CN102215193B - Frequency-domain equilibrium method and device - Google Patents

Abstract

The invention discloses a kind of frequency-domain equilibrium method and device, wherein, frequency-domain equilibrium method includes: use LMMSE equalization algorithm that the uplink antenna data received are carried out frequency domain equalization；In frequency domain equalization process, according to LMMSE algorithm, matrix operations therein is split, the matrix operations after executed in parallel fractionation.By the present invention, solve existing equalization processing method time delay long, the problem that user's experience is low, and then reached minimizing equalization processing method time delay, improve the effect of user's experience.

Description

Frequency Domain Equalization Method and Device

technical field

The present invention relates to the communication field, in particular, to a frequency domain equalization method and device for an LTE (Long Term Evolution, long term evolution) system.

Background technique

OFDM (Orthogonal Frequency Division Multiplexing, Orthogonal Frequency Division Multiple Access) is a special MCM (Multi-carrier Modulation, multi-carrier modulation) technology, its idea is to decompose a high-speed data stream into many low-rate sub-streams Data is transmitted on multiple sub-channels in parallel, providing a way for data to be transmitted at a higher rate on a channel with a greater delay. The MIMO (Multi-Input Multi-Output) system configures multiple antennas at the transceiver end and uses spatial channels to achieve diversity gain or multiplexing gain, which greatly improves the performance of the communication system without increasing the system bandwidth. Capacity and spectrum efficiency are key technologies for next-generation wireless communication systems. Combining OFDM and MIMO can not only achieve high spectrum utilization, but also resist the influence of multipath fading channels.

However, affected by factors such as size and cost, the terminal in the LTE phase generally has one antenna, so that sufficient capacity gain cannot be obtained from multiple transmit antennas. 3GPP LTE provides a solution which is Mu-MIMO (Multi-User MIMO), that is, multi-user MIMO. Multi-user MIMO enables single-antenna users to share the antennas of other users through cooperative communication, and use the same time-frequency resources for communication to improve system performance.

Although OFDM has had some success in increasing the amount of data transmitted over a wireless link through multi-user MIMO, since the signal is transmitted over radio waves, in typical operating environments there are multipaths in the signal received by the base station, which in turn Delay spread due to multipath issues. At the same time, during the transmission process from the sending end to the receiving end, the signal will also receive various effects such as nonlinearity, rain attenuation, and multipath, which will distort the transmitted signal, resulting in waveform distortion, and causing intersymbol interference. At present, equalization technology is generally used to solve inter-symbol interference, so as to compensate signal distortion caused by channel parameter changes, offset channel signal transmission attenuation, and at the same time, it can effectively resist inter-symbol interference, reduce bit errors, and increase transmission rate. There are two basic approaches to equalization, frequency domain equalization and time domain equalization. Frequency domain equalization is to make the frequency transfer function of the entire system meet the conditions for distortion-free transmission. The time-domain equalization is directly considered from the time response, so that the impulse response of the entire system including the equalization module meets the condition of no ISI (Inter-Symbol Interference, inter-symbol interference).

Currently, there are many studies on frequency domain equalization (FDE). Usually, the operation in the time domain is transformed into the frequency domain to compensate the damage caused by the multipath channel with long impulse response, and at the same time reduce the complexity of the receiver. But at the same time, adopting this method leads to a long equalization processing delay time, thereby reducing user experience.

Contents of the invention

The main purpose of the present invention is to provide a frequency domain equalization method and device to at least solve the problems of long time delay and low user experience in the existing equalization processing method.

According to one aspect of the present invention, a frequency domain equalization method is provided, including: using the LMMSE equalization algorithm to perform frequency domain equalization on the received uplink antenna data; Divide, and execute the matrix operation after splitting in parallel.

According to another aspect of the present invention, a frequency domain equalization device is provided, including: an equalization module, configured to perform frequency domain equalization on received uplink antenna data using an LMMSE equalization algorithm; the equalization module includes: a splitting module, configured to In the frequency domain equalization process, the matrix operation is split according to the LMMSE algorithm, and the split matrix operation is executed in parallel.

