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 equilibrium method and device

Technical field

The present invention relates to the communications field, in particular to a kind of LTE (Long Term Evolution, Long Term Evolution) The frequency-domain equilibrium method of system and device.

Background technology

OFDM (Orthogonal Frequency Division Multiplexing, OFDM) is a kind of special Different MCM (Multi-carrier Modulation, multi-carrier modulation) technology, its thought is the data stream of a high speed Resolve into the sub-stream data of many low rates, and upload defeated at many sub-channels in a parallel fashion, it is provided that allow data with relatively High speed is in a kind of approach of the channel of bigger delay.MIMO (Multi-Input Multi-Output, multi input Multi output) system by sending and receiving end configure multiple antennas, utilize space channel to realize diversity gain or spatial multiplexing gain, do not increasing In the case of adding system bandwidth, greatly improve capacity and the spectrum efficiency of communication system, be Next-Generation Wireless Communication Systems Key technology.OFDM and MIMO is combined, the highest availability of frequency spectrum can be realized, the shadow of multidiameter fading channel can be resisted again Ring.

But, being affected by the factor such as size and cost, the antenna of LTE stage terminal is generally one, thus can not Enough capacity gain are obtained from multi-emitting antenna.It is Mu-MIMO (Multi-that 3GPP LTE gives a kind of solution UserMIMO) i.e. multiuser MIMO.Multiuser MIMO makes single-antenna subscriber share the antenna of other users by collaboration communication, Utilize identical running time-frequency resource to communicate, improve the performance of system.

Although by multiuser MIMO, OFDM achieves certain one-tenth in terms of increasing by wireless link transmission data volume Merit, but, owing to transmitting signal by radio wave, therefore in typical operating environment, the signal that base station receives exists Multipath, and then owing to multi-path problem result in delay extension.Meanwhile, signal in the transmitting procedure from transmitting terminal to receiving terminal, Also can receive the various impacts such as non-linear, rain declines, multipath, make the signal of transmission be distorted, cause wave distortion, thus cause Intersymbol interference.The most general employing balancing technique solves intersymbol interference, thus compensates the signal caused by channel parameter change Distortion, offsets channel signal transmission decay, can be effective against intersymbol interference simultaneously, reduce error code, improve transfer rate.Equilibrium Mainly having two Basic Ways, frequency domain equalization and time domain equalization, frequency domain equalization is to make whole system frequency transfer function meet nothing The condition of distortion transmissionAnd time domain equalization is directly to consider from time response, make to include equilibrium model Block meets the bar without ISI (Inter-Symbol Interference, intersymbol interference) at the impulse response of interior whole system Part.

Currently, the research carried out frequency domain equalization (FDE) is the most.Generally by by the operational transformation in time domain to frequency domain In carry out, to compensate there is the infringement that the multipath channel of long shock response is brought, the complexity of receiver can be reduced simultaneously. But meanwhile, use which to result in longer equilibrium treatment delay time, and then reduce user's experience.

Summary of the invention

Present invention is primarily targeted at a kind of frequency-domain equilibrium method of offer and device, at least to solve above-mentioned existing equilibrium Processing method time delay is long, the problem that user's experience is low.

According to an aspect of the invention, it is provided a kind of frequency-domain equilibrium method, including: use LMMSE equalization algorithm pair The uplink antenna data received carry out frequency domain equalization；In frequency domain equalization process, according to LMMSE algorithm to matrix operations therein Split, the matrix operations after executed in parallel fractionation.

According to a further aspect in the invention, it is provided that a kind of frequency domain equalization device, including: balance module, it is used for using LMMSE equalization algorithm carries out frequency domain equalization to the uplink antenna data received；This balance module includes: split module, is used for In frequency domain equalization process, splitting matrix operations therein according to LMMSE algorithm, the matrix after executed in parallel splits is transported Calculate.

