CN110492956A - A kind of error compensation multi-user test method and device for MUSA system - Google Patents
A kind of error compensation multi-user test method and device for MUSA system Download PDFInfo
- Publication number
- CN110492956A CN110492956A CN201910834807.8A CN201910834807A CN110492956A CN 110492956 A CN110492956 A CN 110492956A CN 201910834807 A CN201910834807 A CN 201910834807A CN 110492956 A CN110492956 A CN 110492956A
- Authority
- CN
- China
- Prior art keywords
- matrix
- error
- estimated value
- indicate
- signal
- 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.)
- Granted
Links
- 238000010998 test method Methods 0.000 title claims abstract description 14
- 241000234295 Musa Species 0.000 title claims abstract 12
- 239000011159 matrix material Substances 0.000 claims abstract description 101
- 238000001514 detection method Methods 0.000 claims abstract description 52
- 238000000034 method Methods 0.000 claims abstract description 49
- 238000000354 decomposition reaction Methods 0.000 claims abstract description 23
- 238000004422 calculation algorithm Methods 0.000 claims description 14
- 230000008569 process Effects 0.000 claims description 10
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000005540 biological transmission Effects 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 abstract description 12
- 230000002596 correlated effect Effects 0.000 abstract description 3
- 238000010295 mobile communication Methods 0.000 abstract description 2
- 230000000875 corresponding effect Effects 0.000 description 24
- 230000007480 spreading Effects 0.000 description 17
- 238000004088 simulation Methods 0.000 description 6
- 230000008901 benefit Effects 0.000 description 5
- 238000012545 processing Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000010363 phase shift Effects 0.000 description 2
- 240000007594 Oryza sativa Species 0.000 description 1
- 235000007164 Oryza sativa Nutrition 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 235000021186 dishes Nutrition 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 235000009566 rice Nutrition 0.000 description 1
- 230000011664 signaling Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/309—Measuring or estimating channel quality parameters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/382—Monitoring; Testing of propagation channels for resource allocation, admission control or handover
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/391—Modelling the propagation channel
- H04B17/3912—Simulation models, e.g. distribution of spectral power density or received signal strength indicator [RSSI] for a given geographic region
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/18—Phase-modulated carrier systems, i.e. using phase-shift keying
- H04L27/20—Modulator circuits; Transmitter circuits
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/18—Phase-modulated carrier systems, i.e. using phase-shift keying
- H04L27/22—Demodulator circuits; Receiver circuits
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Quality & Reliability (AREA)
- Spectroscopy & Molecular Physics (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention belongs to mobile communication technology fields, are related to the uplink multiuser detection technique of MUSA system；For a kind of error compensation multi-user test method and device for MUSA system；Method includes after the user data of transmitting terminal is modulated by QPSK, randomly choosing complex spread sequence extension；Receiving end detects the signal received using parallel interference canceller, calculates the estimated value that weight matrix obtains the subscriber signal of transmitting terminal；Eigenvalues Decomposition is carried out to eigenmatrix, selected part decomposition result reconstructs the subscriber signal of transmitting terminal；It is minimized using lagrange's method of multipliers by equivalent noise matrix, readjusts its estimated value；Actual subscriber signal estimated value is calculated using maximal possibility estimation and selects an output.The present invention estimates the error of signal by compensating MMSE criterion, and correlated noise is projected to characteristic vector space, searches for signal on the direction of Noise enhancement, obtains better detection performance, improves detection accuracy etc..
Description
Technical field
The invention belongs to mobile communication technology fields, and in particular to the uplink multiuser detection technique of MUSA system；
Specially a kind of error compensation multi-user test method and device for MUSA system.
Background technique
3GPP R15 is that the first edition of 5G standard mainly focuses the support to scenes such as eMBB and URLLC, non-orthogonal multiple
The alternative for accessing Uplink multiple access scheme of newly eating dishes without rice or wine as 5G, makes mMTC scene in subsequent 5G and being specifically designed
When, its unique advantage will necessarily be shown, to meet requirement of the 5G to different business scene.Therefore need at this stage to it is non-just
Hand over multiple access technology input research.Existing non-orthogonal multiple access technology have SCMA (Sparse Code MultipleAccess,
Sparse CDMA) technology, MUSA (Multiple user shared access, multiple users share access) technology, NOMA
(Non-Orthogonal Multiple Access, non-orthogonal multiple access) technology, PDMA (Pattern
DivisionMultipleAccess, pattern divide multiple access) technology.
Non-orthogonal multiple technology has obvious gain in terms of system uplink handling capacity, access customer number.In emerging public affairs
The MUSA technology that department proposes is a kind of code domain, the non-orthogonal multiple access scheme based on spread spectrum communication, and MUSA technology is due to connecing
The random selection sequence spreading of the user entered, so not needing complicated scheduling access procedure, i.e., this is that one kind exempts from scheduling method,
And this scheduling method of exempting from saves scheduling time, reduces signaling overheads, the magnanimity in this three kinds of scene required for 5G
Connection and low time delay have very big advantage.Its concrete scheme may include the independent selection of multiple users in transmitting terminal access system
The modulation symbol of oneself is extended by plural frequency expansion sequence, and the user data after then extending is sent out in identical running time-frequency resource
It send.Receiving end separates the data of each user by the identification of least mean-square error multiuser detection.
