CN107750445A

CN107750445A - Receiver apparatus and its method

Info

Publication number: CN107750445A
Application number: CN201580080995.XA
Authority: CN
Inventors: 陈建军; 弗雷德里克·鲁塞克; 陈俊仕
Original assignee: Huawei Technologies Co Ltd
Current assignee: Huawei Technologies Co Ltd
Priority date: 2015-07-23
Filing date: 2015-07-23
Publication date: 2018-03-02
Also published as: WO2017012671A1

Abstract

The present invention relates to a kind of receiver apparatus and its method.Methods described 200 includes following key step：Receive signal of communication (communication signal, CS)；Multiple interference source n=1 in the signal of communication (communication signal, CS) are determined ..., N；The covariance that interference source n=1 is obtained for interference source n=1 during iteration i=1 is estimated (referring to Fig. 2 B)；It is interference source n=2 in iteration i=1 ..., N obtains interference source n=2 ..., N covariance is estimated (referring to Fig. 2 C)；In iteration i=2 ..., I, when be the multiple interference source n=1 ..., N obtains iteration i=2 ..., interference source n=1 during I ..., at least one covariance estimations of N are (referring to Fig. 2 D).The multiple interference source n=1 during based on iteration i=I ＞ 1 ..., N covariance estimation calculates total covariance estimation R of (212) described signal of communication (communication signal, CS)_yy.Moreover, it relates to a kind of computer program and a kind of computer program product.

Description

Receiver apparatus and its method

Technical field

The present invention relates to a kind of receiver apparatus.Moreover, it relates to a kind of corresponding method, a kind of computer program A kind of and computer program product.

Background technology

In Long Term Evolution (Long Term Evolution, LTE) system, third generation partner program (3rd Generation Partnership Project, 3GPP) propose frequency reuse for 1.Therefore, in this kind of LTE system, Interference would generally be very high.Descending, user equipment (User Equipment, UE) can by adjacent e-NodeB (e-NodeB, ENB interference).Descending interference can reduce the performance of UE receivers.

Therefore, the problem of LTE is descending generally acknowledged be exactly UE not only in the cell of oneself from eNB reception signals, but also from Multiple neighbor cell reception signals for being referred to as interfered cell.The common methods for solving this interference are to pass through interference rejection combination (Interference Rejection Combining, IRC) technology.In IRC, algorithm does not attempt decoding interfering signals, and Only it is an attempt to white noise and adds interference signal.In this respect, the covariance matrix of the necessary estimated disturbance signals of UE.

Covariance matrix has long abundant history in wireless communication field, and in LTE transmission pattern Under (Transmission Modes, TM) 3 and 4, the problem is simplified to a certain extent, because some portions of covariance matrix It is known to divide.Because the physical channel of interfered cell highlights correlations over time and frequency, actually UE can be well Physical channel is estimated.But eNB changes power amplification (Power Amplification, PA) electricity with speed quickly Gentle pre-coding matrix (Precoding Matrix, PM), so that in the case of known physical channel, can not learn The covariance matrix of interference.

Therefore, the problem of estimating to be converted into PA and PM of detection interfered cell the problem of interference covariance matrix.Another Problem is the PA that eNB is not used in training data position using the PA at PM, and training data and in payload data opening position May be different.Therefore, PA and PM detections can not use training data position.

Except that by using IRC, can also eliminate and suppress in recent LTE features, i.e. network assistance interference Attempted in the environment of (Assisted Interference Cancellation and Suppression, NAICS) to interference The signal of cell carries out actual decoding, then eliminates them.This especially has when at least one interfered cell is stronger than serving cell It is attractive.At this time, it may be necessary to be not interference covariance matrix but effective channel matrix.

However, this has just been summed up in the point that the problem of same, i.e., build association's hard decision by detecting the PA and PM of interfered cell Covariance matrix, or the one group of probable value for the PA and PM for passing through weighted array interfered cell are estimated to build the soft of covariance matrix Meter.

On one public reference signal (Common Reference Signal, CRS) interfered cell to conflict of detection Some previous traditional schemes be present in PA, PM and covariance matrix.However, when the CRS interfered cells of multiple conflicts be present, These traditional schemes do not work.Or the method in these traditional schemes is under all orders to all possible PA values and PM values Using exhaustive search, otherwise assuming all interference signals of multiple interfered cells all has identical PA values or identical PM values, so Substantially reduce the search space of parameter combination.

Above-mentioned traditional scheme mainly has two shortcomings.First shortcoming is that computation complexity is high.When multiple interference sources being present When, the complexity of exhaustive search PA parameters and PM parameters is very high.Second shortcoming is performance, because it was assumed that all interference signals It is unpractical with identical PA level, it is meant that performance reduces.

The content of the invention

The purpose of the embodiment of the present invention is to provide a kind of scheme alleviated or solve the shortcomings that traditional scheme and problem.

In particular, the embodiment of the present invention is it is intended that improved receiver performance provides improved association side in wireless communication system Difference estimation.

"or" in this specification and corresponding claims be interpreted as covering "AND" and "or" mathematically "or" and should not be construed as XOR (different OR).

Above-mentioned purpose and other purposes solve by the theme of independent claims.More beneficial implementations of the present invention Form can be found in the dependent claims.

According to a first aspect of the present invention, the above-mentioned purpose referred to and other purposes pass through the reception for wireless communication system Device equipment is realized.The receiver apparatus includes：

Receiver, it is used for：

Receive signal of communication；

Processor, it is used for：

Multiple interference source n=1 in the signal of communication are determined ..., N；

It is interference source n=1 in iteration i=1：

Calculate power amplification (power amplification, PA) and pre-coding matrix (Precoding Matrix, PM) To each may combination error covariance and corresponding covariance estimate, wherein the error covariance is to be based on the communication Signal calculates,

The measurement of each error covariance is calculated,

The set of computation measure and corresponding covariance estimation is formed,

The first subset is selected in the set of the computation measure and corresponding covariance estimation,

The likelihood of each computation measure in first subset is calculated,

Covariance estimation in first subset is combined with their own likelihood, done with obtaining during iteration i=1 Disturb source n=1 covariance estimation；

It is interference source n=2 in iteration i=1 ..., N：

Calculate PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein the error is assisted Variance is previous interference source m=1 when being based on the signal of communication and iteration i=1 ..., n-1 covariance is estimated to count Calculate,

The measurement of each error covariance is calculated,

The likelihood of each computation measure in first subset is calculated,

Covariance in first subset is estimated that their own likelihood is combined, to obtain interference source n= 2 ..., N covariance estimation；

In iteration i=2 ..., it is the multiple interference source n=1 during I ..., N：

Calculate PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein the error is assisted Previous interference source m=1 when variance is based on the signal of communication, current iteration i ..., n-1 covariance estimation and preceding Subsequent interference source m=n+1 during one iteration i-1 ..., N covariance is estimated to calculate,

The measurement of each calculation error covariance is calculated,

The likelihood of each computation measure in first subset is calculated,

Covariance estimation in first subset is combined with their own likelihood, to obtain iteration i= 2 ..., interference source n=1 during I ..., N at least one covariance estimation；

The multiple interference source n=1 during based on iteration i=I ＞ 1 ..., N covariance is estimated to calculate described lead to Believe total interference covariance estimation of signal

Compared to traditional scheme, this receiver apparatus provides many advantages.Compared to traditional scheme, by providing iterative scheme To estimate interference covariance (matrix) to realize that low computation complexity is possible.In addition, comparing traditional scheme, this programme is also Provide estimation more accurately covariance.Due to can preferably alleviate or reduce interference, so improved receiver performance is can Can.

