CN101499834B - Feedback quantity control method, communication system and related apparatus - Google Patents

Abstract

The embodiment of the invention discloses a feedback amount control method, a communication system and related equipment for reducing the feedback amount. The method of the present invention includes: receiving the data frame sent by the sender; determining the precoding matrix of each pilot subcarrier in the data frame; obtaining the transition code corresponding to the precoding matrix according to the preset codeword state classification code book word information; feeding back the transfer code word information to the sender. The embodiment of the present invention also provides a communication system and related equipment. The embodiment of the present invention can effectively reduce the amount of feedback.

Description

Feedback control method, communication system and related equipment

technical field

The present invention relates to the communication field, in particular to a feedback control method, a communication system and related equipment.

Background technique

With the increasing demand of users for high-speed wireless networks, the new generation of mobile communication systems is expected to provide higher data transmission rates, more reliable and stable communication and wider signal coverage, and support the use of mobility. In the near future, users will be able to obtain the same or even better service quality than today's wired broadband network through wireless network transmission.

Currently, the World Interoperability for Microwave Access (WiMAX, World Interoperability for Microwave Access) system adopts MIMO (Multiple In and Multiple Out) to increase the transmission rate.

The MIMO system can achieve a compromise between spatial multiplexing and spatial diversity by using the antenna arrays at both ends of the transceiver, thereby providing a wireless communication system with greater system capacity and higher quality of service. If the sender knows the channel state information (CSI), that is, the closed-loop system, the sender can take certain preprocessing on the transmitted signal to offset the influence of the wireless channel on the transmitted signal, and achieve better system performance than the open-loop system .

In the closed-loop MIMO-Orthogonal Frequency Division Multiplexing (OFDM, Orthogonal Frequency Division Multiplexing) system, the sender multiplies the transmitted signal by a preprocessing matrix according to the CSI, so that the transmitted signal is transmitted on the sub-channel with a larger gain, but due to The frequency division duplex (FDD, Frequency Division Duplexing) system does not have channel symmetry, and the receiver needs to pass the CSI to the sender through the feedback channel. In this case, the precoding matrix can be designed at the receiver and then fed back to the sender. Therefore, in the MIMO-OFDM system, the sender needs to know the precoding matrix on all subcarriers. If the feedback algorithm is not designed effectively, the amount of feedback will be too large to bear.

Some feedback amount control methods in the prior art are mostly based on the frequency domain correlation on the channel, that is, there is a large correlation between adjacent subcarriers, so that the receiver feeds back the code word of the precoding matrix at the pilot frequency by using this correlation information to reduce the amount of feedback.

Since the feedback amount control method in the above scheme is based on frequency domain correlation, this scheme is limited by frequency correlation, and the number of carriers contained in each subcarrier block cannot be too large, otherwise it may make the calculated non-pilot The distortion between the precoding matrix of the subcarrier at the subcarrier and the precoding matrix of the subcarrier at the pilot is very large, which affects the system performance. However, if the carrier block is not selected too large, the feedback amount of the system cannot be effectively reduced.

Contents of the invention

Embodiments of the present invention provide a feedback amount control method, a communication system and related equipment, which can effectively reduce the feedback amount of the system.

The method for controlling the amount of feedback provided by the embodiment of the present invention includes: receiving the data frame sent by the sender; determining the precoding matrix of each pilot subcarrier in the data frame; transfer codeword information corresponding to the precoding matrix; and feed back the transfer codeword information to the sender.

The communication system provided by the embodiment of the present invention includes: a sender and a receiver, the sender is used to send a data frame to the receiver, and receives the fed-back transition codeword information; the receiver is used to receive the sent The data frame sent by the party determines the precoding matrix of each pilot subcarrier in the data frame, obtains the transition codeword information corresponding to the precoding matrix according to the preset codeword state classification codebook, and sends the The party feeds back the transfer codeword information.

The data transmission network element provided by the embodiment of the present invention includes: a sending unit, used to send data frames and OFDM symbols; a codeword information receiving unit, used to receive codeword information; a query unit, used to The word information queries the corresponding precoding matrix in the preset codeword state classification code book; the precoding matrix processing unit is used to interpolate the precoding matrix on the non-pilot subcarrier according to the queried precoding matrix to obtain the precoding matrix of all subcarriers.

The data transmission network element provided by the embodiment of the present invention includes: a frequency domain feedback control unit, configured to receive an OFDM symbol, obtain an estimated channel value at the pilot frequency according to the pilot sequence in the OFDM symbol, and obtain an estimated value according to the estimated value The precoding matrix of the subcarrier at the pilot frequency is used to query the codeword information of the precoding matrix in the codeword state classification code book; the time domain feedback control unit is used to receive the data frame and determine the data frame in the data frame The precoding matrix of the subcarrier at each pilot frequency is used to obtain the transfer codeword information corresponding to the precoding matrix according to the preset codeword state classification codebook; the feedback unit is used to feed back the frequency domain feedback control unit to the sender The acquired codeword information and the transfer codeword information acquired by the time domain feedback control unit.

It can be seen from the above technical solutions that the embodiments of the present invention have the following advantages:

In the embodiment of the present invention, after receiving the data frame, the receiver determines the precoding matrix of each pilot subcarrier in the data frame, and obtains the precoding matrix corresponding to the precoding matrix according to the preset codeword state classification codebook. Transfer codeword information, and feed back the transfer codeword information to the sender, so it is not necessary to feed back all the codeword information of the transfer state, thereby reducing the amount of system feedback. Therefore, compared with the prior art, the embodiments of the present invention, Provides another way to reduce the amount of feedback.

