CN107086971A - A kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations - Google Patents
A kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations Download PDFInfo
- Publication number
- CN107086971A CN107086971A CN201710180613.1A CN201710180613A CN107086971A CN 107086971 A CN107086971 A CN 107086971A CN 201710180613 A CN201710180613 A CN 201710180613A CN 107086971 A CN107086971 A CN 107086971A
- Authority
- CN
- China
- Prior art keywords
- mrow
- matrix
- msub
- detection
- mtr
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
- H04L25/03006—Arrangements for removing intersymbol interference
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
- H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
- H04B7/0413—MIMO systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/03—Shaping networks in transmitter or receiver, e.g. adaptive shaping networks
- H04L25/03006—Arrangements for removing intersymbol interference
- H04L2025/03592—Adaptation methods
- H04L2025/03598—Algorithms
- H04L2025/03611—Iterative algorithms
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Power Engineering (AREA)
- Radio Transmission System (AREA)
Abstract
The invention discloses a kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations, this method includes：MMSE detection matrix As are constructed according to the channel response matrix H of non-ideal communication channel；According to detection matrix A and channel correlation coefficient to detecting that each row of matrix set threshold value；Incomplete decomposing is carried out to detection matrix A according to the threshold value of setting and obtains preconditioning matrix D and L；According to preconditioning matrix D and L, the receipt signal matrix exported using steepest descent method to received end matched filterCarry out soft detection and obtain transmission signal estimateFast convergence rate of the present invention, complexity is low.
Description
Technical field
The present invention relates to wireless communication technology field, more particularly to a kind of extensive MIMO suitable for a variety of antenna configurations
Soft detection method.
Background technology
Emerging extensive multiple-input and multiple-output " Large Scale-Multiple-Input Multiple-Output "
(LS-MIMO) it is also known as that " Massive MIMO " (M-MIMO) system has become key technology and mesh in next generation mobile communication
The hot issue of preceding moving communicating field.Compared with legacy single-input single output system.In mimo system, there are multiple interference to disappear
Breath/symbol is sent in transmitting terminal, and at receiver, these symbols for being polluted or disturbing by random noise then need to be detected
Survey/decoding.This multiple symbol can be detected individually or jointly.It is different from independent detection, in joint-detection, each symbol
The characteristic of other symbols is must take into consideration in detection.Therefore, joint-detection usually can be realized than individually detection better performance,
Corresponding joint-detection computation complexity is higher.
The joint-detection of multiple symbols is to realize a variety of benefits of MIMO technology in mimo system.Because believing altogether
Road interference (CCI) is the limited most essential reason of communication system.Unfortunately, optimal MIMO test problems are proved to be non-multinomial
(NP-hard) of formula time complexity, therefore, all known optimization algorithm its complexities conceived for solution this problem
It is exponentially increased with the quantity of decision variable.Therefore, most preferably based on maximum likelihood (ML) criterion or based on maximum a posteriori (MAP)
The optimal MIMO detection algorithms of standard because of its too high computation complexity become that extensive mimo system can not be applied to.So
And with the development of semi-conductor industry, hardware computing capability is being sharply increased always for many years, " less pole in some cases
The computational complexity at end " is no longer considered as the bottleneck of practical application.Although it should be noted, however, that transistor is more next
Faster, supply voltage significant can not be reduced in contemporary metal oxide semiconductor (CMOS) technique.Therefore, it is nearly all existing
Maximum integration density can be all limited to for integrated circuit (IC), this limitation is due to that the power consumption and power density of surplus are led
What the too high chip internal temperature band caused was come.In other words, this problem, still limits the IC exploitations of today.Therefore, people
Can not simply rely on Moore's Law, even if as a result, appropriate complexity MIMO detection algorithm still too power consumption.Therefore,
Complexity is low but performance suboptimum MIMO detectors are that practical MIMO application needs algorithm.
