pilot signal to obtain a least square channel estimation result based on the pilot symbol in channel correlation time;

performing linear interpolation on the obtained least square channel estimation result based on the pilot symbols in the channel correlation time to obtain a channel estimation result on a subcarrier where data to be estimated is located;

calculating a noise variance estimation value on a subcarrier where a pilot symbol is located according to a least square channel estimation result based on the pilot symbol in the correlation time;

and acquiring statistical channel state information in relevant time according to the noise variance estimation value, wherein the statistical channel state information comprises the channel mean and the channel covariance.

Based on the first possible implementation manner, in a second possible implementation manner, the performing channel state estimation on pilot symbol positions on the received pilot signals to obtain a least squares channel estimation result based on the pilot symbols in the channel correlation time specifically includes:

n in one sub-frame_pA pilot symbol Y_k,p(i_m)＝X_k,p(i_m)H_k,p(i_m)+W_k,p(i_m),0≤m≤N_p-1

The integration is as follows: y is_k,p＝X_k,pH_k,p+W_k,p

Wherein k represents the kth subcarrier; i.e. i_mIs the position index, N, on the OFDM symbol where the pilot signal is present_pIs the number of pilot symbols on one subcarrier;in order to receive the pilot symbols, the mobile station,pilot frequency sent for a sending end; n is a radical of_rFor receiving the number of antennas, N_tNumber of transmit antennas;is the corresponding frequency domain noise; h_k,p(i_m) Representing the frequency domain response of the MIMO channel on the k sub-carrier, and having

In said Y_k,p＝X_k,pH_k,p+W_k,pIn (1),

according to said Y_k,p＝X_k,pH_k,p+W_k,pObtaining the frequency domain response estimated on the k sub-carrier of the MIMO channelWherein,is X_k,pThe pseudo-inverse of (1);

according to the aboveAcquiring a cascaded receiving pilot matrix in the coherent time:

wherein H_k,p(i_m)≈H_k,p(i₀),m＝1,…,N_c，N_c>N_pIs the equivalent OFDM symbol number corresponding to the channel correlation time, wherein:

according to the cascaded receiving pilot matrix in the coherent time:

the LS channel estimation result based on pilot symbols in the channel correlation time is obtained as follows:

wherein,is composed ofThe pseudo-inverse of (1).

Based on the second possible implementation manner, in a third possible implementation manner, the performing linear interpolation on the obtained least-squares channel estimation result based on the pilot symbols in the channel correlation time to estimate a channel estimation result of the position where the data symbol is located further includes:

the specific steps of channel estimation of the position of the data symbol to be estimated in the time domain are as follows:

H_k,data(l′)＝α₁×H_k,p,LS(l)+β₁×H_k,p,LS(l+Δl)

wherein l 'represents the time domain index of the position of the data to be estimated, l < l' < l + delta l, l represents the position of the pilot signal in the time domain, delta l represents the time domain interval of two adjacent pilot signals in the time domain, k represents the subcarrier serial number of the pilot signal, and the interpolation coefficient α₁，β₁Is calculated as follows

And

the step of channel state information on the frequency domain index of the position of the data to be estimated in the frequency domain specifically comprises the following steps:

H_k′(l)＝α₂×H_k,S(l)+β₂×H_k+Δk,S(l),S∈[p,data]；

k 'represents the frequency domain index of the position of the data to be estimated, k < k' < k + Δ k, k represents the position of the pilot signal in the frequency domain, Δ k represents the frequency interval of two adjacent pilot signals in the frequency domain, and the interpolation coefficient α₂，β₂Is calculated as follows

Based on the third possible implementation manner, in a fourth possible implementation manner, the calculating, according to the least square channel estimation result based on the pilot symbol in the correlation time, a noise variance estimation value on a subcarrier where the pilot symbol is located specifically includes:

wherein H_k,p,LS(i_m+1)-H_k,p,LS(i_m)||_FRepresents the Frobenius norm, i_m+1,i_mRepresenting the serial numbers of adjacent pilot subcarriers on the same pilot OFDM symbol, the final noise variance estimate can be represented by the following formula:

where M is the total number of sub-carriers occupied by pilot symbols.

Based on the fourth possible implementation manner, in a fifth possible implementation manner, the obtaining statistical channel state information within a correlation time according to the noise variance estimation value, where the statistical channel state information includes the channel mean and the channel covariance, and specifically includes:

selecting time domain sliding window length L according to the correlation time of the communication channel, wherein the sliding window length corresponds to the number of pilot symbols included in the subframe in the sliding window, and according to the correlation bandwidth of the communication channel, selecting M included in each coherent bandwidth in the frequency domain₀Calculating time i from the channel estimation results of the pilot subcarriersMean of individual channel state information:

and covariance:

whereinvec (-) represents vectorizing the matrix, (.)^HRepresenting a conjugate transpose.