Description of drawings

FIG. 1 is a schematic diagram of a MIMO-OFDM system in an embodiment of the present invention;

Fig. 2 is a flow chart of the first embodiment of the feedback amount control method in the embodiment of the present invention;

FIG. 3 is a schematic diagram of a precoding interpolation algorithm in an embodiment of the present invention;

Fig. 4 is a flow chart of the second embodiment of the feedback amount control method in the embodiment of the present invention;

5 is a schematic diagram of a Markov codebook model in an embodiment of the present invention;

FIG. 6 is a schematic diagram of an embodiment of a communication system in an embodiment of the present invention;

FIG. 7 is a schematic diagram of a first embodiment of a data transmission network element in an embodiment of the present invention;

FIG. 8 is a schematic diagram of a second embodiment of a data transmission network element in an embodiment of the present invention.

Detailed ways

Embodiments of the present invention provide a feedback amount control method, a communication system and related equipment, which are used to reduce the feedback amount of the system.

In the embodiment of the present invention, since the channel has correlation in the time domain, the codeword transfers from one state to several fixed states, not all states when performing state transition in the time domain. In the embodiment of the present invention, after receiving the data frame, the receiver determines the precoding matrix of each pilot subcarrier in the data frame, and obtains the transition corresponding to the precoding matrix according to the preset codeword state classification codebook The codeword information feeds back the transition codeword information to the sender, so it is not necessary to feed back codeword information of all transition states, thereby reducing the amount of system feedback.

On the basis of the above, in the embodiment of the present invention, the reduction feedback algorithm in the frequency domain and the time domain can also be used at the same time, that is, the time-frequency two-dimensional joint reduction feedback algorithm.

Please refer to FIG. 1, first introduce the MIMO-OFDM system structure in the embodiment of the present invention, the whole system consists of four parts: sender, forward wireless channel, receiver and feedback channel.

Assuming that the system has T transmitting antennas, R receiving antennas and C subcarriers, and the feedback algorithm adopts the codebook-based design method, the number of codeword information is N _B (=2 ^B ), that is, each codeword information uses B Bit means that if no feedback algorithm is used, the amount of feedback each time the receiver receives is C*B.

According to the type of feedback reduction, the feedback amount control method in the embodiment of the present invention is divided into: a time-frequency joint reduction feedback method and a time domain reduction feedback method, which are specifically introduced below:

1. Time-frequency joint reduction feedback method:

In practical applications, sometimes the coherent bandwidth of the channel is relatively small. At this time, the wideband MIMO-OFDM system needs to be modulated on correspondingly more subcarriers to ensure that the channel of each subcarrier is flat-fading.

In this method, both the sending and receiving ends transmit in units of data frames, and each data frame is divided into three parts: the pilot sequence, using the Chu sequence as the pilot, it is understandable that other similar sequences can also be used as the pilot The frequency sequence is used for frame synchronization and channel estimation; the signaling domain includes the number of subcarriers used by OFDM symbols in the data domain of the current frame, and also includes the frame length ΔT of the current frame, the frequency domain interval ΔF of the pilot, and modulation and coding Mode and other fields; data field, that is, the data load to be transmitted in the current frame.

Among them, both the pilot sequence and the signaling domain are transmitted with a fixed number of subcarriers and a reliable modulation method (such as binary phase shift keying (BPSK, Binary Phase Shift Keying)), and the signaling domain can also be properly coded , to ensure the accuracy of its reception.

In this method, the time-frequency two-dimensional correlation of the channel is used, and the sender determines the pilot sequence according to the performance requirements of the system, and uses the time-frequency two-dimensional joint reduction feedback algorithm to calculate the precoding matrix of each subcarrier of the current data block; Correspondingly, the receiver uses the pilot information to estimate the channel information on the subcarriers at the pilots, and then calculates the precoding matrices on the subcarriers at these pilots, and classifies these precoding matrices in the preset codeword state The codeword information in the book is fed back to the sender.

For ease of understanding, detailed description will be given below. Please refer to FIG. 2. The first embodiment of the feedback amount control method in the embodiment of the present invention includes:

201. The sender sets a pilot sequence;

表示通过内插得到的第t个OFDM符号的第n个子载波上的预编码矩阵，即发送方根据接收方反馈的导频处子载波的预编码矩阵内插得到其他非导频处子载波的预编码矩阵。In this embodiment, when the system is initialized, the sender needs to set some pilots to form a pilot sequence according to the system performance requirements, that is, because the channel has a correlation between time and frequency, it can be evenly distributed in the time domain and frequency domain. Some subcarriers are selected, and pilots are inserted on these subcarriers, and the pilots are used to instruct the receiver to feed back the codeword information of the precoding matrix of these subcarriers in the codebook. The specific performance requirements and the setting of the pilot sequence can be determined according to the actual application, which is not limited here, and only an example is used for illustration. The precoding interpolation in this embodiment is shown in Figure 3, and the black block in Figure 3 Indicates the pilot, that is, assuming that a pilot is inserted every ΔF=K subcarriers in the frequency domain, and a pilot is inserted every ΔT=L OFDM symbols in the time domain, as shown in Figure 3, F(n, t) Represents the precoding matrix on the nth subcarrier of the tth OFDM symbol, which is obtained by feedback from the receiver;

Represents the precoding matrix on the nth subcarrier of the tth OFDM symbol obtained by interpolation, that is, the sender interpolates the precoding matrix of the subcarrier at the pilot location fed back by the receiver to obtain the precoding of subcarriers at other non-pilot locations matrix.