Below since system model, the LS-MIMO key issues detected and difficult point are analyzed：Fig. 1 is shown
One typical MIMO model, wherein, base station (B antenna) and (B exemplified by user's (U single-antenna subscriber)>U).Passed up
In defeated (user to base station), because base station possesses powerful computing capability, it is possible to use pilot frequency sequence estimates channel, therefore on
The detection of line link can be carried out directly；In downlink transfer (base station to user), because each user is independent, and base station
Signal be while issuing multiple users, therefore the detection of downlink needs base station end to do precoding in advance to eliminate user
Between disturb, then detected by user terminal.The design mainly studies walking along the street, and its mode can be expressed as：(H is y=Hx+n
B*U channel matrix) so we to solve x will both sides multiply balanced matrix A, the transmission signal estimated:X '=A*y=
A*Hx+A*n.In order to obtain accurate estimate, it is necessary to which the selection of A*H=I. equilibrium matrix As has many kinds, simplest one
It is the pseudoinverse for taking A to be H to plant, and this is famous broken zero balanced (ZF)；In order to improve accuracy, A can also be that satisfaction is minimum
The balanced matrix of square error criterion (MMSE).But either break zero criterion or minimum mean square error criterion, its balanced matrix
The inverse of calculating matrix is required to, for LS-MIMO systems, although this class linear algorithm be it is very simple,
But still complexity is higher.Therefore several difficult points are just brought：1st, the hardware design inverted, it is understood that big matrix is inverted
High is required to hardware cost, while complicated calculating can cause the degradation of handling capacity, that is to say, that turning into herein
Where the bottleneck of MIMO efficiency.2nd, in addition to the algorithm for inverting part, the matrix multiplication of many places has been related in formula itself, this
A little big multiplication of matrices designs also constrain the efficiency of hardware.
It was exactly mathematically a variety of originally for the scheme inverted, and had and accurately seek method, have and method is asked by iterative approximation.
It is needed to carry out investigation and comparison from Detection results, hardware cost and throughput, preferably scheme is drawn.But three are examined
The direction of worry inherently contradiction, it will necessarily be had some thanks on more preferable accuracy so hardware if desired, thus scheme
A lot, but find the equalization point of contradiction but into the key point of problem, how some to be passed through by rational near-optimal
Allusion quotation theoretical framework is improved to reach more preferable effect, also into breach.
Differentiation simultaneously for different frames also needs to rely on an important factor, is exactly channel.Ratio is mentioned in document
More is all Gauss ideal communication channel.But in real life, it is impossible to ensure the desirability of channel, then set up a ginseng
The controllable non-ideal communication channel of number becomes the premise of this research.Under non-ideal communication channel, the performance of detection module and hardware into
This multi-party investigation is more of practical significance.
The content of the invention
Goal of the invention：The problem of present invention exists for prior art is there is provided a kind of suitable for the big of a variety of antenna configurations
The soft detection methods of scale MIMO.The present invention is directed to more common of antenna configuration, in order to improve the convergence rate of iterative algorithm, first
Drop detecting method under a kind of steepest of the LDL pretreatments based on coefficient correlation and user base station antenna ratio is first proposed, and is provided
Its hardware design based on FPGA.In order to reduce the complexity that pretreatment is brought, the present invention is caused in detection matrix using threshold value
Element zero setting, with simplify calculating；For complexity highest LLR (log-likelihood ratio) part in soft detection, the present invention is utilized
Pre-process obtained product directly to be calculated, greatly reduce complexity；In hardware aspect, present invention firstly provides be based on
The steepest of LDL pretreatments declines hardware structure and its optimization of detection.