Through the present invention, in the process of frequency domain equalization, according to the LMMSE algorithm, the matrix operation is appropriately split, and the split matrix operation is calculated in parallel, and the split and parallel calculation are solved. The existing equalization processing method has long time delay and low user experience, and then achieves the effect of reducing the time delay of the equalization processing method and improving user experience.

Description of drawings

The accompanying drawings described here are used to provide a further understanding of the present invention and constitute a part of the application. The schematic embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations to the present invention. In the attached picture:

FIG. 1 is a flowchart of steps of a frequency domain equalization method according to Embodiment 1 of the present invention;

FIG. 2 is a flow chart of steps of a frequency domain equalization method according to Embodiment 2 of the present invention;

FIG. 3 is a flow chart of steps of a frequency domain equalization method according to Embodiment 3 of the present invention;

FIG. 4 is a structural block diagram of a frequency domain equalization device according to Embodiment 4 of the present invention;

FIG. 5 is a schematic structural diagram of a frequency domain equalization device according to Embodiment 5 of the present invention;

FIG. 6 is a schematic diagram of a calculation flow of the frequency domain equalization device of the embodiment shown in FIG. 5;

Fig. 7 is a schematic diagram of a configurable calculation module of the frequency domain equalization device of the embodiment shown in Fig. 5 .

detailed description

Hereinafter, the present invention will be described in detail with reference to the drawings and examples. It should be noted that, in the case of no conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

In LTE, it is necessary to support TDD (Time Division Duplex, time division duplex) and FDD (Frequency Division Duplex, frequency division duplex) two modes, support 1, 2, 4, 8 antenna configurations and two-user virtual MIMO, the present invention implements The example provides a method and device based on LTE uplink receiving LMMSE (linear minimum mean square error) equalization, which is used on the LTE base station side to complete the LTE uplink according to the received antenna data, channel estimation H and the inverse matrix of the noise covariance matrix Rn Receive LMMSE equalization.

Embodiment one

Referring to FIG. 1 , it shows a flowchart of steps of a frequency domain equalization method according to Embodiment 1 of the present invention.

The frequency domain equalization method of this embodiment includes the following steps:

Step S102: using the LMMSE equalization algorithm to perform frequency domain equalization on the received uplink antenna data;

In this embodiment, the uplink antenna data may be multiple-input multiple-output MIMO antenna system data, single-input multiple-output SIMO antenna system data, or single-input single-output SISO antenna system data.

Step S104: During the frequency domain equalization process, the matrix operation is split according to the LMMSE algorithm, and the split matrix operation is executed in parallel.

In related technologies, the equalization processing method has long delay and low user experience. Through this embodiment, in the process of performing frequency domain equalization, according to the LMMSE algorithm, the matrix operation is appropriately split, and the split matrix operation is calculated in parallel, and through the split and parallel calculation, the solution The problem of long time delay and low user experience of the existing equalization processing method is solved, and the effect of reducing the time delay of the equalization processing method and improving user experience is achieved.

Embodiment two

Referring to FIG. 2 , it shows a flowchart of steps of a frequency domain equalization method according to Embodiment 2 of the present invention.

The frequency domain equalization method of this embodiment includes the following steps:

Step S202: Use the LMMSE equalization algorithm to perform frequency domain equalization on the received uplink antenna data.

In this embodiment, the uplink antenna data may be multiple-input multiple-output MIMO antenna system data, single-input multiple-output SIMO antenna system data, or single-input single-output SISO antenna system data.