According to described in a first aspect, in the first possible form of implementation of receiver apparatus, the processor is performing It is additionally operable to before iteration i=1：

According to respective receiving power by the multiple interference source n=1 ..., N sorts in descending order.

The advantages of the first described possible form of implementation is：By according to respective receiving power by the multiple interference source N=1 ..., N sorts in descending order, and more accurately covariance can be estimated.

According to the first aspect the first possible form of implementation or according to as described above in a first aspect, receiving In the possible form of implementation of second of device equipment, the processor is additionally operable to：

Based on Fu Luobin Nice norm of respective channel estimation by the multiple interference source n=1 ..., N is in descending order Sequence.

It is described second may form of implementation the advantages of be：Described second possible form of implementation is easily realized, because only Fu Luobin Nice norm is achieved that by the way that the power of all elements in matrix is added.

In the foregoing possible form of implementation of the first aspect any one or according to as described above in a first aspect, In the third possible form of implementation of receiver apparatus, the likelihood of each computation measure is the energy based on the computation measure Content.

The advantages of the third described possible form of implementation is：The third described possible form of implementation is easily achieved and possessed good Good performance, because the likelihood by using the energy content based on computation measure, it is not necessary to matrix inversion.

In the foregoing possible form of implementation of the first aspect any one or according to as described above in a first aspect, In the 4th kind of possible form of implementation of receiver apparatus, the computation measure is Fu Luobin Nice model of the error covariance Number F.

The advantages of 4th kind of possible form of implementation is：The 4th kind of possible form of implementation is easily achieved and possessed good Good performance, because only achieving that Fu Luobin Nice norm by the way that the power of all elements in matrix is added.

According to the 4th kind of possible form of implementation of the first aspect, the 5th kind in receiver apparatus may implement In form, the PA and PM are to α_n, W_nInterference source n Fu Luobin Nice norm F of the error covariance be according to as inferior Formula calculates：

WhereinFor the signal of communication (communication signal, CS) covariance, R_yy(α_n, W_n) it is total Covariance is estimated.

The advantages of 5th kind of possible form of implementation is：By using this equation, can easily and quickly find really The distance between covariance and estimate covariance.

In the foregoing possible form of implementation of the first aspect any one or according to as described above in a first aspect, In the 6th kind of possible form of implementation of receiver apparatus, each first subset includes the computation measure and corresponding covariance is estimated K minimum of computation measurement in the set of meter.

K is less than the sum of computation measure.

The advantages of 6th kind of possible form of implementation is：Performance is only improved by using K minimum of computation measurement.This Outside, computation complexity is also reduced using the possible form of implementation.

In the foregoing possible form of implementation of the first aspect any one or according to as described above in a first aspect, In the 7th kind of possible form of implementation of receiver apparatus, the processor is additionally operable to calculate the communication letter in the following manner Number total interference covariance estimation

The multiple interference source n=1 during iteration I ..., the estimation of N covariance, serving cell covariance and make an uproar Sound covariance is added.

The advantages of 7th kind of possible form of implementation is：It is low compared to traditional algorithm, computation complexity.Estimated by iteration Meter has also gradually stepped up performance.

According to the 5th kind of possible form of implementation of the first aspect, the 8th kind in receiver apparatus may implement In form, the processor is additionally operable to：

Formation includes the multiple interference source n=1 during iteration I ..., N all K measurements and their corresponding association sides The common set of difference estimation；

At least one measurement and its corresponding covariance in common set based on the measurement and corresponding covariance estimation Estimation calculates total interference covariance estimation of the signal of communication

The advantages of 8th kind of possible form of implementation is：By only taking K^NIt is individual measurement and not all M^NIt is individual to combine, energy Reduce complexity, the quantity for the possibility combination that wherein M is the PA and PM of an interference source.

According to the 8th kind of possible form of implementation of the first aspect, the 9th kind in receiver apparatus may implement In form, the processor is additionally operable to：

Select the yield in the second subset in the measurement and the common set of corresponding covariance estimation；

Total interference association side of the signal of communication is calculated based on the measurement in the yield in the second subset and corresponding covariance estimation Difference estimation

The advantages of 9th kind of possible form of implementation is：Due to have selected the yield in the second subset of measurement, can further reduce Complexity.

According to the 9th kind of possible form of implementation of the first aspect, the tenth kind in receiver apparatus may implement In form, the processor is additionally operable to：

The likelihood each measured in the yield in the second subset is calculated, wherein the likelihood is the energy content based on measurement；

The communication letter is calculated by the way that the covariance estimation in the yield in the second subset is combined with their own likelihood Number total interference covariance estimation

The advantages of described ten kind of possible form of implementation is：By by soft likelihood soft combination, there is provided more preferable covariance Estimation.

According to the 9th kind of possible form of implementation of the first aspect, in a kind of the tenth possible reality of receiver apparatus Apply in form, the processor is additionally operable to：

The likelihood each measured in the yield in the second subset is calculated, wherein the likelihood is the maximum likelihood (Maximum of measurement Likelihood, ML)；

A kind of the advantages of ten possible form of implementation is：Such as compared to the method based on energy, pass through the ML side Method can provide better performance.However, the complexity of the ML methods is higher.

In the foregoing possible form of implementation of the first aspect any one or according to as described above in a first aspect, In the 12nd kind of possible form of implementation of receiver apparatus, in iteration i=1, interference source n=1 covariance estimation is according to such as Lower equation calculates：

WhereinEstimate for weighting covariance,For normalization factor, interference source n=2 during iteration i=1 ..., N covariance estimation plus Upper serving cell covariance and noise covariance calculate according to following equation：

WhereinFor the covariance of serving cell,For the interference source n of current iteration Covariance estimation,For the previous interference source m=1 of current iteration ..., n-1 covariance estimation, R_uuFor The covariance of noise.

