CN103379055B - interference rejection combining method and device - Google Patents
interference rejection combining method and device Download PDFInfo
- Publication number
- CN103379055B CN103379055B CN201210132127.XA CN201210132127A CN103379055B CN 103379055 B CN103379055 B CN 103379055B CN 201210132127 A CN201210132127 A CN 201210132127A CN 103379055 B CN103379055 B CN 103379055B
- Authority
- CN
- China
- Prior art keywords
- interpolation
- factor
- matrix
- data
- precomputation
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Landscapes
- Noise Elimination (AREA)
- Radio Transmission System (AREA)
Abstract
The invention discloses a kind of interference rejection combining method and device, wherein, this method includes：According to the channel estimate matrix of frequency pilot sign and noise jamming matrix, the precomputation factor corresponding to frequency pilot sign is obtained, wherein, it is contemplated that it is that pair data that receive corresponding with frequency pilot sign carry out the used factor when interpolation and balancing operational of amalgamation to calculate the factor；The factor pair data that receive corresponding with frequency pilot sign are calculated on the estimation into row interpolation and balancing operational.The present invention solves first carries out channel estimation interpolation in the prior art, reconstruct merging weight matrix carry out it is balanced and caused by computational complexity increase, the technical problem that the data throughput of system declines and real-time reduces, has reached the technique effect for the real-time for reducing computational complexity, raising throughput of system and data processing.
Description
Technical field
The present invention relates to the communications field, in particular to a kind of interference rejection combining method and device.
Background technology
For cell mobile communication systems, the data rate of center of housing estate position and the data speed of cell edge position
Rate is there is very big difference, this not only limit the capacity of whole system, and the clothes for making user be obtained in different positions
Business quality has very big difference.At present, the new generation broadband wireless communication system researched and developed, such as Long Term Evolution (Long
Term Evolution, referred to as LTE), worldwide interoperability for microwave accesses (Worldwide Interoperability for
Microwave Access, referred to as Wimax) etc., it will all improve main index of the cell-edge performance as system performance
One of.
Moreover, in order to improve the availability of frequency spectrum and ensure that different communication standard cell coexists, LTE system will face
Co-channel interference can be more serious.In order to meet the requirement of the data rate of LTE edge customers, under special scenes, uplink receiving
Need to use AF panel folding in system, so as to suppress inter-cell interference, make the handling capacity of cell-edge data maximum
Change.This is that the correlation matrix estimation that disturb (interference is considered coloured noise) and noise can be calculated the docking collection of letters by one kind
Number handled, so as to play a kind of method of inhibitory action to interference.
Existing AF panel merges (Interference Rejection Combining, referred to as IRC), and algorithm is extremely
Two kinds of criterions can be followed less to be designed.One kind is that (Signal to Noise Ratio, are referred to as based on maximum signal to noise ratio
SNR) criterion, one kind are based on least mean-square error (Minimum Mean Square Error, referred to as MMSE) criterion.Base
It is smaller in the IRC method operands of maximum S/N R criterions, but poor-performing.It is entitled such as Patent No. CN200910244004.3
《A kind of interference rejection combining method and system》Patent be exactly that a kind of AF panel for being designed based on maximum S/N R criterions closes
And system.The performance of MMSE criterions in practical applications is good, but its calculation amount is bigger.
Assuming that by whole LTE combination systems be known as multiple-input and multiple-output (Multiple-Input Multiple-Output,
Referred to as MIMO) system, at present the combination system of the communications field generally include two independent devices：Channel estimating apparatus and
Weigh device, carries out the processing step of AF panel merging and mainly includes：
S1：The channel response of each carrier wave on frequency pilot sign is calculated by channel estimating apparatus, is then rung according to the channel
The channel response on each carrier wave of other symbols should be estimated by ad hoc approach.A kind of application with 2 pilot tones is linearly inserted
The signal processing model of the channel estimation methods of value is as follows：
hl=αlH0+(1-αl)H1
Wherein, l is notation index, αlRepresent the linear interpolation factor of l-th of symbol, H0、H12 pilot tone symbols are represented respectively
Number channel response matrix.
S2：Carried out by balancer balanced.
Specifically, exemplified by based on maximum S/N R criterions, then the signal processing model of Pool-ing Equilibrium device is as follows：
HHRn-1Y
Wherein, Rn is noise and interference matrix, and Y represents the signal matrix received.
Exemplified by based on MMSE criterions, then balancer data processing then can be more complicated, its signal processing model is as follows：
Wherein, I is unit matrix.
Since LTE system is the system of a multicarrier and multi-antenna, application IRC algorithms when will be related to dimension compared with
High matrix operation, calculation amount can dramatically increase, and cause merging device time delay to dramatically increase.
SNR criterions or MMSE criterions are either used, IRC is required for estimating the characteristic of channel.For LTE system
This data structure with frequency pilot sign, current processing method typically first carry out channel estimation interpolation, reconstruct conjunction
And weight matrix carries out equilibrium.This discrete processing method can also dramatically increase the complexity of computing, be filled so as to cause to merge
The data throughput and real-time put are difficult to reach system requirements.
For it is above-mentioned the problem of, not yet propose effective solution at present.
The content of the invention
The present invention provides a kind of interference rejection combining method and device, at least to solve first to carry out channel in the prior art
Estimation interpolation reconstruct merging weight matrix carry out it is balanced and caused by computational complexity increase, the data throughout of system decline
The technical problem that real-time reduces.
According to an aspect of the invention, there is provided a kind of interference rejection combining method, including：According to the letter of frequency pilot sign
Channel estimation matrix and noise jamming matrix, obtain the precomputation factor corresponding to frequency pilot sign, wherein, it is contemplated that calculate the factor be pair with
Used factor when the corresponding reception data of frequency pilot sign are into row interpolation and balancing operational；Calculate that factor pair is above-mentioned connects on the estimation
Receive interpolation and balancing operational that data carry out amalgamation.
Preferably, it is contemplated that it is that pair data that receive corresponding with frequency pilot sign carry out the interpolation of amalgamation and balanced fortune to calculate the factor
The factor at least twice is used during calculation.