Technical scheme：Extensive MIMO soft detection methods of the present invention suitable for a variety of antenna configurations include：
MMSE detection matrix As are constructed according to the channel response matrix H of non-ideal communication channel；
According to detection matrix A and channel correlation coefficient to detecting that each row of matrix set threshold value；
Incomplete decomposing is carried out to detection matrix A according to the threshold value of setting and obtains preconditioning matrix D and L；
According to preconditioning matrix D and L, the reception signal square exported using steepest descent method to received end matched filter
Battle arrayCarry out soft detection and obtain transmission signal estimate
Further, it is described that MMSE detection matrix As are constructed according to the channel response matrix H of non-ideal communication channel, specifically include：
MMSE detection matrix As are constructed according to below equation according to the channel response matrix H of non-ideal communication channel：
In formula, N0It is noise variance, EsIt is the mean power of transmission signal, Ι is unit matrix.
Further, it is described that each row of detection matrix are set according to detection matrix A, matrix size and channel related system
Threshold value is put, is specifically included：
According to detection matrix A and channel correlation coefficient to detecting that each row of matrix set threshold value, wherein, the threshold of the i-th row
It is worth and is：
In formula, i=1,2 ..., n, because A is square formation, n is the dimension of detection matrix A,For channel correlation coefficient, B is
Base station number, U is user terminal antenna number, the element that A (i, i) arranges for the i-th row i-th of detection matrix A.
Further, the threshold value according to setting to detection matrix A carry out incomplete decomposing obtain preconditioning matrix D and
L, is specifically included：
(1) by D (1,1)=A (1,1), wherein, the element of the row * row of shape such as Δ (, *) representing matrix Δ；
(2) i=2 is set；
(3) judge whether A (i, j) is more than or equal to pi, wherein, j=1,2 ..., i-1, if so, then performing step (4)；
(4) calculate according to the following formula：
(5) by i=i+1, and (3) are returned circulated, until the dimension of i=n, n for detection matrix A；
(6) D (i, j) obtained according to calculating is integrated and is obtained diagonal matrix D, and the L (i, j) obtained according to calculating is integrated and obtained
Lower triangular matrix L.
Further, it is described according to preconditioning matrix D and L, it is defeated to received end matched filter using steepest descent method
The receipt signal matrix gone outCarry out soft detection and obtain transmission signal estimateSpecifically include：
(1) initialize：x0=0,
(2) iterations d=1 is set；
(3) calculated according to below equation：
(4) by d=d+1, and it is back to (3), untill preset times m is iterated to, then xmEstimate for transmission signal matrix
Evaluation
Further, this method also includes：
According toL and D, to calculate log-likelihood ratio.
In formula,N0It is noise variance, EsIt is the mean power of transmission signal,Represent
G-th of symbol, D (g, g) be D g rows g row element, q0It is to belong to constellation collection Qb=0 bit is 0 symbol, q1It is
Belong to constellation collection Qb=1 bit is 1 symbol.
Beneficial effect：Compared with prior art, its remarkable advantage is the present invention：1st, the present invention changes from extensive MIMO is improved
Performance for detection algorithm is set out, it is proposed that the method handled in advance coefficient matrix before iteration, improves iteration inspection
The detection performance surveyed under a variety of MIMO scenes expansion of the correlation of channel and system scale (when).2nd, the present invention considers
The raising that performance improves brought complexity is arrived.Complexity is reduced from multiple angles:First, iterative part is from most simple
Steepest decline alternative manner, it is per single-step iteration matrix-vector multiplication fewer than the conjugate gradient method of general concern, and vector is interior
Product is calculated, scalar division；Secondly, this programme is directly using the conclusion of pretreatment, to calculate complexity highest in traditional algorithm
The calculating of likelihood ratio coefficient, greatly reduces complexity.3rd, it is of the invention compared with existing iteration detection method be applied to it is more many
The channel model of change and the system scale constantly expanded, remain to keep the low error rate under low signal-to-noise ratio under severe conditions, more
Fill it up with requirement of the sufficient next generation mobile communication for detection technique.