The advantages of 12nd kind of possible form of implementation is：The previous of other interference sources is started without in alternative manner In the case of information, initial covariance estimation can be provided.

According to the 12nd kind of possible form of implementation of the first aspect, in the 13rd kind of possibility of receiver apparatus In form of implementation, total covariance estimation R of signal of communication during iteration i=I ＞ 1_yyCalculated according to following equation：

WhereinFor the covariance of the serving cell,For the current iteration Interference source n covariance estimation,For the interference source m=1 of the current iteration ..., n-1 covariance is estimated Meter,For the interference source m=n+1 of preceding iteration ..., N covariance estimation, R_uuFor the association of the noise Variance.

The advantages of 13rd kind of possible form of implementation is：Estimated by iteratively updating covariance according to the equation, Improved covariance estimation can be provided.

According to a second aspect of the present invention, above-mentioned referring to and other purposes are by the method for wireless communication system come real It is existing.Methods described includes：

Receive signal of communication；

It is interference source n=1 in iteration i=1：

The measurement of each error covariance is calculated,

The likelihood of each computation measure in first subset is calculated,

It is interference source n=2 in iteration i=1 ..., N：

The measurement of each error covariance is calculated,

The likelihood of each computation measure in first subset is calculated,

Covariance estimation in first subset is combined with their own likelihood, to obtain interference source n= 2 ..., N covariance estimation；

The measurement of each calculation error covariance is calculated,

The likelihood of each computation measure in first subset is calculated,

The multiple interference source n=1 during based on iteration i=I ＞ 1 ..., N covariance estimation calculates the communication Total interference covariance estimation of signal

According to the second aspect, in the first possible form of implementation of method, methods described is performing iteration i=1 Also include before：

The first possible form of implementation or according to second aspect as described above according to the second aspect, in method Second may be in form of implementation, methods described also includes：

In the foregoing possible form of implementation of the second aspect any one or according to second aspect as described above, In the third possible form of implementation of method, the likelihood of each computation measure is the energy content based on the computation measure.

In the foregoing possible form of implementation of the second aspect any one or according to second aspect as described above, In the 4th kind of possible form of implementation of method, the computation measure is Fu Luobin Nice norm F of the error covariance.

According to the 4th kind of possible form of implementation of the second aspect, in the 5th kind of possible form of implementation of method In, the PA and PM are to α_n, W_nInterference source n Fu Luobin Nice norm F of the error covariance be according to following equation meter Calculate：

In the foregoing possible form of implementation of the second aspect any one or according to second aspect as described above, In the 6th kind of possible form of implementation of method, each first subset includes the computation measure and the collection of corresponding covariance estimation K minimum of computation measurement in conjunction.

In the foregoing possible form of implementation of the second aspect any one or according to second aspect as described above, In the 7th kind of possible form of implementation of method, total interference covariance estimation of the signal of communicationBased in the following manner Calculate：

According to the 5th kind of possible form of implementation of the second aspect, in the 8th kind of possible form of implementation of method, Methods described also includes：

According to the 8th kind of possible form of implementation of the second aspect, in the 9th kind of possible form of implementation of method In, methods described also includes：

According to the 9th kind of possible form of implementation of the second aspect, in the tenth kind of possible form of implementation of method In, methods described also includes：

According to the 9th kind of possible form of implementation of the second aspect, in a kind of the tenth possible form of implementation of method In, methods described also includes：

In the foregoing possible form of implementation of the second aspect any one or according to second aspect as described above, In the 12nd kind of possible form of implementation of method, in iteration i=1, interference source n=1 covariance is estimated according to following equation Calculate：

WhereinFor the covariance of serving cell,For the interference source n of current iteration Covariance estimation,For the previous interference source m=1 of current iteration ..., n-1 covariance estimation, R_uuTo make an uproar The covariance of sound.

According to the 12nd kind of possible form of implementation of the second aspect, the 13rd kind in method may implement shape In formula, total covariance estimation R of the signal of communication in iteration i=I ＞ 1_yyCalculated according to following equation：

As the advantages of receiver apparatus that the advantages of method that the second aspect provides provides with the first aspect.

The present invention also relates to a kind of computer program of code form, when running the computer program by processing component So that the processing component performs any method provided by the invention.Further, the present invention also relates to computer program product, Including computer-readable medium and the computer program.Wherein described computer program is included in computer-readable Jie In matter, and including read-only storage (Read-Only Memory, ROM), programmable read only memory (Programmable ROM, PROM), Erasable Programmable Read Only Memory EPROM (Erasable PROM, EPROM), flash memory, electric erazable programmable is read-only deposits One or more of reservoir (Electrically EPROM, EEPROM) and hard disk drive this group.

The other application and advantage of the present invention in following illustrate it will be evident that.

Brief description of the drawings

Accompanying drawing is intended to illustrate and explains every embodiment of the present invention, wherein：

Fig. 1 shows receiver apparatus according to embodiments of the present invention；

Fig. 2A to Fig. 2 D shows method according to embodiments of the present invention；

Fig. 3 shows wireless communication system according to embodiments of the present invention；

Fig. 4 to Fig. 6 shows the results of property of the embodiment of the present invention.

Embodiment

Fig. 1 shows receiver apparatus 100 according to embodiments of the present invention.In the present example embodiment, receiver is set Standby 100 are communicably coupled to the processor 102 of receiver 104 including the use of communication means 108.In Fig. 1, communication means 108 are illustrated as the dotted arrow between processor 102 and receiver 104.Communication means 108 are according to well known in the prior art Technology.For example, communication means 108 can be used for transmitting data or control signaling between processor 102 and receiver 104.At this In specific embodiment, user equipment 100 also includes the control member 110 that processor 102 operates (or control) receiver 104.Should Control member arrives the arrow diagramming of receiver 104 using processor 102.User equipment 100 also includes being coupled to receiver 104 Antenna member 106, to be received within the wireless communication system 300.Alternatively, receiver 104 can be wireless communication system 300 In be used for the part of transceiver that receives and transmit.

According to this programme, receiver 104 is used to receive the signal of communication transmitted in wireless communication system 300 (communication signal, CS), as shown in Figure 1.Processor 102 is used to determine signal of communication (communication Signal, CS) in multiple interference source n=1 ..., N.According to one embodiment, multiple interference source n=1 ..., N It can be the interfered cell in the cellular system such as LTE.Disturbance cell has the unique cell mark of different each cells Know (cell Identity, ID).In LTE system, cell ID be included in synchronizing channel (Synchronous Channel, SCH in).Therefore by detecting SCH, cell ID can be obtained；After cell ID is obtained, the common reference of interfered cell can be obtained Signal (Common Reference Signal, CRS) and corresponding channel estimation.