Based on the fifth possible implementation manner, in a sixth possible implementation manner, the calculating a precoding matrix according to the channel mean and the channel covariance further includes:

the system channel mean value obtained according to estimationChannel covarianceSum noise varianceAdding the effect of noise to the channel meanSum channel covariance Φ_HTo obtain a new channel mean valueSum channel covarianceWherein:

wherein the noise covariance matrix

According toAnd phi_H,nCalculating a matrixWherein

Wherein X_[i,j]The ijth dimension representing X is N_r×N_rThe sub-block of (1);

will matrix Z^(s)Cholesky decomposition is carried out to obtain a matrix L^(s)：

Z^(s)＝L^(s)·(L^(s))^H；

To L^(s)Performing singular value decomposition on the matrix:

to obtain Z^(s)Characteristic value decomposition expression of

The precoding matrix based on the average signal-to-interference-and-noise ratio criterion is calculated according to the following method

Wherein

WhereinIs a matrixAnd mu is a water filling factor. P is the sum of all subcarriers carrying user data at the transmitting end and the transmitting power of the transmitting antenna.

Based on the first aspect, in a seventh possible implementation manner, the signal received by the receiving end after precoding is:

Y_k(l)＝H_k(l)·B·X_k(l)+W_k(l)

wherein

Channel state information H of non-pilot symbol position is obtained through time domain and frequency domain two-dimensional linear interpolation estimation_k(l) And the precoding matrix B is combined with the precoding matrix B,determining an equivalent channel matrix H_E,k(l) I.e. by

H_E,k(l)＝Β·H_k(l)

According to an equivalent joint channel matrix H_E,k(l) The receiving end can perform joint detection and reception of de-precoding and MIMO detection by taking the minimum mean square error as a criterion

In a second aspect, an embodiment of the present invention provides a precoding apparatus, where the apparatus includes:

a receiving unit, configured to receive a pilot signal sent by a sending end;

the estimating unit is used for estimating the channel state information according to the uplink pilot signal and the downlink pilot signal so as to obtain a channel mean value and a channel covariance;

a calculating unit, configured to calculate a precoding matrix according to the channel mean and the channel covariance;

and the feedback unit is used for feeding the precoding matrix back to the sending end when the uplink channel and the downlink channel do not meet the symmetry, so that the sending end can precode the data signals needing to be sent according to the precoding matrix.

And the decoding unit is used for performing de-precoding on the received data signals according to the precoding matrix.

Based on the second aspect, in a first possible implementation manner, the estimating unit is specifically configured to:

By the precoding method provided by the embodiment of the invention, the pilot signal sent by the sending end is received; estimating channel state information according to the pilot signal to obtain a channel mean value and a channel covariance; calculating a precoding matrix according to the channel mean value and the channel covariance; and feeding the precoding matrix back to the sending end so that the sending end can precode the data signals needing to be sent according to the precoding matrix. The MIMO-OFDM precoding based on the average signal-to-interference-and-noise ratio criterion can better process the reliable performance and the MIMO system capacity of the MIMO precoding under the condition of channel state information statistics, and the reliable performance and the MIMO system capacity which are superior to those based on codebook precoding are obtained under the low-correlation environment.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without inventive exercise.

Fig. 1 is a schematic diagram of an implementation of MIMO precoding technology based on an average signal-to-interference-and-noise ratio criterion under TD-LTE system conditions.

Fig. 2 is a schematic block diagram of MIMO precoding reception processing based on an average signal-to-interference-and-noise ratio criterion at a receiving end under TD-LTE system conditions.

FIG. 3 is a flow chart of a precoding method provided by an embodiment of the present invention;

FIG. 4 is a schematic block diagram of a MIMO precoding transmission process based on an average signal-to-interference-and-noise ratio criterion at a transmitting end under TD-LTE system conditions;

fig. 5 is a schematic diagram illustrating the reliable performance of MIMO precoding based on the average signal-to-interference-and-noise ratio criterion under the condition of a highly correlated ETU1 channel;

fig. 6 is a schematic diagram of the capacity of a MIMO precoding system based on the average signal-to-interference-and-noise ratio criterion under the condition of a highly correlated ETU1 channel;

fig. 7 is a diagram illustrating reliable performance of MIMO precoding based on an average signal-to-interference-and-noise ratio criterion under a low correlation ETU1 channel condition;

fig. 8 is a schematic diagram of the capacity of a MIMO precoding system based on the average signal-to-interference-and-noise ratio criterion under the condition of a low correlation ETU1 channel;

fig. 9 is a schematic diagram of reliable performance of MIMO precoding based on an average signal-to-interference-and-noise ratio criterion under a high correlation EVA1 channel condition;

fig. 10 is a schematic diagram of the capacity of a MIMO precoding system based on the average signal-to-interference-and-noise ratio criterion under the condition of a highly correlated EVA1 channel;

fig. 11 is a schematic diagram of reliable performance of MIMO precoding based on an average signal-to-interference-and-noise ratio criterion under a low correlation EVA1 channel condition;

fig. 12 is a schematic diagram of the capacity of a MIMO precoding system based on the average signal-to-interference-and-noise ratio criterion under the condition of a low correlation EVA1 channel;