By the present invention, use during carrying out frequency domain equalization, according to LMMSE algorithm, to matrix operations therein Carry out suitable fractionation, and the matrix operations after splitting is calculated parallel, by splitting and parallel computation, solve existing Equalization processing method time delay is long, the problem that user's experience is low, and then has reached minimizing equalization processing method time delay, improves and uses The effect of family experience.

Accompanying drawing explanation

Accompanying drawing described herein is used for providing a further understanding of the present invention, constitutes the part of the application, this Bright schematic description and description is used for explaining the present invention, is not intended that inappropriate limitation of the present invention.In the accompanying drawings:

Fig. 1 is the flow chart of steps of a kind of frequency-domain equilibrium method of according to embodiments of the present invention；

Fig. 2 is the flow chart of steps of a kind of frequency-domain equilibrium method of according to embodiments of the present invention two；

Fig. 3 be according to embodiments of the present invention three one at the flow chart of steps of frequency-domain equilibrium method；

Fig. 4 is the structured flowchart of a kind of frequency domain equalization device of according to embodiments of the present invention four；

Fig. 5 is the structural representation of a kind of frequency domain equalization device of according to embodiments of the present invention five；

Fig. 6 is the calculation process schematic diagram of the frequency domain equalization device of embodiment illustrated in fig. 5；

Fig. 7 is the configurable computing module schematic diagram of the frequency domain equalization device of embodiment illustrated in fig. 5.

Detailed description of the invention

Below with reference to accompanying drawing and describe the present invention in detail in conjunction with the embodiments.It should be noted that do not conflicting In the case of, the embodiment in the application and the feature in embodiment can be mutually combined.

In LTE, need to support TDD (Time Division Duplex, time division duplex) and FDD (Frequency DivisionDuplex, FDD) both of which, support antenna configurations and two users' virtual MIMO, the present invention of 1,2,4,8 The method and device equalized based on LTE uplink receiving LMMSE (linear minimum mean-squared error) that embodiment is provided, for LTE Base station side is estimated H and the inverse matrix of noise covariance matrix Rn according to the antenna data received, channel, is completed LTE up Receive LMMSE equilibrium.

Embodiment one

With reference to Fig. 1, it is shown that the flow chart of steps of a kind of frequency-domain equilibrium method of according to embodiments of the present invention.

The frequency-domain equilibrium method of the present embodiment comprises the following steps:

Step S102: use LMMSE equalization algorithm that the uplink antenna data received are carried out frequency domain equalization；

In the present embodiment, uplink antenna data can be that how defeated multiple-input and multiple-output mimo antenna system data, single input be Go out SIMO antenna system data or single-input single-output SISO antenna system data.

Step S104: in frequency domain equalization process, splits matrix operations therein according to LMMSE algorithm, parallel Perform the matrix operations after splitting.

In correlation technique, equalization processing method time delay is long, and user's experience is low.By the present embodiment, use and carrying out During frequency domain equalization, according to LMMSE algorithm, matrix operations therein is carried out suitable fractionation, and to the square after splitting Battle array computing calculates parallel, by splitting and parallel computation, solves existing equalization processing method time delay long, and user uses body Test low problem, and then reached minimizing equalization processing method time delay, improve the effect of user's experience.

Embodiment two

With reference to Fig. 2, it is shown that the flow chart of steps of a kind of frequency-domain equilibrium method of two according to embodiments of the present invention.

The frequency-domain equilibrium method of the present embodiment comprises the following steps:

Step S202: use LMMSE equalization algorithm that the uplink antenna data received are proceeded by frequency domain equalization.

In the present embodiment, uplink antenna data can be that how defeated multiple-input and multiple-output mimo antenna system data, single input be Go out SIMO antenna system data or single-input single-output SISO antenna system data.