Processor 102 is additionally operable to perform alternative manner, to provide signal of communication (communication signal, CS) Total interference covariance estimationThe total interference covariance estimation providedIt can be used to reducing, mitigate or suppressing interference, so as to carry High systematic function.For example, this programme can be used for interference rejection combination (Interference Rejection Combining, IRC) algorithm.Therefore, according to embodiments of the present invention, processor is additionally operable to estimate total interference covarianceReduce, subtract as interference Light or restrainable algorithms inputs.

Therefore, the processor 102 of receiving device 100 is additionally operable to：It is interference source n=1 in iteration i=1：

Calculate power amplification PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein missing Poor covariance is calculated based on signal of communication (communication signal, CS)；

Calculate the measurement of each error covariance；

Form the set of computation measure and corresponding covariance estimation；

The first subset is selected in the set of computation measure and corresponding covariance estimation；

Calculate the likelihood of each computation measure in the first subset；

, during the obtaining iteration i=1 interference source combined with their own likelihood by the covariance estimation in the first subset N=1 covariance estimation.

Therefore, the processor 102 of receiving device 100 is additionally operable to：It is interference source n=2 in iteration i=1 ..., N：

Calculate PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein error covariance It is previous interference source m=1 when being based on signal of communication (communication signal, CS) and iteration i=1 ..., n- 1 covariance is estimated to calculate；

Calculate the measurement of each error covariance；

Form the set of computation measure and corresponding covariance estimation；

Calculate the likelihood of each computation measure in the first subset；

Covariance estimation in first subset is combined with their own likelihood, to obtain interference source n= 2 ..., N covariance estimation.

Therefore, the processor 102 of receiving device 100 is additionally operable to：In iteration i=2 ..., it is interference source n=during I 1 ..., N：

Calculate PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein error covariance Previous interference source m=1 when being based on signal of communication, current iteration i ..., n-1 covariance estimation and preceding iteration i- Subsequent interference source m=n+1 when 1 ..., N covariance is estimated to calculate；

Calculate the measurement of each calculation error covariance；

Form the set of computation measure and corresponding covariance estimation；

Calculate the likelihood of each computation measure in the first subset；

Covariance estimation in first subset is combined with their own likelihood, to obtain iteration i=2 ..., Interference source n=1 during I ..., N at least one covariance estimation.

Finally, the processor 102 of receiving device 100 is additionally operable to：Multiple interference source n=during based on iteration i=I ＞ 1 1 ..., N covariance estimation calculates total interference covariance estimation of signal of communication (communication signal, CS)

Fig. 2A to Fig. 2 D shows corresponding method 200.Method 200 can be held in receiver apparatus 100 as shown in Figure 1 OK.This alternative manner can be divided into several main steps including sub-step.That is, step 206：Interference source n is obtained in iteration i=1 =1 covariance estimation, referring to Fig. 2 B.Then, step 208：Interference source n=2 is obtained in iteration i=1 ..., n association Variance evaluation, referring to Fig. 2 C.Then, step 210：In iteration i=2 ..., interference source n=1 is obtained during I ..., n's Covariance is estimated, referring to Fig. 2 D.Finally, total interference covariance of signal of communication (communication signal, CS) is calculated EstimationAccording to one embodiment of the invention, a calculation in three termination algorithms being described in more below by using the present invention Method is estimated to calculate total interference covariance

Therefore, this method 200 includes following key step：

Receive signal of communication (communication signal, CS)；

Multiple interference source n=1 in signal of communication (communication signal, CS) are determined ..., N；

The covariance that interference source n=1 is obtained for interference source n=1 in iteration i=1 is estimated (referring to Fig. 2 B)；

It is interference source n=2 in iteration i=1 ..., N obtains interference source n=2 ..., N covariance estimation (referring to Fig. 2 C)；

In iteration i=2 ..., it is multiple interference source n=1 during I ..., N：When obtaining iteration i=2 ... ..., I Interference source n=1 ..., N at least one covariance is estimated (referring to Fig. 2 D)；

Multiple interference source n=1 during based on iteration i=I ＞ 1 ..., N covariance estimation calculates (212) signal of communication Total covariance estimation R of (communication signal, CS)_yy。

With reference to Fig. 2 B, method 200 is further comprising the steps of：

Calculate the measurement of each error covariance；

Form the set of computation measure and corresponding covariance estimation；

Calculate the likelihood of each computation measure in the first subset；

With reference to Fig. 2 C, method 200 is further comprising the steps of：

Calculate the measurement of each error covariance；

Form the set of computation measure and corresponding covariance estimation；

Calculate the likelihood of each computation measure in the first subset；

, during the obtaining iteration i=1 interference source combined with their own likelihood by the covariance estimation in the first subset N=2 ..., N covariance estimation.

With reference to Fig. 2 D, method 200 is further comprising the steps of：

Calculate the measurement of the error covariance each calculated；

Form the set of computation measure and corresponding covariance estimation；

Calculate the likelihood of each computation measure in the first subset；

In one embodiment of the invention, receiver apparatus 100 can be user equipment (User Equipment, UE), Any one in mobile station (mobile station, MS), wireless terminal or mobile terminal, they can be otherwise referred to as Radio communication is carried out in the wireless communication system of cellular wireless system.UE is also referred to as having the cell phone of wireless capability, honeybee Nest mobile phone, tablet personal computer or notebook computer.Such as UE can be portable, pocket-type, hand-held, computer in the present context Comprising formula or car-mounted mobile devices, it can be entered with another entity such as another receiver or server by Radio Access Network Row voice or data communication.UE can be base station (Station, STA), and it is wireless comprising the connection for meeting IEEE802.11 standards Medium education (Media Access Control, the MAC) interface and physical layer of medium (Wireless Medium, WM) Any equipment of (Physical Layer, PHY) interface.

Fig. 3 shows wireless communication system 500 according to embodiments of the present invention.Receiver apparatus 100 is shown as UE in figure 3 100.UE 100 receives the signal of communication (communication for including useful signal from the serving BS 400a in serving cell Signal, CS).However, in this specific example, signal of communication (communication signal, CS) also includes coming respectively From neighboring interfering base station 400b and 400c interference signal.Due to interference source be present, receiver performance can decline in UE 100.So And according to this programme, by using alternative manner in UE 100, there is provided improved receiver performance.

Base station 400 can be (wireless) network node or access node or access point or base station, such as wireless base station (Radio Base Station, RBS), transmitter, " eNB ", " eNodeB ", " B node ", depending on using are can be described as in some networks Technology and term.Wireless network node can be divided into variety classes, such as grand eNodeB, Home eNodeB or femto base station, be based on Depending on transimission power, thus also based on cell size depending on.Wireless network node can be base station (Station, STA), and it is bag Medium education (Media containing the connection wireless medium (Wireless Medium, WM) for meeting the standards of IEEE 802.11 Access Control, MAC) interface and physical layer (Physical Layer, PHY) interface.

