CN108737032A - A kind of compression superposition sequence C SI feedback methods - Google Patents
A kind of compression superposition sequence C SI feedback methods Download PDFInfo
- Publication number
- CN108737032A CN108737032A CN201810494993.0A CN201810494993A CN108737032A CN 108737032 A CN108737032 A CN 108737032A CN 201810494993 A CN201810494993 A CN 201810494993A CN 108737032 A CN108737032 A CN 108737032A
- Authority
- CN
- China
- Prior art keywords
- sequence
- bit
- csi
- rarefication
- degree
- 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
- 238000000034 method Methods 0.000 title claims abstract description 39
- 230000006835 compression Effects 0.000 title claims abstract description 15
- 238000007906 compression Methods 0.000 title claims abstract description 15
- 238000003892 spreading Methods 0.000 claims abstract description 18
- 238000005516 engineering process Methods 0.000 claims abstract description 12
- 230000007935 neutral effect Effects 0.000 claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 11
- 239000011159 matrix material Substances 0.000 claims description 19
- 230000014509 gene expression Effects 0.000 claims description 11
- 238000001228 spectrum Methods 0.000 claims description 11
- 230000017105 transposition Effects 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims description 9
- 229910002056 binary alloy Inorganic materials 0.000 claims description 4
- 238000011002 quantification Methods 0.000 claims description 4
- GNFTZDOKVXKIBK-UHFFFAOYSA-N 3-(2-methoxyethoxy)benzohydrazide Chemical compound COCCOC1=CC=CC(C(=O)NN)=C1 GNFTZDOKVXKIBK-UHFFFAOYSA-N 0.000 claims description 3
- 238000005259 measurement Methods 0.000 claims 1
- 230000005540 biological transmission Effects 0.000 abstract 1
- 238000013139 quantization Methods 0.000 abstract 1
- 230000003044 adaptive effect Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 4
- 239000000203 mixture Substances 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 2
- 238000004891 communication Methods 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 238000012163 sequencing technique Methods 0.000 description 1
- 238000000638 solvent extraction Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/0001—Systems modifying transmission characteristics according to link quality, e.g. power backoff
- H04L1/0023—Systems modifying transmission characteristics according to link quality, e.g. power backoff characterised by the signalling
- H04L1/0026—Transmission of channel quality indication
-
- 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
- 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
- H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
- H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
- H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
- H04B7/0619—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal using feedback from receiving side
- H04B7/0621—Feedback content
- H04B7/0626—Channel coefficients, e.g. channel state information [CSI]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Quality & Reliability (AREA)
- Compression, Expansion, Code Conversion, And Decoders (AREA)
Abstract
The present invention provides a kind of compressions to be superimposed sequence C SI feedback methods, and this method includes：Position sequence is built, the location index of nonzero element and neutral element in CSI is recorded；Single-bit compressed sensing quantization is carried out to CSI, obtains bit information；Position sequence, bit information and degree of rarefication are constituted into bit sequence；Bit sequence is subjected to spread processing, obtains spread data sequence, then forms sequence spreading after the zero-adding of tail portion；It is superposed to send sequence after sequence spreading and data sequence weighting processing；Receiver receives transmission sequence, and recovers bit sequence using decoding technology, then recovers position sequence, bit information and degree of rarefication；Receiver reconstructs CSI using position sequence, bit information and degree of rarefication.The advantage of the invention is that：Improve the overhead bit needed for feedback；System bandwidth expense is reduced, while also improving the precision of CSI feedback.
Description
Technical field
The present invention relates to channel feedback techniques in communication system in the compression of user terminal, overlap-add procedure and base station end
Signal reconstruction technical field, more particularly to a kind of compression are superimposed sequence C SI feedback methods.
Background technology
Relative to traditional frequency division duplex (FDD, Frequency Division Duplex) multiple-input and multiple-output (MIMO,
Multiple-Input Multiple-Output) system, the extensive mimo system based on FDD can bring power system capacity, steady
The promotion of the performances such as qualitative.However, many operation (such as multi-users for bringing performance boost of the extensive mimo system based on FDD
Scheduling, rate-allocation, transmitting terminal precoding etc.) it is obtained with dependence in accurate CSI.Therefore, can base station end accurately obtain down
Row CSI directly affects the performance of the extensive mimo systems of FDD.