In LTE, need to support TDD (time division multiplex) and FDD (frequency division multiplexing) both of which, support the sky of 1,2,4,8 Line configuration and two users' virtual MIMO, use LMMSE equalization algorithm as follows:

$W={(H^H{R}_n^{-1}H+I_{n_T\×n_T})}^{-1}{H}^H{R}_n^{-1}$

Wherein, W represents equilibrium matrix, $H=\left[\begin{array}{ccc}{H}_{11}& ...& H_{1n_T}\\ .& & .\\ .& & .\\ .& & .\\ H_{n_{R}1}& ...& H_{n_R{n}_T}\end{array}\right],$ It is+a n_R×n_TChannel estimate matrix, n_RRepresent Reception antenna number, n_TRepresent the number launching antenna, H^HRepresent the associate matrix of H,Represent n_R×n_TUnit matrix,Represent R_nInverse matrix.

$R_n=\left[\begin{array}{ccc}{R}_{11}& ...& R_{1n_R}\\ .& & .\\ .& & .\\ .& & .\\ R_{n_{R}1}& ...& R_{n_R{n}_R}\end{array}\right],$ It is a n_R×n_RMatrix.R_nMatrix has two kinds of forms: in the case of MIMO, Do not consider presence of intercell interference, R_nMatrix reduction becomes a diagonal matrix；In the case of SIMO, R_nIt is n_R×n_RMatrix.

Step S204: carry outMatrix operations, obtainsMatrix operation results.

In this step, rightIn matrix operations split, first carry out Matrix operations, obtainsMatrix operation results.

Step S206: rightMatrix operation results and H-matrix carry out multiplying, obtain the same of the first multiplication result Time, carry out parallelThe multiplying of the uplink antenna data of matrix operation results and reception, obtains the second multiplication result.

In this step,Matrix operation results one tunnel is multiplied with H-matrix, obtains the first multiplication result；One tunnel and reception Uplink antenna data be multiplied, obtain the second multiplication result.This two tunnels matrix operations executed in parallel.

Step S208: the first multiplication result and I matrix are carried out additive operation, obtains the first addition results.

Step S210: the first addition results is asked matrix determinant computing, while obtaining determinant result, parallel Carry out the first addition results withMatrix, the i.e. second multiplication result, multiplying, obtain the 3rd multiplication result.

In, r represents the data of reception antenna.

In this step, the first addition results is asked matrix determinant computing, obtain the computing of determinant result, with the One addition results withMultiplication of matrices computing, obtains the computing executed in parallel of the 3rd multiplication result.

Step S212: the 3rd multiplication result and determinant result carry out division arithmetic, obtains equilibrium result.

In this step, use the 3rd multiplication result is divided by determinant result, and then obtains equilibrium result.

In the present embodiment, H and noise covariance matrix Rn is estimated according to the antenna data received, channel in LTE base station side Inverse matrix, complete LTE uplink receiving LMMSE equilibrium.In balancing procedure, by the fractionation of matrix operations, at operand In the case of relatively big, serial arithmetic is carried out suitable carrying out parallel, reduce hard-wired cost.

Embodiment three

With reference to Fig. 3, it is shown that the flow chart of steps of frequency-domain equilibrium method at the one of three according to embodiments of the present invention.

In the present embodiment, the LMMSE equalization algorithm of employing is as shown in embodiment two, it may be assumed that

$W={(H^H{R}_n^{-1}H+I_{n_T\×n_T})}^{-1}{H}^H{R}_n^{-1}.$

The frequency-domain equilibrium method of the present embodiment comprises the following steps:

Step S302:Computing.

In the case of 8 antenna SIMO, the data of 8 antennas completeComputing, by 8 (corresponding to antenna number) bases This matrix multiplication unit carries out computing, and each elementary cell completes matrix and a 4*4 multiplication of matrices, the Ben Ji of a 1*4 This matrix multiplication unit can also configure the matrix of 1*1,1*2 and the matrix multiple of 1*1,2*2.Different H according to configuration Matrix is probably the matrix of 1*1,1*2,1*4,1*8 or 2*2,2*4,2*8, and Rn inverse matrix may be 1*1,2*2,4*4,8*8 Matrix.