Preferably, it is contemplated that calculate the factor be channel estimate matrix and noise jamming matrix are carried out obtain after linear operation because
Son.
Preferably, it is contemplated that calculating the factor is：WithWherein, H0For the channel estimate matrix on time slot 0,
H1For the channel estimate matrix on time slot 1, Rn is noise jamming matrix.
Preferably, factor pair is calculated on the estimation receive interpolation and balancing operational that data carry out amalgamation, including：According to pre-
Calculate the factor and obtain the interpolation merging factor, wherein, interpolation merges the factor and includes：First kind interpolation merges the factor and the second class interpolation
Merge the factor, wherein, it is the precomputation factor and the linear product of channel estimate matrix that first kind interpolation, which merges the factor, and the second class is inserted
It is linear product of the precomputation factor with receiving data that value, which merges the factor,；Merge that factor pair frequency pilot sign is corresponding connects according to interpolation
Receive interpolation and balancing operational that data carry out amalgamation.
Preferably, the second class interpolation merging factor isAnd/orWherein, H0For on time slot 0
Channel estimate matrix, H1For the channel estimate matrix on time slot 1, Rn is noise jamming matrix, wherein, Y is and frequency pilot sign
Corresponding reception data matrix.
Preferably, it is at least one of that first kind interpolation, which merges the factor,：WithWherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise
Interference matrix.
Preferably, factor pair is calculated on the estimation receive interpolation and balancing operational that data carry out amalgamation, including：According to slotting
It is worth formula, first kind interpolation merges the factor and the second class interpolation merges factor pair and receives interpolation and equilibrium that data carry out amalgamation
Computing.
Preferably, the factor is merged according to interpolation formula, first kind interpolation and the second class interpolation merges factor pair and receives data
The interpolation and balancing operational of amalgamation are carried out, including：According to formulaTo receive data into
The interpolation and balancing operational of row amalgamation, wherein,
Wherein, interpolation formula hl=αlH0+(1-αl)H1, l expression time domains
Notation index, αlTo receive the interpolation coefficient on l-th of symbol of data, wherein, H0For the channel estimate matrix on time slot 0, H1
For the channel estimate matrix on time slot 1, Rn is noise jamming matrix, and I is unit matrix.
Preferably, takeConjugate transposition obtain
Preferably, data are received according to the precomputation factor pair into row interpolation and balancing operational, including：According to slotting
It is worth interpolation and balancing operational that formula and the second class interpolation factor carry out the reception data amalgamation.
Preferably, according to interpolation formula and the second class interpolation factor the reception data are carried out with the interpolation of amalgamation
And balancing operational, including：According to formulaAmalgamation is carried out to receiving data
Interpolation and balancing operational, wherein, hl=αlH0+(1-αl)H1, the notation index of l expression time domains, αlTo receive l-th of symbol of data
On interpolation coefficient, wherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise
Interference matrix.
According to another aspect of the present invention, there is provided a kind of AF panel merges device, including：Processing unit, for root
According to the channel estimate matrix and noise jamming matrix of frequency pilot sign, the precomputation factor corresponding to frequency pilot sign is obtained, wherein, in advance
It is pair corresponding with frequency pilot sign used factor when receiving data into row interpolation and balancing operational to calculate the factor；Computing list
Member, the interpolation and balancing operational for receiving data for calculating factor pair on the estimation and carrying out amalgamation.
Preferably, arithmetic element includes：First computing module, the interpolation merging factor is obtained for calculating the factor on the estimation,
Wherein, interpolation merges the factor and includes：First kind interpolation merges the factor and the second class interpolation merges the factor, wherein, first kind interpolation
It is the precomputation factor and the linear product of channel estimate matrix to merge the factor, the second class interpolation merge the factor be the precomputation factor with
Receive the linear product of data；Second computing module, for merging the corresponding reception data of factor pair frequency pilot sign according to interpolation
Carry out the interpolation and balancing operational of amalgamation.
Preferably, arithmetic element includes：3rd computing module and/or the 4th computing module, wherein, the 3rd computing module,
Data progress amalgamation is received for merging factor pair according to interpolation formula, the first kind interpolation merging factor and the second class interpolation
Interpolation and balancing operational；4th computing module, for being joined according to interpolation formula and the second class interpolation factor to receiving data
The interpolation and balancing operational of combination.
In the present invention, used factor during into row interpolation and balancing operational is preferentially calculated to frequency pilot sign, then,
According to the corresponding data that receive of the precomputation factor pair frequency pilot sign being calculated into row interpolation and balancing operational.Pass through above-mentioned side
Originally the interpolation separately handled and balancing operational, are merged the processing for carrying out amalgamation by formula, so as to reduce pair
Signal carries out the operand of AF panel merging, solves and first carries out channel estimation interpolation in the prior art, reconstructs merging power
Value matrix carry out it is balanced and caused by computational complexity increase, the technology that the data throughput of system, which declines real-time, to be reduced asks
Topic, has reached the technique effect for the real-time for reducing computational complexity, raising throughput of system and data processing.
Brief description of the drawings
Attached drawing described herein is used for providing a further understanding of the present invention, forms the part of the application, this hair
Bright schematic description and description is used to explain the present invention, does not form inappropriate limitation of the present invention.In the accompanying drawings：
Fig. 1 is a kind of preferred flow charts of interference rejection combining method according to embodiments of the present invention；
Fig. 2 is another preferred flow charts of interference rejection combining method according to embodiments of the present invention；
Fig. 3 is a kind of preferred structure block diagram that AF panel according to embodiments of the present invention merges device；
Fig. 4 is another preferred structure block diagram that AF panel according to embodiments of the present invention merges device；
Fig. 5 is another preferred structure block diagram led AF panel and merge device according to embodiments of the present invention；
Fig. 6 is a kind of sub-frame of uplink time-frequency structure schematic diagram of LTE system according to embodiments of the present invention；
Fig. 7 is a kind of preferred flow charts of the method according to embodiments of the present invention that IRC equilibriums are carried out with MMSE criterions；
Fig. 8 is a kind of preferred structure block diagram of the corresponding realization device of IRC equalization methods according to embodiments of the present invention；
Fig. 9 is a kind of preferred flow charts of IRC equalization methods according to embodiments of the present invention；
Figure 10 is a kind of preferred flow charts of the method according to embodiments of the present invention that IRC equilibriums are carried out with SNR criterions；
Figure 11 is another preferred flow charts of the method according to embodiments of the present invention that IRC equilibriums are carried out with SNR criterions.