Traditional multiuser detection generally has least mean-square error-serial interference elimination MMSE-SIC algorithm, minimum
Mean square error-Parallel Interference Cancellation MMSE-PIC algorithm and quasi- parallel interference canceller detection algorithm；MMSE-SIC detection method
It carries out how many grades of detections to be determined by number of users, and the detection of every level-one is all needed to weight matrix ωMMSEIt is calculated, weight matrix
Calculating is related to matrix inversion, and the complexity of matrix inversion is high, and complexity is O (M3), when number of users is big, this method is integrally multiple
Miscellaneous degree height causes time delay to increase, and is not suitable for the low time delay scene of 5G.Although and MMSE-PIC detection method detection rank is few, energy
Time delay is enough effectively reduced, but whole detection performance is again poor.Although quasi- parallel interference canceller detection algorithm processing delay
Smaller than MMSE-SIC, performance is better than MMSE-PIC, but detection performance is still poor compared to MMSE-SIC.
Summary of the invention
Based on problem of the existing technology, if the present invention considers that the property of MMSE-PIC detection method can be improved
Can, the accuracy of detection is improved, then may be implemented that the requirement of low time delay can be reached while guaranteeing performance and reduce
Computation complexity；Therefore the invention proposes a kind of error compensation multi-user test methods and device for MUSA system.
A kind of error compensation multi-user test method for MUSA system proposed by the present invention, the method such as Fig. 1 institute
Show, it may include:
S1, transmitting terminal multiple user data be modulated after, random selection complex spread sequence respective modulation is accorded with
It number is extended, and is sent from identical running time-frequency resource；
S2, receiving end carry out parallel interference canceller detection by MMSE detector to the signal y received, calculate weight
Matrix ωMMSE, obtain the modulation symbol estimated value x of the subscriber signal of transmitting terminal；
S3, Eigenvalues Decomposition is carried out to the eigenmatrix in weight matrix, decomposites eigenvalue λl, corresponding feature vector
vl, and the H-matrix of feature vector and noise constructed into equivalent noise al；The use of l={ 1,2 ..., M }, M expression transmitting terminal
Amount mesh；
S4, the biggish eigenvalue λ of top n is chosenk, corresponding feature vector vkAnd equivalent noise akCalculate transmitting terminal
The corresponding every kind of modulation symbol of subscriber signal respectively with the error e between the modulation symbol estimated value that estimatesl(m)；K=
{1,2,...,N}；
S5, using lagrange's method of multipliers by equivalent noise matrix aHA is minimized, and utilizes calculated error and feature
Value λkWith feature vector vk, calculate the estimated value of the corresponding equivalent noise matrix of every kind of modulation symbol
S6, by the way of maximal possibility estimation according to the estimated value of equivalent noise matrixReadjust error el
(m), it and using its error formed to detection algorithm compensates, to calculate actual subscriber signal estimated valueAnd
Select the wherein the smallest actual subscriber signal estimated value of errorAs final output；
H-matrix indicates hermitian Hermitian matrix；M indicates the corresponding symbolic number of modulation system.
For example, when the modulation system used is QPSK modulation, m={ 1,2,3,4 }.
In addition, the invention also provides a kind of error compensation multi-user detection device for MUSA system, described device
Include:
Transmitting antenna: for emitting user data；
Modulator: for being modulated to user data；
Sequence extension module: for carrying out sequence extension to modulation data；
Receiver: for receiving the user data after modulation extends；
MMSE detector: for acquiring weight by minimizing the least mean-square error between transmission vector sum estimate vector
Matrix；
Feature decomposition module: Eigenvalues Decomposition is carried out for the eigenmatrix in weight matrix；Decomposite eigenvalue λi, it is right
The feature vector v answeredi；
Equivalent noise constructs module: the H-matrix of the eigenvectors matrix for being gone out according to eigendecomposition unit decomposition
It is multiplied with noise；
Estimation error module: the biggish characteristic value of top n, corresponding feature vector and equivalent noise is taken to calculate transmitting
The corresponding every kind of modulation symbol of the subscriber signal at end respectively with the error between the modulation symbol estimated value that estimates；
Lagrange's method of multipliers module: it makes an uproar for calculating corresponding equivalence under every kind of symbol according to lagrange's method of multipliers
The estimated value of sound matrix；
Error calculating module: for the estimation according to the calculated each equivalent noise matrix of lagrange's method of multipliers module
Value calculates actual subscriber signal；
Hard decision module: for carrying out hard decision to calculated actual subscriber signal, the smallest work of error is selected
For output.