In the extensive mimo system based on FDD, downlink CSI acquisitions usually rely on channel feedback techniques.Traditional " base
In code book " CSI feedback method be then difficult in practice because antenna number crowd's feedback overhead of extensive mimo system is huge
It is applied.Relative to the CSI feedback method of " being based on code book ", (present invention is referred to as the CSI feedback technology based on compressed sensing
For " traditional compressed sensing CSI feedback " technology), it is fed back after using CSI there is sparse characteristic to be compressed to CSI, it can be one
Determine to improve CSI feedback expense in degree.Even so, " traditional compressed sensing CSI feedback " still feedback overhead is larger for technology,
Certain bandwidth resources are occupied, are remained to be further improved.
Invention content
The present invention in view of the drawbacks of the prior art, provides a kind of method of compression superposition sequence C SI feedback methods.Into
When row channel feedback, by the location index information of the nonzero element of CSI and neutral element, together with the CSI after single-bit quantification, superposition
It is fed back on data information.To so that the CSI feedback bandwidth cost additional without system, and CSI feedback can be improved
Precision.
In order to realize the above goal of the invention, the technical solution adopted by the present invention is as follows：
A kind of compression superposition sequence C SI feedback methods, including：
Behaviour below is made to channel state information (CSI, Channel State Information) H of sparse N × 1
Make：
A) " position sequence " Z that N × 1 is constructed with 0,1 element, records the location index of nonzero element and neutral element in CSI；
B) single-bit quantification is carried out to CSI using single-bit compression sensing method, " the bit letter of M × 1 after being quantified
Breath " Y；
C) " position sequence " Z, " bit information " Y and degree of rarefication K are constituted into NL× 1 " bit sequence " L；The NLIt is full
Sufficient NL=N+2M+NK；
The NKIt indicates required number of bits when degree of rarefication K binary representations, is set according to engineering experience, it is usually full
Sufficient NK< log2N；
D) spread processing is carried out to " bit sequence " L, obtained " spread data sequence "Then by " spread data sequence "Add N in tail portionzeroN is formed after a zeroD× 1 " sequence spreading " W；
" spread data sequence "According to formulaIt obtains, whereinIndicate spread spectrum sequence
Row, symbolIndicate Kronecker products；
The NzeroMeet
The NDIt is chosen according to engineering experience, usually chooses ND> > NL；Wherein, c indicates the length of spread spectrum, also known as
Spreading gain.
E) by " sequence spreading " W and ND× 1 " data sequence " D is according to formula
T=a × W+ (1-a) × D
It is weighted processing, to form ND× 1 " sending sequence " T is sent；
The constant a is selected according to engineering experience, meets 0 a≤1 <；
F) receiver receives " sending sequence " T, and recovers " bit sequence " L using technology is despread, and then restores
Go out " position sequence " Z, " bit information " Y and degree of rarefication K；
G) receiver utilizes " position sequence " Z, " bit information " Y and degree of rarefication K to reconstruct CSI.
Further, " position sequence " Z described in step a), be after reading pre-stored CSI sequences H, will be first in H
0 is labeled as at the position that element is zero, element is that 1 is labeled as at the position of non-zero.
Further, step a) includes：
It is N to read pre-stored length, and the channel state information H and size that degree of rarefication is K are M × N random distributions
Calculation matrix Φ.
Further, the sparse grade K described in step a) is estimated to obtain by user terminal receiver channel；The channel shape
State information H is estimated to obtain by user terminal receiver channel.
Further, the random distribution measures desirable：Gaussian Profile random matrix, bernoulli distribution random matrix,
Subalpine forests random matrix and very sparse projection matrix, etc..
Particularly, step a) and existing channel feedback handle the difference is that, when carrying out CSI feedback, consideration
The non-zero of CSI elements and the location index information of neutral element.
Further, step b) includes：
By the real part H of CSI sequences HrealWith imaginary part HimagIt is compressed to obtain " bit using single-bit compressed sensing principle
The real and imaginary parts of information " Y, " bit information " Y are denoted as Y respectivelyreal, Yimag。
The single-bit compression, obtains the real part Y of " bit information " YrealWith imaginary part Yimag, obtained according to following formula：
Operation the sign () expressions take symbol manipulation, i.e.,：
Further, step c) includes：
Real and imaginary parts, the degree of rarefication K of " bit information " Y that " position sequence " Z, length that length is N are M will be obtained,
Composition length is NL" bit sequence " L, i.e.,：
L=(ZT,(Yreal)T,(Yimag)T, (Kbin)T)T
Wherein, transposition arithmetic operation is sought in subscript " T " expression；
The KbinIt is N to indicate that degree of rarefication K is expressed as the length constituted when binary systemKColumn vector；
Further, step d) includes：
It is describedIndicate that frequency expansion sequence, spreading length c areWherein, symbolIndicate downward rounding
Number, the NzeroMeet
The NDIt is chosen according to engineering experience, usually chooses ND> > NL；The symbolIndicate downward rounding
Number.Wherein, c indicates the length of spread spectrum, also known as spreading gain.
Further, step 0 includes：
Using despreading after technology recovers " bit sequence " L, according in c) position relationship and each sequence
Length recovers " position sequence " Z, " bit information " Y and degree of rarefication K from " bit sequence " L, is then taken to " bit information " Y
Real and imaginary parts.The c) in position relationship be：
L=(ZT,(Yreal)T,(Yimag)T, (Kbin)T)T
Wherein, transposition arithmetic operation is sought in subscript " T " expression；
The KbinIt is N to indicate that degree of rarefication K is expressed as the length constituted when binary systemKColumn vector；
Further, step g) further comprises：The CSI reconstruct, refers to that restructing algorithm is added in " position sequence " Z
Constraints is constituted, and combines the existing reconstructing method based on single-bit compressed sensing.Described is " existing to be based on single-bit
Binary system iteration hard -threshold (BIHT, Binary Iterative Hard can be used in the reconstructing method of compressed sensing "
Thresholding), iteration hard -threshold (IHT, Iterative Hard Thresholding), adaptive 1-Bit compressed sensings
(Adaptive 1-Bit CS)。
Compared with prior art the advantage of the invention is that：
Single-bit compressed sensing technology is introduced on the basis of traditional compressed sensing CSI feedback, is improved needed for feedback
Overhead bit；Meanwhile also introducing " superposition feedback " mode：By the compressed CSI information of single-bit, together with CSI nonzero elements with
The location index information that neutral element is constituted, which is superimposed upon together on upstream data, to be fed back.System bandwidth expense is reduced, simultaneously
Also improve the precision of CSI feedback.
Description of the drawings
Fig. 1 is the CSI feedback method flow schematic diagram of the present invention；
Algorithm flow chart when Fig. 2 is present invention reconstruct channel signal.
Specific implementation mode
To make the objectives, technical solutions, and advantages of the present invention more comprehensible, it develops simultaneously embodiment below in conjunction with attached drawing, it is right
The present invention is described in further details.
As shown in Figure 1, to channel state information (CSI, Channel State Information) H of sparse N × 1
Make operation below：
A) " position sequence " Z that N × 1 is constructed with 0,1 element, records the location index of nonzero element and neutral element in CSI；
A1 it is N) to read pre-stored length, and the channel state information H and size that degree of rarefication is K are that M × N is random
Distribution measuring matrix Φ.
The H is the sparse signal that degree of rarefication is K, i.e. only has K nonzero element in H, and meet K < < N.
To ensure that the reconstruction property of single-bit compressed sensing, the compressed sampling number M usually meet M >=N.
Preferably, random distribution calculation matrix production method includes that Gaussian Profile random matrix, bernoulli are distributed random square
Battle array, subalpine forests random matrix and the unusual existing random distribution calculation matrix production method such as sparse projection matrix.
A2 the location index that nonzero element and neutral element in H) are recorded with 0,1 element, at the position for being zero by element in CSI
Labeled as 0, element is that 1 is labeled as at the position of non-zero, to constitute " position sequence " Z that length is N.
Example 1：Described " position sequence " the Z construction process is as follows：
Assuming that N=16, K=5.Then H is represented by H=(h1,h2,…,h16)T.It illustrates, practical large-scale MIMO
In system, the element in channel state information H is hundreds and thousands of, and N=16 is conveniently taken here for writing.Assuming that the element of H
In remove element " h1,h3,h5,h10,h16" outside 5 elements, remaining element is 0.It can obtain " position sequence " Z, be expressed as Z=
(1010100001000001)T, wherein the transposition arithmetic operation of subscript T representing matrixes.
B) single-bit quantification is carried out to CSI using single-bit compression sensing method, " the bit letter of M × 1 after being quantified
Breath " Y；
Channel state information (CSI, Channel State Information) H is taken into real part HrealWith imaginary part Himag, profit
It is compressed to obtain " bit information " Y with single-bit compressed sensing principle, the real and imaginary parts of " bit information " Y are denoted as respectively
Yreal, Yimag。
Wherein, take real and imaginary parts by following channel state information (CSI, Channel State Information) H
Formula carries out：
Hreal=real (H)
Himag=imag (H)
Wherein, the process compressed using compressed sensing principle is：
Wherein, the operation sign () expressions take symbol manipulation, i.e.,：
The Y obtained at this timerealAnd YimagIt is the vector of M × 1, and vector element is 0 and 1.
C) " position sequence " Z, " bit information " Y and degree of rarefication K are constituted into NL× 1 " bit sequence " L；
The NLMeet NL=N+2M+NK；The NKIndicate required number of bits when degree of rarefication K binary representations；
C1) by obtain length be " position sequence " Z of N, length be M " bit information " Y real part YrealAnd imaginary part
YimagWith degree of rarefication K, composition length is NL" bit sequence " L, i.e., constructed according to following formula：
L=(ZT,(Yreal)T,(Yimag)T, (Kbin)T)T
Wherein, transposition arithmetic operation is sought in subscript " T " expression；The KbinIndicate that degree of rarefication K is expressed as constituting when binary system
Length be NKColumn vector；It is set according to engineering experience, usually meets NK< log2N；
D) spread processing is carried out to " bit sequence " L, obtained " spread data sequence "Then by " spread spectrum data sequence
Row "Add N in tail portionzeroN is formed after a zeroD× 1 " sequence spreading " W；
D1 it is N) to obtain lengthL" bit sequence " L, calculate spreading gain c, obtain frequency expansion sequenceWherein,
Spreading gain c is obtained as follows：
The length N of " data sequence " D to be sentDIt is much larger than " position sequence " Z and " bit information " Y length constituted
NL" bit sequence " L.Wherein, symbolIndicate downward round numbers.
D2 spread processing) is carried out to " bit sequence " L, is obtained " spread data sequence "Wherein spread processing is by following
Formula carries out：
Wherein, symbolIndicate Kronecker products.
D3) by " spread data sequence "Add N in tail portionzeroN is formed after a zeroD× 1 " sequence spreading " W；Described
NzerO meetsThe NDIt is chosen according to engineering experience, usually chooses ND> >
NL；The symbolIndicate downward round numbers.Wherein, c indicates the length of spread spectrum, also known as spreading gain.
" sequence spreading " W can be obtained：
Wherein, transposition arithmetic operation is sought in subscript " T " expression
E) by " sequence spreading " W and ND× 1 " data sequence " D is according to formula：
T=a × W+ (1-a) × D
It is weighted processing, to form ND× 1 " sending sequence " T is sent；The constant a is passed through according to engineering
Test selected, 0 a≤1 < of satisfaction；
F) receiver receives " sending sequence " T, and restores " bit sequence " L using the technology that despreads, and then recovers
" position sequence " Z, " bit information " Y and degree of rarefication K.
F1 it is N that) receiving terminal, which receives length,D" send sequence " T after, using technology is despread, recover that " bit is believed
The length of breath " Y and " position sequence " Z compositions is NL" bit sequence " L.Wherein, the despreading process is as follows：
Under normal conditions, it is N that receiving terminal receives length to f1-1D" send sequence " T,Afterwards, right
It carries out piecemeal processing, is divided into the matrix of m c × 1, wherein m values carry out as follows：
Wherein, the symbolIndicate that downward round numbers, c are Spread Spectrum MatrixSpread spectrum increase.To
It obtains " sending sequence "Partitioning Expression of A be：
Wherein,
F1-2 takes segmentation sequences "In elementUtilize identical spread spectrum square
Battle array is de-spread to obtain " bit sequence " L, L=(l1,…,lm)T, de-spread and carry out as follows：
Wherein, conjugate transposition operation is sought in subscript " H " expression.
F2) according to L=(ZT,(Yreal)T,(Yimag)T, (Kbin)T)T, " position sequence " Z length is N, " bit information " Y's
Real part YrealWith imaginary part YimagLength is M, KbinLength is NK, it is N using lengthL" bit sequence " L can recover " position
Sequence " Z, " bit information " Y and degree of rarefication K.I.e. relationship solves " position sequence " Z, the real part of " bit information " Y according to the following formula
YrealWith imaginary part Yimag, degree of rarefication K：
The wherein transposition arithmetic operation of subscript T representing matrixes.
G) receiver utilizes " position sequence " Z, " bit information " Y and degree of rarefication K to reconstruct CSI.
Reconstruct CSI described in step g) is the reconstruct that " position sequence " Z is utilized and carries out signal.Due to the present invention be by
" position sequence " Z that the nonzero element of CSI and the location index information of neutral element are constituted is felt with existing based on single-bit compression
Auxiliary information is formed in the reconstructing method known.
Binary system iteration hard -threshold can be used in " the existing reconstructing method based on single-bit compressed sensing "
(BIHT, Binary Iterative Hard Thresholding), iteration hard -threshold (IHT, Iterative Hard
Thresholding), adaptive 1-Bit compressed sensings (Adaptive 1-Bit CS), etc..
Embodiment 1
Below by taking BIHT restructing algorithms as an example, as shown in Figure 2.
Initialization data：Location information " position sequence " Z that reconstruct obtains is read, is read " bit information " that reduction obtains
Y real parts YrealWith imaginary part Yimag, degree of rarefication K, calculation matrix Φ, maximum iteration Iteration, step-length β.At the beginning of reconstruction signal
Initial value Xreal=0, Ximag=0, iterations initial value n=0.The maximum iteration Iteration is according to engineering experience
It is set.The step-length β is the step parameter declined for regulating gradient, set by engineering experience, 0 < β of usual value
≤1。
G 1) the update X that is iteratedreal=Xreal+βΦT(Yreal-sign(ΦXreal)), Ximag=Ximag+βΦT
(Yimag-sign(ΦXimag)).Wherein, ΦTExpression does transposition arithmetic operation to matrix Φ.
G2) according to " position sequence " Z, by XlaerAnd XimagCorresponding to " position sequence " Z labeled as member at 0 location index
Plain zero setting.
G3) the update of iterations, n=n+1 return to g2 if n < Iteration)；Otherwise, by XrealWith XimagInto
Row normalized, i.e.,：
Wherein, " | | | |2" indicate that Frobenius norms is taken to operate.Then, according to formula：
MergeWithObtain CSI signalsWherein, i states imaginary number, meets i2=-1.
According to above-mentioned specific example, the present invention has used single-bit compression compared to traditional channel feedback method
The method of perception reduces the overhead bit of system.The position sequence for simultaneously constituting sequence neutral element and nonzero element, with data
Sequence is overlapped feedback so that the auxiliary for having location information when single-bit reconstruction signal greatly improves reconstruction signal
Precision.To reduce the bandwidth cost of system, the availability of frequency spectrum is improved, the precision of CSI feedback is improved.
Those of ordinary skill in the art will understand that in each method embodiment of the present invention, the sequence of each step
The sequencing that number can not be used to limit each step is not paying creative labor for those of ordinary skill in the art
Under the premise of dynamic, the priority of each step is changed also within protection scope of the present invention.The embodiments described herein be in order to
Reader is helped to understand the implementation of the present invention, it should be understood that protection scope of the present invention is not limited to such especially old
It states and embodiment.Those skilled in the art can make various do not depart from according to the technical disclosures disclosed by the invention
Various other specific variations and combinations of present invention essence, these variations and combinations are still within the scope of the present invention.
Claims (10)
1. a kind of compression is superimposed sequence C SI feedback methods, which is characterized in that including：
Operation below is made to the channel state information H of sparse N × 1：
A) " position sequence " Z that N × 1 is constructed with 0,1 element, records the location index of nonzero element and neutral element in CSI；
B) single-bit quantification, " bit information " Y of M × 1 after being quantified are carried out to CSI using single-bit compression sensing method；
C) " position sequence " Z, " bit information " Y and degree of rarefication K are constituted into NL× 1 " bit sequence " L；The NLMeet NL
=N+2M+NK；
The NKIt indicates required number of bits when degree of rarefication K binary representations, is set according to engineering experience, usually meet NK
< log2N；
D) spread processing is carried out to " bit sequence " L, obtained " spread data sequence "Then by " spread data sequence "
Add N in tail portionzeroN is formed after a zeroD× 1 " sequence spreading " W；
" spread data sequence "According to formulaIt obtains, whereinIndicate frequency expansion sequence, symbol
NumberIndicate Kronecker products；
The NzeroMeet
The NDIt is chosen according to engineering experience, usually chooses ND> > NL；Wherein, c indicates the length of spread spectrum, also known as spreads
Gain.
E) by " sequence spreading " W and ND× 1 " data sequence " D is according to formula
T=a × W+ (1-a) × D
It is weighted processing, to form ND× 1 " sending sequence " T is sent；
The constant a is selected according to engineering experience, meets 0 a≤1 <；
F) receiver receives " sending sequence " T, and recovers " bit sequence " L using technology is despread, and then recovers " position
Set sequence " Z, " bit information " Y and degree of rarefication K；
G) receiver utilizes " position sequence " Z, " bit information " Y and degree of rarefication K to reconstruct CSI.
2. according to the method described in claim 1, it is characterized in that：" position sequence " Z described in step a) is to read in advance
After the CSI sequences H of storage, 0 is labeled as at the position for being zero by element in H, element is that 1 is labeled as at the position of non-zero；Step
A) when carrying out CSI feedback, it is contemplated that the non-zero of CSI elements and the location index information of neutral element.
3. according to the method described in claim 2, it is characterized in that：Step a) includes：
It is N to read pre-stored length, and the channel state information H and size that degree of rarefication is K are that M × N random distributions measure
Matrix Φ.
4. according to the method described in claim 3, it is characterized in that：Sparse grade K described in step a) is by user terminal receiver
Channel estimation obtains；The channel state information H is estimated to obtain by user terminal receiver channel.
5. according to the method described in claim 4, it is characterized in that：The random distribution measurement takes：The random square of Gaussian Profile
The one of which of battle array, bernoulli distribution random matrix, subalpine forests random matrix and unusual sparse projection matrix.
6. according to the method described in claim 5, it is characterized in that：Step b) includes：
By the real part H of CSI sequences HrealWith imaginary part HimagIt is compressed to obtain " bit information " using single-bit compressed sensing principle
The real and imaginary parts of Y, " bit information " Y are denoted as Y respectivelyreal, Yimag；
The single-bit compression, obtains the real part Y of " bit information " YrealWith imaginary part Yimag, obtained according to following formula：
Operation the sign () expressions take symbol manipulation, i.e.,：
7. according to the method described in claim 6, it is characterized in that：Step c) includes：
Real and imaginary parts, the degree of rarefication K of " bit information " Y that " position sequence " Z, length that length is N are M will be obtained, constituted
Length is NL" bit sequence " L, i.e.,：
L=(ZT,(Yreal)T,(Yimag)T, (Kbin)T)T
Wherein, transposition arithmetic operation is sought in subscript " T " expression；
The KbinIt is N to indicate that degree of rarefication K is expressed as the length constituted when binary systemKColumn vector.
8. according to the method described in claim 7, it is characterized in that：Step d) includes：
It is describedIndicate that frequency expansion sequence, spreading length c areWherein, symbolIndicate downward round numbers,
The NzeroMeet
The NDIt is chosen according to engineering experience, usually chooses ND> > NL；The symbolIndicate downward round numbers.
Wherein, c indicates the length of spread spectrum, also known as spreading gain.
9. according to the method described in claim 8, it is characterized in that：Step 0 includes：
Using despreading after technology recovers " bit sequence " L, according to the length of position relationship and each sequence in c),
Recover " position sequence " Z, " bit information " Y and degree of rarefication K from " bit sequence " L, then to " bit information " Y take real part and
Imaginary part；The c) in position relationship be：
L=(ZT,(Yreal)T,(Yimag)T, (Kbin)T)T
Wherein, transposition arithmetic operation is sought in subscript " T " expression；
The KbinIt is N to indicate that degree of rarefication K is expressed as the length constituted when binary systemKColumn vector.
10. according to the method described in claim 9, it is characterized in that：Step g) includes：The CSI reconstruct, refers to by " position
Setting sequence " Z is added restructing algorithm and constitutes constraints, and combines the existing reconstructing method based on single-bit compressed sensing；Institute
" the existing reconstructing method based on single-bit compressed sensing " stated can be used binary system iteration hard -threshold, iteration hard -threshold, from
Adapt to 1-Bit compressed sensings.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810494993.0A CN108737032B (en) | 2018-05-22 | 2018-05-22 | Compressed superposition sequence CSI feedback method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810494993.0A CN108737032B (en) | 2018-05-22 | 2018-05-22 | Compressed superposition sequence CSI feedback method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108737032A true CN108737032A (en) | 2018-11-02 |
| CN108737032B CN108737032B (en) | 2020-11-24 |
Family
ID=63937815
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810494993.0A Active CN108737032B (en) | 2018-05-22 | 2018-05-22 | Compressed superposition sequence CSI feedback method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108737032B (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109818645A (en) * | 2019-02-20 | 2019-05-28 | 西华大学 | CSI feedback method is superimposed with what supported collection assisted based on signal detection |
| CN110166089A (en) * | 2019-05-24 | 2019-08-23 | 西华大学 | Supercomposed coding CSI feedback method based on deep learning |
| CN110289898A (en) * | 2019-07-18 | 2019-09-27 | 中国人民解放军空军预警学院 | A kind of channel feedback method based on the perception of 1 bit compression in extensive mimo system |
| CN112564757A (en) * | 2020-12-03 | 2021-03-26 | 西华大学 | Deep learning 1-bit compression superposition channel state information feedback method |
| CN113726376A (en) * | 2021-08-31 | 2021-11-30 | 西华大学 | 1bit compression superposition CSI feedback method based on feature extraction and mutual-difference fusion |
| CN115001629A (en) * | 2022-04-29 | 2022-09-02 | 清华大学 | Channel quantization feedback method and device, electronic equipment and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101572682A (en) * | 2008-05-04 | 2009-11-04 | 华为技术有限公司 | Method and device for acquiring channel information |
| US20110319068A1 (en) * | 2010-06-29 | 2011-12-29 | Samsung Electronics Co. Ltd. | Method and apparatus for transmitting/receiving csi in cellular communication system supporting carrier aggregation |
| CN105206277A (en) * | 2015-08-17 | 2015-12-30 | 西华大学 | Voice compression method base on monobit compression perception |
| CN106209193A (en) * | 2016-08-19 | 2016-12-07 | 西华大学 | CSI feedback method based on the extensive mimo system of compressed sensing |
| CN106841865A (en) * | 2017-01-22 | 2017-06-13 | 西华大学 | The single-bit sampling of Short Duration Power Quality Disturbance signal and reconstructing method |
-
2018
- 2018-05-22 CN CN201810494993.0A patent/CN108737032B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101572682A (en) * | 2008-05-04 | 2009-11-04 | 华为技术有限公司 | Method and device for acquiring channel information |
| US20110319068A1 (en) * | 2010-06-29 | 2011-12-29 | Samsung Electronics Co. Ltd. | Method and apparatus for transmitting/receiving csi in cellular communication system supporting carrier aggregation |
| CN105206277A (en) * | 2015-08-17 | 2015-12-30 | 西华大学 | Voice compression method base on monobit compression perception |
| CN106209193A (en) * | 2016-08-19 | 2016-12-07 | 西华大学 | CSI feedback method based on the extensive mimo system of compressed sensing |
| CN106841865A (en) * | 2017-01-22 | 2017-06-13 | 西华大学 | The single-bit sampling of Short Duration Power Quality Disturbance signal and reconstructing method |
Non-Patent Citations (1)
| Title |
|---|
| DAOFENG XU等: "Feedback of Downlink Channel State Information Based on Superimposed Coding", 《IEEE COMMUNICATIONS LETTER》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109818645A (en) * | 2019-02-20 | 2019-05-28 | 西华大学 | CSI feedback method is superimposed with what supported collection assisted based on signal detection |
| CN110166089A (en) * | 2019-05-24 | 2019-08-23 | 西华大学 | Supercomposed coding CSI feedback method based on deep learning |
| CN110166089B (en) * | 2019-05-24 | 2021-06-04 | 西华大学 | Superposition coding CSI feedback method based on deep learning |
| CN110289898A (en) * | 2019-07-18 | 2019-09-27 | 中国人民解放军空军预警学院 | A kind of channel feedback method based on the perception of 1 bit compression in extensive mimo system |
| CN110289898B (en) * | 2019-07-18 | 2022-05-06 | 中国人民解放军空军预警学院 | Channel feedback method based on 1-bit compressed sensing in large-scale MIMO system |
| CN112564757A (en) * | 2020-12-03 | 2021-03-26 | 西华大学 | Deep learning 1-bit compression superposition channel state information feedback method |
| CN113726376A (en) * | 2021-08-31 | 2021-11-30 | 西华大学 | 1bit compression superposition CSI feedback method based on feature extraction and mutual-difference fusion |
| CN115001629A (en) * | 2022-04-29 | 2022-09-02 | 清华大学 | Channel quantization feedback method and device, electronic equipment and storage medium |
| CN115001629B (en) * | 2022-04-29 | 2024-03-12 | 清华大学 | Channel quantization feedback method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108737032B (en) | 2020-11-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108737032A (en) | A kind of compression superposition sequence C SI feedback methods | |
| CN110350958B (en) | CSI multi-time rate compression feedback method of large-scale MIMO based on neural network | |
| CN101635612B (en) | Precoding code book constructing method and precoding code book constructing device of multi-input multi-output system | |
| CN101032109B (en) | A method of processing received signals in a multi-input multi-output (MIMO) system | |
| CN109560841A (en) | Extensive mimo system channel estimation methods based on improved distributed compression perception algorithm | |
| CN101854236B (en) | Method and system for feeding back channel information | |
| CN101771442B (en) | Method, system and device for selecting precoding vector and mobile terminal | |
| CN102474314B (en) | Precoding method and device | |
| CN101304300A (en) | Method and device for quantizing multiuser MIMO system channel based on limiting feedback | |
| CN108599820A (en) | The extensive mimo system channel estimation methods of matching pursuit algorithm are sampled based on block structure self-adapting compressing | |
| Yin et al. | Deep CSI compression for massive MIMO: A self-information model-driven neural network | |
| CN109818645B (en) | Superposition CSI feedback method based on signal detection and support set assistance | |
| CN106911443B (en) | Pilot tone optimum design method in compressed sensing based M2M communication system | |
| CN108462519A (en) | A kind of channel state information feedback method, UE and access network entity | |
| CN102868479B (en) | Method, device and system for multi-antenna transmission | |
| CN102340336A (en) | User pairing method of MU-MIMO based on SLNR | |
| CN101931512B (en) | A kind of acquisition methods of channel information and device | |
| CN101826945B (en) | The sending method of channel information and device | |
| TWI669921B (en) | Feedback method for use as a channel information based on deep learning | |
| WO2023011472A1 (en) | Method for feeding back channel state information, method for receiving channel state information, and terminal, base station, and computer-readable storage medium | |
| CN107370533A (en) | It is a kind of to be used to carry out the methods, devices and systems for simulating CSI feedback | |
| CN102571180B (en) | The method and apparatus carrying out user scheduling in mimo wireless communication system | |
| CN108418615B (en) | User scheduling method based on MU-MIMO limited feedback system | |
| CN101834698A (en) | Method and device for transmitting channel information | |
| CN101834706A (en) | Codebook quantification and feedback method and system of channel information |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20181102 Assignee: Suining Feidian Cultural Communication Co.,Ltd. Assignor: XIHUA University Contract record no.: X2023510000027 Denomination of invention: A Compressed Stacked Sequence CSI Feedback Method Granted publication date: 20201124 License type: Common License Record date: 20231129 |
|
| EE01 | Entry into force of recordation of patent licensing contract |