Embodiment
Come that the present invention will be described in detail below with reference to attached drawing and in conjunction with the embodiments.It should be noted that do not conflicting
In the case of, the feature in embodiment and embodiment in the application can be mutually combined.
A kind of interference rejection combining method is present embodiments provided, as shown in Figure 1, this method comprises the following steps：
Step S102：According to the channel estimate matrix of frequency pilot sign and noise jamming matrix, obtain corresponding to frequency pilot sign
The precomputation factor, wherein, it is contemplated that it is pair corresponding with frequency pilot sign when receiving data into row interpolation and balancing operational to calculate the factor
The used factor；
Step S104：Interpolation and balancing operational that the above-mentioned reception data of factor pair carry out amalgamation are calculated on the estimation.
In the present embodiment, used factor during into row interpolation and balancing operational is preferentially calculated to frequency pilot sign,
Then, the interpolation of amalgamation and balanced is carried out according to the corresponding data that receive of the precomputation factor pair being calculated and frequency pilot sign
Computing.By the above-mentioned means, the interpolation separately handled originally and balancing operational are merged to the place for carrying out amalgamation
Reason, thus reduce to signal carry out AF panel merging operand, solve in the prior art first carry out channel estimation insert
Value, reconstruct merging weight matrix carry out it is balanced and caused by computational complexity increase, the data throughput of system declines real-time
Property the technical problem that reduces, reached reduce computational complexity, improve throughput of system and data processing real-time technology
Effect.
In a preferred embodiment, it is contemplated that it is that a pair reception data corresponding with the frequency pilot sign are combined to calculate the factor
The factor at least twice is used when the interpolation and balancing operational of change.By first calculating when into row interpolation and balancing operational at least
The factor used twice can be greatly enhanced the efficiency of computing, reduce the complexity of computing.
Can be that basis linearly transports channel estimate matrix and noise jamming matrix for the above-mentioned precomputation factor
The factor obtained after calculation.Preferably, the above-mentioned precomputation factor is：WithWherein, H0For on time slot 0
Channel estimate matrix, H1For the channel estimate matrix on time slot 1, Rn is noise jamming matrix.
After the precomputation factor is obtained, it can first calculate interpolation and merge interpolation and balanced fortune of the factor in progress amalgamation
Calculate, in a preferred embodiment, calculate the factor pair data that receive corresponding with frequency pilot sign on the estimation and carry out amalgamation
The method of interpolation and balancing operational, as shown in Fig. 2, this method includes：
Step S202：The factor is calculated on the estimation obtains the interpolation merging factor, wherein, interpolation merges the factor and includes：The first kind
Interpolation merges the factor and the second class interpolation merges the factor, wherein, it is that the precomputation factor is estimated with channel that first kind interpolation, which merges the factor,
The linear product of matrix is counted, it is linear product of the precomputation factor with receiving data that the second class interpolation, which merges the factor,；
Step S204：According to the corresponding interpolation for receiving data progress amalgamation of interpolation merging factor pair frequency pilot sign and
Weighing apparatus computing.
In a preferred embodiment, the second above-mentioned class interpolation merging factor isAnd/orWherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn does for noise
Matrix is disturbed, Y is reception data matrix corresponding with frequency pilot sign.
It is at least one of that above-mentioned first kind interpolation, which merges the factor,：WithWherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise
Interference matrix.
The interpolation for the data progress amalgamation that factor pair antenna receives is merged based on above-mentioned interpolation and balancing operational can
In a manner of use is following two：
1) mode one：The factor is merged according to interpolation formula, first kind interpolation and the second class interpolation merges factor pair and receives number
According to the interpolation and balancing operational for carrying out amalgamation.
Specifically：According to formulaTo receive data carry out amalgamation interpolation and
Weighing apparatus computing, wherein,
Wherein, interpolation formula hl=αlH0+(1-αl)H1, l expression time domains
Notation index, αlTo receive the interpolation coefficient on l-th of symbol of data, wherein, H0For the channel estimate matrix on time slot 0, H1
For the channel estimate matrix on time slot 1, Rn is noise jamming matrix, and I is unit matrix.
In a preferred embodiment, it can use and takeThe mode of conjugate transposition obtain
So as to reduce calculatingThe calculation amount of Shi Jinhang multiply-add operations.
Pass-through mode one realizes the IRC for the MMSE criterions that the IRC based on MMSE criterions is balanced, is realized using aforesaid way
Equilibrium greatly reduces the complexity of calculating, improves the efficiency of computing,
2) mode two：Interpolation and the equilibrium of amalgamation are carried out to receiving data according to interpolation formula and the second class interpolation factor
Computing.
Specifically：According to formulaTo receiving data into row interpolation and equilibrium
Computing, wherein, hl=αlH0+(1-αl)H1, the notation index of l expression time domains, αlTo receive the interpolation system on l-th of symbol of data
Number, wherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise jamming matrix.
The IRC that pass-through mode two realizes the SNR criterions that the IRC based on SNR criterions is balanced, is realized using aforesaid way is equal
Weighing apparatus greatly reduces the complexity of calculating, improves the efficiency of computing,
A kind of AF panel is additionally provided in the present embodiment and merges device, which is used for realization above-described embodiment and excellent
Embodiment is selected, had carried out repeating no more for explanation.As used below, term " module " " unit " can be realized pre-
Determine the combination of the software and/or hardware of function.Although the described device of following embodiments is preferably with hardware realization, soft
Part, or the realization of the combination of software and hardware is also what may and be contemplated.Fig. 3 is the interpolation balancer of frequency pilot sign
A kind of preferred structure block diagram, as shown in figure 3, including：Processing unit 302 and arithmetic element 304, below say the structure
It is bright.
