CN1832386A - Multi-selection multi-input multi-out put detection method and device based on reducing conditional number - Google Patents
Multi-selection multi-input multi-out put detection method and device based on reducing conditional number Download PDFInfo
- Publication number
- CN1832386A CN1832386A CNA2005100543742A CN200510054374A CN1832386A CN 1832386 A CN1832386 A CN 1832386A CN A2005100543742 A CNA2005100543742 A CN A2005100543742A CN 200510054374 A CN200510054374 A CN 200510054374A CN 1832386 A CN1832386 A CN 1832386A
- Authority
- CN
- China
- Prior art keywords
- signal
- estimation
- matrix
- row
- data
- 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
Images
Classifications
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Radio Transmission System (AREA)
Abstract
This invention relates to a multi-in multi-out test method based on reducing condition numbers including: carrying out channel estimation to received signals to get a channel property matrix H of row M, orderly eliminating the ith row of H to make up of M channel matrix Hi of the new M-1 rows, computing the condition numbers of them and selecting the Hi with the smallest condition numbers as the output, selecting K estimations to the ith transmitted signals to let the received signal r deduct the re-built signal of the signal so as to generate K new received signals, a test method includes utilizing matrix Hi to test the remaining M-1 matrixes to constitute estimation of M data together with the ith data to generate estimation of K groups of data and selecting the optimum data from the data to be output.
Description
Technical field
The present invention relates to the signal detection technique in a kind of multi-aerial radio communication system, particularly, can improve error performance in multiple-input, multiple-output (MIMO) system based on the multiple selection multiple-input and multiple-output detection method and the device that reduce conditional number.
Background technology
Multiple-input, multiple-output (MIMO) technology is the important breakthrough of wireless mobile communications art.The MIMO technology is meant the transmission of data and receives and all adopted many antennas.Studies show that utilize the MIMO technology can improve the capacity of channel, the while also can be improved the reliability of channel, reduces the error rate.The heap(ed) capacity of mimo system or maximum size be linear increasing with the increase of minimum antenna number.And under similarity condition, adopting the common antenna system of many antennas or aerial array at receiving terminal or transmitting terminal, its capacity only increases with the logarithm of antenna number.Comparatively speaking, the MIMO technology has great potentiality for the capacity that improves wireless communication system, is the key technology that the third generation mobile communication system adopts.
Figure 1 shows that the mimo system structural representation of common employing.In this structure, make a start and receiving end adopt n respectively
TAnd n
RIndividual antenna carries out the transmission and the reception of signal.Transmitting terminal comprises serial/parallel converter unit 101 and a plurality of transmitting antenna 102.Receiving terminal comprises a plurality of reception antennas 103, channel estimating unit 104 and detector 105.For simplicity, only show among Fig. 1 and the part that is used to illustrate that its operation is relevant.
At transmitting terminal, data to be sent at first are divided into n through serial/parallel converter unit 101
TIndividual data flow, the corresponding transmitting antenna of each data flow.At receiving terminal, at first by n
R Individual reception antenna 103 receives signal, carries out channel estimating by channel estimating unit 104 according to this received signal then, estimates current characteristic of channel matrix H.MIMO detection module 105 utilizes this characteristic of channel matrix H to detect to received signal, demodulates the information bit of making a start and sending.
Traditional detector mainly adopts the Maximum Likelihood Detection method, and ZF (ZF) method and least mean-square error (MMSE) method detect.
Maximum likelihood detection method can come out by the noise variance direct derivation of abundant statistics vector, but the complexity of Maximum Likelihood Detection is with being exponential increase, is difficult to realize.
The characteristics of ZF detector are the interference of having eliminated fully between each transmitting antenna, and its cost is to have strengthened background noise.The basic thought of MMSE detector is that the data of estimation and the mean square error between the real data are minimized.It has considered the influence of background noise, eliminate between each antenna and strengthen obtain between the background noise one compromise, performance is better than the ZF detector.
In addition, also has BLAST detector (ZF-BLAST and MMSE-BLAST).The BLAST detector comprises two parts: a linear transformation and a serial interference elimination device.At first obtain data decision on the strongest i root transmitting antenna of signal to noise ratio,, rebuild the transmission data of i antenna by these data by linear transformation.And then from received signal, deduct the influence of this symbol.Then the data estimation of calculating on the antenna that signal to noise ratio is the strongest in the remaining data is carried out interference eliminated.Repeat this process then up to the estimation that obtains all data always.