In LTE, two modes of TDD (Time Division Multiplexing) and FDD (Frequency Division Multiplexing) need to be supported, antenna configurations of 1, 2, 4, and 8 and two-user virtual MIMO are supported. The LMMSE equalization algorithm is used as follows:

$\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}}<span\; class="notranslate">\; )</span>}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}$

Among them, W represents the equalization matrix, $h=\left[\begin{array}{ccc}{h}_{11}& ...& h_{1{\mathrm{no}}_T}\\ .& & .\\ .& & .\\ .& & .\\ h_{{\mathrm{no}}_{R}1}& ...& h_{{\mathrm{no}}_R{\mathrm{no}}_T}\end{array}\right],$ is a +n _R ×n _T channel estimation matrix, n _R represents the number of receiving antennas, n _T represents the number of transmitting antennas, H ^H represents the conjugate transposition matrix of H, represents the identity matrix of n _R × n _T , Represents the inverse matrix of R _n .

$R_{\mathrm{no}}=\left[\begin{array}{ccc}{R}_{11}& ...& R_{1{\mathrm{no}}_R}\\ .& & .\\ .& & .\\ .& & .\\ R_{{\mathrm{no}}_{R}1}& ...& R_{{\mathrm{no}}_R{\mathrm{no}}_R}\end{array}\right],$ is an n _R ×n _R matrix. There are two forms of the R _n matrix: In the case of MIMO, the R _n matrix is simplified into a diagonal matrix without considering inter-cell interference; in the case of SIMO, R _n is a matrix of n _R ×n _R.

Step S204: carry out Matrix operation, get Matrix operation result.

In this step, the The matrix operation in the split, first perform Matrix operation, get Matrix operation result.

Step S206: yes The matrix operation result is multiplied by the H matrix, and the first multiplication result is obtained in parallel. The matrix operation result is multiplied by the received uplink antenna data to obtain a second multiplication result.

In this step, The result of the matrix operation is multiplied with the H matrix along one path to obtain the first multiplication result; one path is multiplied with the received uplink antenna data to obtain the second multiplication result. These two matrix operations are executed in parallel.

Step S208: Perform an addition operation on the first multiplication result and the I matrix to obtain a first addition result.

Step S210: Perform a matrix determinant operation on the first addition result, and simultaneously perform the first addition result and matrix, that is, the result of the second multiplication, to obtain the result of the third multiplication.

Among them, r represents the data of the receiving antenna.

In this step, the matrix determinant operation is performed on the first addition result, and the operation to obtain the determinant result is combined with the first addition result and The multiplication operation of the matrix, the operation of obtaining the result of the third multiplication is executed in parallel.

Step S212: Perform a division operation on the third multiplication result and the determinant result to obtain an equalization result.

In this step, the third multiplication result is divided by the determinant result to obtain an equalization result.

In this embodiment, the LTE base station side completes the LTE uplink receiving LMMSE equalization according to the received antenna data, the channel estimation H and the inverse matrix of the noise covariance matrix Rn. In the equalization process, through the splitting of matrix operations, the serial operations are properly parallelized in the case of a large amount of calculations, which reduces the cost of hardware implementation.

Embodiment three

Referring to FIG. 3 , it shows a flow chart of steps of a frequency domain equalization method according to Embodiment 3 of the present invention.

In this embodiment, the LMMSE equalization algorithm adopted is as shown in Embodiment 2, namely:

$\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}}<span\; class="notranslate">\; )</span>}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; .</span>$

The frequency domain equalization method of this embodiment includes the following steps:

Step S302: operation.

In the 8-antenna SIMO case, the data for 8 antennas is complete The operation is performed by 8 (corresponding to the number of antennas) basic matrix multiplication units. Each basic unit completes the multiplication of a 1*4 matrix and a 4*4 matrix. The basic matrix multiplication unit can also be configured to complete 1* 1. Multiply the matrix of 1*2 and the matrix of 1*1 and 2*2. According to different configurations, the H matrix may be 1*1, 1*2, 1*4, 1*8 or 2*2, 2*4, 2*8 matrix, and the Rn inverse matrix may be 1*1, 2*2 , 4*4, 8*8 matrix.

When the cell is configured as a TDD8 antenna or MIMO, each sample value of the channel estimation matrix H ^H is multiplied by the inverse matrix of Rn in two clocks. When configured as a TDD8 antenna, the channel estimation on each RE (resource unit) lasts for two clocks, and the Rn inverse matrix on each RE is input in two clocks; when configured as MIMO, the channel estimation on each RE The channel estimation value is input by two clocks, and the Rn inverse matrix on each RE lasts for two clocks; in other configurations, each clock completes the multiplication of a sample point value and the Rn inverse matrix.

Step S304: The result of the operation is multiplied with H one way, and the other way is multiplied with the received antenna data. The result of the operation is multiplied by H to obtain the first multiplication result; The result of the operation is multiplied by the received antenna data to obtain the second multiplication result. where, using 2 basic matrix multiplication units to complete and antenna data multiplication, done using the basic 4 basic matrix multiplication units and matrix multiplication of H.

Step S306: Add the first multiplication result to the identity matrix I according to the configuration, and obtain the first addition result.

Step S308: Obtain the matrix A according to the result of the first addition, one way to find the matrix determinant unit, that is, to find |A|, and the other way to sum after transformation That is, the second multiplication result is multiplied to obtain the third multiplication result.

Among them, r represents the data of the receiving antenna.

Step S310: use the pipeline divider to divide the third multiplication result by |A|, and the obtained data is equalized data.

In the above equalization process, when one or more antennas fail, assuming that k antennas among the m antennas are broken, then preferably, the Rn matrix and the H matrix are respectively processed as:

$\left[\begin{array}{cc}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; k</span>}}& {\mathrm{<span\; class="notranslate">\; 0</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; k</span>}}\\ {\mathrm{<span\; class="notranslate">\; 0</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>}& {\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; k</span>}}\end{array}\right]<span\; class="notranslate">\; ,</span>\left[\begin{array}{c}{\mathrm{<span\; class="notranslate">\; h</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; 2</span>}}\\ {\mathrm{<span\; class="notranslate">\; 0</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; 2</span>}}\end{array}\right]<span\; class="notranslate">\; .</span>$

Among them, 0 represents the zero matrix of the corresponding dimension, and I is the identity matrix. The processed Rn and H matrices are equalized according to the antenna configuration of the original cell, and the channel equalization is completed by using the configuration when there is no antenna failure, so as to maximize the system performance. performance.

In this embodiment, by splitting the matrix operation and performing parallel operation, a relatively small processing delay is achieved. At the same time, through the multiplexing of resources, a set of equalization circuits for SISO, SIMO, and MIMO equalization is realized, reducing system area and power consumption. In addition, in practical applications, one or several antennas often fail due to various reasons. The current LTE system supports configurations of 1, 2, 4, and 8 antennas in the uplink. Generally, there is no time for antenna failure to occur When repairing in time, you can select the available antenna to continue to work through cell reconfiguration. For example, if one antenna fails in a cell with 8 antennas, it can be configured as a cell with 4 antennas to work, but this will sacrifice the other three antennas that have not failed. gain, and through this embodiment, when the antenna fails, the Rn matrix and the H matrix are processed, so that when the antenna fails, all the antennas that have not failed can continue to work without reconfiguration of the cell, and the maximum guarantee system performance.

Embodiment Four

Referring to FIG. 4 , it shows a structural block diagram of a frequency domain equalization device according to Embodiment 4 of the present invention.

The frequency domain equalization device in this embodiment includes: an equalization module 402, which is used to perform frequency domain equalization on the received uplink antenna data by using the LMMSE equalization algorithm; the equalization module 402 further includes: a splitting module 4022, which is used to perform frequency domain equalization in the frequency domain equalization process In , the matrix operations are split according to the LMMSE algorithm, and the split matrix operations are executed in parallel.

Preferably, the uplink antenna data includes one of the following: MIMO antenna system data, SIMO antenna system data, and SISO antenna system data.

Preferably, the above-mentioned LMMSE equalization algorithm is expressed as: Among them, W represents the equalization matrix, H represents the channel estimation matrix of n _R × n _T , n _R represents the number of receiving antennas, n _T represents the number of transmitting antennas, H ^H represents the conjugate transpose matrix of H, Represents the inverse matrix of R _n , Represents the identity matrix of n _R ×n _T ; when the uplink antenna data is MIMO antenna system data, the R _n matrix is a diagonal matrix; when the uplink antenna data is SIMO antenna system data, the R _n matrix is an n _R ×n _R matrix.

Preferably, the splitting module 4022 includes: a first splitting module for performing Matrix operation, get Matrix operation result; the second splitting module is used for The matrix operation result is multiplied by the H matrix, and the first multiplication result is obtained in parallel. The multiplication operation of the matrix operation result and the received uplink antenna data to obtain the second multiplication result; the third splitting module is used for the first multiplication result Carry out the addition operation with the I matrix to obtain the first addition result; the fourth split module is used to perform the matrix determinant operation on the first addition result, and perform the first addition result and the first addition result in parallel while obtaining the determinant result. The matrix is the multiplication operation of the second multiplication result to obtain the third multiplication result, wherein r represents the data of the receiving antenna; the result module is used to divide the third multiplication result and the determinant result to obtain the equalization result.

Preferably, when the uplink antenna data is SIMO antenna system data, the first splitting module is used to parallelize the uplink antenna data corresponding to each antenna Matrix operation, get Matrix operation result; when the uplink antenna data is MIMO antenna system data, the first splitting module is used to divide each sample point value of the H ^H channel estimation matrix into two clocks to complete and Matrix multiplication operation, get Matrix operation result.

Preferably, the splitting module 4022 also includes: a compensation module, configured to process the R _n matrix according to the following formula when there is one or more antenna failures: $\left[\begin{array}{cc}{R}_{m-k}& 0_{(m-k)\&CenterDot;k}\\ 0_{k\&CenterDot;(m-k)}& I_{k\&CenterDot;k}\end{array}\right];$ The H matrix is processed according to the following formula: $\left[\begin{array}{c}{h}_{(m-k)\&Center\; Dot;2}\\ 0_{k\&Center\; Dot;2}\end{array}\right];$ Among them, m represents the number of all receiving antennas, k represents the number of failed antennas, I represents the identity matrix, and 0 represents the zero matrix.

Through this embodiment, the deficiency of the traditional frequency domain equalizer is overcome; a relatively small processing delay is achieved; at the same time, through the multiplexing of resources, a set of equalization circuits for SISO, SIMO, and MIMO equalization is realized, and the system area is reduced and power consumption; in addition, it also solves the problem of channel equalization when there is an antenna failure, so that when there is an antenna failure, the antenna that has not failed can continue to work without having to reconfigure the cell.

Embodiment five

Referring to FIG. 5 , it shows a schematic structural diagram of a frequency domain equalization device according to Embodiment 5 of the present invention.

The frequency domain equalization device in this embodiment uses the pilot symbols to obtain the channel estimation of the pilot bits, and then obtains the channel estimation value on the data symbol bits, the Rn inverse matrix value and the antenna data, and then completes the equalization of the data. When MIMO and non-MIMO users exist in a cell at the same time, it supports MIMO, SIMO and SISO equalization.

The splitting module in this embodiment is used to split the matrix operation according to the LMMSE algorithm in the process of performing frequency domain equalization on the received uplink antenna data by using the LMMSE equalization algorithm, and execute the split matrix operation in parallel. The split module mainly includes matrix multiplication module 1, matrix multiplication module 2, matrix multiplication module 3, data alignment module 1, data alignment module 2, determinant calculation module, matrix multiplication module 4, and division module. In the base, the matrix multiplication module 1 is equivalent to the first splitting module in the fourth embodiment, the matrix multiplication module 2 and the matrix multiplication module 3 are equivalent to the second splitting module in the fourth embodiment, and the data alignment module 1 and the data alignment Module 2 is equivalent to the third splitting module in embodiment four together, and determinant module and matrix multiplication module 4 are equivalent to the fourth splitting module in embodiment four together, and division module is equivalent to the result module in embodiment four .

The calculation process of the frequency domain equalization device of the present embodiment is shown in Figure 6, the configurable calculation module of the frequency domain equalization device of the present embodiment (such as the configurable calculation module for multiplying multiple matrices in the matrix multiplication module 1, etc. ) as shown in Figure 7.

The frequency domain equalization device of this embodiment needs to support two modes of TDD and FDD in LTE, and support 1, 2, 4, 8 antenna configurations and two-user virtual MIMO. The present embodiment adopts the above-mentioned LMMSE equalization algorithm as follows:

$\mathrm{<span\; class="notranslate">\; W</span>}<span\; class="notranslate">\; =</span>{<span\; class="notranslate">\; (</span>{\mathrm{<span\; class="notranslate">\; h</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}\mathrm{<span\; class="notranslate">\; h</span>}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; I</span>}}_{{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}<span\; class="notranslate">\; \&times;</span>{\mathrm{<span\; class="notranslate">\; no</span>}}_{\mathrm{<span\; class="notranslate">\; T</span>}}}<span\; class="notranslate">\; )</span>}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}^{\mathrm{<span\; class="notranslate">\; h</span>}}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}^{<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; .</span>$

Using the frequency domain equalization device of this embodiment to perform frequency domain equalization on the received antenna data includes the following steps:

step one: operation.

In the case of 8-antenna SIMO, the data of 8 antennas is sent to the matrix multiplication module 1 to complete operation. In the matrix multiplication module 1, there are 8 basic matrix multiplication units, and each basic unit completes the multiplication of a 1*4 matrix and a 4*4 matrix. The basic matrix multiplication unit module can also be configured to multiply 1*1, 1*2 matrices and 1*1, 2*2 matrices. According to different configurations, the H matrix may be 1*1, 1*2, 1*4, 1*8 or 2*2, 2*4, 2*8 matrix, and the Rn inverse matrix may be 1*1, 2*2 , 4*4, 8*8 matrix.

It should be noted that the matrix multiplication module 1 and the configuration of the basic matrix multiplication unit therein in the case of 8 antennas are merely illustrative. Those skilled in the art can refer to the above settings and appropriately set the number of the matrix multiplication module 1 and the basic matrix multiplication units therein (generally an even number) according to the actual situation, which is not limited by the present invention.

When the cell is configured as a TDD8 antenna or MIMO, each sample value of the channel estimation matrix H ^H is multiplied by the inverse matrix of Rn in two clocks. When configured as a TDD8 antenna, the channel estimation on each RE lasts for two clocks, and the Rn inverse matrix on each RE is input in two clocks; when configured as MIMO, the channel estimation value on each RE is divided into Two clock inputs, the Rn inverse matrix on each RE lasts for two clocks; in other configurations, each clock completes the multiplication of a sample point value and the Rn inverse matrix.

Step two: The result of the operation is multiplied with H in the matrix multiplication module 3 one way, and multiplied with the received antenna data in the matrix multiplication module 2 in the other way. Wherein, use 2 basic matrix multiplication units to complete in the matrix multiplication module 23 And the multiplication of the antenna data, the matrix multiplication module 32 uses the basic 4 basic matrix multiplication units to complete and matrix multiplication of H.

Step 3: The data (the first multiplication result) that comes out of the matrix multiplication module 3 is sent to the data addition and alignment module 1, and is processed differently according to the configuration, and is added to the unit matrix I; the data from the matrix multiplication module 2 ( The second multiplication result) is sent to the data addition and alignment module 2, and the data is processed differently according to the configuration.

Step 4: Obtain the matrix A (the first addition result) from the data addition and alignment module 1, and send it to the matrix determinant module to find |A|, and the other way to the matrix multiplication module 4, and after deformation (the result of the second multiplication) is multiplied.

Step 5: The data obtained from the matrix multiplication module 4 (the third multiplication result) and |A| are sent to the division module, and the pipeline divider is used to complete the division, and the obtained data is equalized data.

When one or more antennas fail, assuming that the k antennas out of the m antennas are broken, the Rn matrix and the H matrix are processed as follows: $\left[\begin{array}{cc}{R}_{m-k}& 0_{(m-k)\&CenterDot;k}\\ 0_{k\&Center\; Dot;(m-k)}& I_{k\&Center\; Dot;k}\end{array}\right],\left[\begin{array}{c}{h}_{(m-k)\&CenterDot;2}\\ 0_{k\&Center\; Dot;2}\end{array}\right].$ Among them, 0 represents the zero matrix of the corresponding dimension, and I is the identity matrix. The processed Rn and H matrices are equalized according to the antenna configuration of the original cell, and the channel equalization is completed by using the configuration when there is no antenna failure, so as to maximize the system performance. performance.

In this embodiment, by splitting the matrix operation in the equalization process, the serial operation is properly performed in parallel in the case of a large amount of calculation, thereby reducing the cost of hardware implementation. Moreover, through the configuration of the core computing module, the frequency domain equalization of various antenna configurations is realized. also. In this embodiment, when an antenna failure occurs and there is no time for timely maintenance, the H matrix and the Rn matrix are properly processed, so that all antennas that have not failed can continue to work without reconfiguration of the cell, and the performance of the system is guaranteed to the greatest extent.

From the above description, it can be seen that the technical solution for frequency domain equalization shared by multiple antennas in the LTE (Long Term Evolution) system of the present invention is applicable to either a MIMO system or a multi-antenna SIMO system Compared with the traditional equalization technical solution, through the combination of serial and parallel calculations and the processing of the matrix, the progress of reducing the multiplier and supporting the problem of antenna failure has been achieved, saving power consumption, and the maximum guarantee when there is an antenna failure Improve the performance of the system, improve the competitiveness of the system and so on.

Obviously, those skilled in the art should understand that each module or each step of the above-mentioned present invention can be realized by a general-purpose computing device, and they can be concentrated on a single computing device, or distributed in a network formed by multiple computing devices Alternatively, they may be implemented in program code executable by a computing device so that they may be stored in a storage device to be executed by a computing device, and in some cases in an order different from that shown here The steps shown or described are carried out, or they are separately fabricated into individual integrated circuit modules, or multiple modules or steps among them are fabricated into a single integrated circuit module for implementation. As such, the present invention is not limited to any specific combination of hardware and software.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and changes. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims (11)

1. A frequency domain equalization method, characterized in that, comprising: Use the linear minimum mean square error LMMSE equalization algorithm to perform frequency domain equalization on the received uplink antenna data; In the frequency domain equalization process, the matrix operation is split according to the LMMSE algorithm, and the split matrix operation is executed in parallel; Wherein, the LMMSE equalization algorithm is expressed as: Among them, W represents the equalization matrix, H represents the channel estimation matrix of n _R × n _T , n _R represents the number of receiving antennas, n _T represents the number of transmitting antennas, H ^H represents the conjugate transpose matrix of H, Represents the inverse matrix of R _n , Represents the n _R ×n _T identity matrix. 2. The method according to claim 1, wherein the uplink antenna data includes one of the following: multiple-input multiple-output MIMO antenna system data, single-input multiple-output SIMO antenna system data, single-input single-output SISO antenna system data data. 3. The method of claim 2, wherein, When the uplink antenna data is the MIMO antenna system data, the R _n matrix is a diagonal matrix; when the uplink antenna data is the SIMO antenna system data, the R _n matrix is an n _R A matrix of ×n _R. 4. method according to claim 3, is characterized in that, described matrix operation wherein is split according to LMMSE algorithm, the step of described matrix operation after parallel execution split comprises: conduct Matrix operation, get Matrix operation result; to the said The matrix operation result and the H matrix are multiplied, and while the first multiplication result is obtained, the described multiplying the matrix operation result and the received uplink antenna data to obtain a second multiplication result; performing an addition operation on the first multiplication result and the I matrix to obtain a first addition result; Performing a matrix determinant operation on the first addition result, while obtaining the determinant result, performing a multiplication operation of the first addition result and the second multiplication result in parallel to obtain a third multiplication result; performing a division operation on the third multiplication result and the determinant result to obtain an equalization result. 5. method according to claim 4, is characterized in that, described carrying out Matrix operation, get The steps for matrix operation results include: When the uplink antenna data is the SIMO antenna system data, the uplink antenna data corresponding to each antenna is performed in parallel. Matrix operation, get Matrix operation result; When the uplink antenna data is the MIMO antenna system data, each sample point value of the H ^H channel estimation matrix is divided into two clocks to complete and the Matrix multiplication operation, get Matrix operation result. 6. The method according to claim 3, wherein when there is one or more antenna failures, the method further comprises: The R _n matrix is processed according to the following formula: $\left[\begin{array}{cc}{\mathrm{<span\; class="notranslate">\; R</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; k</span>}}& {\mathrm{<span\; class="notranslate">\; 0</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; k</span>}}\\ {\mathrm{<span\; class="notranslate">\; 0</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Center\; Dot;</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span>}& {\mathrm{<span\; class="notranslate">\; I</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&Center\; Dot;</span>\mathrm{<span\; class="notranslate">\; k</span>}}\end{array}\right]<span\; class="notranslate">\; ;</span>$ The H matrix is processed according to the following formula: $\left[\begin{array}{c}{\mathrm{<span\; class="notranslate">\; h</span>}}_{<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; 2</span>}}\\ {\mathrm{<span\; class="notranslate">\; 0</span>}}_{\mathrm{<span\; class="notranslate">\; k</span>}<span\; class="notranslate">\; \&CenterDot;</span>\mathrm{<span\; class="notranslate">\; 2</span>}}\end{array}\right]<span\; class="notranslate">\; ;</span>$ Among them, m represents the number of all receiving antennas, k represents the number of failed antennas, I represents the identity matrix, and 0 represents the zero matrix. 7. A frequency domain equalization device, characterized in that, comprising: An equalization module, configured to use a linear minimum mean square error LMMSE equalization algorithm to perform frequency domain equalization on the received uplink antenna data; The equalization module includes: a splitting module, configured to split the matrix operation according to the LMMSE algorithm in the frequency domain equalization process, and execute the split matrix operation in parallel; Wherein, the LMMSE equalization algorithm is expressed as: Among them, W represents the equalization matrix, H represents the channel estimation matrix of n _R × n _T , n _R represents the number of receiving antennas, n _T represents the number of transmitting antennas, H ^H represents the conjugate transpose matrix of H, Represents the inverse matrix of R _n , Represents the n _R ×n _T identity matrix. 8. The device according to claim 7, wherein the uplink antenna data includes one of the following: multiple-input multiple-output MIMO antenna system data, single-input multiple-output SIMO antenna system data, single-input single-output SISO antenna system data data. 9. The device of claim 8, wherein: When the uplink antenna data is the MIMO antenna system data, the R _n matrix is a diagonal matrix; when the uplink antenna data is the SIMO antenna system data, the R _n matrix is an n _R A matrix of ×n _R. 10. The device according to claim 9, wherein the splitting module comprises: The first split module is used to conduct Matrix operation, get Matrix operation result; The second split module is used for the The matrix operation result and the H matrix are multiplied, and while the first multiplication result is obtained, the described multiplying the matrix operation result and the received uplink antenna data to obtain a second multiplication result; A third splitting module, configured to perform an addition operation on the first multiplication result and the I matrix to obtain a first addition result; The fourth splitting module is used to perform a matrix determinant operation on the first addition result, and simultaneously perform a multiplication operation on the first addition result and the second multiplication result to obtain the second multiplication result while obtaining the determinant result. triple multiplication result; A result module, configured to perform a division operation on the third multiplication result and the determinant result to obtain an equalization result. 11. The device according to claim 9, wherein the splitting module further comprises: The compensation module is used to process the R _n matrix according to the following formula when there is one or more antenna failures: The H matrix is processed according to the following formula: Among them, m represents the number of all receiving antennas, k represents the number of failed antennas, I represents the identity matrix, and 0 represents the zero matrix.