Processing unit 302, for the channel estimate matrix according to frequency pilot sign and noise jamming matrix, obtains corresponding to and leads
The precomputation factor of frequency symbol, wherein, it is contemplated that it is pair reception data corresponding with frequency pilot sign into row interpolation and balanced to calculate the factor
Used factor during computing；
Arithmetic element 304, the interpolation of amalgamation and balanced fortune are carried out for calculating the above-mentioned reception data of factor pair on the estimation
Calculate.
In a preferred embodiment, as shown in figure 4, arithmetic element 304 includes：First computing module 402, for root
Merge the factor it is expected that calculating the factor and obtaining interpolation, wherein, interpolation merges the factor and includes：First kind interpolation merges the factor and the second class
Interpolation merges the factor, wherein, it is the precomputation factor and the linear product of channel estimate matrix that first kind interpolation, which merges the factor, second
It is linear product of the precomputation factor with receiving data that class interpolation, which merges the factor,；Second computing module 404, for being closed according to interpolation
And the corresponding interpolation and balancing operational for receiving data and carrying out amalgamation of factor pair frequency pilot sign.
In a preferred embodiment, as shown in figure 5, arithmetic element 304 includes：3rd computing module 502 and/or
Four computing modules 504, wherein, the 3rd computing module 502, for merging the factor and second according to interpolation formula, first kind interpolation
Class interpolation merges factor pair and receives interpolation and balancing operational that data carry out amalgamation；4th computing module 504, for according to slotting
It is worth interpolation and balancing operational that formula and the second class interpolation factor carry out amalgamation to receiving data.
Specifically：3rd computing module 502 is according to formulaCarried out to receiving data
The interpolation and balancing operational of amalgamation, wherein,
Wherein, interpolation formula hl=αlH0+(1-αl)H1, l expression time domains
Notation index, αlTo receive the interpolation coefficient on l-th of symbol of data, wherein, H0For the channel estimate matrix on time slot 0, H1
For the channel estimate matrix on time slot 1, Rn is noise jamming matrix, and I is unit matrix.
In a preferred embodiment, it can use and takeThe mode of conjugate transposition obtain
So as to reduce calculatingThe calculation amount of Shi Jinhang multiply-add operations.
Specifically：4th computing module 504 is according to formulaTo receiving number
According to the interpolation and balancing operational for carrying out amalgamation, wherein, hl=αlH0+(1-αl)H1, the notation index of l expression time domains, αlTo connect
The interpolation coefficient on l-th of symbol of data is received, wherein, H0For the channel estimate matrix on time slot 0, H1Estimate for the channel on time slot 1
Matrix is counted, Rn is noise jamming matrix.
The present invention provides a kind of preferred embodiment further to be explained to the present invention, but noticeable
It is that the preferred embodiment is intended merely to the preferably description present invention, does not form and the present invention is improperly limited.
Above-mentioned method and apparatus are further described with several specific embodiments below in conjunction with the accompanying drawings.
Embodiment 1
Illustrated in the preferred embodiment by taking MMSE as an example：
It is a kind of sub-frame of uplink time-frequency structure schematic diagram of LTE system as shown in Figure 6.Wherein, a subframe includes two
Time slot.In time domain, a time slot includes 7 SC-FDMA (single-carrier frequency division multiple access) symbols (l=0 ... 6).On frequency domain, often
A symbol takes 1200 useful information subcarriers (k=0 ... 1999).In the 4th symbol of time slot 0 and the 4th symbol of time slot 1
There are frequency pilot sign on number, for transmitting the reference signal of UE (user equipment).
According to the time-frequency structure schematic diagram shown in Fig. 6, the method that IRC equalization methods are carried out with MMSE criterions, such as Fig. 7 institutes
Show, comprise the following steps：
Step S702：Obtain the channel estimation value on frequency pilot sign.
Wherein, above-mentioned steps S702 includes：
Obtain the coarse channel estimate H on each subcarrier on 2 frequency pilot signsC(iSLOT,lDMARS,k,kaRX) and thin letter
Road estimate HR(iSLOT,lDMARS,k,kaRX), wherein iSLOT=0,1 indexes for time slot, lDMARS=3,10 be pilot tone time-domain symbol
Index, k=0,1 ... 1199 for subcarrier in frequency domain index, kaRX=0,1 ... 7 indexes for reception antenna.
Thin channel estimate matrix H in pilot tone 00With the thin channel estimate matrix H in pilot tone 11Represent as follows：
Dimension is KaRX× 1,Dimension is KaRX×1。
Step S704：Obtain interference and noise covariance value.
Wherein, above-mentioned steps S704, including：
S1：According to the channel estimation value obtained in step S702, calculate on 2 time domain pilot symbols on each subcarrier
Interference and noise figure (that is, noise jamming matrix above) NI (iSLOT,lDMARS,k,kaRX),
Preferably, NI (iSLOT,lDMARS,k,kaRX)=HC(iSLOT,lDMARS,k,kaRX)-HR(iSLOT,lDMARS,k,kaRX)
S2：Interference and noise figure NI are divided into group according to carrier number, carry out data smoothing in group.The index of group is d, every group
Account for L subcarrier.The present embodiment is illustrated by taking L=12 as an example, is carried with the son that 1 RB (resource block) in LTE system includes
Wave number is consistent.Calculate the interference on each frequency pilot sign on each RB and noise matrix (i.e. above-mentioned noise jamming square
Battle array)：
That is, i-thSLOTOn time slot, the pth row of noise and interference matrix Rn on d-th of RB, the element on q row is：
Its dimension is KaRX×KaRX。
It is worth noting that the mode of above-mentioned acquisition channel estimate matrix and noise jamming matrix is only as a kind of preferred
Embodiment, other modes generation channel estimate matrix and noise jamming matrix can also be used, the present invention not limit this
It is fixed.
Step S706：Calculate interpolation and merge the factor.