Beneficial effects of the present invention:
The present invention be based on MMSE criterion, and on this basis by the error of signal that MMSE detection is estimated into
Row compensation, correlated noise is projected in the feature vector of matrix R, while searching for signal on the direction of Noise enhancement, is improved
The accuracy of detection, this method avoid the matrix inversion operations that the multistage of serial sensing method is complicated, so that complexity is big
It is big to reduce, time delay is reduced, the scene of 5G low time delay more is suitable for.It is shown by analysis and simulation result, it is of the invention
Detection performance is better than MMSE-SIC detection method, is even more to greatly improve compared to MMSE-PIC detection method performance.
Detailed description of the invention
Fig. 1 is system flow chart used by the method for the present invention；
Fig. 2 is the MUSA system block diagram that the present invention uses；
Fig. 3 is the planisphere of ternary complex sequences element；
Fig. 4 is the flow chart of sequence spreading selection；
Fig. 5 is simulation result performance chart.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with attached drawing to of the invention real
The technical solution applied in example is clearly and completely described, it is clear that described embodiment is only that present invention a part is implemented
Example, instead of all the embodiments.
The present embodiment is by taking Fig. 2 as an example, it is assumed that the number of users of transmitting terminal is M, and each customer traffic is each by coding
And modulated process；The selection of each user's independent random meets the complex spread sequence of balance criterion and correlation criterion,
Modulation symbol is extended with respective sequence spreading.
In one embodiment, the cataloged procedure is by the way of differential encoding；The modulated process uses quadrature phase shift key
Control the modulation system of (Quadrature Phase Shift Keying, QPSK).The length of sequence spreading is N, then identical
Running time-frequency resource on send, i.e. channel shared by multiple users passively produces noise after the transmission, and receiving end passes through multi-purpose
Each user data is isolated in family detection.
As a kind of achievable mode, the present invention can also take 16QAM, 64QPSK etc. modulation system.
The real and imaginary parts of multiple sequence spreading are taken respectively from a ternary set { -1,0,1 }, constellation in present embodiment
Figure is as shown in Figure 3.The process of user data respective independent choice sequence spreading after QPSK is modulated, sequence spreading selection can be such as figure
Shown in 4.
Each independent selection sequence spreading of user, when initial, enables i=1；
I-th of user generates sequence spreading at random, judges whether sequence spreading meets balance criterion and auto-correlation criterion,
If satisfied, then enabling i=i+1；Otherwise continue to generate sequence spreading at random to i-th of user；
After i=i+1, judge whether i≤M is true, if so, sequence spreading is also generated at random to i-th of user；It is no
Then, independent this process of selection sequence spreading of all users is completed；
Judge whether meet cross-correlation criterion between two two users, if satisfied, then terminating whole flow process, otherwise from initial shape
State starts again to the independent selection sequence spreading of each user.
In one embodiment, receiving end isolates the process of each user data by Multiuser Detection can include:
Step 1, receiving end carry out parallel interference canceller detection by MMSE detector to the signal y received first；
In this process, it needs to calculate weight matrix and estimates the modulation of transmitting end subscriber according to weight matrix
Symbol；
Weight matrix indicates are as follows: ωMMSE=(HHH+σ2I)-1HH, and remember R=(HHH+σ2I)-1；
The modulation symbol estimated indicates are as follows:
Wherein, H is the channel matrix of the user of transmitting terminal；HHFor the Hermitian matrix of channel matrix H；σ2Indicate noise
Variance；I indicates unit matrix.
Receiving signal y indicates are as follows:
Y=Hx+n；
It is indicated by scalarPopularization obtains, wherein SmFor the sequence spreading of m-th of user, gmIt is m-th
The channel gain of user, xmFor the modulation symbol of m-th of user.
In its vector expression, y=(y1,y2,y3,......,yN)H, H is equivalent comprising sequence spreading and channel gain
Channel matrix is obtained by channel estimation, x=(x1,x2,x3,......,xM)H, n=(n1,n2,n3,......,nN)H, nkIt indicates
K-th of noise signal, N are sequence spreading length, and M is number of users.
Matrix R is carried out characteristic value eigendecomposition by step 2, obtains R=VDVH；Specifically decomposite eigenvalue λi, it is right
The feature vector v answeredi, the H-matrix of feature vector and noise are constructed into equivalent noise ai；I={ 1,2 ..., M }, M indicate hair
Penetrate the number of users at end；
Wherein, V indicates that eigenvectors matrix includes M × M dimension tenth of the twelve Earthly Branches that M mutually orthogonal M × 1 tie up column feature vector composition
Matrix, V=[v1,v2,......,vM]；D indicate eigenvalue matrix include M eigenvalue cluster at M × M diagonal matrix, D=diag
[λ1,λ2,......,λM]；VHIt indicates the Hermitian matrix of eigenvectors matrix, the equivalence including M equivalent noise is enabled to make an uproar
Sound matrix is a=VHN', n'=(n'1,n'2,...,n'M)H, n'iIndicate i-th of the noise signal assumed, subscript H is indicated
Hermitian matrix, so its each rank statistical property is constant, that is to say its mean value, variance and n=since a is equivalent noise
(n1,n2,n3,......,nN)HIt is identical.