202. Send OFDM symbols including pilot sequences;

When the system is initialized, the data is transmitted in units of blocks, and the pilot sequence is evenly inserted in the first OFDM symbol of the first data block every certain interval, and the data block containing the pilot sequence is sent to the receiver .

203. The receiver obtains the estimated channel value at the pilot according to the pilot sequence;

In this embodiment, after the receiver obtains the data block, it reads the pilot sequence from the first OFDM symbol of the data block, and obtains the channel estimate at the pilot according to the pilot sequence. The specific acquisition process It is a prior art, and will not be described in detail here.

After the receiver acquires at least two pilot sequences, it acquires channel information corresponding to at least two time points according to the pilot sequences, performs statistics on the channel information to obtain a transition probability matrix, and according to the transition probability The matrix constructs a codebook for classifying codeword states, and at the same time sends the transition probability matrix to the sender, and the sender also constructs a codebook for classifying codeword states based on the transition probability matrix.

204. The receiver selects the subcarrier precoding matrix at the pilot and feeds back the matrix codeword information; after obtaining the channel estimate at the pilot, the receiver selects the precoding matrix of the subcarrier at the pilot according to the preset precoding matrix selection rules , and feed back the codeword information corresponding to the selected precoding matrix in the codeword state classification codebook to the sender.

First, the precoding strategy in the embodiment of the present invention is described. The Grassmann precoding strategy is adopted in the embodiment of the present invention. Specifically, in the MIMO-OFDM system, when the length of the cyclic prefix is greater than the channel length, the inter-symbol interference The influence can be completely eliminated, then the input-output relationship on the nth carrier is Y _n ＝H _n X _n +N _n , where N _n is an N _r- dimensional column whose elements obey the CN(0, N ₀ ) distribution vector, X _n is the N _t- dimensional column vector of transmitted signals, and the total transmitted power is ε _s , namely $\mathrm{E.}\left[x_{\mathrm{no}}^h{x}_{\mathrm{no}}\right]\≤{\mathrm{\ϵ}}_{\mathrm{the\; s}},$ Y _n is the N _r dimensional received signal column vector, and H _n is the N _r ×N _t frequency domain channel matrix. In a linear precoding system, X _n can be expressed as X _n =F _n S _n , where F _n is an N _t ×M precoding matrix on subcarrier n, and S _n is an M-dimensional pretransmission signal column vector , since the maximum multiplexing number of the MIMO system is the minimum number of transmitting and receiving antennas, M≤min{N _t , N _r }. Due to the limited transmit power, the precoding must satisfy $f_{\mathrm{no}}^h{f}_{\mathrm{no}}=I_m,$ Therefore, F _n is selected as a N _t ×M quasi-unitary matrix. If the receiver uses the minimum mean square error detector, the receiver multiplies the received signal Y _n by a detection matrix G _n , at this time $G_{\mathrm{no}}=[f_{\mathrm{no}}^h{h}_{\mathrm{no}}^h{h}_{\mathrm{no}}{f}_{\mathrm{no}}+({\mathrm{MN}}_0/{\mathrm{\ϵ}}_{\mathrm{the\; s}})I_m{]}^{-1}{f}_{\mathrm{no}}^h{h}_{\mathrm{no}}^h,$ Then the detected signal is ${\hat{x}}_{\mathrm{no}}=G_{\mathrm{no}}{Y}_{\mathrm{no}}.$

The precoding strategy in this embodiment is introduced above. It can be understood that the above introduction is only for ease of understanding, and other precoding strategies can also be used in practical applications, which are not limited here.

The following describes the rules for selecting the precoding matrix in this embodiment. There are many specific rules for selecting the precoding matrix, which are not limited here, and only two examples are used for illustration. In the following two examples and subsequent embodiments Both take the precoding matrix selected by the rule of minimizing the system bit error rate and/or maximizing the system capacity as an example for illustration:

(1) Minimize the system bit error rate rule:

When the system adopts different modulation methods, the bit error rate of the system can be written as a function of the signal-to-noise ratio of the received signal, so the maximum signal-to-noise ratio of the received signal can be used as the criterion for selecting the precoding matrix. The signal-to-noise ratio of the received signal of the minimum mean square error detection receiver can be expressed as

${\mathrm{<span\; class="notranslate">\; SNR</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\frac{{\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}}{{\mathrm{<span\; class="notranslate">\; MN</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; +</span><span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; MN</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}<span\; class="notranslate">\; /</span>{\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}{<span\; class="notranslate">\; |</span>}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}$

Then the standard of maximizing the signal-to-noise ratio of the received signal is to select a precoding matrix F _n such that

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>\underset{\underset{\mathrm{<span\; class="notranslate">\; f</span>}}{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; f</span>}}{\mathrm{<span\; class="notranslate">\; arg</span>}\mathrm{<span\; class="notranslate">\; max</span>}}\mathrm{<span\; class="notranslate">\; SN</span>}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>$

by the formula ${\hat{x}}_{\mathrm{no}}=G_{\mathrm{no}}{Y}_{\mathrm{no}}$ It can be seen that the precoding matrix based on the criterion of maximizing the signal-to-noise ratio of the received signal is not unique. For any U∈U(M, M), SNR _n (F _n )=SNR _n (F _n U), that is, the precoding After the matrix is multiplied by an M-order unitary matrix to the right, the signal-to-noise ratio of the received signal remains unchanged, which means that several precoding matrices on the subcarrier at the pilot can be obtained.