Wherein, above-mentioned steps S706, including：
S1：The precomputation factor of 2 pilot tones is calculated, that is, calculates the precomputation factor of the 1st pilot toneWith the 2nd
The precomputation factor of a pilot toneThe dimension of two matrixes is all 1 × KaRX。
S2：Calculate 3 first kind interpolation and merge the factor3 matrixes
Dimension all be 1 × 1.
S3：As the Y data notation index in time domain changes, 2 the second class interpolation are calculated respectively and merge the factor Dimension is 1 × 1.
Wherein, Y is reception data matrix (that is, the multi-antenna data that base station receives) corresponding with frequency pilot sign.
Preferably, Y=[Y (1,1) ... Y (KaRX,1)]H, dimension KaRX×1。
Step S708：Calculating is merged according to interpolation factor.
Preferably, according to linear interpolation formula hl=αlH0+(1-αl)H1Calculating is combined to multiple interpolation factors, its
In, l represents the notation index of time domain, αlIt is the interpolation coefficient on l-th of symbol, specific above-mentioned steps S708 includes：
S1：The factor is merged according to above-mentioned obtained first kind interpolation and calculates being worth as follows in each time-domain symbol：
In above-mentioned steps S1, due to Rn-1For hermitian (Hermitian) matrix, therefore can willConjugation is taken to turn
Putting to directly obtainSo as to avoid complex multiplication and add operation.
S2：The factor is merged according to above-mentioned obtained second class interpolation and calculates being worth as follows in each time-domain symbol：
The dimension of result of calculation is 1 × 1.
S3：Calculate the merging coefficient that dimension is 1 × 1And finally calculate each time-domain symbol
On following value：
GainedTerminal data as after multiple antennas IRC equilibriums.
The above embodiment be time-domain data symbols number using LTE system as 12, when reception antenna number is 8 exemplified by carry out
Illustrate, as described above, the complex multiplication operation number using the present embodiment the method is only the 0.2 of existing method,
Complex addition operations number is only the 0.36 of existing method.It can be seen that the method for the embodiment of the present invention significantly reduces operand.
IRC equalization methods the present embodiment according to Fig. 7 additionally provides a kind of preferable realization device, as shown in figure 8,
Including following 5 modules：
Shared storage and control unit 802, for being responsible for Read-write Catrol, input and the output of all modules, intermediate result
The storage of data, mainly includes RAM storage arrays, control circuit and time domain notation index counter.Wherein, RAM storage circuits
It is multiplexing with OFDM demodulation storage unit, for storing the antenna data of the N number of symbol of Ka antenna, and is multiplexed Ka day of storage
The channel evaluation data of line.
Channel estimation value computing unit 804 (part for processing unit i.e. above), based on carrying out channel estimation
Calculate.
Interference and noise covariance value computing unit 806 (part in processing unit i.e. above), for being done
Disturb the calculating and storage with noise covariance matrix.
Interpolation merges factor calculating unit 808 (the first computing module 402 i.e. above), for merging the factor into row interpolation
Calculating and storage, mainly include input data cross selection circuit, array multiplies/power up road and storage array.Storage array
Including precomputation factor storage array, first kind interpolation merges factor storage array and the second class interpolation merges factor storage battle array
Row.
Interpolation combining unit 810 (the second computing module 404, the 3rd computing module 502 and/or the 4th computing module 504),
Merge the combination of the factor for completing interpolation according to the notation index of current data to calculate, mainly including divider, multiplier with
And adder circuit.
Based on above-mentioned IRC balanced realizations device carry out IRC equilibriums step as shown in figure 9, including：
Step S902：Each antenna data that shared storage and control unit 802 receive receiver serially arrive by storage
In its internal multiple memory space；After storage to be shared and control unit 802 store the data of full 1 subframe, channel estimation
It is worth computing unit 804 and the antenna data of 2 frequency pilot signs is taken out into progress channel estimation calculation.The channel estimation value one calculated
Road former address is stored back into shared storage and 802 memory space of control unit, is input to interference all the way and noise covariance value calculates
Unit 806；
Step S904：Interference and noise covariance value computing unit 806 carry out interference and noise square according to channel estimation value
The calculating of battle array；While interference and noise covariance value 806 output data of computing unit, from shared storage and control unit 802
Respective stored space in read out the channel estimation values of two frequency pilot signs parallel, pass through interference and noise covariance value calculate
Data cross selection circuit inside unit 806 switches over selection to 2 input operands of array multiplicaton addition unit, counts first
The precomputation factor is calculated to be stored in precomputation factor storage array；Then list is stored from shared storage and control unit 802 again
The channel estimation value of 2 frequency pilot signs, and the data feedback of combined calculation factor storage array are read out in member parallel, is calculated
1st class interpolation merges the factor and is stored in the 1st class interpolation merging factor storage storage array；
Step S906：Under the control of read address counter inside shared storage and control unit 802, sequentially read out
The antenna data of each symbol, is output to interpolation and merges factor calculating unit 808, by controlling data cross selection circuit to be poised for battle
2 input operands of row multiplicaton addition unit switch over selection and calculate the 2nd class interpolation merging factor and store；
Step S908：Interpolation combining unit 810 indexes letter according to the time-domain symbol that shared storage and control unit 802 export
Breath calculates interpolation coefficient, and root and interpolation merge the 1st class interpolation that factor calculating unit 808 calculates and merge the factor and the 1st class
Interpolation merges the factor and carries out relevant multiply-add operation, obtains the single channel output data after IRC equilibriums.
The method for realizing IRC equilibriums by the above-mentioned interpolation to channel estimate matrix and balancing combine computing, significantly reduces
The operands of IRC equilibriums based on MMSE criterions, will effectively reduce computational complexity and time delay.To complete N carrier wave L symbols
The LTE system of M antennas is assessment models, using the prior art in background technology, i.e., carries out channel first with pilot tone and estimates
Interpolation is counted, reconstructs the method that merging weight matrix carries out IRC equilibriums, calculation amount is as shown in table 1：
Table 1
Using the method for the invention, calculation amount (does not enumerate the operand with prior art scheme identical calculations step)
As shown in table 2：
Table 2
As it can be seen that mode when using the present invention carries out used by equilibrium complex multiplication operation number as original (N*
L* (2M+3/4)+N*M* (2M+3))/N*L*M* (2*M+5/2), complex addition number is original (N*L* (2M-1)+N* (M-
1)*(2M+3))/N*L*(M*M+M-1).When the time domain symbolic number L of system is more, reception antenna number M is more, the computing of saving
Amount will be apparent from.