Step 3 chooses wherein a eigenvalue λ of biggish N (N < M) from M characteristic valuek(k={ 1,2 ..., N }), with
And the corresponding feature vector v of this N number of characteristic valuek, form new matrix V and matrix D, due to actual user send signal s with
There are error between estimated value x, relationship isTherefore, new matrix V is M × N-dimensional matrix, and new matrix D is N
Rank matrix sends signal to actual user and estimates, is expressed asAt this point, due to not calculating
Equivalent noise matrix a, so next needing to estimate a.
Step 4, using lagrange's method of multipliers by equivalent noise matrix aHA is minimized, and is calculated often by constraint condition
The estimated value of corresponding equivalent noise matrix under kind symbol
The estimated value of the corresponding equivalent noise matrix of every kind of modulation symbolCalculation formula are as follows:
Wherein, C=[c1,c2,...,cM],(·)↑It indicates
Moore-Penrose pseudoinverse；E=[e1 *(m),e2 *(m),...,eM *(m)]；bl(m) l is indicated
M-th of symbol that a transmitting end subscriber may be sent, Indicate first of transmitting
The modulation symbol estimated value that end subscriber is sent；(vk)lIndicate feature vector vkFirst of element；Subscript H indicates Hermitian square
Battle array.
Need to calculate the error between the corresponding transmitting symbol of every kind of symbol and modulation symbol in this processSince the value of the QPSK modulation system that the present embodiment uses, and each transmitting symbol has four
Kind symbol, thenbl(m) m-th of symbol that first of transmitting end subscriber may be sent is indicated.
In a preferred embodiment, the estimated value of the corresponding equivalent noise matrix of every kind of modulation symbolCalculation formula
It further include with the corresponding every kind of modulation symbol b of the subscriber signal of transmitting terminall(m) respectively with the modulation symbol estimated value x that estimates
Between error as constraint condition, minimize cost function f [a] using lagrange's method of multipliers, two transmitting terminals selected to use
Family l1And l2To estimated valueEstimated；Wherein, the process of lagrange's method of multipliers minimum cost function includes:
s.t.
Constraint condition 1:
Constraint condition 2:
Constraint condition 3:
Wherein, f [a] indicates estimated valueCost function；γ1And γ2Indicate Lagrange coefficient；el1(m) indicate real
L in the subscriber signal on border1The corresponding error of m-th of symbol that a transmitting end subscriber may be sent, el2(m) actual use is indicated
L in the signal of family2The corresponding error of m-th of symbol that a transmitting end subscriber may be sent, * indicate conjugate of symbol； Subscriptl1And subscriptl2Expression belongs to two different transmittings of { 1,2 ..., M }
End subscriber；(v1)l1Indicate feature vector v1L1A element；(v1)l2Indicate feature vector v1L2A element.
In addition, the generation process of constraint condition 1 is by a*It seeks local derviation and obtains:
Believe it is understood that only only used two users about transmitting terminal during above-mentioned lagrange's method of multipliers
Number corresponding every kind of modulation symbol bl(m) constraint condition with the error between the modulation symbol estimated value x that estimates respectively, i.e.,
Two corresponding constraint conditions of transmitting end subscriber have been selected from C:And(2 peace treaty of constraint condition
Beam condition 3)；Correspondingly, E=[el1 *(m),el2 *(m)]H；If in order to enable estimated valueEstimated result it is more accurate, can be with
More constraint conditions are selected from M transmitting end subscriber, such as also increase constraint condition:l3It indicates to belong to
Transmitting end subscriber in { 1,2 ..., M }；E=[e at this timel1 *(m),el2 *(m),el3 *(m)]H；Certainly transmitting terminal at this time is used
Family l1、l2And l3And by calculate SINR sequence after select come；Specifically refer to following procedure:
Emit end subscriber l1And l2Selection course include by calculate it is each transmitting end subscriber Signal to Interference plus Noise Ratio SINR, and
It is ranked up, selects the smallest two users of wherein SINR as transmitting end subscriberl1With transmitting end subscriberl2；Wherein
The calculation formula of SINR indicates are as follows:
Wherein ωi,MMSEThat indicate is the i-th row of weight matrix, hiThat indicate is the i-th column of equivalent channel matrix H, σ2It is
Noise variance, | | | | represent Frobenius norm.
Step 5, by the way of maximal possibility estimation according to the estimated value of equivalent noise matrixCalculate actual user
Signal estimated valueAnd select the wherein the smallest actual subscriber signal estimated value of errorAs final output.