Yet, above-mentioned MIMO detector exist or complexity higher, or strengthen defective such as background noise.
Summary of the invention
In view of the above problems, made the present invention.The purpose of this invention is to provide a kind of method and apparatus that detects based on the multiple selection MIMO that reduces conditional number, can improve error performance.
In order to realize purpose of the present invention, according to an aspect of the present invention, provide a kind of based on the multiple selection multiple-input and multiple-output detection method that reduces conditional number, comprise step: receiving end/sending end is by the signal of a plurality of antenna transmission, the signal that receives is carried out channel estimating, estimate current, have the characteristic of channel matrix H of M row, wherein M is a natural number; Remove the i row of channel matrix H successively, constitute the channel matrix H of M new M-1 row
i, wherein i is a natural number; Calculate the channel matrix H of M new M-1 row
iConditional number, and the channel matrix H of selector bar number of packages minimum
iAs output; Select K to i estimation that sends signal, make received signal r deduct the reconstruction signal r of this signal
j=r-H (:, i)
i(j), thereby produce K new received signal, wherein K is a natural number, H (:, the i) i of representing matrix H row,
i(j) expression is to j estimation of i signal that sends; Select a kind of detection method, utilize matrix H
i, detect M-1 remaining matrix, and, form the estimation of this M data with i columns certificate, generate the estimation of K group data.Utilize the method for maximum likelihood, from these K group data, select the conduct output of optimum data.
According to another aspect of the present invention, provide a kind of, comprising based on the multiple selection multiple-input and multiple-output checkout gear that reduces conditional number: matrix conditional number calculation element, be used for successively the M row channel matrix H that estimates being removed the i row, constitute new channel matrix H
i, produce M matrix H
i, and calculate this M H
iConditional number; Select output device, according to a described M conditional number, a H of selector bar number of packages minimum
iOutput to the parallel detection device; The parallel detection device is according to H
i, H (:, i), to the estimation of i signal
i(j) and received signal, estimate M-1 remaining transmission signal, and and
i(j) constitute together sending the estimation of signal, produce the K group and estimate, and this K group is estimated that sending into maximum likelihood relatively installs; Maximum likelihood relatively installs, and is used for estimating to select a best conduct output from this K group.
The method according to this invention can improve performance, but complexity improves seldom.In this method is bright, have or not the method performance of interference eliminated very approaching, and in traditional method, performance have evident difference.So, use method of the present invention, can not select the method for interference eliminated for use, like this, can make complexity low.
Description of drawings
By below in conjunction with description of drawings the preferred embodiments of the present invention, will make above-mentioned and other purpose of the present invention, feature and advantage clearer, wherein:
Fig. 1 is the structural representation of existing mimo system;
Fig. 2 is the constellation schematic diagram of normalized 16QAM modulation;
Fig. 3 realizes carrying out the block diagram of the system of multiple selection multiple-input, multiple-output detection by reducing conditional number according to the embodiment of the invention;
Fig. 4 is the flow chart that the minimizing conditional number is carried out the method for multiple selection multiple-input, multiple-output detection that passes through according to the embodiment of the invention;
Fig. 5 is the emulation schematic diagram relatively according to MIMO detection method of the present invention and existing various detection methods;
Fig. 6 is the comparative result schematic diagram of expression according to MIMO detection method M-ZF under different K values of the embodiment of the invention; With
Fig. 7 is the MIMO detection method of expression according to the embodiment of the invention, the comparative result schematic diagram of LR-ZF and LR-SIC.
Embodiment
With reference to the accompanying drawings embodiments of the invention are described in detail, having omitted in the description process is unnecessary details and function for the present invention, obscures to prevent that the understanding of the present invention from causing.
Illustrate at first that below based on the multiple selection MIMO detection method that reduces conditional number, receiving terminal is by a plurality of antenna (n
RIndividual) signal that receiving end/sending end sends, carry out channel estimating by channel estimating unit according to this received signal then, estimate current characteristic of channel matrix H.We can suppose that channel matrix H has the M row.In order to reduce conditional number, channel matrix H is removed the i row, constitute new channel matrix H
iFor example, since first row, deletion first row constitute new matrix H
1, delete secondary series then and constitute H
2, and the like.Suppose that H has four row, then 2 of H, 3,4 row constitute H
11,3,4 row of H constitute H
21,2,4 row of H constitute H
31,2,3 row of H constitute H
4Like this, one have 4 such H
iTotal like this M H
i, calculate this M H
iConditional number, a H of selector bar number of packages minimum
i(being equivalent to leave out the i row of H).
The conditional number of matrix is meant the maximum singular value of matrix and the ratio between the minimum singular value.The conditional number of channel matrix is big more, and correlation between channels is big more.The performance of detector is also just poor more.Compare with Maximum Likelihood Detection, when the conditional number of channel matrix is big more, ZF, the detector that MMSE etc. are traditional and the poor performance of Maximum Likelihood Detection are big more.When each row quadrature of channel, condition number of channel is 1, and this moment, ZF or matched filter were equivalent to maximum likelihood, but complexity is very low.Therefore, if determine a kind of detection method, condition number of channel is more little, and performance is good more.At this M H
iIn, the H of selector bar number of packages minimum
iBe in order in testing process subsequently, to improve performance.
Select K to i estimation that sends signal, can make received signal r deduct the reconstruction signal (r of this signal
j=r-H (:, i)
i(j)), like this, one meets generation together K new received signal.Then, select a kind of detection method, utilize matrix H
i, detect M-1 remaining data, and and i data supposing together, form the estimation of this M data.Like this, the estimation of total K group data.In these K group data, utilize the method for maximum likelihood, select best conduct output.
According to the modulation system such as N-QAM, the selection of i signal there is the N kind.For example for the 16-QAM modulation, each signal has 16 kinds of possibilities.Select K to i estimation that sends signal, be meant and in N optional value, select K as i estimation that sends signal.The criterion of selecting is in K value, chooses i real transmission signal with big as far as possible probability.Below enumerate two kinds of methods of selecting K
1) under the situation of K=N, real transmission signal is in selected set;
2) when K＜N, before selection, should use certain traditional detector, ZF for example, MMSE determines K estimation choosing, wherein K is a natural number.For example obtain the estimation of i preceding signal of demodulation with the ZF method.In modulation constellation shown in Figure 2, select K nearest with it constellation point as candidate value.
Selected H
iAfterwards, promptly H has deleted the matrix that the i row constitute.If received signal is r, H (:, i) the i row of expression H,
i(j) expression is to j estimation of i signal that sends.Then make r
j=r-H (:, i)
i(j).One total K r
jIf
i(j) estimate correctly, then r
jIn just removed s
iInterference to remaining data.At this moment, just can use r
j, H
iDetect a remaining M-1 data.This M-1 signal estimated and
i(j) together, constituted estimation to all M data.Such data estimation one total K is individual.In this K candidate solution, utilize the method for maximum likelihood, select best exporting as demodulation.
The model of mimo system is described below.
Can make s=[s
1..., s
Nt]
TExpression transmits nt * 1 dimensional vector of symbol.S wherein
iIt is the symbol of i root antenna transmission.The signal vector of corresponding nr * 1 reception antenna is r=[r
1..., r
Nr]
T, can use following expression (1) expression;
R=Hs+n (1) is n=[n wherein
1..., n
Nr]
TBe illustrated in the zero-mean on the nr root reception antenna, variance is σ
2White Gauss noise.H is nr * nt channel matrix.The purpose of MIMO receiver detector is to recover to send symbol s from receive vector x.
Below in conjunction with Fig. 3 and 4 explanation the specific embodiment of the present invention.
Initial condition at first is set, supposes to adopt mimo system structure shown in Figure 1.Wherein, transmitting terminal and receiving terminal adopt n respectively
TAnd n
RIndividual antenna carries out the transmission and the reception of signal.At transmitting terminal, data to be sent at first are divided into n through serial/parallel converter unit 101
TIndividual data flow, the corresponding transmitting antenna of each data flow.At receiving terminal, at first by n
RThe signal that individual reception antenna 103 receiving end/sending ends send.Then, carry out channel estimating according to this received signal, estimate current characteristic of channel matrix H by channel estimating unit 104.
Fig. 3 illustrates the block diagram based on the multiple selection MIMO checkout gear that reduces conditional number according to the embodiment of the invention.
As shown in Figure 3, the MIMO checkout gear according to present embodiment comprises matrix conditional number computing unit 200, selection output unit 201, estimation generation unit 202, parallel detection unit 203 and maximum likelihood comparison unit 204.