When cell configuration is TDD8 antenna or MIMO when, channel estimate matrix H^HEach sample value divide two Clock completes and the multiplication of Rn inverse matrix.When being configured to TDD8 antenna when, the channel on each RE (Resource Unit) is estimated Continuing two clocks, the Rn inverse matrix on each RE divides two clock inputs；Letter when being configured to MIMO when, on each RE Road estimated value divides two clock inputs, and the Rn inverse matrix on each RE continues two clocks；The when of other configurations, each clock Complete a sample value and the multiplication of Rn inverse matrix.

Step S304:Result one tunnel of computing is multiplied with H, and an other road is multiplied with the antenna data received.The result of computing is multiplied with H, obtains the first multiplication result；The result of computing and the antenna data phase received Take advantage of, obtain the second multiplication result.Wherein, 2 fundamental matrix multiplication units are used to completeWith the multiplication of antenna data, make Complete with basic 4 fundamental matrix multiplication unitsMatrix multiplication with H.

Step S306: the first multiplication result is added according to configuration and unit matrix I, obtains the first addition results.

Step S308: obtain matrix A according to the first addition results, a road seeks matrix determinant unit, i.e. seeks | A |, another Road through deformation after andI.e. second multiplication result is multiplied, it is thus achieved that the 3rd multiplication result.

Wherein, r represents the data of reception antenna.

Step S310: utilizing flowing water divider to complete the division of the 3rd multiplication result and | A |, the data obtained are equilibrium After data.

In above-mentioned balancing procedure, when there is one or more antenna failure when, it is assumed that k root antenna in m root antenna It is broken, the most preferably, Rn matrix and H-matrix is processed as respectively:

$\left[\begin{array}{cc}{R}_{m-k}& 0_{(m-k)\·k}\\ 0_{k\·(m-k)}& I_{k\·k}\end{array}\right],\left[\begin{array}{c}{H}_{(m-k)\·2}\\ 0_{k\·2}\end{array}\right].$

Wherein 0 null matrix representing corresponding dimension, I is unit matrix, by the Rn after processing and H-matrix according to the least The antenna configurations in district carries out equilibrium treatment, and configuration when being used without antenna failure, to complete channel equalization, improves to greatest extent The performance of system.

In the present embodiment, by the fractionation of matrix operations and carry out concurrent operation, reach smaller process time delay, Meanwhile, by the multiplexing to resource, it is achieved that SISO, SIMO, MIMO equilibrium shares a set of equalizing circuit, reduces system area And power consumption.It addition, in actual applications, often due to a variety of causes can occur the problem of one or several antenna failures, mesh The configuration of up support 1,2,4,8 antenna in front LTE system, the typically on-call maintenance that is not able to do in time at generation antenna failure When, the antenna that can use can be selected to work on by cell re-configuration, the community of such as 8 antennas is if it occur that a sky Line lost efficacy, and work is carried out in the community that can be made into 4 antennas, but so sacrifices the gain of other three antennas that do not lost efficacy in vain, And by the present embodiment, when antenna failure, Rn matrix and H-matrix are processed such that it is able to support at antenna failure Time, using all antennas not lost efficacy to work on, community need not be carried out and reconfigure, maximum ensure that the property of system Energy.

Embodiment four

With reference to Fig. 4, it is shown that the structured flowchart of a kind of frequency domain equalization device of four according to embodiments of the present invention.

The frequency domain equalization device of the present embodiment includes: balance module 402, for using LMMSE equalization algorithm to reception Uplink antenna data carry out frequency domain equalization；This balance module 402 farther includes: split module 4022, at frequency domain equalization During, according to LMMSE algorithm, matrix operations therein is split, the matrix operations after executed in parallel fractionation.

Preferably, uplink antenna data include one below: mimo antenna system data, SIMO antenna system data, SISO antenna system data.