(2) Maximize system capacity rules:

For a given H _n and F _n , the system capacity of an incoherent complex Gaussian source is

${\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; lo</span>}{\mathrm{<span\; class="notranslate">\; g</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; det</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; +</span>\frac{{\mathrm{<span\; class="notranslate">\; \&epsiv;</span>}}_{\mathrm{<span\; class="notranslate">\; the\; s</span>}}}{{\mathrm{<span\; class="notranslate">\; MN</span>}}_{\mathrm{<span\; class="notranslate">\; 0</span>}}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; )</span>$

Maximizing the system capacity criterion is to choose a precoding matrix F _n such that

${\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>\underset{{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; \&Element;</span>\mathrm{<span\; class="notranslate">\; f</span>}}{\mathrm{<span\; class="notranslate">\; arg</span>}\mathrm{<span\; class="notranslate">\; max</span>}}{\mathrm{<span\; class="notranslate">\; C</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ,</span>\mathrm{<span\; class="notranslate">\; i</span>}}<span\; class="notranslate">\; )</span>$

Similarly, the precoding matrix based on maximizing system capacity is not unique, it is also invariant, for any U∈U(M, M), C _n (F _n )=C _n (F _n U) There are several precoding matrices.

Several precoding matrices can be obtained from the two precoding matrix selection criteria introduced above, and these precoding matrices have the same column space. Therefore, the selection criteria of the precoding matrix depends on the column space of F _n .

The set of M-dimensional subspaces spanned by all matrices on U(N _r , M) is the Grassmann manifold, denoted as G(N _r , M). Since the Grassmann manifold is a quotient space, any point F∈G(N _r , M) in the space represents the equivalent class of an N _r ×M orthogonal matrix. If the column spaces of the two matrices are the same, then the two The matrices are equivalent. For example, the equivalent class of F can be expressed as [F]={FU: U∈U(M, M)}. Therefore, the precoding matrix is a point on the Grassmann manifold.

When the number of codeword information N _B is given, the codebook design method based on Grassmann manifold is to maximize the minimum codeword information distance. In the Grassmann manifold, there are many definitions of the distance between subspaces, and the three main subspace distances are defined as: chord distance, norm projection distance and Fubini-Study distance. The chordal distances of subspaces P _F1 and P _F2 are defined as

The norm projection distance of subspace P _F1 and P _F2 is

${\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; proj</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; |</span><span\; class="notranslate">\; |</span>}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; =</span>\sqrt{\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; l</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}^{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; \{</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; *</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; \}</span>}$

And the Fubini-Study distance of subspace P _F1 and P _F2 is defined as

${\mathrm{<span\; class="notranslate">\; d</span>}}_{\mathrm{<span\; class="notranslate">\; FS</span>}}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; ,</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; arccos</span>}<span\; class="notranslate">\; |</span>\mathrm{<span\; class="notranslate">\; det</span>}<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}^{<span\; class="notranslate">\; *</span>}{\mathrm{<span\; class="notranslate">\; f</span>}}_{\mathrm{<span\; class="notranslate">\; 2</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; |</span>$

即由不同的距离定义，可以得到不同的码书。make $S=\{P_{f_1},P_{f_2},K,P_{f_{N_B}}\}$ is the subspace set of codeword information, where P _Fi is the column space of codeword information F _i , then the codebook design criterion based on Grassmann manifold is

That is, defined by different distances, different codebooks can be obtained.

After the precoding matrix of the subcarrier at the pilot is determined, the codeword information corresponding to the precoding matrix is queried in the previously obtained codeword state classification codebook, and the codeword information is fed back to the sender.

It can be understood that the above two precoding matrix selection rules can also be used in combination, and the specific manner is not limited here.

205. The sender interpolates precoding matrices of all subcarriers according to the codeword information;

After the sender obtains the codeword information of the precoding matrix of the subcarrier at the pilot frequency fed back by the receiver through the feedback channel, it finds the corresponding precoding matrix in the codebook according to the codeword information, and according to the frequency domain reduction feedback algorithm, the internal Insert the precoding matrix on other carriers of this OFDM symbol. Since the channel is block fading, the precoding matrix on other OFDM symbols in this data block is the same as the precoding matrix on the corresponding subcarrier of the pilot symbol, so The sender can obtain the precoding matrix of all subcarriers according to the codeword information fed back by the receiver.

In this embodiment, the process of the sender interpolating the precoding matrix on other carriers of the OFDM symbol according to the received codeword information is similar to the process in the frequency domain reduction feedback mode in the prior art, and will not be repeated here.

206. The receiver calculates the precoding matrix of the subcarrier at each pilot;

In this embodiment, when the receiver receives the next symbol, it calculates the precoding matrix of the subcarrier at each pilot in the symbol, and feeds back the corresponding codeword information to the sender according to the time domain reduction feedback algorithm.

In this embodiment, if the symbol received by the receiver is not the first symbol, the precoding matrix will be calculated for each pilot subcarrier in the symbol, the specific calculation method and the precoding matrix of the aforementioned calculation pilot subcarrier method is similar.

The time domain reduction feedback algorithm described in this embodiment is specifically introduced below:

In the whole system, the change of the precoding matrix on each subcarrier in the time domain is modeled as a finite-state discrete-time Markov chain, and its state J _n is related to each codeword information in the codebook There is a one-to-one correspondence with F _n .

An important parameter of the discrete-time finite-state Markov chain is its transition probability matrix Pr(J _n =i|J _n =j). Since the codebook is a quantization of the precoding matrix, it is difficult to give the transition probability The closed-form expression of the matrix, but the transition probability matrix of the codebook can be obtained through Monte Carlo simulation. When the codeword is in state transition, it is transferred from one state to several fixed states, not all states. Therefore, in the time domain feedback, since the sender knows the precoding matrix at the previous pilot, the receiver only needs to send the codeword information of the state where the previous codeword state may transition, and the amount of feedback can be reduced.