Since the value symbol of b (m) is different, calculate under all situations, E=[e1 *(m),e2 *(m)]H,
And it calculatesIt obtains in various situations(·)↑Indicate that Moore-Penrose is pseudo-
It is inverse, final output result is calculated from following formula:
Wherein, argmin is indicatedObtain corresponding independent variable when minimum valueQ () indicates hard decision；|
| | | represent Frobenius norm.
In one embodiment, the present invention gives a kind of error compensation Multiuser Detection dress for MUSA system
It sets, described device includes:
Transmitting antenna: for emitting user data；
Modulator: for being modulated to user data；
Sequence extension module: for carrying out sequence extension to modulation data；
Receiver: for receiving the user data after modulation extends；
MMSE detector: for acquiring weight by minimizing the least mean-square error between transmission vector sum estimate vector
Matrix；
Feature decomposition module: Eigenvalues Decomposition is carried out for the eigenmatrix in weight matrix；Decomposite eigenvalue λi, it is right
The feature vector v answeredi；
Equivalent noise constructs module: the H-matrix of the eigenvectors matrix for being gone out according to eigendecomposition unit decomposition
It is multiplied with noise；
Estimation error module: the biggish characteristic value of top n, corresponding feature vector and equivalent noise is taken to calculate transmitting
The corresponding every kind of modulation symbol of the subscriber signal at end respectively with the error between the modulation symbol estimated value that estimates；
Lagrange's method of multipliers module: it makes an uproar for calculating corresponding equivalence under every kind of symbol according to lagrange's method of multipliers
The estimated value of sound matrix；
Error calculating module: for the estimation according to the calculated each equivalent noise matrix of lagrange's method of multipliers module
Value calculates actual subscriber signal；
Hard decision module: for carrying out hard decision to calculated actual subscriber signal, the smallest work of error is selected
For output.
Wherein, the feature decomposition module includes Eigenvalues Decomposition unit and eigendecomposition unit；The characteristic value
Decomposition unit be used for decomposite M eigenvalue cluster at M × M diagonal matrix；Described eigenvector decomposition unit is for decompositing M
M × M dimension unitary matrice M that mutually orthogonal M × 1 ties up column feature vector composition indicates the number of users of transmitting terminal.
The present apparatus emits user data to receiver by transmitting antenna, before emitting data to receiver, passes through tune
The data that device processed emits transmitting antenna are modulated, and are extended, are being passed through to modulated data using sequence extension module
After crossing multiple users share channel transfer, receiver receives user data, seeks weight matrix by MMSE detector, utilizes spy
Sign decomposing module decomposes the eigenmatrix in weight matrix, to decomposite eigenmatrix and characteristic value；And pass through
Valence noise building module is multiplied to the H-matrix of the eigenvectors matrix decomposited with noise, constructs equivalent noise matrix；Pass through
Result and equivalent noise matrix after the decomposition of estimation error module selected section is every kind corresponding to the subscriber signal of transmitting terminal
Modulation symbol is estimated with the error between the modulation symbol estimated value that estimates respectively；To the subscriber signal to transmitting terminal
It is reconstructed；The estimated value that equivalent noise is calculated using lagrange's method of multipliers module, eventually by error calculating module root
According to calculating the estimated value of equivalent noise, to calculate corresponding actual subscriber signal estimated value, hard decision module is from each
A actual subscriber signal estimated value selects one as output by way of hard decision.
The present embodiment combines specific data, is detected using Matlab to traditional MMSE-SIC detection method, MMSE-PIC
Method and the present invention BER error performance in awgn channel carry out emulation and comparative analysis, and simulation parameter setting is as shown in table 1,
Performance Simulation Results are as shown in Figure 5.According to simulation result, it can be seen that, the error compensation multi-user of the embodiment of the present invention changes parallel
Into detection method performance when Signal to Noise Ratio (SNR) is greater than 10dB better than the performance of MMSE-SIC detection method, and and MMSE-PIC
Detection performance, which is compared, to have greatly improved.With the increase of signal-to-noise ratio, innovatory algorithm is also more accurate.MMSE-SIC detection method
The complexity of middle matrix inversion is O (M3), the complexity of entire algorithm is proportional to O (M4+M3N), and in improved method, matrix
The complexity of characteristic value eigendecomposition is O (M3), various situationsCalculating be related to MN multiplication, a of each case
~calculating be related to 8+8N multiplication, so the complexity of entire innovatory algorithm is proportional to O (M3).When smaller for number of users,
Advantage of the innovatory algorithm in complexity is little, but when number of users increases, the complexity of improved method will be will be greatly reduced,
The complexity that improved method calculates will reduce by a magnitude than the complexity of MMSE-SIC method, this will be advantageous to 5G's
Magnanimity connection and low time delay scene.