FIG. 13 is a diagram illustrating reliable performance of MIMO precoding based on average SINR criterion under highly correlated EPA channel conditions;

FIG. 14 is a diagram illustrating the capacity of a MIMO precoding system based on the average SINR criterion under highly correlated EPA channel conditions;

FIG. 15 is a diagram illustrating the reliable performance of MIMO precoding based on the average SINR criterion under low-correlation EPA channel conditions;

FIG. 16 is a diagram illustrating the capacity of a MIMO precoding system based on the average SINR criterion under low EPA-related channel conditions;

fig. 17 is a schematic structural diagram of a precoding apparatus according to an embodiment of the present invention;

fig. 18 is a schematic structural diagram of a precoding apparatus according to an embodiment of the present invention.

Detailed Description

The technical solution of the present invention is further described in detail by the accompanying drawings and embodiments.

The technical scheme provided by the embodiment of the invention can be applied to various wireless communication networks, such as: a global system for mobile communication (GSM) system, a Code Division Multiple Access (CDMA) system, a Wideband Code Division Multiple Access (WCDMA) system, a Universal Mobile Telecommunications (UMTS) system, a General Packet Radio Service (GPRS) system, a Long Term Evolution (LTE) system, a long term evolution advanced (LTE-a) system, a Worldwide Interoperability for Microwave Access (WiMAX) system, and the like. The terms "network" and "system" are used interchangeably.

In the embodiment of the present invention, a Base Station (BS) may be a device that communicates with a User Equipment (UE) or other communication stations, such as a relay station, and the base station may provide communication coverage in a specific physical area. For example, the base station may specifically be a Base Transceiver Station (BTS) or a Base Station Controller (BSC) in GSM or CDMA; or Node B (NB) in UMTS or Radio Network Controller (RNC) in UMTS; the ue may also be an evolved node B (ENB or eNodeB) in LTE; alternatively, the present invention is not limited to this, and the present invention may be other access network devices that provide access services in a wireless communication network.

In the embodiments of the present invention, the UEs may be distributed throughout the wireless network, and each UE may be static or mobile. A UE may be referred to as a terminal (terminal), a mobile station (mobile station), a subscriber unit (subscriber unit), a station (station), etc. The UE may be a cellular phone (cellular phone), a Personal Digital Assistant (PDA), a wireless modem (modem), a wireless communication device, a handheld device (hand-held), a laptop computer (laptop computer), a cordless phone (cordless phone), a Wireless Local Loop (WLL) station, or the like.

Fig. 1 shows a schematic diagram of an embodiment of implementing MIMO precoding based on an average signal-to-interference-and-noise ratio criterion proposed by the present invention in a TD-LTE system. In order to implement MIMO precoding based on the average signal-to-interference-and-noise ratio criterion, a sending end information source signal is sent to a sending end precoding module after scrambling and modulating processing, a determined precoding matrix is calculated according to a precoding matrix calculation module, transmission precoding processing is carried out on sending data, then resource particle mapping is sent, an OFDM symbol is generated after the OFDM signal generation module, and the OFDM symbol is sent to a sending antenna port to be transmitted through an MIMO-OFDM channel. In order to implement MIMO precoding based on the average signal-to-interference-and-noise ratio criterion, a receiving end of a TD-LTE system receives signals through a receiving antenna port, the signals are converted back to frequency domain signals from time domain signals after CP (channel control protocol) and OFDM (orthogonal frequency division multiplexing) demodulation, then resource particle mapping is removed, joint detection of MIMO and precoding is executed through a precoding removing module, and finally information source signals sent by a base station end are restored after demodulation and descrambling. In the embodiment of the present invention, both the base station and the user equipment may be used as main bodies for executing the method provided by the present invention, for example, when the base station is used as a receiving end, the base station receives a pilot signal sent by the user equipment through an uplink channel, and obtains a channel mean and a channel covariance, and calculates a precoding matrix. On the contrary, if the base station is used as the sending end, the user equipment is used as the receiving end, and the user equipment receives the pilot signal sent by the base station through the downlink channel and calculates the precoding matrix.

Fig. 2 shows a schematic block diagram of MIMO precoding reception processing based on an average signal-to-interference-and-noise ratio criterion at a receiving end under TD-LTE system conditions. A receiving end receives signals through a receiving antenna port, removes a Cyclic Prefix (CP), and converts time domain signals back to frequency domain signals after OFDM demodulation; then, resource particle mapping is solved; the channel estimation module estimates the channel gain of the pilot symbol position according to the received pilot signal, and counts the channel mean value and the channel covariance, then the precoding calculation module calculates and determines the precoding matrix according to the channel mean value and the channel covariance, and when the sending end can not estimate the channel state information by using the channel reciprocity, the receiving end feeds back the precoding matrix to the sending end through the feedback channel so as to implement precoding processing. And the precoding matrix and the channel estimation of the data symbol position obtained by implementing two-dimensional linear interpolation calculation according to the channel estimation of the pilot frequency symbol position are sent to a precoding and MIMO combined detection module. And after the joint detection result of the precoding removal and the MIMO is subjected to layer de-mapping, demodulation and descrambling, restoring the joint detection result to the sending end to send the message.

Based on the foregoing system, an embodiment of the present invention provides a precoding method, where in this embodiment, a sending end may be a mobile phone or a base station, and correspondingly, a receiving end may be a base station or a mobile phone, where the sending end is a mobile phone, the receiving end is a base station, and the receiving end is a mobile phone, as shown in fig. 3, where the method includes:

301, a receiving end receives an uplink/downlink pilot signal sent by a sending end;

302, the receiving end estimates the channel state information according to the uplink/downlink pilot signals to obtain the channel mean and the channel covariance;

more specifically, the step further comprises: the receiving end carries out channel state estimation of the pilot frequency symbol position on the received pilot frequency signal so as to obtain a least square channel estimation result based on the pilot frequency symbol in the channel correlation time;

performing linear interpolation on the obtained least square channel estimation result based on the pilot symbols in the channel correlation time to obtain a channel estimation result of the data to be estimated on the subcarrier where the pilot signal is located;

303, the receiving end calculates a precoding matrix according to the channel mean value and the channel covariance;

304, when the uplink/downlink channel does not satisfy the symmetry, the receiving end feeds back the precoding matrix to the transmitting end, so that the transmitting end performs precoding for the data signal to be transmitted according to the precoding matrix;

305, the receiving end de-precodes the received data signal according to the precoding matrix.

Specifically, the channel state estimation of the pilot symbol position is performed on the received pilot signal to obtain the least square channel estimation result based on the pilot symbol in the channel correlation time, which may be implemented in the following manner:

the receiving end of the TD-LTE system receives the signal through the receiving antenna port, and the frequency domain pilot signal received after the CP and OFDM demodulation are removed is

Y_k,p(i_m)＝X_k,p(i_m)H_k,p(i_m)+W_k,p(i_m),0≤m≤N_p-1 (1)

Wherein i_mIs the position index, N, on the OFDM symbol where the pilot signal is present_pIs the number of pilot symbols on one subcarrier;for received pilot symbols, N_rK represents the kth subcarrier for the number of receive antennas;pilot matrix, N, for transmit side transmission_tNumber of transmit antennas;is the corresponding frequency domain noise; h_k,p(i_m) Representing the frequency response of the MIMO channel at DFT frequency k, and having

N in one sub-frame_pThe pilot symbols are integrated

Y_k,p＝X_k,pH_k,p+W_k,p(3)

Wherein

Based on the measurement signal of formula (3), the LS channel estimation result of the pilot symbol can be obtained as follows:

whereinIs X_k,pThe pseudo-inverse of (1). Consider that in practical systems the channel gain does not change significantly within the channel coherence time, i.e. H_k,p(i_m)≈H_k,p(i₀),m＝1,…,N_cIn which N is_c>N_pFor the equivalent OFDM symbol number corresponding to the channel correlation time, equation (3) can be rewritten as the following cascade observation signal sequence:

wherein

Based on the measurement signal of equation (5), the LS channel estimation result based on the pilot symbols in the channel correlation time can be obtained as follows:

whereinIs composed ofThe pseudo-inverse of (1).

On the basis of estimating the channel state information of the pilot frequency symbol position, the channel state information of the non-pilot frequency symbol position can be deduced through time domain and frequency domain two-dimensional linear interpolation; namely, linear interpolation is carried out on the obtained least square channel estimation result based on the pilot frequency symbol in the channel correlation time so as to obtain the channel estimation result of the data to be estimated on the subcarrier of the pilot frequency signal;

firstly, the step of obtaining the channel state information on the time domain index of the position of the to-be-homed data in the time domain specifically comprises:

H_k,data(l′)＝α₁×H_k,p,LS(l)+β₁×H_k,p,LS(l+Δl) (7)

wherein l 'represents the time domain index of the position of the data to be estimated and has l < l' < l + delta l, l represents the position of the pilot signal in the time domain, delta l represents the time domain interval of two adjacent reference signals in the time domain, k represents the subcarrier serial number of the pilot signal, and the interpolation coefficient α₁，β₁Is calculated as follows

And obtaining the channel estimation result of the subcarrier where the pilot symbol is located at the position of the non-pilot symbol in the time domain after time domain interpolation.

Then, the step of performing channel state information on the time domain index of the to-be-corrupted data position in the frequency domain specifically includes:

H_k′(l)＝α₂×H_k,S(l)+β₂×H_k+Δk,S(l),S∈[p,data](9)

k 'represents the frequency domain index of the position of the data to be estimated and k < k' < k + Δ k, k represents the position of the pilot signal in the frequency domain, Δ k represents the frequency interval of two adjacent reference signals in the frequency domain, and the interpolation coefficient α₂，β₂Is calculated as follows

And estimating to obtain channel estimation results of all data symbol positions after frequency domain interpolation.

The formula (6) shows

Then, according to the least square channel estimation result based on the pilot frequency symbol in the correlation time, calculating the noise variance estimation value of the sub-carrier where the pilot frequency symbol is located;

this is achieved in particular by the fact that the gain of the channel on adjacent pilot subcarriers on the same OFDM symbol is very close, i.e. H_k,p(i_m)≈H_k,p(i_m+1), the noise variance estimation on the sub-carrier k where the pilot symbol is located can be directly derived by using the following method

Wherein H_k,p,LS(i_m+1)-H_k,p,LS(i_m)||_FRepresenting the Frobenius norm. The final total noise variance estimate may be represented by:

where M represents the total number of sub-carriers occupied by pilot symbols.

Finally, the process is carried out in a batch,according to the correlation time of communication channel, reasonably selecting time domain sliding window length L (corresponding to the number of pilot symbols included in sub-frame in sliding window), according to the correlation bandwidth of channel, in frequency domain, from M contained in each coherent bandwidth₀The channel estimation result of each pilot frequency subcarrier is estimated and calculated at the time i by the following methodMean value of individual channel state informationSum covarianceWhere N is the total number of subcarriers carrying user data.

Wherein,vec (·) represents the vectorization of the matrix, (·)^HRepresenting a conjugate transpose.

The above means for calculating the channel state informationSum covarianceAccording to adjacent M within the coherence bandwidth of the frequency domain₀The channel estimation result of the sub-carrier where the pilot frequency symbol is located is counted and calculated to obtain the average value of the channel state information in the frequency domain coherent bandwidthSum covarianceSince the channel characteristics of different subcarriers within the coherence bandwidth are approximately the same, it can be considered that the mean values of the channel state information corresponding to each subcarrier within the coherence bandwidth are all the sameCovariance of

In the MIMO precoding system, B represents the precoding matrix, and H represents N_r×N_tOf MIMO channels, wherein N_rFor receiving the number of antennas, N_tFor the number of transmitting antennas, A represents the receiving detection matrix, and the corresponding error vectorHas a covariance matrix of

Wherein

R_y＝E{yy^H}＝HBB^HH^H+R_n(18)

Covariance matrix if complex white gaussian noiseThe receiving end adopts a linear minimum mean square error LMMSE detection matrix, namely

The covariance matrix can be further written as:

the signal to interference plus noise ratio SINR of the corresponding ith transmission symbol_iAnd its corresponding MSE_iThe following relationships exist:

wherein

Wherein b is_iThe ith column vector, [ X ], representing matrix B]_i,jI rows and j columns of elements of the matrix X are represented. The sum of the signal to interference and noise ratios of all transmitted symbols can be expressed as follows:

under the constraint condition of total transmission power, the optimization target can be set to maximize the sum of the signal-to-interference-and-noise ratios of all the transmitted symbols to determine the precoding matrix B, i.e. the corresponding precoding optimization design problem can be described as the following optimization problem

Where P/N denotes the total transmit power allocated to each subcarrier by the transmitting end and Tr (-) denotes the trace. From the exchange properties of the trace, tr (ab) ═ tr (ba), the objective function can be equivalently as follows:

E_H{tr(B^HH^HHB)}＝E_H{tr(HBB^HH^H)}＝E_H{tr(HQH^H)} (25)

wherein Q ═ BB^HThe autocorrelation matrix representing the transmit precoding. Covariance matrix if complex white gaussian noiseThe form of the objective function remains unchanged, and only the following replacement is needed:

whereinRepresenting the Kronecker product.

The solution to the equation (24) optimization problem may calculate and determine the optimal precoding matrix B under the average signal-to-interference-and-noise ratio criterion. Definition ofWhere unec (-) is the inverse operator of vec (-) and utilizes a variableBased on the statistical DCSIT channel model in equation (16), the objective function in equation (24) can be further written as:

wherein

Wherein (·)^TRepresenting the transpose of a matrix, here using the trace function property Tr (A)^HB)＝(vec(A))^Hvec (B) in combination with the channel covariance Φ_HThe definition of (1). Substituting equation (28) into equation (27), and according to the operational property of Kronecker product, the objective function can be transformed into:

wherein (·)^*The conjugate of the matrix is represented as,(29) the equation property of trace Tr (A) ═ Tr (A) is used^T)。N_rN_t×N_rN_tDimension matrixN of (i, j)_r×N_rThe sub-matrices being proportional unit matrices, i.e.Wherein q is_ijI row and j column elements, delta, representing Q_rsIs the Kronecker delta function (delta)_rs＝1,r＝s;δ_rs0, r ≠ s) according toThe special form of the matrix and the definition that the trace of the matrix is equal to the sum of the diagonal elements of the matrix. (29) The formula can be further simplified as follows:

wherein X_[i,j]N of (i, j) th representing X_r×N_rA sub-matrix. Therefore, the objective function in equation (24) can be rewritten as

Because of the elements of the Z matrixThe prime is the trace of a sub-block of the semi-positive definite matrix, and the Z matrix is also semi-positive definite. It can be shown that the basis of the matrix Q corresponds to Z, i.e. ifRepresenting the eigenvalue decomposition of Z, the eigenvalue of Q is decomposed into

Wherein, Λ_QA matrix of eigenvalues representing Q. Substituting the eigenvalue decomposition of Z and the above formula into the objective function, Λ_QBecomes an optimization problem as follows

Wherein λ_Zi，λ_QiRespectively representing the characteristic values of Z and Q. Since the Z matrix is represented by the channel mean and the channel covariance. If the sub-channels are considered to distribute the transmission power equally, i.e.' A_QIs equal toIf the transmission power is taken into accountDistributed to each transmit antenna by power flooding, then Λ_QIs calculated as follows

Wherein,is a matrixAnd mu is a water filling factor.

In particular, the average value of the MIMO-OFDM system channel obtained according to estimationChannel covarianceSum noise varianceAdding the influence of noise into the channel mean and the channel covariance to obtain a new channel meanSum channel covarianceWherein

Wherein the noise covariance matrixAccording toAnd phi_H,nCalculating a matrixWherein:

wherein X_[i,j]The ijth dimension of X is N_r×N_rThe sub-blocks of (1). Will matrix Z^(s)Cholesky decomposition is carried out to obtain a matrix L^(s)

Z^(s)＝L^(s)·(L^(s))^H(38)

Then to L^(s)Singular value decomposition of matrix

Thereby obtaining Z^(s)Characteristic value decomposition expression of

The precoding matrix based on the average signal-to-interference-and-noise ratio criterion is calculated as follows

Wherein

Considering the transmission power P to be equally distributed to each transmission antennaIs calculated as follows

If the transmission power is taken into accountIs distributed to each transmitting antenna by power water filling, thenIs calculated as follows

WhereinIs a matrixAnd mu is a water filling factor.

The form of the precoding matrix B is observed,for the power allocation matrix, the power allocation situation is reflected,is a matrix Z^(s)The unitary matrix of (3) can be regarded as a multimode beamforming matrix by decomposing the multiplexed signal into several mutually orthogonal directions for transmission due to the characteristic that the column vectors are mutually orthogonal, and the input forming matrix can be regarded as a unit matrix I. And limit conditionsThe total power constant characteristic is satisfied. Obviously, with the precoding matrix shown in formula (41), the average signal-to-interference-and-noise ratio at the receiving end is

From the foregoing analysis, it can be seen that when each subcarrier precoding matrix is computationally determined from statistical DCSIT information, the precoding scheme is primarily dependent on the channel mean of each subcarrier channelChannel covarianceChannel mean value considering different sub-carrier channels within frequency domain coherence bandwidthChannel covarianceBasically, when the precoding matrix of different subcarrier channels is determined by actual calculation, the channels can be correspondingly divided into frequency domain coherent bandwidthAnd (4) calculating a group to determine a corresponding precoding matrix.

At the receiving end, after receiving the data sent by the sending end by adopting the fed-back precoding matrix, the receiving end receives signals after adopting precoding:

Y_k(l)＝H_k(l)·B·X_k(l)+W_k(l) (46)

whereinOn the basis of estimating the channel state information of the pilot frequency symbol position, the receiving end obtains the channel state information H of the non-pilot frequency symbol position through two-dimensional linear interpolation estimation of a time domain and a frequency domain_k(l) Combining the pre-coding matrix B determined by calculation, an equivalent channel matrix H can be calculated and determined_E,k(l) I.e. by

H_E,k(l)＝Β·H_k(l) (47)

According to an equivalent joint channel matrix H_E,k(l) Receiving joint detection receiving capable of performing de-precoding and MIMO detection by taking minimum mean square error as criterion

Fig. 4 shows a schematic block diagram of a MIMO precoding transmitting end process under TD-LTE system conditions to implement the average signal-to-interference-and-noise ratio criterion. When the uplink and the downlink satisfy reciprocity, a sending end receives signals through an antenna port, after SC-FDMA demodulation, resource particle mapping is solved, a channel estimation module estimates channel gain of a pilot frequency symbol position according to received uplink pilot frequency signals, channel mean values and channel covariance are counted, and then a precoding calculation module calculates and determines a downlink precoding matrix according to the channel mean values and the channel covariance. When the sending end can not estimate the channel state information by using the channel reciprocity, the sending end receives the precoding matrix determined by the receiving end through the feedback channel so that the sending end can carry out precoding processing.

Fig. 5-16 compare error rate performance and system capacity under ETU1, EVA1 and EPA channel conditions of SVD linear precoding (average power allocation SVD-ave and power water-filling SVD-wf), MIMO precoding based on average signal-to-interference-and-noise ratio (SINR mean) criterion under DCSIT condition and codebook precoding based on maximized SINR codebook selection (average power allocation SINRave and power water-filling SINRwf) under DCSIT condition, where feedback delay of channel state information is transmission time of one subframe, feedback granularity is one subframe, time domain feedback granularity of PMI in closed-loop codebook feedback scheme is one subframe, frequency domain feedback granularity is 2 resource blocks, time domain feedback granularity of linear precoding SVD is one OFDM symbol, frequency domain feedback granularity is 1, and simulation conditions are shown in table 1. From fig. 5-16 the following conclusions can be drawn:

on the first hand, from the perspective of precoding reliability and system traversal capacity achieved by precoding, the SVD-wf can obtain the best compromise between precoding performance and system capacity, and the precoding performance of the SVD-ave, signal-to-interference-and-noise ratio mean SINRave and SINRwf is basically equivalent to codebook precoding based on SINR criterion.

In a second aspect, under low correlation channel conditions, the precoding scheme based on the signal-to-interference-and-noise ratio averages SINRave and SINRwf can obtain reliable performance close to that of the SVD precoding scheme, and the system capacity is equivalent to that of codebook precoding based on the SINR criterion. This shows that the precoding scheme based on the signal to interference plus noise ratio (SINR) mean value provides a feasible technical scheme for precoding implementation under the condition of low correlation channel.

In a third aspect, in a low signal-to-noise ratio region, the reliable performance of the SINRave precoding is better than that of the SINRwf precoding, while the system capacity is slightly worse than that of the SINRwf precoding, and in a high signal-to-noise ratio region, the precoding performances of the two power allocation schemes are equivalent. This shows that the power water filling scheme has little performance gain for precoding based on the average SINR criterion, and the implementation of average allocated power is simple, and SINRave precoding is a simple and beneficial technical scheme.

Considering the reliability and system capacity of the system, and the requirement of the system for channel state information, the precoding scheme based on the SINR mean value criterion is a reasonable precoding technical scheme under the DCSIT condition.

Correspondingly, an embodiment of the present invention further provides a precoding apparatus, where the apparatus may be a base station or a user equipment, and when the apparatus is a base station, the base station is a receiving end and the transmitting end is a user equipment, on the contrary, if the apparatus is a user equipment, the base station is a transmitting end and the receiving end is a user equipment, as can be seen from fig. 17, the apparatus includes:

a receiving unit 701, configured to receive a pilot signal sent by a sending end;

an estimating unit 702, configured to perform channel state information estimation according to the uplink and downlink pilot signals to obtain a channel mean and a channel covariance;

a calculating unit 703, configured to calculate a precoding matrix according to the channel mean and the channel covariance;

a feedback unit 704, configured to feed back the precoding matrix to the sending end when the uplink and downlink channels do not satisfy symmetry, so that the sending end performs precoding on a data signal that needs to be sent according to the precoding matrix.

A decoding unit 705, configured to perform de-precoding on the received data signal according to the precoding matrix.

Further, the estimation unit is specifically configured to:

For how the precoding apparatus calculates the channel mean and the covariance in this embodiment, and the method for obtaining the precoding matrix, reference may be made to the foregoing embodiments, which are not repeated herein.

As shown in fig. 18, an embodiment of the present invention further provides a precoding apparatus, where the apparatus may be a base station or a user equipment, and when the apparatus is a base station, the base station is a receiving end and the transmitting end is a user equipment, on the contrary, if the apparatus is a user equipment, the base station is a transmitting end and the receiving end is a user equipment, as seen from fig. 18, the apparatus includes a transmitter 182, a receiver 181, a memory 183, and a processor 184 connected to the transmitter 182, the receiver 181, and the memory 183, respectively. Of course, the base station may further include general components such as an antenna, a baseband processing component, a medium radio frequency processing component, and an input/output device, and the embodiments of the present invention are not limited in any way here.

Wherein the memory 183 stores a set of program codes, and the processor 184 is configured to call the program codes stored in the memory, and is configured to perform the following operations:

receiving an uplink/downlink pilot signal sent by a sending end through a receiver;

feeding the precoding matrix back to the sending end when the uplink/downlink channel does not meet the symmetry through a transmitter, so that the sending end can precode the data signals needing to be sent according to the precoding matrix;

It should be noted that the precoding apparatus shown in fig. 17 and fig. 18 can be used to implement any one of the methods provided by the above method embodiments, and details are not repeated here.

Those of skill would further appreciate that the various illustrative components and algorithm steps described in connection with the embodiments disclosed herein may be implemented as electronic hardware, computer software, or combinations of both, and that the various illustrative components and steps have been described above generally in terms of their functionality in order to clearly illustrate this interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. 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 present invention.

The steps of a method or algorithm described in connection with the embodiments disclosed herein may be embodied in hardware, a software module executed by a processor, or a combination of the two. A software module may reside in Random Access Memory (RAM), memory, Read Only Memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art.

The above-mentioned embodiments are intended to illustrate the objects, technical solutions and advantages of the present invention in further detail, and it should be understood that the above-mentioned embodiments are merely exemplary embodiments of the present invention, and are not intended to limit the scope of the present invention, and any modifications, equivalent substitutions, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims

1. A method of precoding, the method comprising:

receiving an uplink/downlink pilot signal sent by a sending end;

2. The precoding method of claim 1, wherein the estimating of channel state information based on the uplink/downlink pilot signals to obtain a channel mean and a channel covariance, further comprises:

3. The precoding method of claim 2, wherein the performing channel state estimation of pilot symbol positions on the received pilot signals to obtain a least-squares channel estimation result based on the pilot symbols within the channel correlation time comprises:

n in one sub-frame_pA pilot symbol Y_k,p(i_m)＝X_k,p(i_m)H_k,p(i_m)+W_k,p(i_m),0≤m≤N_p-1 is integrated as: y is_k,p＝X_k,pH_k,p+W_k,p

Wherein k represents the kth subcarrier; i.e. i_mIs on an OFDM symbol in which a pilot signal is presentPosition index of (1), N_pIs the number of pilot symbols on one subcarrier;in order to receive the pilot symbols, the mobile station,pilot frequency sent for a sending end; n is a radical of_rFor receiving the number of antennas, N_tNumber of transmit antennas;is the corresponding frequency domain noise; h_k,p(i_m) Representing the frequency domain response of the MIMO channel on the k sub-carrier, and having

In said Y_k,p＝X_k,pH_k,p+W_k,pIn (1),

according to the cascaded receiving pilot matrix in the coherent time:

wherein,is composed ofThe pseudo-inverse of (1).

4. The method of claim 3, wherein said performing linear interpolation on said obtained least squares channel estimation based on pilot symbols over channel correlation time to estimate channel estimation for the location of the data symbol, further comprises:

H_k,data(l′)＝α₁×H_k,p,LS(l)+β₁×H_k,p,LS(l+Δl)

And

H_k′(l)＝α₂×H_k,S(l)+β₂×H_k+Δk,S(l),S∈[p,data]；

5. The method as claimed in claim 4, wherein said calculating the noise variance estimation value on the sub-carrier where the pilot symbol is located according to the least square channel estimation result based on the pilot symbol in the correlation time comprises:

wherein H_k,p,LS(i_m+1)-H_k,p,LS(i_m)||_FRepresents the Frobenius norm, i_m+1,i_mRepresenting the order of adjacent pilot sub-carriers on the same pilot OFDM symbolThe final noise variance estimate is represented by:

where M is the total number of sub-carriers occupied by pilot symbols.

6. The method as claimed in claim 5, wherein said obtaining statistical channel state information within a correlation time according to the noise variance estimation value, the statistical channel state information including the channel mean and the channel covariance, comprises:

and covariance:

7. The method of claim 6, wherein the computing a precoding matrix based on the channel mean and channel covariance further comprises:

obtaining the mean value of the system channel according to the estimationChannel covarianceSum noise varianceAdding the effect of noise to the channel meanSum channel covariance Φ_HTo obtain a new channel mean valueSum channel covarianceWherein:

wherein the noise covariance matrix

According toAnd phi_H,nCalculating a matrixWherein

Z^(s)＝L^(s)·(L^(s))^H；

To L^(s)Performing singular value decomposition on the matrix:

to obtain Z^(s)Characteristic value decomposition expression of

Wherein

WhereinIs a matrixIs the r non-zero eigenvalues, mu is the water filling factor, P is all the children of the sending end carrying user dataThe sum of the carrier and transmit antenna transmit power.

8. The method of claim 1, wherein the signals received by the receiving end after precoding are:

Y_k(l)＝H_k(l)·B·X_k(l)+W_k(l)

wherein

Channel state information H of non-pilot symbol position is obtained through time domain and frequency domain two-dimensional linear interpolation estimation_k(l) Determining an equivalent channel matrix H in combination with the precoding matrix B_E,k(l) I.e. by

H_E,k(l)＝Β·H_k(l)

According to an equivalent joint channel matrix H_E,k(l) Using minimum mean square error as criterion, the receiving end carries out joint detection receiving of de-precoding and MIMO detection

9. A precoding apparatus, characterized in that the apparatus comprises:

a receiving unit, configured to receive a pilot signal sent by a sending end;

a feedback unit, configured to feed back the precoding matrix to the sending end when the uplink and downlink channels do not satisfy symmetry, so that the sending end performs precoding on a data signal to be sent according to the precoding matrix;

10. The precoding apparatus of claim 9, wherein the estimating unit is specifically configured to: