CN110247686A - A kind of Limited Feedback quantization method based on code book rotation - Google Patents
A kind of Limited Feedback quantization method based on code book rotation Download PDFInfo
- Publication number
- CN110247686A CN110247686A CN201910504147.7A CN201910504147A CN110247686A CN 110247686 A CN110247686 A CN 110247686A CN 201910504147 A CN201910504147 A CN 201910504147A CN 110247686 A CN110247686 A CN 110247686A
- Authority
- CN
- China
- Prior art keywords
- code book
- current time
- code word
- vector
- rotation
- 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
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
- H04B7/0417—Feedback 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
- H04B7/0636—Feedback format
- H04B7/0639—Using selective indices, e.g. of a codebook, e.g. pre-distortion matrix index [PMI] or for beam selection
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Radio Transmission System (AREA)
Abstract
The invention discloses a kind of Limited Feedback quantization methods based on code book rotation of wireless communication technology field, it is intended to solve the problems, such as in extensive mimo system in the prior art number of bits of feedback it is larger and caused by feedback overhead it is excessive.Described method includes following steps: obtaining the last moment optimal codes that user terminal is chosen；On the basis of last moment optimal codes, each d code word of its left and right side is chosen, forms current time virtual code book；User terminal chooses the optimal codes in current time virtual code book, is fed back to base station end.
Description
Technical field
The present invention relates to a kind of Limited Feedback quantization methods based on code book rotation, belong to wireless communication technology field.
Background technique
In MIMO (Multiple-Input Multiple-Output, multi-emitting antenna and more receiving antennas) system,
More data can be transmitted by spatial reuse.In order to make full use of given mimo channel, base station end must be generated suitably
Beam-former.CSI (Channel State Information, channel state information) is needed to construct wave beam appropriate
Shaper.The feedback of CSI to base station end is the important component of mimo system, because it is most important to precoding.So
And, it is difficult to perfect CSI is provided to base station end, and user terminal is sent out usually using the CSI that low-rate feedback link will quantify
It is sent to base station end.Therefore, base station end and user terminal should be provided simultaneously with the code book for quantifying required channel state information.This germline
Collectively referred to as MIMO Systems with Limited Feedback.
As the quantity of antenna increases, feedback information volume is consequently increased.Therefore, anti-under extensive mimo system
Feedback scheme becomes the hot issue of current research.Increased by the performance of the design improvement system to code book scheme in antenna number
When by feedback overhead limitation in a certain range.
Jihoon Sung and Wonjin Sung are directed in extensive mimo system and ask with the increased feedback overhead of antenna number
Topic, proposes a kind of scheme that channel training is combined with wave beam forming.The program is based on LTE (Long Term
Evolution, long term evolution) iterative feedback of code book combination and CSI in system carries out, i.e., to the small ruler in each LTE
The PMI (Payload Missing Indication, payload lose instruction) of very little code book is fed back, and base station end is accumulative each
PMI information to formed a large scale beam vector, to reduce the feedback overhead of system.
Li Jingyu and Su Xin propose a kind of new double codebook design methods based on Kronecker theory.This method
The case where using uniform rectangular array for base station, code book is made of two kinds of code books, is by the reflection long period characteristic of channel respectively
Code book and reflection short-term channel behaviors code book.Wherein, it is introduced in the code book and Release 10 of reflection short cycle characteristic
It is consistent based on the short cycle code book in DFT dicode sheet, and indicate that the code book of the long period characteristic of channel is then different, it be by
Reflect the code book of the horizontal direction long period characteristic of channel and the code book of the reflection vertical direction long period characteristic of channel, passes through
What Kronecker product generated.This method achieves preferable in the case where not increasing system complexity and feedback overhead
System performance, the design for subsequent code book provide theoretical foundation.
In conclusion Limited Feedback technology can obtain channel state information for traditional feedback technique
While greatly reduce the system bandwidth of occupancy, however to Limited Feedback technology is applied to multi-user Massive MIMO
Still there are several critical issues to need to study in system.Due to the limitation of system bandwidth, reducing feedback overhead becomes the head to be solved
Want problem.
The CSI information content of extensive MIMO communication system is because the increase of antenna number and number of users is larger, and actual wireless
The feedback channel capacity of communication system is limited, and therefore, how to be realized and is reduced to the greatest extent under limited feedback channel capacity instead
Bit number is presented, so that reducing overhead is a urgent problem to be solved.
Summary of the invention
It is an object of the invention to overcome deficiency in the prior art, provide a kind of based on the limited anti-of code book rotation
Quantization method is presented, is included the following steps:
Obtain the last moment optimal codes that user terminal is chosen；
On the basis of last moment optimal codes, each d code word of its left and right side is chosen, composition current time is virtual
Code book；
User terminal chooses the optimal codes in current time virtual code book, is fed back to base station end.
Further, the last moment optimal codes that user terminal is chosen are obtained, comprising:
The direction vector generated when extracting user terminal last moment progress Limited Feedback quantization inceptive, constitutes code book collection
It closes, the direction vector is as code word；
Compare the inner product in current time channel vector and codebook set between direction vector；
The smallest direction vector of inner product is obtained, using its corresponding code word as last moment optimal codes.
Further, the direction vector shares 2BIt is a, it is uniformly generated according to direction, B is system feedback bit number.
Further, on the basis of last moment optimal codes, each d code word of its left and right side is chosen, composition is current
Moment virtual code book, comprising:
Obtain the corresponding direction vector of d-th of code word and right side (d-1) a code word on the left of last moment optimal codes
The inner product of corresponding direction vector, as maximum channel difference；
Compare the inner product between the subsequent time channel vector at the current time and the current time channel vector；
If the inner product between the subsequent time channel vector at the current time and the current time channel vector is small
In maximum channel difference, codebook set is rotated forward to 2d code word；
Preceding 2d code word in codebook set after choosing rotation constitutes current time virtual code book.
Further, further includes: if the subsequent time channel vector at the current time and the current time channel
Inner product between vector is not less than maximum channel difference, and codebook set is fed back to base station end.
Further, the code word in virtual code book is according to clockwise direction sequencing numbers, the code that do not choose in codebook set
Word is connected according to clockwise direction and is numbered, and pseudo code epicycle goes to the initial position of codebook set.
Further, the code word in codebook set uniformly generates, and according to clockwise direction sequencing numbers.
Further, user terminal chooses the optimal codes in current time virtual code book, comprising:
Compare between the subsequent time channel vector direction vector corresponding with code word in virtual code book at the current time
Inner product；
The smallest direction vector of inner product is obtained, using its corresponding code word as the optimum code in current time virtual code book
Word.
Further, the value range of d is [1,2B-1], B is system feedback bit number.
Compared with prior art, advantageous effects of the invention: by reduce quantization code book in number of codewords come
Number of bits of feedback is reduced to reduce feedback overhead, improves system performance, while reducing system feedback process and calculating complexity
Degree.
Detailed description of the invention
Fig. 1 is a kind of code book rotation process schematic provided in an embodiment of the present invention；
Fig. 2 is a kind of Limited Feedback quantization method flow chart based on code book rotation provided in an embodiment of the present invention；
Fig. 3 is algorithm provided in an embodiment of the present invention and traditional algorithm channel capacity contrast schematic diagram；
Fig. 4 is the different user number channel capacity contrast schematic diagram using algorithm provided in an embodiment of the present invention；
System channel capacity comparison is shown when Fig. 5 is the different number of bits of feedback using algorithm provided in an embodiment of the present invention
It is intended to；
Fig. 6 is a kind of different execution algorithm number of users channel capacity contrast schematic diagrams provided in an embodiment of the present invention；
Computation complexity comparison signal when Fig. 7 is different execution inventive algorithm numbers of users provided in an embodiment of the present invention
Figure；
Feedback bits contrast schematic diagram when Fig. 8 is the different rotation code word numbers of the embodiment of the present invention；
Computation complexity contrast schematic diagram when Fig. 9 is the different rotation code word numbers of the embodiment of the present invention.
Specific embodiment
The invention will be further described below in conjunction with the accompanying drawings.Following embodiment is only used for clearly illustrating this hair
Bright technical solution, and not intended to limit the protection scope of the present invention.
Embodiment one:
Consider downlink system in an extensive mimo system, system is single base station model, base station end antenna number for M (
In practice 100) this antenna amount is typically larger than, and the number of users of random distribution is K in cell, then i-th user (i=1,
2 ..., K) the signal y that receivesi, as shown in formula (1):
In formula, hiFor the channel vector of i-th of user, it is assumed that channel is that block declines and the decline between block and block is
Independent, channel obeys Ruili decline and is flat channel；gi、gjRespectively the precoding of i-th of user and jth user to
Amount, andWherein H is the conjugate transposition of channel, | | | |FIndicate the norm of vector；xiFor to i-th user's
Send information vector；niFor the additive white Gaussian noise that mean value is 0, variance is 1.
The quantization code book that channel information amount is converted to B bit is fed back to base station end by each user terminal, and the B is to be
System number of bits of feedback；Each quantization code book includes 2BA unit direction vector, it may be assumed that C=v1,v2,v3,...,v2B, in formula, C is
Codebook set, v1For first direction vector in codebook set,For the 2nd in codebook setBA direction vector；User terminal
2BIt is chosen and the immediate direction vector v of channel vector in a vectorclostAs optimal codes, then by optimal codes
Code book label feeds back to base station end, and base station end can just obtain channel state information.It is code word that feedback optimal codes, which are usually fed back,
Code book label in the codebook, code book label is usually by interior product representation, wherein the code book label R of i-th of useri, such as formula
(2) shown in:
In formula, max indicates maximum value, and min indicates minimum value, and arg min (f (s)) is indicated so that f (s) obtains maximum value
Corresponding variable point s, ∠ indicates the angle of two vectors.
Then, by code book label RiBase station end is fed back to, base station end selects the code word of corresponding label to believe as channel status
Breath, base station end carry out precoding according to the channel state information that quantization obtains.
The signal-to-noise ratio SINR of i-th of useri, as shown in formula (3):
In formula,The precoding vector that respectively i-th and j-th of user are obtained by quantized channel vector,For
Channel vector after i-th of user's quantization, meets E | x2|≤P, wherein E is to seek desired value, and P is the transmitting function of base station
Rate, x are the transmission information vector to user.
Power system capacity C ', as shown in formula (4):
For solve the problems, such as in extensive mimo system number of bits of feedback it is larger and caused by feedback overhead it is excessive, this hair
It is bright to propose the Limited Feedback quantization method based on code book rotation.
As shown in Figure 1, being a kind of code book rotation process schematic provided in an embodiment of the present invention, this method is limited anti-
In the quantizing process of feedback, code book takes rotation mechanism, centered on the optimal codes selected by user's last moment, with its front and back d
Code word wheel switchs to the code book at current time, is quantified to obtain the optimal codes at current time in current code book, passes through reduction
Number of codewords is in quantizing process to reduce number of bits of feedback.
As shown in Fig. 2, being a kind of Limited Feedback quantization method process based on code book rotation provided in an embodiment of the present invention
Figure, the method includes the steps:
Step 1, initialization, uniformly generates 2 according to directionBA direction vector v, 2BA direction vector constitutes codebook setDirection vector in codebook set is also referred to as code word；
Code word in codebook set C uniformly generates, and according to successively sorting clockwise, is generating 2d code word
Virtual code book when, code word is still according to clockwise direction sequencing numbers in virtual code book, and remainder codewords are according to clockwise
Continue to number, and pseudo code epicycle goes to the initial position of codebook set C；
The size of d can be changed as required in virtual code book, for the definite value that could dictate that, work as desirable system
Capacity increases, while when number of bits of feedback increase, can suitably increase the value of d；When in desirable system number of bits of feedback reduce,
When simultaneity factor capacity increases, the value of d can be suitably reduced, the value range of d is [1,2B-1]。
Step 2, user KiChannel state information when being quantified at the T=t moment, user KiIn code book collection when quantization
It closesIn, compare channel vector hiInner product Q between each code wordj, as shown in formula (5):
User terminal has channel information, does not need to obtain channel information by transmission, base station end is obtained by client feeds back
To channel state information；
Choose the smallest direction vector v of inner productjmin, as the optimal codes of last moment, with the last moment
Optimal codes on the basis of；
Simultaneously by optimal codes vjminD-th of right side direction vector vmin-dWith optimal codes vjminThe d-1, right side direction
Vector vmin+d-1Inner product be set as maximum channel differenceAs shown in formula (6):
Step 3, by user the T=t+1 moment channel vector ht+1Quantified, compares and obtain channel vector ht+1With
Channel vector h of the user at the T=t momenttBetween inner product Qh, as shown in formula (7):
Qh=sin2(∠(ht+1,ht))(7)
As inner product QhValue be less than maximum channel difference QhmaxWhen, execute step 4；Otherwise, then actuating code epicycle is abandoned
Turn, is still quantified according to original code book；This step can effectively prevent two users' channel vector still to use this calculation when differing farther out
Method and the appearance for causing bigger quantization error.
Step 4, by codebook set2d code word is rotated forward, by vmin-dWheel goes to first code word
Position, vmin+d-1Wheel switchs to the 2d code word of current code book, it follows that the code book after rotation
Step 5, in quantizing process, the code book of actual quantization is the code book of preceding 2d code word composition, in this algorithm
Referred to as virtual code book.That is, 2d code word constitutes current time virtual code book C before choosing in codebook sett+1=vmin-
d,...,vmin-1,vmin,vmin+1,…,vmin+d-1。
Step 6 re-flags current time virtual code book are as follows:
Step 7, in current time virtual code book Ct+1In carry out the selections of optimal codes again, compare channel vector ht+1
With codebook set Ct+1Inner product between middle direction vector obtains the smallest direction vector v of inner productjmin, and minimum inner product Qjmin
Corresponding optimal codes vjmin。
Step 8 returns to step 3.
Embodiment two:
Number of bits of feedback and computation complexity are analyzed.
As user KiIt is poor to be less than maximum channel in the channel vector difference of T=t+1 moment channel vector and last moment
When, choosing its preceding 2d code word after rotating is virtual code book, and the code word number in code book is by 2 in original code bookBA reduction
For in virtual code bookWherein:
Bt+1=log2(2d) (8)
Therefore, compared with traditional feedback method, the number of bits of feedback Δ B of reduction, as shown in formula (9):
Δ B=B-log2(2d) (9)
K user of synchronization carries out Limited Feedback simultaneously in extensive mimo system, and each user is at current time
Number of bits of feedback can be reduced by performing effectively algorithm, but to be unsatisfactory for code book there is also current time and be rotated condition, can not be executed
The user of inventive algorithm；At this point, the number of bits of feedback total amount B of systemsum, as shown in formula (10):
Bsum=log2(2d)·X+B·(K-X) (10)
In formula, X indicates that current time can effectively use the number of users of code book round robin algorithm, it is known that 1≤X≤K, then
System feedback bit number total amount BsumVariation range, as shown in formula (11):
log2(2d)·K≤Bsum≤B·K (11)
When the number of bits of feedback variation in system, the computation complexity during Limited Feedback also changes therewith.It uses
When inventive algorithm, system-computed complexity are as follows:
By formula (9) it is found that due to the error occurred in quantizing process in system, each user under Limited Feedback with
Base station end has the power system capacity difference DELTA C ' in the case of complete channel status information, as shown in formula (13):
The power system capacity difference that user uses inventive algorithm at current time can be obtained:
Formula (10) are substituted into formula (13), it is known that as the user for using inventive algorithm in system at current time
When for X, the total system capacity difference DELTA C ' of systemsum, as shown in formula (15):
When rotating mono- timing of area size d in algorithm, power system capacity difference executes inventive algorithm user with current time
Number increases and increases, i.e., system capacity loss increases, because when executing the increase of inventive algorithm number of users, the feedback ratio of system
Special number reduces, and channel state information quantization error increases, therefore power system capacity reduces, so as to cause the increase of power system capacity difference.
When current time execution inventive algorithm number of users X mono- periodically, rotates area in power system capacity difference arithmetic in algorithm
Domain size d increases and reduces, and due to rotating the increase in region, makes to quantify code word number in code book and increases, miss to reduce quantization
Difference increases power system capacity, eventually leads to the reduction of power system capacity error.
In order to observe the performance of algorithm of the invention, emulation experiment has been carried out with Matlab software, has been constructed extensive
Limited feedback systems under MIMO, and used in quantizing process based on code book rotation grouping limited feedback method, and
It the use of power system capacity is to measure the index of system performance；The parameter being arranged in system be in single base station, in coverage area with
Machine is distributed 4 users, and user is close to stationary state, and mobility is smaller；Base station range is 40m, and specific simulation parameter is such as
Shown in table 1:
Table 1: simulation parameter
As shown in figure 3, being algorithm provided in an embodiment of the present invention and traditional algorithm channel capacity contrast schematic diagram, in day
Line number is 100, and initial feedback bit number B is that 4, d value is 2, and number of users K is emulated in the case where being 4, is rotated using based on code book
Limited Feedback quantization method power system capacity than original quantisation method system capacity reduce, due to reducing number of bits of feedback, code
Number of codewords is reduced in this, and the quantization error when channel capacity chooses optimal codes increases, and power system capacity is caused to reduce.
As shown in figure 4, being the different user number channel capacity comparison signal using algorithm provided in an embodiment of the present invention
Figure, in order to study influence of the number of users to code book grouping quantization method, simulating rotation region respectively is 4, performs effectively this
When invention algorithm user is constant, total number of users K is situation of change of 2,4,8 power system capacities with signal-to-noise ratio, it can be seen that when
When number of users is gradually incremented by, power system capacity gradually increases, and is total power system capacity, user due to calculating in algorithm
Number increases, and total system capacity increases.
As shown in figure 5, system channel capacity when being the different number of bits of feedback using algorithm provided in an embodiment of the present invention
Contrast schematic diagram, it can be seen that when number of bits of feedback is respectively set to 2,4,8 bits, the power system capacity of system is with feedback
Bit number increases and increases, and in systems when number of bits of feedback increases, does not execute inventive algorithm due to existing in system
User, therefore code word number increases in the quantization code book of such user, makes the reduction of system total quantization error, to improve system
Capacity.
As shown in fig. 6, being that a kind of different algorithm number of users channel capacities that execute provided in an embodiment of the present invention compare signal
Figure, it can be seen that it is when increasing that synchronization, which executes the user of inventive algorithm, in system, and the number of bits of feedback of system reduces,
And when rotation code word number increases in the codebook, system feedback bit number is reduced, total number of users in the system being arranged in emulation
It is 4, therefore when execution inventive algorithm number of users reaches 4, i.e., all users are all made of calculation of the present invention in current time system
Method, number of bits of feedback reaches minimum in system at this time.For user when executing inventive algorithm, what is obtained after rotation is virtual
Quantified in code book, code word number is reduced in code book, therefore the number of bits of feedback drop of the channel state information obtained after quantifying
It is low.
As shown in fig. 7, computation complexity pair when being different execution inventive algorithm numbers of users provided in an embodiment of the present invention
Than schematic diagram, when rotating code word number in setting system is 2, the computation complexity of observing system is used with inventive algorithm is executed
The variation of amount, it can be seen that computation complexity drop when executing the increase of inventive algorithm number of users in algorithm, in system
It is low, and when number of users increases to 1 by 0, i.e., the initial calculating for system reduction when execution inventive algorithm user occur is complicated
Spend it is more, when user gradually increases, close in system when maximum number of user, system Limited Feedback computation complexity reduction tends to
Gently.And when antenna number gradually increases, the computation complexity of system is gradually increased, and thus will also realize that extensive MIMO system
Increase system Limited Feedback computation complexity with antenna number in system to increase.And when whole users execute inventive algorithm,
It is close when system-computed complexity is more less than execution inventive algorithm user.Mainly due to number of bits of feedback in system to calculating
The influence of complexity is exponential type and antenna number is linear.
As shown in figure 8, feedback bits contrast schematic diagram when being the different rotation code word numbers of the embodiment of the present invention, to table
Show the influence in different rotation code word numbers to system feedback bit number；As can be seen that the system feedback when rotating code word number and increasing
Bit number increases, and as rotation code word number increases, the increased rate of system feedback bit number is reduced, and is executing sheet
When invention algorithm number increases, system feedback bit number is reduced, and when code word number is constant in code book original in system, takes turns transcoding
It is that code word number in virtual code book increases that word, which increases, and user passes through the channel state information institute that quantifies in virtual code book
Bit number is occupied to increase.When rotating the increase of code word number to a certain extent, rotation region is overlapped with original code book is then calculated with tradition
Method is identical, and number of bits of feedback is no longer reduced.
As shown in figure 9, computation complexity contrast schematic diagram when being the different rotation code word number of the embodiment of the present invention, when being
When code word number is 32 in original code book in system, it can be seen that when rotating code word number and increasing, computation complexity increases in system,
When rotation code word number reaches 16, the computation complexity of system is than rotating three times when code word is 8, it is known that rotation code word number pair
The computation complexity of system Limited Feedback is affected.And when rotating the increase of code word number, the growth of system-computed complexity
Rate increases.Since rotation code word number increases, the virtual code book for executing inventive algorithm user increases, then includes in code book
Code word number increases, therefore the computation complexity of system Limited Feedback increases.When antenna number increases in system, system-computed is complicated
Degree increases
In conclusion the embodiment of the present invention considers under extensive mimo system, close to static or mobile in single base station
Property lesser multiple users, and in the case of considering inter-user interference, obtained after at a time subscriber channel status information quantization
Optimal codes are numbered in code book, are compared to subsequent time channel vector and last moment channel vector, when being less than present invention provide that
Minimum channel difference when, the total 2d code word wheels of each d in left and right centered on last moment optimal codes go to current time and measure
The beginning for changing code book, is referred to as current time virtual code book, current time user terminal is quantified in virtual code book.This
The number of bits of feedback in quantizing process can be effectively reduced in invention, while reducing system feedback process computation complexity.But it is same
When, due to the reduction of number of bits of feedback, number of codewords is reduced, and quantization error increase reduces power system capacity.It can be as needed
Weighed, obtains optimal system performance.
The above is only a preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, without departing from the technical principles of the invention, several improvement and deformations can also be made, these improve and become
Shape also should be regarded as protection scope of the present invention.
Claims (9)
1. a kind of Limited Feedback quantization method based on code book rotation, characterized in that include the following steps:
Obtain the last moment optimal codes that user terminal is chosen；
On the basis of last moment optimal codes, each d code word of its left and right side is chosen, forms current time virtual code book；
User terminal chooses the optimal codes in current time virtual code book, is fed back to base station end.
2. the Limited Feedback quantization method according to claim 1 based on code book rotation, characterized in that obtain user terminal choosing
The last moment optimal codes taken, comprising:
The direction vector generated when extracting user terminal last moment progress Limited Feedback quantization inceptive, constitutes codebook set, institute
Direction vector is stated as code word；
Compare the inner product in current time channel vector and codebook set between direction vector；
The smallest direction vector of inner product is obtained, using its corresponding code word as last moment optimal codes.
3. the Limited Feedback quantization method according to claim 2 based on code book rotation, characterized in that the direction vector
Share 2BIt is a, it is uniformly generated according to direction, B is system feedback bit number.
4. the Limited Feedback quantization method according to claim 2 based on code book rotation, characterized in that most with last moment
On the basis of good code word, each d code word of its left and right side is chosen, forms current time virtual code book, comprising:
It is corresponding with right side (d-1) a code word to obtain the corresponding direction vector of d-th of code word on the left of last moment optimal codes
The inner product of direction vector, as maximum channel difference；
Compare the inner product between the subsequent time channel vector at the current time and the current time channel vector；
If the inner product between the subsequent time channel vector at the current time and the current time channel vector is less than most
Codebook set is rotated forward 2d code word by big channel differences；
Preceding 2d code word in codebook set after choosing rotation constitutes current time virtual code book.
5. the Limited Feedback quantization method according to claim 4 based on code book rotation, characterized in that further include: if
Inner product between the subsequent time channel vector at the current time and the current time channel vector is not less than maximum channel
Codebook set is fed back to base station end by difference.
6. the Limited Feedback quantization method according to claim 4 based on code book rotation, characterized in that in virtual code book
Code word is according to clockwise direction sequencing numbers, and the code word that do not choose in codebook set is connected according to clockwise direction numbers, virtually
Code book wheel goes to the initial position of codebook set.
7. the Limited Feedback quantization method based on code book rotation according to any one of claim 2 to 6, characterized in that
Code word in codebook set uniformly generates, and according to clockwise direction sequencing numbers.
8. the Limited Feedback quantization method according to any one of claim 1 to 6 based on code book rotation, characterized in that
User terminal chooses the optimal codes in current time virtual code book, comprising:
Compare interior between the subsequent time channel vector direction vector corresponding with code word in virtual code book at the current time
Product；
The smallest direction vector of inner product is obtained, using its corresponding code word as the optimal codes in current time virtual code book.
9. the Limited Feedback quantization method according to any one of claim 1 to 6 based on code book rotation, characterized in that d
Value range be [1,2B-1], B is system feedback bit number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910504147.7A CN110247686B (en) | 2019-06-12 | 2019-06-12 | Codebook rotation-based finite feedback quantization method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910504147.7A CN110247686B (en) | 2019-06-12 | 2019-06-12 | Codebook rotation-based finite feedback quantization method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110247686A true CN110247686A (en) | 2019-09-17 |
| CN110247686B CN110247686B (en) | 2021-04-09 |
Family
ID=67886641
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910504147.7A Active CN110247686B (en) | 2019-06-12 | 2019-06-12 | Codebook rotation-based finite feedback quantization method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110247686B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115053465A (en) * | 2020-02-07 | 2022-09-13 | 华为技术有限公司 | Information transmission method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103731244A (en) * | 2014-01-03 | 2014-04-16 | 东南大学 | Method for designing double-codebook difference in LTE-A terminal feedback system |
| CN104468054A (en) * | 2014-11-10 | 2015-03-25 | 上海交通大学 | Limited feedback precoding method based on short wave multi-antenna system |
| CN104734813A (en) * | 2015-03-24 | 2015-06-24 | 江苏省电力公司信息通信分公司 | Upstream pre-coding method based on rotary codebook for fixed electric power communication terminal |
| US20160365910A1 (en) * | 2008-09-02 | 2016-12-15 | Intel Corporation | Mimo beamforming method and method of constructing a differential codebook for a wireless network |
| CN108880642A (en) * | 2018-09-06 | 2018-11-23 | 华北电力大学（保定） | Channel vector Limited Feedback quantization method based on bearing mark |
| CN109327918A (en) * | 2018-11-19 | 2019-02-12 | 东南大学 | A kind of extensive MIMO down channel method for reconstructing of the FDD of low overhead |
-
2019
- 2019-06-12 CN CN201910504147.7A patent/CN110247686B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160365910A1 (en) * | 2008-09-02 | 2016-12-15 | Intel Corporation | Mimo beamforming method and method of constructing a differential codebook for a wireless network |
| CN103731244A (en) * | 2014-01-03 | 2014-04-16 | 东南大学 | Method for designing double-codebook difference in LTE-A terminal feedback system |
| CN104468054A (en) * | 2014-11-10 | 2015-03-25 | 上海交通大学 | Limited feedback precoding method based on short wave multi-antenna system |
| CN104734813A (en) * | 2015-03-24 | 2015-06-24 | 江苏省电力公司信息通信分公司 | Upstream pre-coding method based on rotary codebook for fixed electric power communication terminal |
| CN108880642A (en) * | 2018-09-06 | 2018-11-23 | 华北电力大学（保定） | Channel vector Limited Feedback quantization method based on bearing mark |
| CN109327918A (en) * | 2018-11-19 | 2019-02-12 | 东南大学 | A kind of extensive MIMO down channel method for reconstructing of the FDD of low overhead |
Non-Patent Citations (2)
| Title |
|---|
| JIHOON SUNG 等: "Large-scale MIMO beamforming using successive channel state estimation and codebook extension", 《2013 IEEE 24TH ANNUAL INTERNATIONAL SYMPOSIUM ON PERSONAL, INDOOR, AND MOBILE RADIO COMMUNICATIONS (PIMRC)》 * |
| NAYAN FANG等: "Effective user selection algorithm for quantized precoding in massive MIMO", 《2013 8TH INTERNATIONAL CONFERENCE ON COMMUNICATIONS AND NETWORKING IN CHINA (CHINACOM)》 * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115053465A (en) * | 2020-02-07 | 2022-09-13 | 华为技术有限公司 | Information transmission method and device |
| CN115053465B (en) * | 2020-02-07 | 2024-01-02 | 华为技术有限公司 | Information transmission method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110247686B (en) | 2021-04-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8284855B2 (en) | Multi-user downlink linear MIMO precoding system | |
| CN106060950B (en) | It is a kind of that data transmission method in the cellular downlink channel of alignment is interfered based on chance | |
| US9124322B2 (en) | Method and apparatus for constructing codebook, and method, apparatus and system for precoding | |
| CN104321983B (en) | Method, user equipment and the base station of transmitting pre-encoding matrix | |
| CN106130614B (en) | A kind of method, user equipment and the base station of low overhead feedback | |
| CN103152140A (en) | Limiting feedback method of three-dimensional multi-user MIMO (multiple input multiple output) based on direct-product code books | |
| CN101369871A (en) | Multi-user MIMO system precoding method and apparatus based on limited feedback | |
| CN107888264A (en) | The feedback method and device of channel information | |
| CN104144036B (en) | PU2RC-based MU-MIMO user selection method | |
| CN107809275B (en) | Finite feedback hybrid precoding method based on millimeter wave MIMO system | |
| CN105959046B (en) | A kind of beam form-endowing method, signal transmitting equipment and signal receiver | |
| CN102111352A (en) | Method, device and system for information feedback in multi-point coordinated joint transmission network | |
| CN105721029B (en) | Multi-user dispatching method based on this Limited Feedback of dicode in 3D MU-MIMO FDD system | |
| JP5706514B2 (en) | Method and system for correlation matrix feedback for antenna arrays | |
| Samanta et al. | Codebook adaptation for quantized MIMO beamforming systems | |
| CN106028461B (en) | Relay cooperative distributed multi-user dispatching method | |
| CN110247686A (en) | A kind of Limited Feedback quantization method based on code book rotation | |
| Wang et al. | Joint user association and hybrid beamforming designs for Cell-Free mmWave MIMO communications | |
| WO2023246618A1 (en) | Channel matrix processing method and apparatus, and terminal and network side device | |
| CN109361438A (en) | A kind of letter leakage noise ratio method for precoding of Filled function matching leakage weighting | |
| CN102710308B (en) | Codebook designing method for reducing feedback information cost in pre-coding MIMO (Multiple Input Multiple Output) system | |
| CN105208572B (en) | A kind of beam-forming method and base station | |
| CN102082635B (en) | Codebook storage and information transceiving method and device using codebook | |
| CN101834706B (en) | The codebook quantification feedback method of a kind of channel information and system | |
| CN105262524B (en) | The multi-user dispatching method of mimo system based on feedback compensation |
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 |