Embodiment 2
Illustrated in the preferred embodiment by taking the IRC of SNR methods merges as an example：
In the present embodiment still be as shown in Figure 6 a kind of sub-frame of uplink time-frequency structure schematic diagram of LTE system in order to
Illustrate.
According to the time-frequency structure schematic diagram shown in Fig. 6, the method that IRC equilibriums are carried out in SNR methods, as shown in Figure 10, bag
Include following steps：
Step S1002：Obtain the channel estimation value on frequency pilot sign.
Wherein, above-mentioned steps S1002 includes：
Obtain the coarse channel estimate H on each subcarrier on 2 frequency pilot signsC(iSLOT,lDMARS,k,kaRX) and thin letter
Road estimate HR(iSLOT,lDMARS,k,kaRX), wherein iSLOT=0,1 indexes for time slot, lDMARS=3,10 be pilot tone time-domain symbol
Index, k=0,1 ... 1199 for subcarrier in frequency domain index, kaRX=0,1 ... 7 indexes for reception antenna.
Thin channel estimate matrix H in pilot tone 00With the thin channel estimate matrix H in pilot tone 11Represent as follows：
Dimension is KaRX× 1,Dimension is KaRX×1。
Step S1004：Obtain interference and noise covariance value
Wherein, above-mentioned steps S1004, including：
S1：According to the channel estimation value obtained in step S702, calculate on 2 time domain pilot symbols on each subcarrier
Interference and noise figure (that is, noise jamming matrix above) NI (iSLOT,lDMARS,k,kaRX),
Preferably, NI (iSLOT,lDMARS,k,kaRX)=HC(iSLOT,lDMARS,k,kaRX)-HR(iSLOT,lDMARS,k,kaRX)
S2：Interference and noise figure NI are divided into group according to carrier number, carry out data smoothing in group.The index of group is d, every group
Account for L subcarrier.The present embodiment is illustrated by taking L=12 as an example, is carried with the son that 1 RB (resource block) in LTE system includes
Wave number is consistent.Calculate the interference on each frequency pilot sign on each RB and noise matrix (i.e. above-mentioned noise jamming square
Battle array)：
That is, i-thSLOTOn time slot, the pth row of noise and interference matrix Rn on d-th of RB, the element on q row is：
Its dimension is KaRX×KaRX。
It is worth noting that the mode of above-mentioned acquisition channel estimate matrix and noise jamming matrix is only as a kind of preferred
Embodiment, other modes generation channel estimate matrix and noise jamming matrix can also be used, the present invention not limit this
It is fixed.
Step S1006：Calculate interpolation and merge the factor.
Wherein, above-mentioned steps S1006, including：
S1：The precomputation factor of 2 pilot tones is calculated, that is, calculates the precomputation factor of the 1st pilot toneWith the 2nd
The precomputation factor of a pilot toneThe dimension of two matrixes is all 1 × KaRX。
S2：As the Y data notation index in time domain changes, 2 the second class interpolation are calculated respectively and merge the factor Dimension is 1 × 1.
Wherein, Y is reception data matrix (that is, the multi-antenna data that base station receives) corresponding with frequency pilot sign,
Preferably, Y=[Y (1,1) ... Y (KaRX,1)]H, dimension KaRX×1。
Step S1008：Calculating is merged according to interpolation factor.
Preferably, merged according to above-mentioned obtained second class interpolation as follows in each time-domain symbol of factor calculating
Value：
The dimension of result of calculation is 1 × 1, wherein, when l is represented
The notation index in domain, αlIt is the interpolation coefficient on l-th of symbol.As
Terminal data after the multiple antennas IRC equilibriums of SNR methods.
The method for being carried out IRC equilibriums in SNR methods based on the device of Fig. 8, as shown in figure 11, is comprised the following steps：
Step S1102：Each antenna data that shared storage and control unit 802 receive receiver serially arrive by storage
In its internal multiple memory space；After storage to be shared and control unit 802 store the data of full 1 subframe, channel estimation
It is worth computing unit 804 and the antenna data of 2 frequency pilot signs is taken out into progress channel estimation calculation.The channel estimation value that will be calculated
Former address is stored back into shared storage and 802 memory space of control unit all the way, is input to interference and noise covariance value meter all the way
Calculate unit 806；
Step S1104：Interference and noise covariance value computing unit 806 carry out interference and noise square according to channel estimation value
The calculating of battle array；While treating interference and noise covariance value 806 output data of computing unit, from shared storage and control unit
The channel estimation value of two frequency pilot signs is read out in 802 respective stored spaces parallel, passes through interference and noise covariance value meter
Calculate the data cross selection circuit inside unit 806 and selection is switched over to 2 input operands of array multiplicaton addition unit, first
Calculate the precomputation factor to be stored in precomputation factor storage array, without carrying out when the method for IRC equilibriums is carried out
1st class interpolation merges the calculating of the factor.
Step S1106：Under the control of read address counter inside shared storage and control unit 802, order is read
Go out the antenna data of each symbol, be output to interpolation and merge factor calculating unit 808, by controlling data cross selection circuit pair
2 input operands of array multiplicaton addition unit switch over selection and calculate the 2nd class interpolation merging factor and store；
Step S1108：Interpolation combining unit 810 calculates interpolation coefficient according to the time-domain symbol index information of output, and
Interpolation merges the 2nd class interpolation merging factor that factor calculating unit 808 calculates and only carries out simple accumulation operations, without calculating
Merge coefficient and directly obtain the single channel output data after IRC equilibriums.
Shared fractional hardware computing resource has been reached by above-mentioned fusion channel estimation interpolation and balanced two modules
Purpose, reduces hardware consumption.