Preferably, above-mentioned LMMSE equalization algorithm is expressed as:Wherein, W represents equal Weighing apparatus matrix, H represents n_R×n_TChannel estimate matrix, n_RRepresent reception antenna number, n_TRepresent the number launching antenna, H^HRepresent H Associate matrix,Represent R_nInverse matrix,Represent n_R×n_TUnit matrix；When uplink antenna data are MIMO During antenna system data, R_nMatrix is a diagonal matrix；When uplink antenna data are SIMO antenna system data, R_nMatrix It is a n_R×n_RMatrix.

Preferably, split module 4022 and include: first splits module, is used for carrying outMatrix operations, obtains Matrix operation results；Second splits module, for rightMatrix operation results and H-matrix carry out multiplying, obtain first While multiplication result, carry out parallelThe multiplying of the uplink antenna data of matrix operation results and reception, obtains the Square law result；3rd splits module, for the first multiplication resultCarry out additive operation with I matrix, obtain first Addition results,；4th splits module, for the first addition results is asked matrix determinant computing, obtains determinant result While, carry out parallel the first addition results withMatrix, the multiplying of the i.e. second multiplication result, obtain the 3rd multiplication As a result, wherein, r represents the data of reception antenna；Object module, for carrying out division to the 3rd multiplication result and determinant result Computing, obtains equilibrium result.

Preferably, when uplink antenna data are SIMO antenna system data, first splits module is used for would correspond to often The uplink antenna data parallel of one antenna is carried outMatrix operations, obtainsMatrix operation results；Work as uplink antenna When data are mimo antenna system data, first splits module for by H^HWhen each sample value of channel estimate matrix divides two Clock complete andMultiplication of matrices computing, obtainsMatrix operation results.

Preferably, split module 4022 and also include: compensating module, for when there is one or more antenna failure, right R_nMatrix processes according to below equation: $\left[\begin{array}{cc}{R}_{m-k}& 0_{(m-k)\·k}\\ 0_{k\·(m-k)}& I_{k\·k}\end{array}\right];$ H-matrix is processed according to below equation: $\left[\begin{array}{c}{H}_{(m-k)\·2}\\ 0_{k\·2}\end{array}\right];$ Wherein, m represents the number of all reception antennas, and k represents the number of inefficacy antenna, I representation unit matrix, and 0 represents Null matrix.

By the present embodiment, overcome the deficiency of traditional frequency domain equalizer；Reach smaller process time delay；Lead to simultaneously Cross the multiplexing to resource, it is achieved that SISO, SIMO, MIMO equilibrium shares a set of equalizing circuit, reduces system area and power consumption； It addition, also solve the channel equalization problem when of there is antenna failure, thus can be in order to there is antenna failure when Work on, without re-starting cell configuration with there is no the antenna lost efficacy.

Embodiment five

With reference to Fig. 5, it is shown that the structural representation of a kind of frequency domain equalization device of five according to embodiments of the present invention.

The channel that frequency domain equalization device in the present embodiment uses frequency pilot sign to obtain pilot bit is estimated, and then the number obtained According to the channel estimation value on sign bit, Rn inverse matrix value and antenna data, and then the equilibrium of complete paired data.When a community The when of there is MIMO and non-MIMO user, support MIMO, SIMO and SISO equilibrium simultaneously.

The fractionation module of the present embodiment is for using LMMSE equalization algorithm that the uplink antenna data received are carried out frequency domain During equilibrium, according to LMMSE algorithm, matrix operations therein is split, the matrix operations after executed in parallel fractionation. This fractionation module mainly includes Matrix Multiplication module 1, Matrix Multiplication module 2, Matrix Multiplication module 3, alignment of data module 1, alignment of data Module 2, seek determinant module, Matrix Multiplication module 4, dividing module totally eight modules.In base, Matrix Multiplication module 1 is equivalent to implement The first fractionation module in example four, Matrix Multiplication module 2 is equivalent to the second fractionation mould in embodiment four together with Matrix Multiplication module 3 Block, alignment of data module 1 is equivalent to the 3rd fractionation module in embodiment four together with data alignment module 2, seeks Norm of Determinant Block is equivalent to the 4th fractionation module in embodiment four together with Matrix Multiplication module 4, and dividing module is equivalent in embodiment four Object module.