In addition, in following discloses, the embodiment of the present invention is described using LTE contexts.Therefore, LTE arts have been used Language, system concept etc..However, it should be understood that this programme is not limited to LTE system, but it can be used for any suitable channel radio Letter system.In addition, in the embodiment being described below, this receiver apparatus 100 is expressed as UE.

The scene that serving cell is influenceed by N+M interfered cell is contemplated herein.N number of cell in these cells is conflict Interference source, that is, refer to the training symbol of these cells and overlapped over time and frequency with the training symbol of serving cell.M interference is small Area is non conflicting, that is, the training symbol of the CRS training symbols and serving cell that refer to these cells does not weigh over time and frequency Close.In any resource particle (Resource Element, RE), the reception signal of the UE in serving cell can be described as：

In equation (1), index and (k, l) is represented in given a pair of resource blocks (Pair of Resource Block, PRB) The 1st RE in k-th of RE and frequency in the middle time.Variable α₀、α_nWith α '_mRespectively represent eNB be used for serving cell, n-th Conflict districts, the PA values of m-th non conflicting cell.Matrix H represents to represent from eNB to serving cell UE channel matrix, matrix W In the PM selected by eNB, vector s represents the vector of transmission, and w has covariance matrix N for complicated₀I Gaussian noise.

In addition assume that all channel matrixes keep constant in whole PRB in equation (1).Due to LTE specification, according to fixed Justice, PM and PA are constant in a PRB.For the sake of simplicity, it should only consider that all eNB are equipped with 4 in wireless communication system The scene of individual antenna.However, those skilled in the art are it is appreciated that the scheme proposed can also be applied to other antennas matches somebody with somebody Put.

Noise and non conflicting interference are expressed as：

So, equation (1) can be written as：

It may be assumed that w '_{K, l}Covariance matrix be completely known, herein by R_uuRepresent, and assume channel matrix H_nTo 0 ≤ n≤N is also known.I.e. channel matrix is known for serving cell, is also known to all conflict interfered cells.

Assuming that interference channel and interference covariance matrix R_uuInterference sections be known because channel matrix is slowly to send out Raw change, therefore interpolating method can be used in several PRB.In training symbol position because all training symbols to UE and Speech is known, can realize iterative interference cancellation technology according to embodiments of the present invention.This kind of alternative manner can be with very high accurate Degree estimates all channel H_{N=0,1 ..., N}, so as to which this is not estimation interference covariance matrix R_uuBottleneck.Enter when to all channels Go estimation, their influences to training data position can be cancelled, then can be to covariance matrix R_uuEstimated.The essence of estimation Exactness is high, because remaining signal only has w ' after offsetting.

Although interference channel matrix may be assumed that to be known, PA and PM are not.Because these can dash forward on PRB borders So change, UE is restricted to single PRB to carry out their monitoring by this.The PA values and PM values of serving cell, i.e. α₀And W₀For It is known, because they are embedded in control data.

This programme embodiment provides the PA values and PM values for detecting interfered cell, i.e. 1≤n≤N α_nAnd W_n, equipment And method.As it was earlier mentioned, the PA in training data position is different from the PA in payload data position, training data is not used PM.Therefore detection PA and PM is only limited to payload data position herein.

In IRC, care is not PA and PM, but the covariance matrix of interference signal, i.e.,

Estimate covariance matrix is not same problem with estimation PA and PM, because covariance matrix R best estimate may It is not corresponding with PA and PM any efficient set.Method will be investigated, most probable PA and PM are attached to estimation association In variance matrix R.

According to 3GPP specifications, PA values can take one in 8 values.But interfered cell can be the free time, so as to α_n=0. In training symbol, interfered cell is never idle, but has to be by known non-zero PA values transmission.Therefore, saying below is simultaneously Not contradiction：N value is known, but regardless of α_n=0 whether the value for being true N is unknown.Promote over time, interfered cell Quantity can be stablized gradually, but in a PRB, conflict interfered cell may not transmit any payload data.Therefore, palpus will α_n=0 this value is taken into account as possibility, selects to have obtained from：

α∈{0、-6dB、-4.77dB、-3dB、-1.77dB、0dB、1dB、2dB、3dB}

For TM4 4x4,64 PM be present, but for all order 4PMI,Therefore 49 are only existed It is individual differentIn view of possible PA quantity, in the case of the non-transmission signal of some cells, then PA and PM may group The total quantity of conjunction is：

(8×49+1)^N

That is a conflict districts are 393, and two conflict districts are 154449.

Method based on energy during N=1

For the sake of simplicity, the method based on energy when this part discusses N=1.Then use is based in this iterative algorithm The method computation measure of energy.Pay attention to, the method based on energy can also be used during N ＞ 1.

In the RE comprising payload data, as the unknown α₁And W₁The covariance matrix of function be equal to

WhereinFor the covariance matrix of serving cell.

Sample mean covariance matrix based on payload data is：

Wherein it is the payload data T to carrying RE_dSummation.

Method based on energy is given by

Wherein

Should it is found out that, for full order precoder W₁, quantity can be constructedTherefore these PM can not be mutual Distinguish, so as to its enough 8 × 49+1=393 measurement of limit.This is feasible, therefore will not pursue any reduction complexity Method.The calculating for reusing this 393 measurement estimations is implicitly present in tremendous potential.Specifically, can be high for given PM Find optimal PA values to effect.

According to the above, the estimation of the covariance matrix of interference signal can be obtained directly according to following formula：

By carrying out soft merging to most probable PA and PM, the estimation can be improved, explanation sees below.Norm F (α₁, W₁) can Writing：

Wherein ∈_{I, j}For error co-variance matrixElement.If as can be seen that basis association Variance matrix R_yyDiagonal entry it is the same, thenAll elements there is the same variance.To make problem mathematically good Processing, it is assumed that this still sets up.

In addition in this case, R_yyIt is unknown, because PA and PM are unknown.To continue in next step, it is assumed that all Diagonal entry is the same and error term ∈_{I, j}For Gaussian Profile, then in view of R_yy(α₁, W₁), error { ∈_{I, j}Seemingly { ∈ can so be passed through_{I, j}Be expressed as：

Wherein δ²For the variance of diagonal entry, γ is normaliztion constant, can be omitted in later step, to reach association side The soft merging of poor Matrix Estimation.Due to the rapid decay of exponential function, only including several has minimum F (α₁, W₁) measurement item just It is enough.Assuming that including T minimum metric, then the soft merging estimation of covariance matrix can be by drawing as follows：

WhereinTo generate t-th of minimum metric F (α₁, W₁) PA values and PM values combination.How variance is selected δ²Desired value have it is to be discussed.Selection this value beTwice of average value of diagonal entry, average value is by right The number T of diagonal element divided by data RE_dQuantity obtain.

Maximum likelihood method during N=1

For the sake of simplicity, this part discusses maximum likelihood (Maximum Likelihood, ML) side in a case of n=1 Method.Then ML method computation measures are used in this iterative algorithm.Pay attention to, ML methods can also be used during N ＞ 1.

Likelihood in method based on energy is not reception signal { y_{K, l}Likelihood, but obtain error { ∈_{I, j}Seemingly So.Detection can be produced into more preferable result based on real likelihood, it will be clear that also resulting in higher complexity.Such as Transmission signal is similar to complicated Gaussian Profile by fruit, then in view of PA and PM couples, reception signal { y_{K, l}Likelihood be equal to：

Wherein κ is normaliztion constant.Equally, log-likelihood function is equal to：

In view of likelihood function, PA and PM ML detections can be achieved by the following procedure：

For the optimal PA of K and PM pairs, interference covariance matrix estimation can be by drawing as follows：

Content as has been mentioned, it is N=1 for single interference source, possible PA and PM pairs of sum are 393, limit It is a feasible number.But in the method based on ML, each pair needs matrix inversion.For actually realizing, it can not recognize Complexity for ML methods is small.Therefore proposition first calculates the measurement F (α based on energy of this 393 couple₁, W₁), select K most preferably , then using this K to carrying out the PA based on ML and PM detections (or covariance estimation).

This iterative algorithm

The situation of this iterative scheme is used during concern N ＞ 1 below.As described above, exhaustive search needs to test (8*49+ 1)^N) individual combination.For N=2, it means that 154449 combinations, be evident as an infeasible numeral in practice.Below Iterative algorithm is proposed, the iterative algorithm is described for the method based on energy.Same iterative algorithm can be used for likelihood cost in the same old way Function, finally two methods are combined to reduce complexity.

As can be seen that the interference source relatively low compared to power, the parameter of the interference source of powerful can be estimated more accurate. The first step of algorithm is that interference source is sorted according to respective power.Because not obtaining their PA values, this is needed based on letter Road matrix H_nFu Luobin Nice norm complete.Assuming that this has been resolved, then | | H₁| | ＞ | | H₂| | ＞ ... | | H_N||。

Iteration i=1

As before, in first time iterative step, allowRepresent the estimation association side corresponding with n-th of interference source Poor matrix.In the first iteration, these covariance matrixes are unknown, it is therefore assumed that being all zero when they start, by the beginning of it Beginning turns to1≤n≤N (because not knowing any information of interference source in this stage).Now by R_yy(α_n, W_n) weight It is newly defined as：

WhereinIt is serving cell,It is interference source 's.

Then the PA and PM of the interference source of limit first are started to (α₁, W₁) all 393 may combination, and based on not sieve 393 measurements of guest Nice norm calculation.

In one embodiment of the invention, the likelihood of each computation measure is the energy content based on computation measure.

In another embodiment of the present invention, computation measure is Fu Luobin Nice norm F of error covariance.

Then these measurements are sorted, obtains K with PA and PM to (α₁, W₁) most probable combine corresponding minimum degree Amount.Optimal right based on this K, i.e. pair with minimum degree value updates covariance matrix according to equation (3) For：

Whereinδ²ForIt is diagonal The variance of line element, { ∈_{I, j}BeElement.

Then this step is repeated to remaining N-1 interference source in signal of communication (communication signal, CS).

For n-th of interference source, first by following renewal R_yy(α_n, W_n)：

WhereinIt is serving cell,It is interference source 's.Pay attention to, m ＜ n covariance matrixIt has been completed that, and and not equal to zero.

Then the PA and PM of n-th of interference source of limit are to (α_n, W_n) it is all 393 may combination, obtain F (α_n, W_n) K minimum metric, then update covariance matrix by mode same in first step

For convenience of calculation, it is pointed out here that for each W_n, can extremely efficiently calculate all α_nMeasurement because α_nIt is Scalar.

Above-mentioned steps require to calculate 393N measurement F (α_n, W_n), it is desirable to N number of estimate covariance matrix of each interference source1≤n≤N result.

In one embodiment of the invention, processor 102 is additionally operable to according to respective reception before iteration i=1 is performed Power by multiple interference source n=1 ..., N sorts in descending order.In an alternative selection, based on respective channel estimation Fu Luobin Nice norm by multiple interference source n=1 ..., N sorts in descending order.

Iterative step i=2 ..., N

In iterative step i ＞ 1, the step of repeating iteration 1 of above-mentioned part, difference is as limit interference source n, Interference source m ＞ n interference covariance matrix is obtained from preceding iteration i -1.R after updating herein_yy(α_n, W_n) by being expressed as below For：

WhereinIt is serving cell, It is interfered cell.

Generally, need to terminate alternative manner in a certain stage, then need the output in more applications or use communication letter Total interference covariance estimation of number (communication signal, CS)Hereinafter, the embodiments of the invention provide three The different algorithm of kind terminates mode.

Algorithm terminates mode 1

As last time iteration I to be terminated, each interference source n=1 ..., N has 393 PA and PM pairs.For knot Beam iterative algorithm, foregoing many approach are contemplated that.One simple termination mode is to be estimated as working as by total interference covariance matrix The summation of preceding estimation, i.e.,

Algorithm terminates mode 2

The another method for terminating this iterative algorithm is by the optimal to limit together of all interference sources.Assuming that each interference source Select K optimal PA and PM pairs.PA and PM is to a total of K^NIndividual combination.Obtained total covariance square will be established from the 1st combination Matrix representation is：

Then can be by K^NIndividual metric calculation is：

Then K optimal PA and PM pairs are preserved., now can be to these K by the direct extension of structure in peer-to-peer (3) Optimal PA and PM is to carrying out soft merging, to obtain being estimated as total interference covariance matrix：

Algorithm terminates mode 3

The another method for terminating this iterative algorithm be using ML frameworks, it is similar to N=1 when performed foregoing description.This Kind in the case of likelihood be：

Based on these likelihoods, total interference covariance matrix can be formedSoft estimation.Pay attention to, because each combination is related to square Battle array is inverted, so digital K must very little.

Equally, log-likelihood function is equal to：

In view of likelihood function, the ML detections of PA and PM pairs can be according to drawing as follows：

For the optimal PA of K and PM pairs, interference covariance matrix estimation can be according to drawing as follows：

The numeric results in the case of two conflict interfered cell i.e. N=2 are presented below in conjunction with Fig. 4 to Fig. 6.Described Fig. 4 X-axis into Fig. 6 represents the signal to noise ratio (Signal-to-Noise Ratio, SNR) of serving cell.Y-axis represents Fig. 4, figure respectively Error matrix norm (between real covariance matrix and estimate covariance matrix) PM detection probabilities, PA detections are general in 5 and Fig. 6 Rate and average error variance (Mean Square Error, MSE).Assuming that covariance matrix R_uuFor the unit matrix of scalarization, from And avoid specifying non conflicting interference source M quantity.In addition, the situation of N=2 i.e. two interfered cell is paid close attention to herein.Two Interfered cell is divided into the average SNR for being normalized to 12dB and 10dB, it is assumed that channel template delivers for the type-A extension of pedestrian's speed Instrument (Extended Vehicle A, EVA).

Fig. 4 and Fig. 5 shows the PM and PA of stronger two interfered cells detection probability.In all situations, including take All cells including business cell and interfered cell all employ 2 transmission of being lost, the random unified selection interference in possible 9 class value PA values.It should be noted that with serving cell SNR increase, testing result is worse and worse.Because interfered cell " flooding " In serving cell signal.Although H0, W0 and α 0 of serving cell is known, payload data is unknown, serving cell letter Number equivalent to estimation interference parameter noise.When signal and interference contribution increase, total covariance matrix R_yyElement into magnitude Increase.Therefore when the power of the power ratio interference signal of service signal is much larger, from R_yyThe data of estimated disturbance signal become It is more difficult.

As described above, Fig. 4 shows the PM testing results of two conflict interfered cells.Thick line is respectively the side based on energy The exhaustive search of method and ML methods.Result during an iteration i=1 shown in phantom, when solid line shows iteration i=3 three times Result.Fig. 5 shows the PA testing results that with Fig. 1 there is same simulation to set.Fig. 6 shows estimation interference covariance matrix MSE, simulation set such as Fig. 1.

Further, it can be seen that limit ML search than the method implementation effect based on energy (although complicated from Fig. 4 to Fig. 6 Spend highest) good a lot (about 4dB-6dB).In all situations, K values elect K=8 as, that is, calculate the most probable of two interfered cells PM and PA couples of K²=64 combinations.It can be seen that algorithm terminates the alternative manner (K based on ML of mode 3²=64 most Good combination) implementation effect is close to optimal.

It can be seen that algorithms of different terminates the influence of mode from Fig. 4 to Fig. 6.(it is based on it can be seen that algorithm terminates mode 3 ML it is) optimal, but it is also more complicated, is inverted because being related to 64 submatrixs.Algorithm terminates mode 2 (being based on energy) and compared Algorithm terminate mode 1 (summation) be have it is improved, then can at 64 because the K=8 that it considers each interference source is most preferably right Optimal joint combination can be generated in combination.In fact, terminating mode 1 using algorithm, 3 ratios of iteration terminate mode using algorithm The implementation effect of 2 iteration 1 time is poor.It can be concluded that from this：Terminated by joint assessment interference source critically important.

Although Fig. 4 and Fig. 5 consider hard output PM and PA testing results, total interference covariance square is considered in figure 6 Battle arrayEstimation.The measurement of proposition is MSE, i.e.,All parameters of this test cases are with Fig. 4 such as Fig. 5.Together Sample, it is well more many than limit ENERGY METHOD for this measurement, the implementation effect of limit ML search (although complexity highest).Enter One step, it is good than the exhaustive search implementation effect based on energy using the alternative manner of algorithm termination mode 3 (being based on ML), it is single It is especially true during secondary iteration I=1.It can be concluded that from this：To obtain low MSE, using ML frameworks rather than energy is based only on The method of amount is critically important.

Further, any method provided by the invention can it is a kind of with the computer program of code form in realize, Processing mode is caused to perform method and step when running computer program by processing mode.Computer program is included in computer journey Among the computer-readable medium of sequence product.Computer-readable medium can include any memory, such as read-only storage substantially (Read-Only Memory, ROM), programmable read only memory (Programmable Read-Only Memory, PROM), Erasable Programmable Read Only Memory EPROM (Erasable PROM, PROM), flash memory, Electrically Erasable Read Only Memory (Electrically Erasable PROM, EEPROM) and hard disk drive.

In addition, technical staff will realize that the receiver apparatus of the present invention is included such as function, component, unit, element The required communication capacity of form is for execution the solution of the present invention.Other such a components, unit, the example of element and function For：Processor, memory, buffer, logic control, encoder, rate matchers, derate adaptation, map unit, multiplexing Device, decision package, selecting unit, interchanger, interleaver, deinterleaver, modulator, demodulator, input, output, antenna, amplification Device, acceptor unit, transmitter unit, DSP, MSD, TCM encoder, TCM decoders, power supply unit, charger, communication connect Mouth, communication protocol etc., they, which are properly placed in, comes together to perform the program.

In particular, the processor of receiver apparatus of the invention may include such as CPU (Central Processing Unit, CPU), processing unit, process circuit, processor, application specific integrated circuit (Application Specific Integrated Circuit, ASIC), microprocessor or interpretable and execute instruction other processing logics One or more examples.Term " processor " therefore the process circuit for including multiple process circuits, the multiple processing can be represented Practical circuit is any, some or all of items listed above.The process circuit can further perform data processing work( Can, input, output and processing data, the function includes data buffering and device control function, for example, call treatment control System, user interface control etc..

Finally, it should be understood that the invention is not limited in above-described embodiment, but be related to simultaneously and be incorporated to appended independent right All embodiments in the range of claim.

Claims

1. one kind is used for the receiver apparatus of wireless communication system (300), it is characterised in that receiver apparatus (100) bag Include：

Receiver (104), is used for：

Receive signal of communication (communication signal, CS)；

Processor (102), is used for：

Multiple interference source n=1 in the signal of communication (communication signal, CS) are determined ..., N；

It is interference source n=1 in iteration i=1：

Calculate power amplification (power amplification, PA) and pre-coding matrix (Precoding Matrix, PM) to The error covariance and corresponding covariance that may each combine are estimated, wherein the error covariance is to be based on the signal of communication (communication signal, CS) come what is calculated,

The measurement of each error covariance is calculated,

The likelihood of each computation measure in first subset is calculated,

, during the obtaining iteration i=1 interference source combined with their own likelihood by the covariance estimation in first subset N=1 covariance estimation；

It is interference source n=2 in iteration i=1 ..., N：

Calculate PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein the error covariance It is previous interference source m=when being based on the signal of communication (communication signal, CS) and iteration i=1 1 ..., n-1 covariance is estimated to calculate,

The measurement of each error covariance is calculated,

The likelihood of each computation measure in first subset is calculated,

Calculate PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein the error covariance Previous interference source m=1 when being based on the signal of communication (communication signal, CS), current iteration i ..., N-1 covariance estimation and subsequent interference source m=n+1 during preceding iteration i-1 ..., N covariance is estimated to calculate ,

The measurement of each calculation error covariance is calculated,

The likelihood of each computation measure in first subset is calculated,

Covariance estimation in first subset is combined with their own likelihood, to obtain iteration i=2 ..., Interference source n=1 during I ..., N at least one covariance estimation；

The multiple interference source n=1 during based on iteration i=I ＞ 1 ..., N covariance estimation calculates the signal of communication Total interference covariance estimation of (communication signal, CS)

2. receiver apparatus (100) according to claim 1, it is characterised in that the processor (102) is performing iteration It is additionally operable to before i=1：

3. receiver apparatus (100) according to claim 1 or 2, it is characterised in that the processor (102) is additionally operable to：

Based on Fu Luobin Nice norm of respective channel estimation by the multiple interference source n=1 ..., N is arranged in descending order Sequence.

4. the receiver apparatus (100) according to preceding claims any one, it is characterised in that each calculating degree The likelihood of amount is the energy content based on the computation measure.

5. the receiver apparatus (100) according to preceding claims any one, it is characterised in that the computation measure is Fu Luobin Nice norm F of the error covariance.

6. receiver apparatus (100) according to claim 5, it is characterised in that the PA and PM are to α_n, W_nInterference source n The error covariance Fu Luobin Nice norm F be according to following equation calculate：

WhereinFor the signal of communication (communication signal, CS) covariance, R_yy(α_n, W_n) it is total covariance Estimation.

7. the receiver apparatus (100) according to preceding claims any one, it is characterised in that each first subset bag Include the K minimum of computation measurement in the set of the computation measure and corresponding covariance estimation.

8. the receiver apparatus (100) according to preceding claims any one, it is characterised in that the processor (102) it is additionally operable to calculate the signal of communication (communication signal, CS) total interference association side in the following manner Difference estimation

The multiple interference source n=1 during iteration I ..., N covariance estimation, serving cell covariance and noise association Variance Addition.

9. receiver apparatus (100) according to claim 6, it is characterised in that the processor (102) is additionally operable to：

Formation includes the multiple interference source n=1 during iteration I ..., N all K measurements and their corresponding covariances are estimated The common set of meter；

At least one measurement and its corresponding covariance in common set based on the measurement and corresponding covariance estimation are estimated Meter calculates total interference covariance estimation of the signal of communication (communication signal, CS)

10. receiver apparatus (100) according to claim 9, it is characterised in that the processor (102) is additionally operable to：

Signal of communication (the communication is calculated based on the measurement in the yield in the second subset and corresponding covariance estimation Signal, CS) total interference covariance estimation

11. receiver apparatus (100) according to claim 10, it is characterised in that the processor (102) is additionally operable to：

The signal of communication is calculated by the way that the covariance estimation in the yield in the second subset is combined with their own likelihood Total interference covariance estimation of (communication signal, CS)

12. receiver apparatus (100) according to claim 10, it is characterised in that the processor (102) is additionally operable to：

13. the receiver apparatus (100) according to preceding claims any one, it is characterised in that in iteration i=1 Interference source n=1 covariance estimation calculates according to following equation：

WhereinEstimate for weighting covariance, For normalization factor, the interference source n=2 in iteration i=1 ..., the estimation of N covariance plus serving cell covariance and Noise covariance calculates according to following equation：

WhereinFor the covariance of serving cell,For the interference source n of current iteration association side Difference estimation,For the previous interference source m=1 of current iteration ..., n-1 covariance estimation, R_uuFor noise Covariance.

14. receiver apparatus (100) according to claim 13, it is characterised in that the communication in iteration i=I ＞ 1 Total covariance estimation R of signal (communication signal, CS)_yyCalculated according to following equation：

WhereinFor the covariance of the serving cell,For the interference of the current iteration Source n covariance estimation,For the interference source m=1 of the current iteration ..., n-1 covariance estimation,For the interference source m=n+1 of preceding iteration ..., N covariance estimation, R_uuFor the association side of the noise Difference.

15. one kind is used for the method for wireless communication system (300), it is characterised in that methods described (200) includes：

Receive (202) signal of communication (communication signal, CS)；

(204) multiple interference source n=1 in the signal of communication (communication signal, CS) are determined ..., N；

It is interference source n=1 in iteration i=1：

Calculate (206a) power amplification (power amplification, PA) and pre-coding matrix (Precoding Matrix, PM) to each may combination error covariance and corresponding covariance estimate, wherein the error covariance is based on described Signal of communication (communication signal, CS) come what is calculated,

The measurement of (206b) each error covariance is calculated,

The set of (206c) computation measure and corresponding covariance estimation is formed,

(206d) first subset is selected in the set of the computation measure and corresponding covariance estimation,

The likelihood of each computation measure in (206e) described first subset is calculated,

Covariance estimation in first subset and their own likelihood is combined (206f), during obtaining iteration i= Interference source n=1 covariance estimation；

It is interference source n=2 in iteration i=1 ..., N：

Calculate (208a) PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein the error Covariance is previous interference source m=when being based on the signal of communication (communication signal, CS) and iteration i=1 1 ..., n-1 covariance is estimated to calculate,

The measurement of (208b) each error covariance is calculated,

The set of (208c) computation measure and corresponding covariance estimation is formed,

(208d) first subset is selected in the set of the computation measure and corresponding covariance estimation,

The likelihood of each computation measure in (208e) described first subset is calculated,

Covariance estimation in first subset and their own likelihood is combined (208f), to obtain interference source n= 2, ..., N covariance estimation；

Calculate (210a) PA and PM pairs of the error covariance that may each combine and corresponding covariance is estimated, wherein the error Previous interference source m=when covariance is based on the signal of communication (communication signal, CS), current iteration i 1......, n-1, covariance estimation and subsequent interference source m=n+1 during preceding iteration i-1 ..., N covariance is estimated Count to calculate,

The measurement of (210b) each calculation error covariance is calculated,

The set of (210c) computation measure and corresponding covariance estimation is formed,

(210d) first subset is selected in the set of the computation measure and corresponding covariance estimation,

The likelihood of each computation measure in (210e) described first subset is calculated,

Covariance estimation in first subset and their own likelihood is combined (210f), to obtain iteration i =2 ..., interference source n=1 during I ..., N at least one covariance estimation；

The multiple interference source n=1 during based on iteration i=I ＞ 1 ..., N covariance estimation calculates (212) described communication Total interference covariance estimation of signal (communication signal, CS)

16. a kind of computer program with program code, it is characterised in that when the computer program is run on computers When, for the method described in perform claim requirement 15.