The setting of 1 simulation parameter of table
The present invention can solve existing alternative multi-user test method there are the problem of, MMSE-SIC detection algorithm is related to
Matrix inversion, algorithm complexity are high, and when number of users is big, detection calculations amount is big, causes processing delay big；MMSE-PIC detection
Algorithm detection number of levels is few, though time delay is small, detection performance is poor；And quasi- parallel interference canceller detection algorithm, although processing
Time delay ratio MMSE-SIC is small, and performance is better than MMSE-PIC, but detection performance is still poor compared to MMSE-SIC.The present invention is
The improvement of MMSE-PIC detection method, MMSE-PIC detection method only carry out level-one detection, and performance is poor, and the present invention passes through compensation
MMSE criterion estimates the error of signal, and correlated noise is projected to characteristic vector space, searches on the direction of Noise enhancement
Signal improves detection accuracy to obtain better detection performance, reduces the bit error rate.The present invention applies to the more of MUSA system
The detection efficiency for the system that improves is realized green communications by user's detection.
Those of ordinary skill in the art will appreciate that all or part of the steps in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can be stored in a computer readable storage medium, storage
Medium may include: ROM, RAM, disk or CD etc..
Embodiment provided above has carried out further detailed description, institute to the object, technical solutions and advantages of the present invention
It should be understood that embodiment provided above is only the preferred embodiment of the present invention, be not intended to limit the invention, it is all
Any modification, equivalent substitution, improvement and etc. made for the present invention, should be included in the present invention within the spirit and principles in the present invention
Protection scope within.
Claims (9)
1. a kind of error compensation multi-user test method for MUSA system, which is characterized in that the described method includes:
S1, transmitting terminal multiple user data be modulated after, random selection complex spread sequence to respective modulation symbol into
Row extension, and sent from identical running time-frequency resource；
S2, receiving end carry out parallel interference canceller detection by MMSE detector to the signal y received, calculate weight matrix
ωMMSE, obtain the modulation symbol estimated value x of the subscriber signal of transmitting terminal；
S3, Eigenvalues Decomposition is carried out to the eigenmatrix in weight matrix, decomposites eigenvalue λl, corresponding feature vector vl, with
And the H-matrix of feature vector and noise are constructed into equivalent noise al；The number of users of l={ 1,2 ..., M }, M expression transmitting terminal
Mesh；
S4, the biggish eigenvalue λ of top n is chosenk, corresponding feature vector vkAnd equivalent noise akCalculate the use of transmitting terminal
The corresponding every kind of modulation symbol of family signal respectively with the error e between the modulation symbol estimated value that estimatesl(m)；K=1,
2,...,N}；
S5, using lagrange's method of multipliers by equivalent noise matrix aHA is minimized, and utilizes calculated error and eigenvalue λk
With feature vector vk, calculate the estimated value of the corresponding equivalent noise matrix of every kind of modulation symbol
S6, by the way of maximal possibility estimation according to the estimated value of equivalent noise matrixReadjust error el(m), and it is sharp
It is compensated with its error formed to detection algorithm, to calculate actual subscriber signal estimated valueAnd it selects wherein
The smallest actual subscriber signal estimated value of errorAs final output；
H-matrix indicates hermitian Hermitian matrix；M indicates the corresponding symbolic number of modulation system.
2. a kind of error compensation multi-user test method for MUSA system according to claim 1, which is characterized in that
In the step S3 in weight matrix eigenmatrix carry out Eigenvalues Decomposition include weight matrix be ωMMSE=(HHH+σ2I)-1HH, enable eigenmatrix R=(HHH+σ2I)-1, eigenmatrix R is subjected to characteristic value eigendecomposition, obtains R=VDVH；H
For the channel matrix of the user of transmitting terminal；HHFor the Hermitian matrix of channel matrix H；σ2Indicate noise variance；I indicates unit
Matrix；V indicates that eigenvectors matrix includes M × M dimension unitary matrice that M mutually orthogonal M × 1 tie up column feature vector composition；D table
Show eigenvalue matrix include M eigenvalue cluster at M × M diagonal matrix；VHIndicate the Hermitian matrix of eigenvectors matrix,
Enable includes that M ties up the equivalent noise matrix of equivalent noise as a=VHN', n'=(n'1,n'2,...,n'M)H, n'iIndicate assume the
The noise signal that l transmitting end subscriber has, subscript H indicate Hermitian matrix.
3. a kind of error compensation multi-user test method for MUSA system according to claim 1, which is characterized in that
The corresponding every kind of modulation symbol of the subscriber signal of transmitting terminal is indicated with the error between the modulation symbol estimated value x that estimates respectively
Forel(m) when indicating m-th of symbol that first of transmitting end subscriber is sent,
With the corresponding modulation symbol estimated value estimatedBetween error；bl(m) indicate what first of transmitting end subscriber may be sent
M-th of symbol；Indicate the modulation symbol estimated value that first of transmitting end subscriber is sent；(vk)lIndicate feature vector vkL
A element.