Wherein matrix conditional number computing unit 200 removes the i row with the channel matrix H (channel matrix has the M row) that estimates, and constitutes new channel matrix H
i, total like this M matrix H
iCalculate this M H
iConditional number.Select output unit 201 according to this M conditional number, a H of selector bar number of packages minimum
iOutput to parallel detection unit 203.Simultaneously, estimate that 202 pairs i of generation unit sends signal and make K estimation, and be provided to parallel detection unit 203.Parallel detection unit 203 is according to received signal H
i, H (:, i), to the estimation of i signal
i(j) and received signal, estimate M-1 remaining transmission signal, and and
i(j) constitute together sending the estimation of signal.Producing the K group like this, altogether estimates.This K group is estimated to send into relatively unit 204 of maximum likelihood.Relatively from estimating, this K group selects a best conduct output in unit 204 by maximum likelihood.
Describe based on the multiple selection MIMO detection method that reduces conditional number referring to Fig. 4.
At first, in step S301, input channel H (H has the M row), received signal r.Set the number K of parallel processing.Then, in step S302, channel matrix H (channel matrix has the M row) is removed the i row, constitute new channel matrix H
i, total like this M H
iFor example, channel H has 4 row, 4 row, channel H=
-0.1416 - 0.0035i - 0.5992 - 0.3845i - 0.2550 - 0.2926i 0.1165 +0.2207i
-0.1131 - 0.2360i 0.1427 - 0.1723i - 0.0363 - 0.0946i -0.1911 -0.1952i
0.0948 - 0.8188i -0.0546 - 0.2316i 0.2448 - 0.9284i 0.2447 +0.3439i
-0.4797 + 0.3428i 0.6783 - 0.2910i 0.7938 + 0.0299i 0.0726 +0.1419i
Then delete first row and constitute new matrix H
1=
-0.5992-0.3845i -0.2550-0.2926i 0.1165+0.2207i
0.1427 -0.1723i -0.0363-0.0946i -0.1911-0.1952i
-0.0546-0.2316i 0.2448-0.9284i 0.2447+0.3439i
0.6783-0.2910i 0.7938+0.0299i 0.0726+0.1419i
The deletion secondary series constitutes new matrix H
2=
-0.1416-0.0035i -0.2550-0.2926i 0.1165+0.2207i
-0.1131-0.2360i -0.0363-0.0946i -0.1911-0.1952i
0.0948-0.8188i 0.2448-0.9284i 0.2447+0.3439i
-0.4797+0.3428i 0.7938+0.0299i 0.0726+0.1419i
Delete the 3rd row and constitute new matrix H
3=
-0.1416-0.0035i -0.5992-0.3845i 0.1165+0.2207i
-0.1131-0.2360i 0.1427-0.1723i -0.1911-0.1952i
0.0948-0.8188i -0.0546-0.2316i 0.2447+0.3439i
-0.4797+0.3428i 0.6783-0.2910i 0.0726+0.1419i
Delete the 4th row and constitute new matrix H
4=
-0.1416-0.0035i -0.5992-0.3845i -0.2550-0.2926i
-0.1131-0.2360i 0.1427-0.1723i -0.0363-0.0946i
0.0948-0.8188i -0.0546-0.2316i 0.2448-0.9284i
-0.4797+0.3428i 0.6783-0.2910i 0.7938+0.0299i
At step S303, calculate this M H
iConditional number, select the then H of conditional number minimum
iAs output.
For example, the conditional number of four matrixes among the step S302 is as shown in table 1
Table 1:H
iConditional number
And the conditional number of matrix H is 10.7163.At H
iIn, the conditional number minimum be H
4, promptly H deletes the matrix that the 4th row structure goes out.So H
4Be the best output.
| H 1 | H 2 | H 3 | H 4 | |
| Conditional number | 4.0836 | 4.9784 | 3.5753 | 2.4459 |
After this, at step S304, select K i estimation s that sends signal
i(j), calculate r
j=r-H (:, i) s
i(j).For example,, send signal, can make if can obtain the 4th according to formula (1)
r
i=r-H (:, i) s (i)=H
iIn the superincumbent formula of s '+n (2) (2), i of s (i) expression sends signal.I signal vector that sends signal removed in s ' expression.H (:, i) the i row of expression H.In this example, i=4.At this moment, condition number of channel is 2.4459, compares with original condition number of channel 10.7163, and it is a lot of to descend.This is that the data in the formula (2) are detected, and no matter uses which kind of method, and performance all can better than detecting with H (two kinds of the methods of maximum likelihood be all the same).Following table 2 has provided the situation of the combination of antennas of varying number (from 2 antenna to 8 antennas) to the minimizing of average condition number.