The calculation process of the frequency domain equalization device of the present embodiment as shown in Figure 6, the frequency domain equalization device of the present embodiment can Configuration computing module (as being used for the configurable computing module etc. of multiple matrix multiples in Matrix Multiplication module 1) is as shown in Figure 7.

The frequency domain equalization device of the present embodiment, in LTE, needs to support TDD and FDD both of which, supports 1,2,4,8 Antenna configurations and two users' virtual MIMO.The present embodiment uses above-mentioned LMMSE equalization algorithm as follows:

$W={(H^H{R}_n^{-1}H+I_{n_T\×n_T})}^{-1}{H}^H{R}_n^{-1}.$

The frequency domain equalization device using the present embodiment carries out frequency domain equalization to the antenna data received and comprises the following steps:

Step one:Computing.

In the case of 8 antenna SIMO, the data of 8 antennas are sent to Matrix Multiplication module 1 and completeComputing.At matrix Taking advantage of by 8 fundamental matrix multiplication units inside module 1, each elementary cell completes the matrix of a 1*4 and a 4*4 matrix Multiplication.This fundamental matrix multiplication unit module can also configure the matrix of 1*1,1*2 and the matrix multiple of 1*1,2*2.Root Be probably the matrix of 1*1,1*2,1*4,1*8 or 2*2,2*4,2*8 according to the different H-matrix of configuration, Rn inverse matrix may be 1* 1, the matrix of 2*2,4*4,8*8.

It should be noted that Matrix Multiplication module 1 and fundamental matrix multiplication unit therein during above-mentioned 8 antenna arrange and are only Exemplary illustration.Those skilled in the art are referred to above-mentioned setting, according to practical situation be appropriately arranged with Matrix Multiplication module 1 and its In the number (generally even number) of fundamental matrix multiplication unit, the invention is not limited in this regard.

When cell configuration is TDD8 antenna or MIMO when, channel estimate matrix H^HEach sample value divide two Clock completes and the multiplication of Rn inverse matrix.When being configured to TDD8 antenna when, when the channel on each RE estimates to continue two Clock, the Rn inverse matrix on each RE divides two clock inputs；When being configured to MIMO when, the channel estimation value on each RE divides Two clock inputs, the Rn inverse matrix on each RE continues two clocks；The when of other configurations, each clock completes a sample Point value and the multiplication of Rn inverse matrix.

Step 2:Result one tunnel of computing is multiplied with H in Matrix Multiplication module 3, and an other road is in Matrix Multiplication module 2 It is multiplied with the antenna data received.Wherein, Matrix Multiplication module 23 use 2 fundamental matrixs take advantage of unit to completeAnd sky The multiplication of line data, uses basic 4 fundamental matrixs to take advantage of unit to complete inside Matrix Multiplication module 32Matrix multiplication with H.

Step 3: inside Matrix Multiplication module 3, data (the first multiplication result) out are sent to data and are added alignment module 1, carry out different process, and and unit matrix I addition according to configuration；Matrix Multiplication module 2 data (the second multiplication knot out Really) it is sent to data and is added alignment module 2, according to configuration, data are carried out different process.

Step 4: being added alignment module 1 from data and obtain matrix A (the first addition results), a road is delivered to seek matrix ranks Formula module, seeks | A |, and Matrix Multiplication module 4 is delivered on another road, through deformation after and(the second multiplication result) is multiplied.

Step 5: data (the 3rd multiplication result) and | A | of obtaining from Matrix Multiplication module 4 deliver to dividing module, utilizes Flowing water divider completes division, the data obtained i.e. be equalized after data.

When there is one or more antenna failure when, it is assumed that after in m root antenna, k root antenna is broken, to Rn matrix It is processed as respectively with H-matrix: $\left[\begin{array}{cc}{R}_{m-k}& 0_{(m-k)\·k}\\ 0_{k\·(m-k)}& I_{k\·k}\end{array}\right],\left[\begin{array}{c}{H}_{(m-k)\·2}\\ 0_{k\·2}\end{array}\right].$ Wherein 0 null matrix representing corresponding dimension, I is unit Matrix, carries out equilibrium treatment by the Rn after processing and H-matrix according to the antenna configurations of original community, is used without antenna failure Time configuration complete channel equalization, improve the performance of system to greatest extent.

The present embodiment is by the fractionation of matrix operations in balancing procedure, by serial arithmetic in the case of operand is relatively big Carry out suitable carrying out parallel, reduce hard-wired cost.Further, by the configuration to its main operational module, it is achieved that The frequency domain equalization of various antenna configurations.In addition.The present embodiment occur antenna failure and be not able to do in time on-call maintenance when, H-matrix and Rn matrix being carried out suitable process, so that working on all antennas not lost efficacy, community need not be carried out Reconfiguring, maximum ensure that the performance of system.

As can be seen from the above description, LTE (Long Term Evolution, the Long Term Evolution) system of the present invention The technical scheme of frequency domain equalization that multiple antennas shares, with otherwise be applicable to mimo system, or be applicable to multiple antennas SIMO system The technical scheme of the tradition equilibrium in system is compared, and the process by the way of string calculations incorporated and to matrix achieves and subtracts Lack multiplier and support the progress of antenna failure problem, saving power consumption, and maximum ensures there is antenna failure when The performance of system, competitiveness that improve system etc..

Obviously, those skilled in the art should be understood that each module of the above-mentioned present invention or each step can be with general Calculating device realize, they can concentrate on single calculating device, or be distributed in multiple calculating device and formed Network on, alternatively, they can with calculate the executable program code of device realize, it is thus possible to by they store Performed by calculating device in the storage device, and in some cases, can perform with the order being different from herein shown The step gone out or describe, or they are fabricated to respectively each integrated circuit modules, or by the multiple modules in them or Step is fabricated to single integrated circuit module and realizes.So, the present invention is not restricted to the combination of any specific hardware and software.

The foregoing is only the preferred embodiments of the present invention, be not limited to the present invention, for the skill of this area For art personnel, the present invention can have various modifications and variations.All within the spirit and principles in the present invention, that is made any repaiies Change, equivalent, improvement etc., should be included within the scope of the present invention.

Claims (11)

1. a frequency-domain equilibrium method, it is characterised in that including:

Use linear minimum mean-squared error LMMSE equalization algorithm that the uplink antenna data received are carried out frequency domain equalization；

In described frequency domain equalization process, splitting matrix operations therein according to described LMMSE algorithm, executed in parallel is torn open Described matrix operations after Fen；

Wherein, described LMMSE equalization algorithm is expressed as:

Wherein, W represents equilibrium matrix, and H represents n_R×n_TChannel estimate matrix, n_RRepresent reception antenna number, n_TRepresent and launch sky The number of line, H^HRepresent the associate matrix of H,Represent R_nInverse matrix,Represent n_R×n_TUnit matrix.

Method the most according to claim 1, it is characterised in that described uplink antenna data include one below: multi input Multi output mimo antenna system data, single input and multi-output SIMO antenna system data, single-input single-output SISO antenna system Data.

Method the most according to claim 2, it is characterised in that

When described uplink antenna data are described mimo antenna system data, described R_nMatrix is a diagonal matrix；When described When uplink antenna data are described SIMO antenna system data, described R_nMatrix is a n_R×n_RMatrix.

Method the most according to claim 3, it is characterised in that described according to LMMSE algorithm, matrix operations therein is entered Row splits, and the step of the described matrix operations after executed in parallel fractionation includes:

Carry outMatrix operations, obtainsMatrix operation results；

To describedMatrix operation results and described H-matrix carry out multiplying, while obtaining the first multiplication result, and Row carries out describedThe multiplying of the uplink antenna data of matrix operation results and described reception, obtains the second multiplication knot Really；

Described first multiplication result and described I matrix are carried out additive operation, obtains the first addition results；

Described first addition results is asked matrix determinant computing, while obtaining determinant result, carries out described parallel First addition results and the multiplying of described second multiplication result, obtain the 3rd multiplication result；

Described 3rd multiplication result and described determinant result are carried out division arithmetic, obtains equilibrium result.

Method the most according to claim 4, it is characterised in that described in carry outMatrix operations, obtainsMatrix The step of operation result includes:

When described uplink antenna data are described SIMO antenna system data, would correspond to the described up sky of each antenna Line data parallel carries out describedMatrix operations, obtainsMatrix operation results；

When described uplink antenna data are described mimo antenna system data, by described H^HEach sampling point of channel estimate matrix Two clocks of value point complete with describedMultiplication of matrices computing, obtainsMatrix operation results.

Method the most according to claim 3, it is characterised in that when there is one or more antenna failure, described method Also include:

To R_nMatrix processes according to below equation:

$\left[\begin{array}{cc}{R}_{m-k}& 0_{(m-k)\·k}\\ 0_{k\·(m-k)}& I_{k\·k}\end{array}\right];$

H-matrix is processed according to below equation:

$\left[\begin{array}{c}{H}_{(m-k)\·2}\\ 0_{k\·2}\end{array}\right];$

Wherein, m represents the number of all reception antennas, and k represents the number of inefficacy antenna, I representation unit matrix, and 0 represents zero moment Battle array.

7. a frequency domain equalization device, it is characterised in that including:

Balance module, for using linear minimum mean-squared error LMMSE equalization algorithm that the uplink antenna data received are carried out frequency Territory equalizes；

Described balance module includes: split module, in described frequency domain equalization process, according to described LMMSE algorithm to it In matrix operations split, executed in parallel split after described matrix operations；

Wherein, described LMMSE equalization algorithm is expressed as:

Wherein, W represents equilibrium matrix, and H represents n_R×n_TChannel estimate matrix, n_RRepresent reception antenna number, n_TRepresent and launch sky The number of line, H^HRepresent the associate matrix of H,Represent R_nInverse matrix,Represent n_R×n_TUnit matrix.

Device the most according to claim 7, it is characterised in that described uplink antenna data include one below: multi input Multi output mimo antenna system data, single input and multi-output SIMO antenna system data, single-input single-output SISO antenna system Data.

Device the most according to claim 8, it is characterised in that

When described uplink antenna data are described mimo antenna system data, described R_nMatrix is a diagonal matrix；When described When uplink antenna data are described SIMO antenna system data, described R_nMatrix is a n_R×n_RMatrix.

Device the most according to claim 9, it is characterised in that described fractionation module includes:

First splits module, is used for carrying outMatrix operations, obtainsMatrix operation results；

Second splits module, for describedMatrix operation results and described H-matrix carry out multiplying, obtain first While multiplication result, carry out described parallelThe multiplication fortune of the uplink antenna data of matrix operation results and described reception Calculate, obtain the second multiplication result；

3rd splits module, for described first multiplication result and described I matrix are carried out additive operation, obtains the first addition knot Really；

4th splits module, for described first addition results is asked matrix determinant computing, obtains determinant result Meanwhile, carry out the multiplying of described first addition results and described second multiplication result parallel, obtain the 3rd multiplication result；

Object module, for described 3rd multiplication result and described determinant result carry out division arithmetic, obtains equilibrium result.

11. devices according to claim 9, it is characterised in that described fractionation module also includes:

, for when there is one or more antenna failure, to R in compensating module_nMatrix processes according to below equation:H-matrix is processed according to below equation:Wherein, m represents the number of all reception antennas Mesh, k represents the number of inefficacy antenna, I representation unit matrix, and 0 represents null matrix.