4. a kind of error compensation multi-user test method for MUSA system according to claim 1, which is characterized in that
The estimated value of the corresponding equivalent noise matrix of every kind of modulation symbolCalculation formula are as follows:
Wherein, C=[c1,c2,...,cM],(·)↑It indicates
Moore-Penrose pseudoinverse；E=[e1 *(m),e2 *(m),...,eM *(m)]H；* conjugate of symbol is indicated；bl(m) m-th of symbol that first of transmitting end subscriber may be sent is indicated, Indicate the modulation symbol estimated value that first of transmitting end subscriber is sent；(vk)l
Indicate feature vector vkFirst of element；Subscript H indicates Hermitian matrix.
5. a kind of error compensation multi-user test method for MUSA system according to claim 4, which is characterized in that
The estimated value of the corresponding equivalent noise matrix of every kind of modulation symbolCalculation formula further include subscriber signal pair with transmitting terminal
The every kind of modulation symbol b answeredl(m) it respectively with the error between the modulation symbol estimated value x that estimates as constraint condition, utilizes
Lagrange's method of multipliers minimizes cost function f [a], selects two transmitting end subscriber l1And l2To estimated valueEstimated；Its
In, the process that lagrange's method of multipliers minimizes cost function includes:
s.t.
Wherein, f [a] indicates estimated valueCost function；γ1And γ2Indicate Lagrange coefficient；el1(m) indicate actual
L in subscriber signal1The corresponding error of m-th of symbol that a transmitting end subscriber may be sent, el2(m) actual user's letter is indicated
L in number2The corresponding error of m-th of symbol that a transmitting end subscriber may be sent, * indicate conjugate of symbol； Subscriptl1And subscriptl2Expression belongs to two different transmittings of { 1,2 ..., M }
End subscriber；(v1)l1Indicate feature vector v1L1A element；(v1)l2Indicate feature vector v1L2A element.
6. a kind of error compensation multi-user test method for MUSA system according to claim 5, which is characterized in that
Emit end subscriber l1And l2Selection course include Signal to Interference plus Noise Ratio SINR by calculating each transmitting end subscriber, and it is carried out
Sequence selects the smallest two users of wherein SINR as transmitting end subscriber l1With transmitting end subscriber l2；The wherein calculating of SINR
Formula indicates are as follows:
Wherein, ωi,MMSEThat indicate is the i-th row of weight matrix, hiThat indicate is the i-th column of equivalent channel matrix H, σ2It is noise
Variance, | | | | represent Frobenius norm.
7. a kind of error compensation multi-user test method for MUSA system according to claim 1, which is characterized in that
Actual user's signal estimated value of final output in step S6It indicates are as follows:
Wherein,Indicate estimation calculated value；Argmin is indicatedObtain corresponding independent variable when minimum valueQ
() indicates hard decision；| | | | represent Frobenius norm.
8. a kind of error compensation multi-user detection device for MUSA system, which is characterized in that described device includes:
Transmitting antenna: for emitting user data；
Modulator: for being modulated to user data；
Sequence extension module: for carrying out sequence extension to modulation data；
Receiver: for receiving the user data after modulation extends；
MMSE detector: for acquiring weight square by minimizing the least mean-square error between transmission vector sum estimate vector
Battle array；
Feature decomposition module: Eigenvalues Decomposition is carried out for the eigenmatrix in weight matrix；Decomposite eigenvalue λi, it is corresponding
Feature vector vi；
Equivalent noise construct module: for according to eigendecomposition unit decomposition go out eigenvectors matrix H-matrix with make an uproar
Sound is multiplied；
Estimation error module: the biggish characteristic value of top n, corresponding feature vector and equivalent noise is taken to calculate transmitting terminal
The corresponding every kind of modulation symbol of subscriber signal respectively with the error between the modulation symbol estimated value that estimates；
Lagrange's method of multipliers module: for calculating corresponding equivalent noise square under every kind of symbol according to lagrange's method of multipliers
The estimated value of battle array；
Error calculating module: based on the estimated value of each equivalent noise matrix calculated according to lagrange's method of multipliers module
Calculate actual subscriber signal；
Hard decision module: for carrying out hard decision to calculated actual subscriber signal, it is the smallest as defeated to select error
Out.