In another embodiment, a kind of software is additionally provided, which is used to perform above-described embodiment and preferred reality
Apply the technical solution described in mode.
In another embodiment, a kind of storage medium is additionally provided, above-mentioned software is stored with the storage medium, should
Storage medium includes but not limited to：CD, floppy disk, hard disk, scratch pad memory etc..
It can be seen from the above description that the present invention realizes following technique effect：Preferentially calculate to frequency pilot sign
Used factor during into row interpolation and balancing operational, then, carries out according to the precomputation factor pair frequency pilot sign arrived of calculating
Interpolation and balancing operational.By the above-mentioned means, the interpolation separately handled originally and balancing operational are merged into progress
The processing of amalgamation, so as to reduce the operand that AF panel merging is carried out to signal, solves and first carries out in the prior art
Channel estimation interpolation, reconstruct merging weight matrix carry out it is balanced and caused by computational complexity increase, the data throughput of system
Rate declines the technical problem that real-time reduces, and has reached the reality for reducing computational complexity, improving throughput of system and data processing
The technique effect of when property.
Obviously, those skilled in the art should be understood that above-mentioned each module of the invention or each step can be with general
Computing device realize that they can be concentrated on single computing device, or be distributed in multiple computing devices and formed
Network on, alternatively, they can be realized with the program code that computing device can perform, it is thus possible to which they are stored
Performed in the storage device by computing device, and in some cases, can be with different from shown in order execution herein
The step of going out or describing, they are either fabricated to each integrated circuit modules respectively or by multiple modules in them or
Step is fabricated to single integrated circuit module to realize.Combined in this way, the present invention is not restricted to any specific hardware and software.
The foregoing is only a preferred embodiment of the present invention, is not intended to limit the invention, for the skill of this area
For art personnel, the invention may be variously modified and varied.Within the spirit and principles of the invention, that is made any repaiies
Change, equivalent substitution, improvement etc., should all be included in the protection scope of the present invention.
Claims (11)
- A kind of 1. interference rejection combining method, it is characterised in that including：According to the channel estimate matrix of frequency pilot sign and noise jamming matrix, obtain precomputation corresponding to the frequency pilot sign because Son, wherein, the precomputation factor is pair data that receive corresponding with the frequency pilot sign into row interpolation and balancing operational when institute The factor used；The interpolation and balancing operational that data carry out amalgamation are received according to the precomputation factor pair；The interpolation and balancing operational that data carry out amalgamation are received according to the precomputation factor pair, including：According to formulaThe reception data are carried out with the interpolation and balancing operational of amalgamation,Wherein,Wherein, interpolation formula hl=αlH0+(1-αl)H1, the symbol of l expression time domains Number index, αlFor it is described reception l-th of symbol of data on interpolation coefficient, wherein, H0For the channel estimate matrix on time slot 0, H1 For the channel estimate matrix on time slot 1, Rn is noise jamming matrix, and I is unit matrix；And according to formulaThe reception data are carried out with the interpolation of amalgamation And balancing operational,Wherein, hl=αlH0+(1-αl)H1, the notation index of l expression time domains, αlFor inserting on described reception l-th of symbol of data Value coefficient, wherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise jamming square Battle array.
- 2. according to the method described in claim 1, it is characterized in that, the precomputation factor is pair corresponding with the frequency pilot sign Reception data into row interpolation and balancing operational when use the factor at least twice.
- 3. method according to claim 1 or 2, it is characterised in that the precomputation factor is to the channel estimation square Battle array and noise jamming matrix carry out the factor obtained after linear operation.
- 4. according to the method described in claim 3, it is characterized in that, the precomputation factor is：WithWherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise jamming matrix.
- 5. according to the method any one of claim 1,2,4, it is characterised in that according to the precomputation factor pair Interpolation and balancing operational that data carry out amalgamation are received, including：Interpolation is obtained according to the precomputation factor and merges the factor, wherein, the interpolation merges the factor and includes：First kind interpolation is closed And the factor and the second class interpolation merge the factor, wherein, it is the precomputation factor and channel that the first kind interpolation, which merges the factor, The linear product of estimated matrix, it is that the precomputation factor receives the linear of data with described that the second class interpolation, which merges the factor, Product；The corresponding interpolation for receiving data progress amalgamation of the frequency pilot sign according to the interpolation merges factor pair and equilibrium are transported Calculate.
- 6. according to the method described in claim 5, it is characterized in that, the second class interpolation merging factor is And/orWherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise jamming matrix, Y is the reception data matrix corresponding with the frequency pilot sign.
- 7. according to the method described in claim 5, it is characterized in that, the first kind interpolation merge the factor for it is following at least it One：WithWherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise jamming matrix.
- 8. according to the method described in claim 1, it is characterized in that, take describedConjugate transposition obtain it is described
- 9. a kind of AF panel merges device, it is characterised in that including：Processing unit, for the channel estimate matrix according to frequency pilot sign and noise jamming matrix, obtains corresponding to the pilot tone The precomputation factor of symbol, wherein, the precomputation factor is pair data that receive corresponding with the frequency pilot sign into row interpolation Used factor during with balancing operational；Arithmetic element, the interpolation and balancing operational for carrying out amalgamation for receiving data according to the precomputation factor pair；The interpolation and balancing operational that data carry out amalgamation are received according to the precomputation factor pair, including：According to formulaThe reception data are carried out with the interpolation and balancing operational of amalgamation,Wherein,Wherein, interpolation formula hl=αlH0+(1-αl)H1, the symbol of l expression time domains Number index, αlFor it is described reception l-th of symbol of data on interpolation coefficient, wherein, H0For the channel estimate matrix on time slot 0, H1 For the channel estimate matrix on time slot 1, Rn is noise jamming matrix, and I is unit matrix；And according to formulaThe reception data are carried out with the interpolation of amalgamation And balancing operational,Wherein, hl=αlH0+(1-αl)H1, the notation index of l expression time domains, αlFor inserting on described reception l-th of symbol of data Value coefficient, wherein, H0For the channel estimate matrix on time slot 0, H1For the channel estimate matrix on time slot 1, Rn is noise jamming square Battle array.