Brief description of the drawings
Fig. 1 is mimo channel illustraton of model；
Fig. 2 is the method Organization Chart of the present invention；
Fig. 3 is that cholesky directly inverts method in different channels condition to the inventive method with conjugation Gradient Iteration method
Under the bit error rate performance compare figure；
Fig. 4 is that cholesky directly inverts method in different system load to the inventive method with conjugation Gradient Iteration method
Under the bit error rate performance compare figure；
The complexity that Fig. 5 is the present invention and traditional conjugate gradient algorithms and the accurate inversion technique of Cholesky factorization is contrasted
Figure.
Embodiment
A kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations are present embodiments provided, its framework is as schemed
Shown in 2.The present embodiment is described in detail below.
1st, channel model
The present embodiment is Kronecker models using a correlated MIMO channel model being widely known by the people, and is described as
WhereinWithThe respectively correlation matrix of reception antenna and transmission antenna,
For independent identically distributed rayleigh fading channel.The two make is identical, and the correlation matrix RRx that receiving terminal is given below specifically gives birth to
Production mode：
Wherein ζ (0≤ζ≤1) represents to continuously transmit the order of magnitude of coefficient correlation between antenna, is preferable letter as ζ=0
Road model, as ζ=1, characterizes the transmission situation of correlation maximum, i.e. worst channel, ω is given phase, and it is simultaneously
The performance to whole system is not influenceed.RRx(p, q) representing matrix RRxPth row q row element value.
2nd, extensive MIMO MMSE detection models
In extensive mimo system, typically there are B ＞ ＞ U (antenna for base station number B is much larger than number of users U).X is allowed to represent U × 1
The signal vector of transmitting terminal transmitting, x contains the transmission symbol produced from U user.H represents channel response matrix, therefore base station
The received signal vector at end can be expressed as
Y=Hx+n0
Wherein n0It is the additive white Gaussian noise vector that a B × 1 is tieed up, its element is obeyed
The multiuser signal detection task of base station is exactly from the plus noise signal vector y estimation transmission signal codes received
x.In this embodiment it is assumed that receiver is known to channel matrix.Managed using least mean-square error (MMSE) linearity test
By the estimation to transmission signal vectors is expressed as
I.e.
3rd, pre-process
Due to being found by mathematical knowledge, the convergence rate of iterative algorithm depends on the conditional number of coefficient matrix, in this hair
Bright middle homography A.It is specifically that the bigger convergence rate of conditional number is slower.If also, with number of users and antenna for base station number
Ratio increase, or channel correlation coefficient increase, matrix conditional number increase.Based on this, the present invention proposes such as lower threshold value：
In formula, i=1,2 ..., n, n is detection square formation A dimension,For channel correlation coefficient, B is base station number, U
The element arranged for user terminal antenna number, A (i, i) for the i-th row i-th of detection matrix A.
Based on this threshold value, incomplete decomposing is carried out to detection matrix A and obtains preconditioning matrix D and L, is specifically included：
(1) by D (1,1)=A (1,1), wherein, the element of the row * row of shape such as Δ (, *) representing matrix Δ；
(2) i=2 is set；
(3) judge whether A (i, j) is more than or equal to pi, wherein, j=1,2 ..., i-1, if so, then performing step (4)；
(4) calculate according to the following formula：
(5) by i=i+1, and (3) are returned circulated, until i=n, n be the dimension of detection matrix A (A is square formation)；
(6) D (i, j) obtained according to calculating is integrated and is obtained diagonal matrix D, and the L (i, j) obtained according to calculating is integrated and obtained
Lower triangular matrix L.
Pretreatment can improve the performance and constringency performance of the iterative algorithm, and the present embodiment is using incomplete LDL points
Solution produces preconditioning matrix.It is that the present embodiment is in order to reduce complexity, based on channel with traditional LDL difference decomposed
Coefficient correlation and antenna proportioning propose threshold value, to the data zero setting less than threshold value so that original coefficient matrix is able to sparse
Change.And traditional LDL is decomposed, the result for a sparse matrix decomposition is often dense matrix, but in this programme, for
In original matrix zero it is original do not use decomposition algorithm, decomposition algorithm is used only for nonzero element, so as to ensure that before decomposition
Matrix is openness identical afterwards, reduces complexity.So as to using the method for zero setting reduction complexity.Zero setting will not be by with element
Preserve and participate in calculate.On the other hand, the result of LDL pretreatments is used directly for calculating the Soft Inform ation (log-likelihood of detection
Ratio, LLR).
4th, the soft detection of steepest descent method
According to preconditioning matrix D and L, the reception signal square exported using steepest descent method to received end matched filter
Battle arrayCarry out soft detection and obtain transmission signal estimateSpecifically include：
(1) initialize：x0=0,
(2) iterations d=1 is set；
(3) calculated according to below equation：
(4) by d=d+1, and it is back to (3), untill preset times m is iterated to, then xmEstimate for transmission signal matrix
Evaluation
The present embodiment uses the more simple steepest Descent iteration method compared with the conjugate gradient method generally studied, each time
A matrix-vector multiplication, two inner product of vectors operations, a scalar division are reduced in iterative process.
5th, log-likelihood ratio LLR is calculated
Calculated according to following formula
In formula,RepresentG-th of symbol, D (g, g) be D g rows g row element, q0 is to belong to constellation collection Qb
=0 bit is 0 symbol, q1It is to belong to constellation collection Qb=1 bit is 1 symbol.As can be seen here, LLR is calculated to still need to obtain
Detect inverse of a matrix.In the present invention, because the result that preprocessing part is obtained can regard the near of detection matrix inverse matrix as
Seemingly, therefore can as follows it be calculated using the obtained lower triangle coefficient matrix L of pretreatment：
Finally to the method for the present embodiment, the cholesky directly method of inverting is contrasted with conjugation Gradient Iteration method,
As a result as shown in Fig. 3,4,5, as can be seen from the figure in performance, in the present invention can obtain and be tall in various channel conditions
The bit error rate performance that the Si Ji directly methods of inverting are closer to, is 10 in the bit error rate-3When 0.5 arrive 1dB's compared with conjugate gradient method
Gain.Under different system loading condition, three kinds of methods show close under a low load, but in middle load and high load condition
The lower present invention will be better than conjugate gradient method.In terms of complexity, the present invention is also optimal.
Above disclosed is only a kind of preferred embodiment of the invention, it is impossible to the right model of the present invention is limited with this
Enclose, therefore the equivalent variations made according to the claims in the present invention, still belong to the scope that the present invention is covered.
Claims (6)
1. a kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations, it is characterised in that this method includes：
MMSE detection matrix As are constructed according to the channel response matrix H of non-ideal communication channel；
According to detection matrix A and channel correlation coefficient to detecting that each row of matrix set threshold value；
Incomplete decomposing is carried out to detection matrix A according to the threshold value of setting and obtains preconditioning matrix D and L；
According to preconditioning matrix D and L, the receipt signal matrix exported using steepest descent method to received end matched filter
Carry out soft detection and obtain transmission signal estimate
2. the extensive MIMO soft detection methods according to claim 1 suitable for a variety of antenna configurations, it is characterised in that：
It is described that MMSE detection matrix As are constructed according to the channel response matrix H of non-ideal communication channel, specifically include：
MMSE detection matrix As are constructed according to below equation according to the channel response matrix H of non-ideal communication channel：
<mrow>
<mi>A</mi>
<mo>=</mo>
<msup>
<mi>H</mi>
<mi>H</mi>
</msup>
<mi>H</mi>
<mo>+</mo>
<mfrac>
<msub>
<mi>N</mi>
<mn>0</mn>
</msub>
<msub>
<mi>E</mi>
<mi>s</mi>
</msub>
</mfrac>
<mi>I</mi>
</mrow>
In formula, N0It is noise variance, EsIt is the mean power of transmission signal, Ι is unit matrix.
3. the extensive MIMO soft detection methods according to claim 1 suitable for a variety of antenna configurations, it is characterised in that：
It is described that threshold value is set to each row of detection matrix according to detection matrix A, matrix size and channel related system, specifically include：
According to detection matrix A and channel correlation coefficient to detecting that each row of matrix set threshold value, wherein, the threshold value of the i-th row is：
In formula, i=1,2 ..., n, because A is square formation, n is the dimension of detection matrix A,For channel correlation coefficient, B is base station
Number, U is user terminal antenna number, the element that A (i, i) arranges for the i-th row i-th of detection matrix A.
4. the extensive MIMO soft detection methods according to claim 1 suitable for a variety of antenna configurations, it is characterised in that：
The threshold value according to setting carries out incomplete decomposing to detection matrix A and obtains preconditioning matrix D and L, specifically includes：
(1) by D (1,1)=A (1,1), wherein, the element of the row * row of shape such as Δ (, *) representing matrix Δ；
(2) i=2 is set；
(3) judge whether A (i, j) is more than or equal to pi, wherein, j=1,2 ..., i-1, if so, then performing step (4)；
(4) calculate according to the following formula：
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<mi>t</mi>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
<mo>=</mo>
<mi>A</mi>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
<mo>-</mo>
<mstyle>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<mi>k</mi>
<mo><</mo>
<mi>j</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</munderover>
<mrow>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
<mi>L</mi>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
</mrow>
</mstyle>
</mtd>
</mtr>
<mtr>
<mtd>
<mi>L</mi>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
<mo>=</mo>
<mi>t</mi>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
<mo>/</mo>
<mi>D</mi>
<mo>(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
</mtd>
</mtr>
<mtr>
<mtd>
<mi>D</mi>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>j</mi>
<mo>)</mo>
<mo>=</mo>
<mi>A</mi>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>i</mi>
<mo>)</mo>
<mo>-</mo>
<mstyle>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>k</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mrow>
<mi>k</mi>
<mo><</mo>
<mi>i</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</munderover>
<mrow>
<mi>t</mi>
<mrow>
<mo>(</mo>
<mi>i</mi>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
<mi>L</mi>
<mrow>
<mo>(</mo>
<mi>j</mi>
<mo>,</mo>
<mi>k</mi>
<mo>)</mo>
</mrow>
</mrow>
</mstyle>
</mtd>
</mtr>
</mtable>
</mfenced>
(5) by i=i+1, and (3) are returned circulated, until the dimension of i=n, n for detection matrix A；
(6) D (i, j) obtained according to calculating is integrated and is obtained diagonal matrix D, and the L (i, j) obtained according to calculating is integrated and obtained down three
Angle matrix L.
5. the extensive MIMO soft detection methods according to claim 1 suitable for a variety of antenna configurations, it is characterised in that：
It is described according to preconditioning matrix D and L, the receipt signal matrix exported using steepest descent method to received end matched filter
Carry out soft detection and obtain transmission signal estimateSpecifically include：
(1) initialize：x0=0,
(2) iterations d=1 is set；
(3) calculated according to below equation：
<mfenced open = "{" close = "">
<mtable>
<mtr>
<mtd>
<msub>
<mi>x</mi>
<mi>d</mi>
</msub>
<mo>=</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>+</mo>
<mfrac>
<mrow>
<mo>(</mo>
<msub>
<mi>z</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mover>
<mi>y</mi>
<mo>&OverBar;</mo>
</mover>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mrow>
<mo>(</mo>
<msub>
<mi>Az</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>z</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
</mfrac>
<msub>
<mi>z</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mover>
<mi>y</mi>
<mo>&OverBar;</mo>
</mover>
<mi>d</mi>
</msub>
<mo>=</mo>
<msub>
<mover>
<mi>y</mi>
<mo>&OverBar;</mo>
</mover>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>-</mo>
<mfrac>
<mrow>
<mo>(</mo>
<msub>
<mi>z</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mover>
<mi>y</mi>
<mo>&OverBar;</mo>
</mover>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
<mrow>
<mo>(</mo>
<msub>
<mi>Az</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>z</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
<mo>)</mo>
</mrow>
</mfrac>
<msub>
<mi>Az</mi>
<mrow>
<mi>d</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msub>
</mrow>
</mtd>
</mtr>
<mtr>
<mtd>
<mrow>
<msub>
<mi>z</mi>
<mi>d</mi>
</msub>
<mo>=</mo>
<msup>
<mrow>
<mo>(</mo>
<msup>
<mi>LDL</mi>
<mi>H</mi>
</msup>
<mo>)</mo>
</mrow>
<mrow>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msup>
<msub>
<mover>
<mi>y</mi>
<mo>&OverBar;</mo>
</mover>
<mi>d</mi>
</msub>
</mrow>
</mtd>
</mtr>
</mtable>
</mfenced>
(4) by d=d+1, and it is back to (3), untill preset times m is iterated to, then xmFor transmission signal Matrix Estimation value
6. the extensive MIMO soft detection methods according to claim 5 suitable for a variety of antenna configurations, it is characterised in that
This method also includes：
According toL and D, to calculate log-likelihood ratio.
In formula,N0It is noise variance, EsIt is the mean power of transmission signal,RepresentG
Individual symbol, D (g, g) is the element of D g rows g row, q0It is to belong to constellation collection Qb=0 bit is 0 symbol, q1It is to belong to
Constellation collection Qb=1 bit is 1 symbol.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710180613.1A CN107086971A (en) | 2017-03-24 | 2017-03-24 | A kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710180613.1A CN107086971A (en) | 2017-03-24 | 2017-03-24 | A kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107086971A true CN107086971A (en) | 2017-08-22 |
Family
ID=59614220
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710180613.1A Pending CN107086971A (en) | 2017-03-24 | 2017-03-24 | A kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107086971A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108736935A (en) * | 2018-05-18 | 2018-11-02 | 东南大学 | A kind of general down and out options method for extensive mimo system signal detection |
| CN109245804A (en) * | 2018-08-27 | 2019-01-18 | 江南大学 | Extensive MIMO signal detection method based on Jacobi iteration |
| CN109257076A (en) * | 2018-09-20 | 2019-01-22 | 东南大学 | Compression Landweber detection method and framework based on extensive MIMO |
| CN113395092A (en) * | 2021-05-21 | 2021-09-14 | 西安电子科技大学广州研究院 | Method and system for improving detection capability of Massive MIMO system |
| CN114389655A (en) * | 2022-01-21 | 2022-04-22 | 郑州大学 | Detection method for incoherent coding of large-scale MIMO system under related channel |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100031113A1 (en) * | 2008-07-30 | 2010-02-04 | Hsie-Chia Chang | Method and apparatus of candidate list augmentation for channel coding system |
| CN106332291A (en) * | 2016-09-14 | 2017-01-11 | 西安交通大学 | Quality of service-oriented downlink radio resource allocation method of user-centered network |
| CN106357309A (en) * | 2016-08-15 | 2017-01-25 | 东南大学 | Method of large scale MIMO linear iterative detection under non-ideal channel |
-
2017
- 2017-03-24 CN CN201710180613.1A patent/CN107086971A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100031113A1 (en) * | 2008-07-30 | 2010-02-04 | Hsie-Chia Chang | Method and apparatus of candidate list augmentation for channel coding system |
| CN106357309A (en) * | 2016-08-15 | 2017-01-25 | 东南大学 | Method of large scale MIMO linear iterative detection under non-ideal channel |
| CN106332291A (en) * | 2016-09-14 | 2017-01-11 | 西安交通大学 | Quality of service-oriented downlink radio resource allocation method of user-centered network |
Non-Patent Citations (1)
| Title |
|---|
| YE XUE ET AL: "Steepest Descent Method Based Soft-Output Detection for Massive MIMO Uplink", 《2016 IEEE INTERNATIONAL WORKSHOP ON SIGNAL PROCESSIG SYSTEMS(SIPS)》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108736935A (en) * | 2018-05-18 | 2018-11-02 | 东南大学 | A kind of general down and out options method for extensive mimo system signal detection |
| CN108736935B (en) * | 2018-05-18 | 2021-02-26 | 东南大学 | Universal descent search method for large-scale MIMO system signal detection |
| CN109245804A (en) * | 2018-08-27 | 2019-01-18 | 江南大学 | Extensive MIMO signal detection method based on Jacobi iteration |
| CN109245804B (en) * | 2018-08-27 | 2020-10-09 | 江南大学 | Large-scale MIMO signal detection method based on Jacobi iteration |
| CN109257076A (en) * | 2018-09-20 | 2019-01-22 | 东南大学 | Compression Landweber detection method and framework based on extensive MIMO |
| CN109257076B (en) * | 2018-09-20 | 2020-06-30 | 东南大学 | Large-scale MIMO-based compressed Landweber detection method and system |
| CN113395092A (en) * | 2021-05-21 | 2021-09-14 | 西安电子科技大学广州研究院 | Method and system for improving detection capability of Massive MIMO system |
| CN114389655A (en) * | 2022-01-21 | 2022-04-22 | 郑州大学 | Detection method for incoherent coding of large-scale MIMO system under related channel |
| CN114389655B (en) * | 2022-01-21 | 2022-11-08 | 郑州大学 | Detection method for incoherent coding of large-scale MIMO system under related channel |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Younis et al. | Generalised sphere decoding for spatial modulation | |
| CN107086971A (en) | A kind of soft detection methods of extensive MIMO suitable for a variety of antenna configurations | |
| Prabhu et al. | Hardware efficient approximative matrix inversion for linear pre-coding in massive MIMO | |
| CN103548310B (en) | MIMO receiver using lattice reduction and k- best detection | |
| Peng et al. | A 1.58 Gbps/W 0.40 Gbps/mm2 ASIC Implementation of MMSE Detection for $128\times 8~ 64$-QAM Massive MIMO in 65 nm CMOS | |
| CN109951214B (en) | Signal detection method suitable for large-scale MIMO system | |
| CN101499840B (en) | Iteration detection method for MIMO system | |
| US9001924B1 (en) | Matrix inversion | |
| Zhang et al. | Large-scale MIMO detection design and FPGA implementations using SOR method | |
| CN105915477A (en) | Large-scale MIMO detection method based on GS method, and hardware configuration | |
| CN103746728B (en) | The MIMO of a kind of mixed self-adapting receives detection method | |
| Chataut et al. | Huber fitting based ADMM detection for uplink 5G massive MIMO systems | |
| CN106357309A (en) | Method of large scale MIMO linear iterative detection under non-ideal channel | |
| Lee et al. | Symbol detection in V-BLAST architectures under channel estimation errors | |
| Hua et al. | Signal detection in uplink pilot-assisted multi-user MIMO systems with deep learning | |
| Rahman et al. | Deep learning based pilot assisted channel estimation for Rician fading massive MIMO uplink communication system | |
| US8107546B2 (en) | Detection method of space domain maximum posteriori probability in a wireless communication system | |
| Jang et al. | Combined deep learning and SOR detection technique for high reliability in massive MIMO systems | |
| CN108736934A (en) | A kind of efficient extensive mimo system signal detecting method | |
| CN107222246A (en) | The efficient extensive MIMO detection method and system of a kind of approximated MMSE-based performance | |
| CN106357318B (en) | The adjustable extensive MIMO iteration detection method of rate of convergence | |
| CN107733487B (en) | Signal detection method and device for large-scale multi-input multi-output system | |
| Liu et al. | Alternative normalized-preconditioning for scalable iterative large-MIMO detection | |
| CN107196686A (en) | A kind of extensive mimo system signal detecting method with pretreatment operation | |
| Moezzi-Madani et al. | A low-area flexible MIMO detector for WiFi/WiMAX standards |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170822 |