Table 2: the average condition number of different antennae combination.
| 2×2 | 3×3 | 4×4 | 5×5 | 6×6 | 7×7 | 8×8 | |
| H | 4.3 | 7.2 | 10.6 | 15 | 16.8 | 20.6 | 24.5 |
| H i(the conditional number minimum) | 1 | 1.7 | 2.7 | 3.6 | 4.5 | 5.4 | 6.4 |
As seen from Table 2, conditional number approximately is 1/4 when not removing the i row.
Select K to i estimation that sends signal.For example, modulation system can adopt 16-QAM (Fig. 2 shows modulation constellation), and each signal has 16 kinds of possibilities.Select K estimation to be meant and in 16 optional values, select K as i estimation that sends signal to i transmission signal.The criterion of selecting is in K value, chooses i real transmission signal with big as far as possible probability.
For example, under the situation of K=16, real transmission signal is in selected set.In K＜16, for example during K=4, before selection, should use certain traditional detector, such as ZF, MMSE determines K the estimation of choosing.For example try to achieve the estimation of i preceding signal of demodulation with the ZF method.In modulation constellation, select K nearest with it constellation point as candidate value.In this example, for example, the output valve of ZF is 0.4037+0.6564i.Then choosing opposes the 4th with its 4 nearest constellation point sends the estimation of signal, is respectively 0.3162+0.9487i, 0.3162+0.3162i, 0.9487+0.9487i, and 0.9487+0.3162i.
Next, in step S305, utilize H
i, r
j, select a kind of detection method to detect M-1 remaining data.This M-1 data and s
i(j) constitute whole data estimation.Total K such signal estimated.Be described as follows:
Can make received signal r deduct the reconstruction signal formula (2) of this signal.Like this, common property is given birth to K new received signal r
j, 1＜=j＜=K.J represents that the estimation of i signal is j in the candidate collection.
Then, select a kind of detection method, utilize matrix H
i, detect M-1 remaining data, with
i(j) constitute together sending the estimation of signal.Like this, total K group estimation.For example, select the method for ZF or grid minimizing (lattice reduction), to each r
j=H
iS ' obtains the estimation of s '.For example, i=4, j=1, K=4 is 0.3162+0.9487i to the 4th candidate collection that sends data, 0.3162+0.3162i, 0.9487+0.9487i, 0.9487+0.3162i selects.Obtain estimating that by ZF 1-3 estimation that sends data is (0.3162+0.3162i-0.9487+0.9487i-0.9487+0.9487i), the estimation of then whole transmission data is (0.3162+0.3162i-0.9487+0.9487i-0.9487+0.9487i 0.3162+0.9487i).Like this, have 4 groups of such estimations.
At step S306, in this K candidate solution, utilize the method for maximum likelihood, select separating of a best as output.Specifically, in S305, obtained K candidate solution, the corresponding estimation of each candidate solution
Calculate each selection
At this K
In, select to make
Minimum
As output, obtain the estimation of signal
Promptly
The following describes in the method that reduces computation complexity in based on the multiple selection multi-input and multi-output detection method that reduces conditional number according to the present invention.
Calculating H
iWhen middle, need the design conditions number, a kind of method is the conditional number of compute matrix.The line number of building matrix H equals columns and equals M.Another kind method is the singular value decomposition with matrix, and general complexity is O (M
4) (O represents exponent number here).Calculate H
iThe time, needing to calculate M, then total complexity is O (M
5).But in fact, the conditional number of compute matrix only need know that maximum characteristic value and minimum characteristic value get final product, and do not need whole characteristic values.Like this, maximum characteristic value can be tried to achieve by power method, and minimum characteristic value can be tried to achieve by inverse power method.Their complexity is O (M
2), calculate M, complexity is O (M
3).So just greatly reduce the complexity of calculating.