- 10. device according to claim 9, it is characterised in that the arithmetic element includes：First computing module, merges the factor for obtaining interpolation according to the precomputation factor, wherein, the interpolation merges the factor Including：First kind interpolation merges the factor and the second class interpolation merges the factor, wherein, it is described that the first kind interpolation, which merges the factor, The linear product of the precomputation factor and channel estimate matrix, it is the precomputation factor and institute that the second class interpolation, which merges the factor, State the linear product for receiving data；Second computing module, is combined for the corresponding data that receive of the frequency pilot sign according to interpolation merging factor pair The interpolation and balancing operational of change.
- 11. device according to claim 10, it is characterised in that the arithmetic element includes：3rd computing module and/or 4th computing module, wherein,3rd computing module, for merging the factor and the second class interpolation according to interpolation formula, the first kind interpolation Merge and the interpolation and balancing operational that data carry out amalgamation are received described in factor pair；4th computing module, for being joined according to interpolation formula and the second class interpolation factor to the reception data The interpolation and balancing operational of combination.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210132127.XA CN103379055B (en) | 2012-04-28 | 2012-04-28 | interference rejection combining method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210132127.XA CN103379055B (en) | 2012-04-28 | 2012-04-28 | interference rejection combining method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103379055A CN103379055A (en) | 2013-10-30 |
| CN103379055B true CN103379055B (en) | 2018-05-15 |
Family
ID=49463626
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210132127.XA Active CN103379055B (en) | 2012-04-28 | 2012-04-28 | interference rejection combining method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103379055B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103916229B (en) * | 2014-04-29 | 2017-12-08 | 电信科学技术研究院 | A kind of disturbance restraining method and device |
| CN113037658B (en) * | 2019-12-09 | 2024-04-30 | 中兴通讯股份有限公司 | Signal detection method, signal detection device, communication equipment and computer storage medium |
| CN113014525B (en) * | 2021-03-15 | 2022-06-24 | 之江实验室 | Interference suppression method in intelligent reflecting surface system |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8401106B2 (en) * | 2008-07-14 | 2013-03-19 | Telefonaktiebolaget Lm Ericsson (Publ) | Interference rejection combining for multi-user MIMO telecommunications system |
| CN102104562B (en) * | 2009-12-17 | 2013-08-21 | 中兴通讯股份有限公司 | Multi-antenna interference rejection combining method and device |
| CN101753176B (en) * | 2009-12-24 | 2012-12-19 | 北京北方烽火科技有限公司 | Interference rejection combining method and system |
-
2012
- 2012-04-28 CN CN201210132127.XA patent/CN103379055B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN103379055A (en) | 2013-10-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103312389B (en) | A kind of multiuser interference suppression method, terminal and base station | |
| CN108234364B (en) | Channel estimation method based on cell reference signal in LTE-A system | |
| CN103475397B (en) | A kind of method, communication station and the movement station of three-dimensional wave beam forming | |
| CN104620527B (en) | Determine method, receiving device and the sending device of pre-coding matrix instruction | |
| CN107852207A (en) | Dispensing device, reception device, control station, communication system and transmission method for precoding | |
| CN102480444B (en) | Method and corresponding system for broadband co-frequency interference noise estimation and interference suppression | |
| CN108141259A (en) | Scalable extensive MIMO receiver | |
| CN106856461B (en) | Interference estimation method and equipment | |
| CN102546088A (en) | BD (block diagonalization) pre-coding method and device | |
| CN104683005B (en) | Disturbance restraining method and device | |
| CN105099610B (en) | The method and device of signal processing | |
| CN103379055B (en) | interference rejection combining method and device | |
| CN104488214B (en) | For combining the method and apparatus for performing channel estimation and Interference Estimation in a wireless communication system | |
| CN102546483B (en) | Method for broadband common-frequency interference noise estimation and interference suppression and corresponding system | |
| EP2713542A1 (en) | Method and transmitter apparatus for composing a transmit signal and method and receiver apparatus for channel estimation | |
| US8363756B2 (en) | Wireless reception device, wireless communication system and wireless communication method | |
| CN104253639B (en) | Obtain the method and device of channel quality instruction | |
| CN107046431A (en) | Transmission, method of reseptance and the device of information | |
| CN102480453B (en) | Method and corresponding system for broadband co-frequency interference noise estimation and interference suppression | |
| JP4299302B2 (en) | Apparatus and method for precoding multi-carrier signals | |
| Lin et al. | Dual-mode low-complexity codebook searching algorithm and VLSI architecture for LTE/LTE-advanced systems | |
| Al-Hussaibi et al. | Extending the user capacity of MU-MIMO systems with low detection complexity and receive diversity | |
| Wei et al. | Cooperative communication with partial channel-state information in multiuser MIMO systems | |
| Zhang et al. | Residual-based detections and unified architecture for massive MIMO uplink | |
| CN102480440B (en) | Channel estimation methods under the co-channel interference environment of a kind of broadband and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C41 | Transfer of patent application or patent right or utility model | ||
| TA01 | Transfer of patent application right |
Effective date of registration: 20151116 Address after: 518057 Nanshan District Guangdong high tech Industrial Park, South Road, science and technology, ZTE building, Ministry of Justice Applicant after: ZTE Corp. Applicant after: SANECHIPS TECHNOLOGY Co.,Ltd. Address before: 518057 Nanshan District science and technology, Guangdong Province, South Road, No. 55, No. Applicant before: ZTE Corp. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20221202 Address after: 518055 Zhongxing Industrial Park, Liuxian Avenue, Xili street, Nanshan District, Shenzhen City, Guangdong Province Patentee after: SANECHIPS TECHNOLOGY Co.,Ltd. Address before: 518057 Ministry of justice, Zhongxing building, South Science and technology road, Nanshan District hi tech Industrial Park, Shenzhen, Guangdong Patentee before: ZTE Corp. Patentee before: SANECHIPS TECHNOLOGY Co.,Ltd. |