9. a kind of error compensation multi-user detection device for MUSA system according to claim 8, which is characterized in that
The feature decomposition module includes Eigenvalues Decomposition unit and eigendecomposition unit；The Eigenvalues Decomposition unit is used for
Decomposite M eigenvalue cluster at M × M diagonal matrix；Described eigenvector decomposition unit is used to decomposite M mutually orthogonal M
M × M of × 1 dimension column feature vector composition ties up unitary matrice；The number of users of M expression transmitting terminal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910834807.8A CN110492956B (en) | 2019-09-05 | 2019-09-05 | Error compensation multi-user detection method and device for MUSA (multiple input multiple output) system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910834807.8A CN110492956B (en) | 2019-09-05 | 2019-09-05 | Error compensation multi-user detection method and device for MUSA (multiple input multiple output) system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110492956A true CN110492956A (en) | 2019-11-22 |
| CN110492956B CN110492956B (en) | 2021-07-06 |
Family
ID=68556589
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910834807.8A Active CN110492956B (en) | 2019-09-05 | 2019-09-05 | Error compensation multi-user detection method and device for MUSA (multiple input multiple output) system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110492956B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114665934A (en) * | 2022-03-18 | 2022-06-24 | 五邑大学 | Signal detection method, system, device and medium based on decomposition factor matrix |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102891815A (en) * | 2012-09-19 | 2013-01-23 | 北京航空航天大学 | Estimation method for low-complexity channel in time division duplex (TDD) multi-base station cooperative system |
| CN106059968A (en) * | 2016-05-27 | 2016-10-26 | 重庆邮电大学 | MUSA system multistage portion parallel interference elimination multi-user detection method |
| CN107809299A (en) * | 2017-10-23 | 2018-03-16 | 哈尔滨工业大学 | After the first string of multiple users share access technology up-link and multi-user test method |
| US20190229791A1 (en) * | 2018-01-19 | 2019-07-25 | Lenovo (Singapore) Pte. Ltd. | Channel compression matrix parameters |
-
2019
- 2019-09-05 CN CN201910834807.8A patent/CN110492956B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102891815A (en) * | 2012-09-19 | 2013-01-23 | 北京航空航天大学 | Estimation method for low-complexity channel in time division duplex (TDD) multi-base station cooperative system |
| CN106059968A (en) * | 2016-05-27 | 2016-10-26 | 重庆邮电大学 | MUSA system multistage portion parallel interference elimination multi-user detection method |
| CN107809299A (en) * | 2017-10-23 | 2018-03-16 | 哈尔滨工业大学 | After the first string of multiple users share access technology up-link and multi-user test method |
| US20190229791A1 (en) * | 2018-01-19 | 2019-07-25 | Lenovo (Singapore) Pte. Ltd. | Channel compression matrix parameters |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114665934A (en) * | 2022-03-18 | 2022-06-24 | 五邑大学 | Signal detection method, system, device and medium based on decomposition factor matrix |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110492956B (en) | 2021-07-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101578831B (en) | Systems and methods for enhanced channel estimation in wireless communication systems | |
| US8619918B2 (en) | Sparse channel estimation for MIMO OFDM systems | |
| RU2321949C1 (en) | Device and method for transmitting data using selected own vector in mimo-system of closed contour mobile communications | |
| US6987819B2 (en) | Method and device for multiple input/multiple output transmit and receive weights for equal-rate data streams | |
| US7420916B2 (en) | Fourier-transform based linear equalization for MIMO CDMA downlink | |
| US8599969B2 (en) | Communications channel estimation | |
| CN110166401B (en) | Phase noise suppression method of large-scale MIMO orthogonal frequency division multiplexing system | |
| US9031123B2 (en) | Communication system and method using subspace interference cancellation | |
| EP3952544A1 (en) | Method, apparatus and device for improving multi-user multiplexing performance, and storage medium | |
| Abebe et al. | Multi-sequence spreading random access (MSRA) for compressive sensing-based grant-free communication | |
| US20080240272A1 (en) | Systems and methods for n-dimensional leaf-node prediction for mimo detection | |
| CN110492956A (en) | A kind of error compensation multi-user test method and device for MUSA system | |
| CN112822697A (en) | Multi-user detection method and device for cell-free MIMO | |
| US9755860B2 (en) | Method of performing uplink channel estimation and base station using the same | |
| WO2016192619A1 (en) | New transmission schemes for device communications | |
| Horlin et al. | Optimal training sequences for low complexity ML multi-channel estimation in multi-user MIMO OFDM-based communications | |
| Torbatian et al. | Asynchronous interference alignment | |
| Xiao et al. | Achieving antenna and multipath diversities in GLRT-based burst packet detection | |
| JP6484205B2 (en) | Base station apparatus, perturbation vector search method and program | |
| Fukuma et al. | Space-time code division multiple access based on spatial modulation | |
| CN111212013A (en) | Extended sequence generation method for MUSA system | |
| CN110753012A (en) | Multi-user detection algorithm of time reversal multiple access system | |
| Wei et al. | Optimal binary/quaternary adaptive signature design for code-division multiplexing | |
| Sangeetha et al. | Performance analysis of subspace based downlink channel estimation for W-CDMA systems using chaotic codes | |
| Lei et al. | Analysis of joint transmit and receive antenna selection in CPM MIMO systems |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240507 Address after: 230000 B-1015, wo Yuan Garden, 81 Ganquan Road, Shushan District, Hefei, Anhui. Patentee after: HEFEI MINGLONG ELECTRONIC TECHNOLOGY Co.,Ltd. Country or region after: China Address before: 400065 Chongwen Road, Nanshan Street, Nanan District, Chongqing Patentee before: CHONGQING University OF POSTS AND TELECOMMUNICATIONS Country or region before: China |
|
| TR01 | Transfer of patent right |