About code reuse, if linear detector, as ZF, MMSE, its weighting matrix of lattice reduction W can only calculate once.Like this, in K demodulation, W can reuse.If the method for interference eliminated is decomposed or the like such as QR, then QR decomposes and can only calculate once, needs calculating K but utilize QR to solve an equation and organize.
Calculate
Complexity is O (M
2), need calculating K, so complexity is O (KM
2).Calculate
Be only to calculate ‖ r
i-H
iS ＇ ‖, two is equivalent.But the minimizing of amount of calculation is few, so whichever complexity all is O (KM
2).
Be that ZF (MMSE) and grid reduce its complexity is described below with the detector.
1. detector is ZF (MMSE)
At step S303, design conditions number, complexity O (M
3) irrelevant with the selection of detector.At step S305, calculate weighting matrix
(subscript+expression pseudoinverse), complexity O (M
3).Calculate Wr
j, each complexity is that (M* (M-1) is O (M
2).
Divide K=N and K＜N that (N is an order of modulation) (step S304, S305) is discussed below.
Under the situation of K=N, complexity is O (NM
2).Under the situation of K＜N, because K＜N needs to use extra computation W1=H
+(O (M
3)), and calculate W1r, complexity M
2So total complexity is O (M
3)+O ((K+1) M
2).
Calculate
Complexity O (KM
2) (step S306).Total complexity is above sum, greatly about O (M
3)～O (M
4), detect roughly at an order of magnitude with BLAST.
2. grid reduces
Reducing for grid, generally use the method for LLL (Lenstra-Lenstra-Lovasz-Reduced), because its complexity is lower, is 3 powers of matrix dimension.
At first, design conditions number, complexity O (M
3) (as top step S303).Computing grid reduces, and complexity is greatly about O (M
3) (S305).
Each demodulation, complexity is O (M
2), when K=N, complexity is O (NM
2), the same with ZF when K＜N, need extra computation, total complexity also is O (M
3)+O ((K+1) M
2) (as step S304, S305).Calculate
Complexity is O (KM
2) (as step S306).
Total complexity is above sum, approximately is O (M
3)～O (M
4), detect also roughly at an order of magnitude with BLAST.
Fig. 5 and Fig. 6 have provided the simulation result based on the multiple selection multi-input and multi-output detection method that reduces conditional number.In emulation, Nt=Nr=4, modulation system is 16-QAM (as shown in Figure 2), each element of channel matrix H produces at random, is multiple Gaussian Profile.In Fig. 5, K=16 promptly necessarily comprises the correct estimation to s (i).Among Fig. 5, M-LR, M represent multiple selection of the present invention, and LR represents H
iUse the method for grid minimizing (lattice reduction), utilize the data of method detection except that s (i) of ZF then; M-ZF represents that multiple selection use ZF of the present invention is as detector; SD represents that ball detects; M-LRSIC represents H
iThe method of using grid to reduce, the method with similar interference eliminated detects data then.As can be seen from Figure 5, method of the present invention is with respect to the method for BLAST, and performance improves greatly.From figure also as can be seen, to H
iCarry out the method that grid reduces again, the method ten minutes that performance and ball detect approaching.And, the method performance of M-LRSIC generally speaking, performance also slightly is better than ball decoding.And because the method that ball detects is commonly considered as very the method near maximum likelihood.But the complexity of ball decoding is higher, and the selection of initial radium is difficult.If select greatly, the space of search is too big.Otherwise, select for a short time, might not have again and separate, and complexity is uncertain, just average complexity is polynomial complexity.And method complexity of the present invention is determined, according to the difference of K value, and complexity max (O (M
3), O (KM
2)), greatly about O (M
3)～O (M
4) between, roughly can be considered as O (M
4), with BLAST approaching complexity is arranged, but performance is near maximum likelihood.
When Fig. 6 shows K and gets different value, the comparison of the performance of M-ZF.The performance that the grid of the method that the use conventional mesh that shows Fig. 7 reduces and the thought of utilization interference eliminated reduces between (LR-SIC) compares.Comparison diagram 7 and Fig. 5, as can be seen, it is a lot of to adopt method of the present invention that performance is improved, but complexity improves seldom.If relatively have or not interference eliminated simultaneously, for example LR ZF and LR-SIC.As can be seen, in the method for the invention, have or not the method performance of interference eliminated very approaching, and in traditional method, performance have evident difference.So, use method of the present invention, can not select the method for interference eliminated for use.Like this, can make complexity low.