The specific process can be as follows: the receiver determines all transition states corresponding to the precoding matrix of the subcarrier at the pilot according to the transition probability matrix obtained by previous statistics, and queries the codeword information corresponding to the transition state according to the codeword state classification code book , that is, determine the codeword information to which the codeword information may be transferred, and feed back the codeword information to the sender.

It is understandable that, under the premise of not affecting the system performance as much as possible, in order to further improve the effect of reducing feedback, the specific codeword information transfer probability can also be obtained, and only when the codeword information transfer probability is greater than a certain threshold value Only when the codeword information is fed back to the sender, that is, the codeword information that is unlikely to be transferred is not fed back, can further improve the effect of reducing feedback.

On the basis of the above, the codeword information encoding process can also be improved, that is, the codeword information is divided according to the size of the transition probability, and the codeword information with a high transition probability is processed with fewer bits according to the Huffman method. For codeword information with low transition probability, according to the Huffman method, more bits are used for coding, and when there are many transition states, the effect of reducing feedback can be further improved. The above-mentioned Huffman coding method is only It is an optional scheme, and different encoding schemes can be selected according to the actual situation in specific applications.

207. Feed back the corresponding precoding matrix according to the time-domain reduction feedback algorithm.

After the time-domain reduction feedback algorithm is performed, the obtained transition codeword information is fed back to the sender.

After step 207, the sender can obtain the precoding matrices on all subcarriers, multiply these precoding matrices by the data domain data in the data frame, and send the precoded data to the receiver.

After the receiver receives the data, since the receiver knows the channel information on each subcarrier, the receiver also uses the frequency domain reduction feedback algorithm to interpolate the precoding matrix on all subcarriers, so that the received data can be Perform zero-forcing or minimum mean square error detection to restore the original data.

The following is an overall introduction to the time-frequency joint reduction feedback method:

In this method, when the system is initialized, the sender continuously transmits multiple pilot sequences, and the receiver obtains channel information at multiple time points according to these pilot sequences, and statistically analyzes these information to obtain the transition probability of the codebook Matrix, and pass this transition probability matrix to the sender. After knowing the probability transition matrix, the sending and receiving ends construct a codebook classified according to the codeword state at the previous moment.

When starting to transmit data, the sender forms a data frame containing the pilot sequence, signaling field and data field with the bit rate to be sent. Each data frame is composed of ΔT OFDM symbols, and every ΔF of the first OFDM symbol of the data frame Subcarriers are evenly inserted into the pilot sequence.

After receiving the pilot symbols, the receiver uses the channel estimation method to estimate the channel information on the carrier where the pilot is located. If the current transmission is the first data frame, since there is no codeword information at the previous moment, the time domain is not used. To reduce the feedback algorithm, it is necessary to feed back the sequence numbers of the precoding matrices of these carriers in the traditional codebook to the sender; if the current transmission is not the first data frame, the receiver classifies the codebook according to the codeword state and calculates the derived The codeword information of the precoding matrix on the subcarrier where the frequency is located, and feeds back the codeword information of the precoding matrix to the sender.

If the feedback channel is error-free and delay-free, after receiving the feedback information, the sender finds the precoding matrix on these subcarriers from the codebook, and then interpolates the non-pilot subcarriers according to the frequency domain reduction feedback algorithm The precoding matrix on the carrier, so that the precoding matrix on all subcarriers can be obtained, and after multiplying these precoding matrices with the data field data in the data frame, these precoded data are sent to the receiver.

Since the receiver knows the channel information on each subcarrier, and also uses the frequency domain reduction feedback algorithm to interpolate the precoding matrix on the non-pilot subcarrier, it can perform zero-forcing or minimum mean square error on the received data Detect and restore the original data.

In the above embodiment, when the cyclic prefix of the OFDM system is greater than the channel length, there is a correlation between adjacent subcarriers, and the correlation between subcarriers can be measured by frequency coherence coefficients. The closer the subcarriers are, the closer the coherence coefficients are. The larger the value, the greater the correlation. Since the precoding matrix is a function of the channel response, the precoding matrix has correlation in the frequency domain. The sender can use the precoding matrix fed back from some subcarriers to calculate the precoding matrix. Similarly, the channel also has correlation in the time domain, and the correlation in the time domain is related to the Doppler frequency spread of the channel. Internal correlation, otherwise, the channel changes slowly and remains correlated for a long time. Since the channel has correlation in the time domain, there is also a correlation between the precoding matrices of two adjacent data blocks. Knowing the previous data The precoding matrix of the block, the range of the possible precoding matrix of the next data block can be greatly reduced, so that the feedback amount of the system can be reduced.

The above introduces the time-frequency joint reduction feedback method, which is mainly applied to the situation where many subcarriers need to be modulated. If the number of subcarriers modulated by OFDM is relatively small, the time domain reduction feedback method can be used:

2. Time domain reduction feedback method:

When the number of subcarriers of OFDM modulation is relatively small compared with the number of subcarriers in the aforementioned time-frequency joint reduction feedback method, only the time-domain reduction feedback algorithm can be used, and at this time it can achieve approximately the same performance as the time-frequency full feedback algorithm.

In this method, the sending and receiving ends transmit in units of data frames, and each data frame is divided into three parts: pilot sequence, using Chu sequence as pilot, it is understandable that other similar sequences can also be used as pilot Sequence, used for frame synchronization and channel estimation; signaling domain, including the number of subcarriers used by the OFDM symbol in the data domain of the current frame, as well as the frame length ΔT of the current frame, the frequency domain interval ΔF of the pilot, and the modulation and coding method and other fields; the data field is the data payload to be transmitted in the current frame.

Among them, both the pilot sequence and the signaling field are transmitted with a fixed number of subcarriers and a reliable modulation method (such as BPSK), and the signaling field can also be properly coded to ensure the accuracy of its reception.

In this way, the sender and the receiver assist in the completion. The sender is responsible for the preprocessing of the transmitted data frame, including the insertion of the pilot, the formation of the signaling field and the selection of the precoding matrix according to the feedback information. The receiver mainly implements channel estimation, according to The time domain reduction feedback algorithm calculates the information that needs to be fed back and transmits the precoding information to the sender.

For ease of understanding, detailed description will be given below. Please refer to FIG. 4. The second embodiment of the feedback amount control method in the embodiment of the present invention includes:

401. Receive the data frame sent by the sender;

The specific system initialization process is the same as that in the foregoing embodiments, and will not be repeated in this embodiment, that is, this embodiment directly starts with the step of receiving a data frame.

402. Calculate a precoding matrix of subcarriers at each pilot;

After receiving the data frame sent by the sender, the receiver calculates the precoding matrix of each subcarrier in the data frame. The specific calculation process is consistent with the process of calculating the precoding matrix described in the above embodiment, and will not be repeated here repeat.

403. Obtain a transition state corresponding to the precoding matrix;

During the system initialization process, the receiver calculates the transition probability matrix, that is, it can know the possible transition status of the codeword information of the precoding matrix of the current subcarrier, that is, to which codeword information the current codeword information may be transferred to.

In the system of this embodiment, the change of the precoding matrix on each subcarrier in the time domain is modeled as a finite-state discrete-time Markov chain, and its state J _n is related to each One code word information F _n corresponds to one to one.

An important parameter of the discrete-time finite-state Markov chain is its transition probability matrix Pr(J _n =i|J _n =j). Since the codebook is a quantization of the precoding matrix, it is difficult to give the transition probability The closed-form expression of the matrix, but the transition probability matrix of the codebook can be obtained through Monte Carlo simulation. Please refer to Figure 5. S1 to S6 in Figure 5 are different codeword information in a codebook. It can be seen from Figure 5 that when the codeword information is in state transition, it will only transfer from one state to several fixed states, and Not all states. For example, S4 in Figure 5 may be transferred to S6 or S3, but not to other codeword information. Therefore, during time domain feedback, since the sender knows the precoding matrix at the previous pilot, The receiver only needs to send the codeword information of the state that the previous codeword state may transfer, so that the amount of feedback can be reduced.

For example, the number of codes is N _B or 2 ^B , and the number of possible transition states of a certain codeword information is T. If the time-domain reduction feedback algorithm is not used, the feedback amount is N _c × log ₂ (B), and this After this algorithm, the amount of feedback becomes N _c ×log ₂ (T). When the number of codebooks is large, the amount of feedback can be effectively reduced.

404. Obtain transfer codeword information corresponding to the transfer state according to preset rules;

In this embodiment, the receiver needs to feed back the codeword information to which the current codeword information may be transferred to the sender, that is, the receiver determines all transition states corresponding to the precoding matrix of the subcarrier at the pilot according to the transition probability matrix, and The codeword information corresponding to the transition state is queried according to the codeword state classification codebook.

The receiver can feed back the queried codeword information to the sender. It is understandable that in order to further reduce the amount of feedback, the transition probability of the acquired codeword information can be set and a threshold value is set, and only the transition probability greater than the threshold The code word information of the limit value is fed back to the sender;

When there are many transition states, in order to further reduce the amount of feedback, the codeword information can also be classified. For the codeword information with a high transition probability, use less bits to encode according to the Huffman method. For the codeword information with a small transition probability The information is coded with more bits according to the Huffman method, and when there are many transition states, the effect of reducing feedback can be further improved.

405. Feed back the transfer codeword information to the sender;

After determining the transfer codeword information, the receiver feeds back the determined transfer codeword information to the sender.

406. The sender performs data processing and sends the processed data.

After step 406, since the receiver has calculated the precoding matrices of all subcarriers, the sender can know the precoding matrices on all subcarriers, and after multiplying these precoding matrices with the data field data in the data frame, these The precoded data is sent to the receiver.

After receiving the data, the receiver can perform zero-forcing or minimum mean square error detection on the received data due to the calculation of the precoding matrix of all subcarriers, thereby recovering the original data.

The following is an overall introduction to the time domain reduction feedback method:

When the system is initialized, the sender sends multiple pilot sequences continuously, and the receiver obtains channel information at multiple time points according to these pilot sequences, and statistically analyzes these information to obtain the transition probability matrix of the codebook, and put this The transition probability matrix is passed to the sender. After knowing the probability transition matrix, the sending and receiving ends construct a codebook classified according to the codeword state at the previous moment.

When starting to transmit data, the sender forms a data frame including a pilot sequence, a signaling field and a data field with the bit rate to be sent, and each data frame consists of ΔT OFDM symbols.

After receiving the pilot symbols and signaling information, the receiver estimates the channel information of each subcarrier through channel estimation. If the current transmission is the first data frame, since it does not have the codeword information of the previous moment, it will not The time-domain reduction feedback algorithm is used instead of the traditional time-frequency full feedback algorithm; if the current transmission is not the first data frame, the receiver classifies the codebook according to the codeword state and calculates the precoding matrix on each subcarrier codeword information, obtain transfer codeword information of the codeword information, and feed back the transfer codeword information to the sender.

If the feedback channel is error-free and delay-free, after receiving the feedback information, the sender finds the precoding matrix on each subcarrier from the codebook, multiplies it with the data field data in the data frame, and sends it to the receiver These precoded data.

Since the receiving side knows the channel information and precoding matrix on each subcarrier, it can perform zero-forcing or minimum mean square error detection on the received data to restore the original data.

In the embodiment of the present invention, since the channel has correlation in the time domain, the codeword transfers from one state to several fixed states, not all states when performing state transition in the time domain. In the embodiment of the present invention, after receiving the data frame, the receiver determines the precoding matrix of each pilot subcarrier in the data frame, and obtains the transition corresponding to the precoding matrix according to the preset codeword state classification codebook The codeword information feeds back the transition codeword information to the sender, so it is not necessary to feed back codeword information of all transition states, thereby reducing the amount of system feedback.

Please refer to FIG. 6, the communication system embodiment in the embodiment of the present invention includes:

sender 601 and receiver 602,

The sender 601 is configured to send a data frame to the receiver 602, and receive the fed back transition codeword information;

The receiver 602 is used to receive the data frame sent by the sender 601, determine the precoding matrix of each pilot subcarrier in the data frame, and obtain the precoding matrix according to the preset codeword state classification codebook. transfer codeword information corresponding to the matrix, and feed back the transfer codeword information to the sender 601 .

In this embodiment, the sender 601 is further configured to send an OFDM symbol including a pilot sequence to the receiver 602, and query the corresponding precoding matrix in the codebook state classification codebook according to the fed back codeword information, Interpolate the precoding matrix on the non-pilot subcarrier according to the precoding matrix to obtain the precoding matrix of all subcarriers, and multiply the precoding matrix of all subcarriers with the data in the preset data field Obtain precoded data, and send the precoded data to the receiver 602;

The receiver 602 is further configured to receive the OFDM symbol sent by the sender 601, obtain an estimated channel value at the pilot position according to the pilot sequence in the OFDM symbol, and obtain a pre-defined value of the subcarrier at the pilot position according to the estimated value. An encoding matrix, querying the codeword information of the precoding matrix in the codeword state classification codebook, and feeding back the codeword information to the sender 601 .

Please refer to FIG. 7, the first embodiment of the data transmission network element in the embodiment of the present invention includes:

A sending unit 701, configured to send data frames and OFDM symbols;

A codeword information receiving unit 702, configured to receive codeword information after the sending unit 701 sends the data frame or OFDM symbol;

The query unit 703 is configured to query the corresponding precoding matrix in the preset codeword state classification codebook according to the received codeword information;

The precoding matrix processing unit 704 is configured to interpolate precoding matrices on non-pilot subcarriers according to the queried precoding matrix to obtain precoding matrices of all subcarriers.

The data transmission network element in this embodiment also includes:

The data processing unit 705 is configured to multiply the precoding matrix of all subcarriers by the data in the preset data field to obtain precoding data, and send the precoding data.

The data transmission network element described in the foregoing embodiments may serve as the sender in the foregoing method embodiments.

Please refer to FIG. 8, the second embodiment of the data transmission network element in the embodiment of the present invention includes:

The frequency domain feedback control unit 801 is configured to receive an OFDM symbol, obtain an estimated channel value at the pilot location according to the pilot sequence in the OFDM symbol, and obtain a precoding matrix of the subcarrier at the pilot location according to the estimated value, and then Querying the codeword information of the precoding matrix in the codeword state classification codebook;

The time-domain feedback control unit 803 is configured to receive the data frame, determine the precoding matrix of each pilot subcarrier in the data frame, and obtain the transition code corresponding to the precoding matrix according to the preset codeword state classification codebook word information;

The feedback unit 802 is configured to feed back the codeword information obtained by the frequency domain feedback control unit 801 and the transfer codeword information obtained by the time domain feedback control unit 803 to the sender.

The data transmission network element described in the foregoing embodiments may serve as the receiver in the foregoing method embodiments.

Those of ordinary skill in the art can understand that all or part of the steps in the method of the above-mentioned embodiments can be completed by instructing related hardware through a program. The program can be stored in a computer-readable storage medium, and the program can be executed when executed , including the following steps:

Receive the data frame sent by the sender;

determining a precoding matrix of subcarriers at each pilot in the data frame;

Acquiring transition codeword information corresponding to the precoding matrix according to a preset codeword state classification codebook;

Feedback the transfer codeword information to the sender.

The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, and the like.

A feedback amount control method, a communication system and related equipment provided by the present invention have been introduced in detail above. For those skilled in the art, based on the idea of the embodiment of the present invention, there will be some specific implementation methods and application ranges. Changes, in summary, the contents of this specification should not be construed as limiting the present invention.

Claims (11)

1. a feedback quantity control method is characterized in that, comprising:

Receive the Frame that transmit leg sends;

Determine the pre-coding matrix of each pilot tone virgin carrier wave in the described Frame;

Obtain transfer codeword information corresponding to described pre-coding matrix according to the code word state classification code book that presets;

Feed back described transfer codeword information to described transmit leg;

Comprise before the step of the Frame that described reception transmit leg sends:

Receive at least two pilot frequency sequences that transmit leg sends;

Obtain channel information at least two corresponding time points according to described pilot frequency sequence;

Described channel information added up obtain transition probability matrix;

Make up code word state classification code book according to described transition probability matrix, send described transition probability matrix to described transmit leg.

2. method according to claim 1 is characterized in that, describedly comprises after described transmit leg sends the step of described transition probability matrix:

Transmit leg makes up code word state classification code book according to the transition probability matrix that receives.

3. method according to claim 1 and 2 is characterized in that, the step of the pre-coding matrix of each pilot tone virgin carrier wave comprises in described definite described Frame:

The recipient determines the pre-coding matrix at each pilot tone place in the described Frame according to minimization system error rate criterion and/or maximized system capacity criterion.

4. method according to claim 1 and 2 is characterized in that, the step that the code word state classification code book that described basis presets obtains transfer codeword information corresponding to described pre-coding matrix comprises:

The recipient determines all transfering states corresponding to pre-coding matrix of pilot tone virgin carrier wave according to described transition probability matrix;

Inquire about codeword information corresponding to described transfering state according to described code word state classification code book;

With described codeword information as shifting codeword information.

5. method according to claim 4 is characterized in that, describedly comprises before inquiring about the step of codeword information corresponding to described transfering state according to described code word state classification code book:

Obtain the transition probability of described all transfering states according to described transition probability matrix;

Described step of inquiring about codeword information corresponding to described transfering state according to described code word state classification code book comprises:

Determine that described all transfering state transition probabilities are more than or equal to the selected transfering state of the threshold value that presets;

Inquire about codeword information corresponding to described selected transfering state according to described code word state classification code book.

6. method according to claim 4 is characterized in that, describedly comprises before inquiring about the step of codeword information corresponding to described transfering state according to described code word state classification code book:

Obtain the transition probability of described all transfering states according to described transition probability matrix;

Described step of inquiring about codeword information corresponding to described transfering state according to described code word state classification code book comprises:

According to described transition probability all transfering states are divided into the first state class and the second state class, the mean transferred probability of described the first state class is greater than the mean transferred probability of described the second state class;

Adopt the huffman coding mode that described the first state class and described the second state class are encoded, the bit number that described the first state class coding is adopted is less than the bit number that described the second state class coding is adopted.

7. method according to claim 1 and 2 is characterized in that, described method also comprises:

Receive first orthogonal frequency division multiplex OFDM symbol that transmit leg sends;

Obtain the channel estimation value at pilot tone place according to the pilot frequency sequence in described first OFDM symbol;

Obtain the pre-coding matrix of described pilot tone virgin's carrier wave according to described estimated value;

The codeword information of the described pre-coding matrix of inquiry in described code word state classification code book;

Feed back described codeword information to described transmit leg.

8. method according to claim 7 is characterized in that, describedly comprises after described transmit leg feeds back the step of described codeword information:

Transmit leg is according to the pre-coding matrix of described codeword information in described code word state classification code book inquiry correspondence;

Go out pre-coding matrix on the non-pilot sub-carrier to know the pre-coding matrix of all subcarriers according to described pre-coding matrix interpolation;

The pre-coding matrix of described all subcarriers and data in the data field that presets are multiplied each other obtain pre-code data;

Send described pre-code data to the recipient.

9. a communication system is characterized in that, comprising:

Transmit leg and recipient,

Described transmit leg is used for sending Frame to described recipient, receives the transfer codeword information of feedback;

Described recipient is used for receiving the Frame that described transmit leg sends, determine the pre-coding matrix of each pilot tone virgin carrier wave in the described Frame, obtain transfer codeword information corresponding to described pre-coding matrix according to the code word state classification code book that presets, feed back described transfer codeword information to described transmit leg;

Described recipient also is used for receiving at least two pilot frequency sequences that transmit leg sends, obtain channel information at least two corresponding time points according to described pilot frequency sequence, described channel information added up obtain transition probability matrix, make up code word state classification code book according to described transition probability matrix, send described transition probability matrix to described transmit leg.

10. communication system according to claim 9 is characterized in that,

Described transmit leg also is used for sending the OFDM symbol that comprises pilot frequency sequence to described recipient, codeword information according to feedback is inquired about corresponding pre-coding matrix at described code word state classification code book, go out pre-coding matrix on the non-pilot sub-carrier to know the pre-coding matrix of all subcarriers according to described pre-coding matrix interpolation, the pre-coding matrix of described all subcarriers and data in the data field that presets are multiplied each other obtains pre-code data, sends described pre-code data to the recipient;

Described recipient also is used for receiving the OFDM symbol that transmit leg sends, obtain the channel estimation value at pilot tone place according to the pilot frequency sequence in the described OFDM symbol, obtain the pre-coding matrix of described pilot tone virgin's carrier wave according to described estimated value, the codeword information of the described pre-coding matrix of inquiry is fed back described codeword information to described transmit leg in described code word state classification code book.

11. a transfer of data network element is characterized in that, comprising:

The frequency domain feedback control unit, be used for receiving the OFDM symbol, obtain the channel estimation value at pilot tone place according to the pilot frequency sequence in the described OFDM symbol, obtain the pre-coding matrix of described pilot tone virgin's carrier wave according to described estimated value, the codeword information of the described pre-coding matrix of inquiry in described code word state classification code book;

The time domain feedback control unit is used for receiving data frames, determines the pre-coding matrix of each pilot tone virgin carrier wave in the described Frame, obtains transfer codeword information corresponding to described pre-coding matrix according to the code word state classification code book that presets;

Feedback unit is used for feeding back the transfer codeword information that codeword information that described frequency domain feedback control unit obtains and described time domain feedback control unit obtain to transmit leg.

Images