Invention has been described in conjunction with the preferred embodiments above.It should be appreciated by those skilled in the art that under the situation that does not break away from the spirit and scope of the present invention, can carry out various other change, replacement and interpolations.Therefore, scope of the present invention should not be understood that to be limited to above-mentioned specific embodiment, and should be limited by claims.
Claims (11)
1. one kind based on the multiple selection multiple-input and multiple-output detection method that reduces conditional number, comprises step:
Receiving end/sending end is by the signal of a plurality of antenna transmission, and the signal that receives is carried out channel estimating, estimate current, have the characteristic of channel matrix H of M row, wherein M is a natural number;
Remove the i row of channel matrix H successively, constitute the channel matrix H of M new M-1 row
i, wherein i is a natural number;
Calculate the channel matrix H of M new M-1 row
iConditional number, and the channel matrix H of selector bar number of packages minimum
iAs output;
Select K to i estimation that sends signal, make received signal r deduct the reconstruction signal of this signal
Thereby produce K new received signal, wherein K is a natural number, H (:, the i) i of representing matrix H row,
Expression is to j estimation of i signal that sends;
Select a kind of detection method, utilize matrix H
i, detect M-1 remaining data, and, form the estimation of this M data with i columns certificate, generate the estimation of K group data; With
Utilize the method for maximum likelihood, from these K group data, select the conduct output of optimum data.
2. method according to claim 1 wherein according to the modulation system of N-QAM, has the N kind to the selection of i signal.
3. method according to claim 1 and 2 is wherein selected K to i estimation that sends signal, is meant to select K as i estimation that sends signal in N optional value.
4. method according to claim 3 is wherein selected the step of K value to comprise with big as far as possible probability and is chosen real i to send signal.
5. method according to claim 1 is wherein selected the step of the conduct output of optimum data further to comprise from K group data and is obtained K candidate solution, estimation of each candidate solution correspondence
(1≤i≤K) calculates each selection
At this K
In, select to make
Minimum
As output, obtain the estimation of signal
Step.
6. method according to claim 1 is wherein calculated the channel matrix H that M new M-1 is listed as
iThe step of conditional number comprise with the singular value of matrix and carry out step of decomposition.
7. method according to claim 1, the method that wherein said detection method adopts grid to reduce.
8. method according to claim 1, wherein said detection method adopts the ZF detection method.
9. method according to claim 1, wherein said detection method adopts the least mean-square error detection method.
10. one kind based on the multiple selection multiple-input and multiple-output checkout gear that reduces conditional number, comprising:
Matrix conditional number calculation element is used for successively the M row channel matrix H that estimates being removed the i row, constitutes new channel matrix H
i, produce M matrix H
i, and calculate this M H
iConditional number;
Select output device, according to a described M conditional number, a H of selector bar number of packages minimum
iOutput to the parallel detection device;
The parallel detection device, according to received signal, H
i, H (:, i), to the estimation of i signal
Estimate M-1 remaining transmission signal, and with
Constitute together sending the estimation of signal, produce the K group and estimate, and this K group is estimated that sending into maximum likelihood relatively installs;
Maximum likelihood relatively installs, and is used for estimating to select a best conduct output from this K group.
11. device according to claim 10 is estimated that wherein generation unit also sends signal to i and makes K estimation, and is provided to the parallel detection device.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2005100543742A CN1832386A (en) | 2005-03-10 | 2005-03-10 | Multi-selection multi-input multi-out put detection method and device based on reducing conditional number |
| PCT/JP2006/304799 WO2006095873A1 (en) | 2005-03-10 | 2006-03-10 | Mimo detection control apparatus and mimo detection control method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNA2005100543742A CN1832386A (en) | 2005-03-10 | 2005-03-10 | Multi-selection multi-input multi-out put detection method and device based on reducing conditional number |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN1832386A true CN1832386A (en) | 2006-09-13 |
Family
ID=36953462
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNA2005100543742A Pending CN1832386A (en) | 2005-03-10 | 2005-03-10 | Multi-selection multi-input multi-out put detection method and device based on reducing conditional number |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN1832386A (en) |
| WO (1) | WO2006095873A1 (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102098242A (en) * | 2010-12-21 | 2011-06-15 | 山东大学 | Iterative detection method of unitary space-time codes (USTCs) in multiple input multiple output (MIMO) system |
| CN103746731B (en) * | 2014-01-21 | 2017-03-15 | 电子科技大学 | Multiple-input and multiple-output detector and detection method based on probability calculation |
| CN116843656A (en) * | 2023-07-06 | 2023-10-03 | 滁州正汇科技有限公司 | Plastic coating control method and system for steel belt pipe |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008062587A1 (en) * | 2006-11-22 | 2008-05-29 | Fujitsu Limited | Mimo-ofdm communication system and mimo-ofdm communication method |
| JP5496020B2 (en) * | 2010-08-25 | 2014-05-21 | 三菱電機株式会社 | Demodulator and demodulation method |
| US9671445B2 (en) * | 2013-03-15 | 2017-06-06 | Litepoint Corporation | System and method for testing radio frequency wireless signal transceivers using wireless test signals |
-
2005
- 2005-03-10 CN CNA2005100543742A patent/CN1832386A/en active Pending
-
2006
- 2006-03-10 WO PCT/JP2006/304799 patent/WO2006095873A1/en active Application Filing
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102098242A (en) * | 2010-12-21 | 2011-06-15 | 山东大学 | Iterative detection method of unitary space-time codes (USTCs) in multiple input multiple output (MIMO) system |
| CN102098242B (en) * | 2010-12-21 | 2013-08-14 | 山东大学 | Iterative detection method of unitary space-time codes (USTCs) in multiple input multiple output (MIMO) system |
| CN103746731B (en) * | 2014-01-21 | 2017-03-15 | 电子科技大学 | Multiple-input and multiple-output detector and detection method based on probability calculation |
| CN116843656A (en) * | 2023-07-06 | 2023-10-03 | 滁州正汇科技有限公司 | Plastic coating control method and system for steel belt pipe |
| CN116843656B (en) * | 2023-07-06 | 2024-03-15 | 安徽正汇汽配股份有限公司 | Plastic coating control method and system for steel belt pipe |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2006095873A1 (en) | 2006-09-14 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN1297076C (en) | Transmission/reception apparatus for wireless system with three transmission antennas | |
| CN1835425A (en) | Self-adaptive modulation method based on multi-user precode | |
| CN101039137A (en) | Method and system for reducing codebook search-based precoding feedback bits of MIMO-OFDM system | |
| CN1581725A (en) | Method and apparatus for determining a shuffling pattern in a double space-time transmit diversity system | |
| CN101032109A (en) | A method of processing received signals in a multi-input multi-output (MIMO) system | |
| CN1832386A (en) | Multi-selection multi-input multi-out put detection method and device based on reducing conditional number | |
| CN1926779A (en) | CPICH processing for SINR estimation in W-CDMA system | |
| CN1841961A (en) | Method and apparatus for selecting transmitting antenna in multi antenna wireless communication system | |
| CN1941660A (en) | Multi-user diversity method and system in multi-antenna radio communication system | |
| CN1636419A (en) | Burst detector | |
| CN101043220A (en) | Apparatus and method for other cell interference cancellation in broadband wireless communication system | |
| CN1941664A (en) | Transmission antenna selecting method and apparatus based on judge feedback in radio communication system | |
| CN101064579A (en) | Method for detecting low-complexity globular decoding | |
| CN1633051A (en) | A low-complexity MIMO detector approximating maximum likelihood detection performance | |
| CN106788626B (en) | Improved orthogonal space modulation transmission method capable of obtaining second-order transmit diversity | |
| CN1968067A (en) | Simplified maximum likelihood test method and device | |
| CN102158311A (en) | Iteration detection method for optimizing serial interference elimination sequence | |
| CN102237950B (en) | A kind of subscriber equipment, base station and channel information feedback method | |
| CN1968043A (en) | Transmitting diversity method and MIMO communication channel | |
| CN1822531A (en) | Airspace filter detecting method for multiple antenna radio communication system | |
| CN101888287A (en) | Signal detection method and device for multi-input multi-output receiver | |
| CN1702979A (en) | Equalization technique and associated detection technique combined receiver and receiving method thereof | |
| CN1886956A (en) | Method and device of multiple antenna receiver | |
| CN1649291A (en) | Receiving and detecting method of vertical layered space-time system based on self adaptive modulation | |
| CN1819571A (en) | Inspection of blank field maximum rear-proving probability in wireless communication system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |