CN107135023A - Three-dimensional training code book design method and beam alignment for millimeter-wave communication system - Google Patents

Three-dimensional training code book design method and beam alignment for millimeter-wave communication system Download PDF

Info

Publication number
CN107135023A
CN107135023A CN201710293707.XA CN201710293707A CN107135023A CN 107135023 A CN107135023 A CN 107135023A CN 201710293707 A CN201710293707 A CN 201710293707A CN 107135023 A CN107135023 A CN 107135023A
Authority
CN
China
Prior art keywords
code book
dimensional training
training
dimensional
training code
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
Application number
CN201710293707.XA
Other languages
Chinese (zh)
Other versions
CN107135023B (en
Inventor
黄永明
苏敏华
章建军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Southeast University
Original Assignee
Southeast University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Southeast University filed Critical Southeast University
Priority to CN201710293707.XA priority Critical patent/CN107135023B/en
Publication of CN107135023A publication Critical patent/CN107135023A/en
Application granted granted Critical
Publication of CN107135023B publication Critical patent/CN107135023B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04BTRANSMISSION
    • H04B7/00Radio transmission systems, i.e. using radiation field
    • H04B7/02Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413MIMO systems
    • H04B7/0456Selection of precoding matrices or codebooks, e.g. using matrices antenna weighting
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04BTRANSMISSION
    • H04B7/00Radio transmission systems, i.e. using radiation field
    • H04B7/02Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/06Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613Diversity 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
    • H04B7/0615Diversity 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
    • H04B7/0617Diversity 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 for beam forming
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04BTRANSMISSION
    • H04B7/00Radio transmission systems, i.e. using radiation field
    • H04B7/02Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/06Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613Diversity 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
    • H04B7/0615Diversity 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
    • H04B7/0619Diversity 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
    • H04B7/0621Feedback content
    • H04B7/0632Channel quality parameters, e.g. channel quality indicator [CQI]

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Radio Transmission System (AREA)
  • Radar Systems Or Details Thereof (AREA)

Abstract

The invention discloses a kind of three-dimensional training code book design method and beam alignment for millimeter-wave communication system, wherein three-dimensional training code book design method comprises the following steps:1st, according to resolution ratio and beam spaceScope to set up depth be the three-dimensional training code book tree construction that S, degree are N;2nd, willIt is divided into N number of rectangular areaIt is designated as set There is beam setCorrespond to therewith;The root node of tree construction isWithCombination;3rd, b-th of node C of s layers is determineds,b, Cs,bForWith beam setCombination;Each node in 2 to S layers is determined successively4th, the beamforming vectors that each node correspondence beam set is included are solvedObtain three-dimensional training code book, wherein q=1 ..., N.The three-dimensional training code book of this method generation can be used for realizing high-precision wave beam alignment and channel estimation, and can significantly reduce the training expense of system.

Description

Three-dimensional training code book design method and wave beam alignment for millimeter-wave communication system Method
Technical field
The invention belongs to wireless communication technology field, and in particular to millimeter wave of the sending and receiving end antenna in uniform surface battle array arrangement The three-dimensional training code book design method of one kind of communication system.
Background technology
With continuing to develop for wireless communication technology, high-speed data service and ubiquitous access demand just show A kind of explosive growth.5G mobile communication technologies of future generation, to the demand of capacity, energy consumption and bandwidth by more and more higher.Work In the millimetre-wave attenuator technology of 30-300GHz relative free frequency ranges, because of the bandwidth of a large amount of unauthorizeds included in its working frequency range It is considered as one of key technology of WLAN of future generation and mobile communication, because directly spread spectrum bandwidth is for lifting It is both simple and effective for power system capacity.There are some researches show millimetre-wave attenuator can realize 10Gbps hardware circuit solution.
However, compared to traditional microwave frequency range, millimeter wave transmission is faced with bigger path loss, thus communication distance with Coverage is extremely limited.Fortunately the wavelength of millimeter-wave signal is extremely short, can encapsulate a large amount of antennas with reduced size, Jin Er great Scale array antenna is combined with numerical model analysis beam forming technique, and its array gain provided and space division multiplexing gain can overcome the disadvantages that The part decay of system, improves system transfer rate and transmission quality.In addition, it is preferably to obtain array gain, logical Letter start when to transmitting-receiving wave beam carry out wave beam alignment it is also necessary, high-precision wave beam be aligned in set up reliable millimeter wave lead to Key effect is played in terms of transmission data and expansion cell communication coverage needed for letter link, acquisition.Pass through estimation Including relevant parameters such as angle of arrival (AoA), departure angle (AoD) and path gains, accurate wave beam alignment can also be used to improve The channel estimating performance of millimeter-wave systems.
In actual millimeter-wave communication system, to realize that accurate wave beam alignment has certain difficulty.First, millimeter wave The high frequency band of section means that channel may quickly change within a short period of time, and wave beam alignment is needed in extremely short channel coherency time It is interior to complete, therefore the beam search algorithm of limit herein and do not apply to.Second, to make full use of the array of large-scale antenna array to increase Benefit, training wave beam should be sufficiently narrow, and this can undoubtedly increase the complexity of wave beam alignment, therefore propose efficient wave beam codebook design Method and beam search algorithm are necessary.To reduce the training expense of wave beam alignment, a kind of effective method is to use base The tree search algorithm of code book is trained in layering multiresolution.Order training method code book is generally made up of the subcode book of different stage, In high-level, the training wave beam of a small amount of low resolution is included in subcode book, predetermined angular range is covered;On low order layer, son The training wave beam included in code book increases, and resolution ratio there has also been certain raising.
Although hierarchical search can significantly reduce the training expense of system, its performance, which is depended greatly on, to be made Level trains code book.Research for level code book design method is a lot, but these researchs are concentrated mainly on uniformly Linear array (ULA) structure, and the research to uniform planar array (UPA) structure is considerably less.To realize the three-dimensional of variable precision Wave cover, and bigger beam forming gain is obtained, the present invention proposes a kind of suitable for the three of millimeter-wave communication system Dimension training code book design method, the design problem for training wave beam is described as the optimization problem of beam pattern by this method, and is proposed A kind of effective iterative algorithm solves this optimization problem.In the present invention, the overshoot and main secondary lobe of training beam pattern Ripple suitably constrained, therefore each training wave beam has the amplitude response of relatively flat and ideal transition Frequency band.The three-dimensional training code book that the present invention is designed, all the time can be in millimeter-wave communication system under Low SNR Realize high-precision wave beam alignment and channel estimation.
The content of the invention
Goal of the invention:For problems of the prior art, it is used for millimeter-wave communication system the invention discloses one kind Three-dimensional training code book design method, the three-dimensional training code book of this method generation can be used for realizing high-precision wave beam alignment and believes Road is estimated, and can significantly reduce the training expense of system.
Technical scheme:One aspect of the present invention discloses a kind of three-dimensional training Codebook Design side for millimeter-wave communication system Method, comprises the following steps:
(1) according to resolution ratio rs and beam spaceScope to set up depth be that the three-dimensional training code book tree that S, degree are N is tied Structure;
If resolution ratio rs=ve×va, beam spaceWidth in e directions is We, the width in a directions is Wa, i.e.,For We×WaRectangular area, the degree N=N of tree constructione×Na, depth S and degree N meet condition:And
Each node has N number of child node in 1 to S-1 layer in the three-dimensional training code book tree construction;
(2) by beam spaceN is carried out in e directionseDecile, N is carried out in a directionsaDecile, willIt is divided into N number of rectangle RegionN=Ne×Na, it is designated as setEach rectangular areaThere are beamforming vectorsTherewith Correspondence, i=1 ..., N, N number of beamforming vectors constitute beam setI.e.
The root node C of the three-dimensional training code book tree construction1,1Gather for rectangular areaWith beam setCombination
(3) b-th of node C of s layers of the three-dimensional training code book tree construction is determineds,b, Cs,bGather for rectangular area With beam setCombination
WhereinWith s-1 layers of nodeWith following relation:
Wherein s=2 ..., S, b=1 ..., Ns-1,For the computing that rounds up, % is complementation;
It is rightN is carried out in e directionseDecile, N is carried out in a directionsaDecile, obtains N number of sub- rectangular areaPer height Rectangular areaThere are beamforming vectorsCorrespond to therewith, i=1 ..., N, N number of beamforming vectors constitute beam setI.e.
Determine each node in 2 to S layers successively by above-mentioned steps
(4) beamforming vectors that each node is included in the three-dimensional training code book tree construction are solvedObtain three-dimensional Train code book, wherein q=1 ..., N.
Preferably, the beamforming vectors that each node is included in the three-dimensional training code book tree constructionIn main lobe and side Valve has amplitude response value that is constant but differing, i.e.,
Wherein ψ=(ψea)=(sin (θe),sin(θa)) it is pitching angle thetaeWith azimuth angle thetaaSine value combination,For the antenna-array response vector a in horizontal directionh(2π/λdhsin(θe)cos(θa)) with vertical direction on antenna Array response vector av(2π/λdvsin(θa)) Kronecker products,It is beamforming vectorsAmplitude response value; Wherein q=1 ..., N.
Preferably, beamforming vectors are solved in step (4)For solving-optimizing problem:
Wherein ε is the ripple of the training main secondary lobe of wave beam, is a minimum arithmetic number,It is main secondary lobe respectively Corresponding region.
Preferably, it is rightWithCarry out sample variance and turn to discrete beam master, secondary lobe corresponding regionWithArea Domain.
To avoid the generation of overshoot phenomenon after discrete sampling, constraints is introducedRipple is solved in step (4) Beam shaping vectorFor solving-optimizing problem:
Wherein ε is the ripple of the training main secondary lobe of wave beam, is a minimum arithmetic number;EsFor constant, value is with s increasing Reduce greatly;WithRespectively discrete beam master, secondary lobe corresponding region.
Preferably, by constraints slacking, using the concavo-convex process iterative algorithm solving-optimizing problem of constraint, step is such as Under:
(6.1) iteration initial value f is determined0, build following constrained optimization problem:
Wherein Expression takes the real part of numerical value in bracket,Represent respectively discrete The main secondary lobe corresponding region changed;
(6.2) optimization problem in iterative step (6.1), checks whether the r values that current iteration is obtained meet convergence Standard, if meeting, the optimal solution f of current iterationnAs last solution f*;If it is not satisfied, according to the optimal solution f of current iterationn, then Optimization problem in secondary solution procedure (6.1);
(6.3) output last solution f*, the three-dimensional training wave beam needed for obtaining.
Preferably, initial value f in step (6.1)0Determined by solving following optimization problem:
Another aspect of the present invention discloses a kind of beam alignment, comprises the following steps:
(8.1) parameter configuration and initialization:Sending and receiving end is according to any of the above-described kind of three-dimensional training code book design method design three Dimension training code book;
If designed three-dimensional training code book contains S training subcode bookS training rank is corresponded to respectively Section;The degree of designed three-dimensional training code book tree construction is N;Beamforming vectorsIn subscript value be initialized as:S=1, bf =1;
(8.2) transmitting terminal is continuously using the low resolution training beam vectors in first layer subcode bookTo pass Defeated training signal z, and each training wave beam repetition training MsThe secondary received signal to noise ratio to increase system;
(8.3) receiving terminal is according to corresponding reception signalSelection is produced most Height receives the beamforming vectors of energy, i.e.,
WhereinP is transmit power, and h is channel matrix, hHFor h associate matrix,For additive Gaussian noise;
Receiving terminal is by index value q*Feed back to transmitting terminal, thus, transmitting terminal according toAnd its corresponding rectangular area Primarily determine that the corresponding departure angle AoD regions of most high-amplitude wave beam;
(8.4) transmitting terminal is according to index value q*, select one group to have compared with high score in next straton code book of three-dimensional code book The beam forming vector of resolutionQ=1 ..., N, to transmit training signal, so as to further more accurately determine AoD institutes In region;
Wherein bf+1=bf*N-N+q*
(8.5) repeat step (8.3)~(8.4), the beamforming vectors of highest resolution requirement are met until finding, The vectorial corresponding rectangular domain is AoD regions.
Beneficial effect:What the present invention was provided is applied to the three-dimensional training code book design method of millimeter-wave communication system, can use In realizing high-precision wave beam alignment and channel estimation, its advantage is:
(1) existing research is in the millimeter-wave systems of ULA arrangements mainly for antenna, and the present invention has taken into full account UPA knots Structure, compensate for the deficiency of existing research.More transmission antennas can be configured under UPA structures, in limited two-dimensional space, because This, which can be detected, intercepts wider array of geographic area, realizes the three-dimensional wave cover of variable precision, and obtain bigger wave beam into Shape gain.
(2) present invention is handled the overshoot phenomenon that may occur on non-sampled point in training beam pattern, simultaneously Also the ripple of main secondary lobe is suitably constrained, thus each training wave beam have relatively flat amplitude response and Ideal transition band, is conducive to improving wave beam alignment precision.
(3) the three-dimensional training code book that the present invention is generated is applied to the wave beam alignment algorithm based on tree search algorithm, can be significantly The training expense of reduction system, and can be realized in the millimeter-wave communication system based on UPA structures the alignment of high-precision wave beam and Channel estimating performance.
Brief description of the drawings
Fig. 1 is the structural representation of uniform planar array (UPA);
Fig. 2 is the structural representation of the three-dimensional training code book generated in embodiment;
Fig. 3 is that main sidelobe magnitudes fluctuate schematic diagram;
Fig. 4 is generate 4 training beam vectors in embodimentAmplitude response figure;
Fig. 5 is the training beam pattern under two kinds of Codebook Design methods (BPSA, LSA);
Fig. 6 is used for ripple after being generated for the particular flow sheet and code book of three-dimensional training code book design method proposed by the present invention Shu Xunlian flow chart;
Fig. 7 is under two kinds of Codebook Design methods (BPSA, LSA), average wave beam alignment error rate becomes with the change of signal to noise ratio Gesture figure.
Embodiment
With reference to the accompanying drawings and detailed description, the present invention is furture elucidated.
As shown in figure 1, three-dimensional training code book design method disclosed by the invention is applied to antenna in uniform surface battle array arrangement Millimeter-wave systems, by taking transmitting terminal as an example, N in figureh、NvThe antenna number on horizontally and vertically, therefore total hair are represented respectively Penetrate antenna number NT=NhNv, dhAnd dvSpacing respectively between horizontal direction, vertically adjacent two antennas.
Beam spaceFor a rectangular area, if e directions are vertical both direction with a directions,Width in e directions Spend for We, the width in a directions is Wa, i.e.,For We×WaRectangular area.
The present embodiment considers millimeter-wave communication system of the transmitting terminal antenna in uniform surface battle array arrangement, wherein Nh=Nv=32, phase Spacing d between adjacent two antennash=dv=3 λ/8, λ are the wavelength of millimeter-wave signal.For the sake of simplicity it is assumed that receiving terminal is only matched somebody with somebody Have single antenna.Although it should be noted that this example is only designed as example, this side with the three-dimensional training code book of transmitting terminal Method is equally applicable to the Codebook Design of receiving terminal.
A kind of three-dimensional training code book design method for millimeter-wave communication system, comprises the following steps:
Step 1, according to resolution ratio rs and beam spaceScope to set up depth be the three-dimensional training code book tree that S, degree are N Structure;
If resolution ratio rs=ve×va;If the pitching angle theta of launching beame, azimuth angle thetaaIt is located at angular range respectivelyWithin, and the angle of pitch is independent uncorrelated to azimuth Variable, can be represented, then beam space in the present embodiment with vertical both direction e directions with the variable in a directionsIt can determine Justice is sin (θe)、sin(θa) correspondence scope product, i.e., It is in (sin (θe),sin(θa)) plane is a rectangular domain, i.e. We=2sin (Θe), Wa=2sin (Θa)。
The degree N=N of three-dimensional training code book tree constructione×Na, depth S and degree N meet condition:AndI.e. By beam spaceIt is divided into NSBehind individual small rectangular area, each small rectangular area is less than equal to resolution ratio.
Each node has N number of child node in 1 to S-1 layer in the three-dimensional training code book tree construction;
Step 2, by beam spaceN is carried out in e directionseDecile, N is carried out in a directionsaDecile, willIt is divided into N number of Rectangular areaN=Ne×Na, it is designated as setEach rectangular areaThere are beamforming vectors Correspond to therewith, i=1 ..., N, N number of beamforming vectors constitute beam setI.e.
The root node C of the three-dimensional training code book tree construction1,1Gather for rectangular areaWith beam setCombination
(3) b-th of node C of s layers of the three-dimensional training code book tree construction is determineds,b, Cs,bGather for rectangular area With beam setCombination
WhereinWith s-1 layers of nodeWith following relation:
Wherein s=2 ..., S, b=1 ..., Ns-1,For the computing that rounds up, % is complementation;
It is rightN is carried out in e directionseDecile, N is carried out in a directionsaDecile, obtains N number of sub- rectangular areaPer height Rectangular areaThere are beamforming vectorsCorrespond to therewith, i=1 ..., N, N number of beamforming vectors constitute beam setI.e.AndWithCorrespondence;
Determine each node in 2 to S layers successively by above-mentioned steps
(4) beamforming vectors that each node is included in the three-dimensional training code book tree construction are solvedObtain three-dimensional Train code book, wherein q=1 ..., N.
The tree construction of three-dimensional training code book (CodeBook) has S layers, and each layer is a sub- code book (Sub-CodeBook), One training stage of correspondence;S layers of subcode book has Ns-1Individual node, each node includes N number of code word (CodeWord), each Node is by beam setAnd its corresponding rectangular areaCarry out table Show;And its corresponding rectangular areaCombinationFor a code word.
As shown in Fig. 2 the three-dimensional training code book generated in the present embodiment is a quad-tree structure, i.e., three-dimensional training code book The degree N=N of tree constructione×Na=2 × 2.It is made up of the S sub- code books of S training stage of correspondence;For s-th of training stage Subcode bookIn, altogether comprising 4sThe individual beamforming vectors with identical main lobe width, and these vectors constitute 4s-1Individual ripple Constriction is closedWherein b-th setInclude beamforming vectorsB=1 ..., 4s -1;Subcode bookIn beamforming vectors number be last layer subcode book4 times of middle training wave beam number.In s-th of instruction Practice the stage, beam space is divided into 4s-1The rectangular domain of individual formed objectsWherein b-th rectangular domain With beam setCorrelation, andWhile beamforming vectors Respectively withCorrespond;In next training stage, rectangular domainIt is divided into 4 smaller again Subregion, j ∈ { LL, LR, RL, RR }, i.e.,
To improve the wave beam alignment performance of millimetre-wave attenuator, designed beamforming vectorsQ=1 ..., N, respectively There is amplitude response value that is constant but differing on main lobe and secondary lobe, i.e.,:
Wherein ψ=(ψea)=(sin (θe),sin(θa)) it is pitching angle thetaeWith azimuth angle thetaaSine value combination,For the antenna-array response vector a in horizontal direction, vertical directionh(2π/λdhsin(θe)cos(θa)) and av(2π/λ dvsin(θa)) Kronecker products, dhAnd dvBetween respectively between horizontal direction, vertically adjacent two antennas Away from, λ is the wavelength of millimeter-wave signal,It is beam vectorsAmplitude response value.
If the ripple ε of the training main secondary lobe of wave beam is strictly limited as 0, it is difficult to the design for successfully realizing three-dimensional code book, is This infeasibility is avoided, while to ensure high-precision wave beam alignment performance, as shown in figure 3, the master of beam pattern can be allowed There is fluctuation within a narrow range in secondary lobe, while in the case where ripple ε value is fixed, the amplitude response value of main lobe should be as big as possible, because The design of this training wave beam can be expressed as optimization problem:
Wherein ε is a minimum arithmetic number,Respectively based on, secondary lobe corresponding region,For rectangular areaIt is interior,ForOutside.Due toIt is continuous noncountable, it is therefore necessary to which that the main secondary lobe corresponding region of wave beam is adopted Sample or discretization;Again due to sampled point limited amount, overshoot phenomenon is there may be in unsampled region, to avoid discrete sampling The generation of overshoot phenomenon afterwards, introduces constraints | | f | |≤Es, EsValue reduce with s increase, | | | | to ask 2 norms of vector, i.e. optimization problem is:
The present embodiment is using following limited scatter channel model:
Wherein L is total number of channels, αlIt is the complex gain of l paths, β is average path loss, It is antenna-array response vector, and ab(b ∈ { h, v }) has Following form: Accorded with for Kronecker product calculations.
With the beamforming vectors in a training code word in three-dimensional training code bookExemplified by, the code word is in discrete wave Beam regionAmplitude response value it is larger, in other regionsAmplitude response it is minimum, the design of the training wave beam is asked Topic is a non-convex optimization problem, is solved difficult.By the way that by part non-convex constraints slacking, the non-convex optimization problem can quilt Convex optimization problem is converted into, and is solved using concavo-convex process (CCCP) iterative algorithm of constraint.If fnRepresent that nth iteration is obtained Optimal solution, then next iteration optimal solution fn+1It can be obtained by solving following problem:
Wherein Expression takes the real part of numerical value in bracket,Represent respectively discrete The main secondary lobe corresponding region changed.
There is considerable influence to CCCP convergence of algorithm performances due to initializing, to obtain a good initial value, can ask Solve following optimization problem:
This problem is a Second-order cone programming problem (SOCP), can be asked by the CVX tool boxes in MATLAB emulation platforms Solution.
Obtaining good initial value f0Afterwards, the three-dimensional training beam designing method based on CCCP algorithms specifically include as Lower step:
(6.1) build and solve following constrained optimization problem:
Wherein Expression takes the real part of numerical value in bracket,Represent respectively discrete The main secondary lobe corresponding region changed;
(6.2) optimization problem in iterative step (6.1), checks whether the r values that current iteration is obtained meet convergence Standard, if meeting, the optimal solution f of current iterationnAs last solution f*;If it is not satisfied, according to the optimal solution f of current iterationn, then Optimization problem in secondary solution procedure (6.1);
(6.3) output last solution f*, the three-dimensional training wave beam needed for obtaining.
Fig. 4 gives 4 of the design method generation using the present invention by taking the first layer subcode book in three-dimensional code book as an example Train code wordAmplitude response figure.It can be found that the beam pattern of these code words all has relatively flat Amplitude response and ideal transition band, and there is no overshoot phenomenon on non-sampled point yet.Fig. 5 also incite somebody to action this Invent compared with the beam designing method based on least square (LS), wherein BPSA represents that three-dimensional code book proposed by the present invention is set Calculating method, LSA represents least square beam designing algorithm.It can be seen that the side of the training wave beam generated using LS methods Valve has relatively large fluctuation, in addition, the width of its intermediate zone is also larger, this further embodies three-dimensional instruction proposed by the present invention Practice the superiority of Codebook Design method.
The three-dimensional training code book that the present invention is generated can be used for realizing the wave beam alignment based on tree search algorithm, detect single footpath The most high-amplitude wave beam of millimeter wave channel, it can also be used to estimate the relevant parameter of millimetre-wave attenuator channel, such as departure angle (AoD) and arrival Angle (AoA), specifically includes following steps:
(8.1) parameter configuration and initialization:Design three-dimensional according to the method any one of claim 1-7 in sending and receiving end Train code book;
If designed three-dimensional S layers of code book tree construction of training, i.e., containing S training subcode bookRespectively The S training stage of correspondence;The degree of designed three-dimensional training code book tree construction is N;Beamforming vectorsIn subscript value at the beginning of Beginning turns to:S=1, bf=1;
(8.2) transmitting terminal is continuously using the low resolution training beam vectors in first layer subcode bookTo pass Defeated training signal z, and each training wave beam repetition training MsThe secondary received signal to noise ratio to increase system;
(8.3) receiving terminal is according to corresponding reception signalSelection is produced most Height receives the beamforming vectors of energy, i.e.,
WhereinP is transmit power, and h is channel matrix, hHFor h associate matrix,For additive Gaussian noise;
Receiving terminal is by index value q*Feed back to transmitting terminal, thus, transmitting terminal according toAnd its corresponding rectangular area Primarily determine that the corresponding departure angle AoD regions of most high-amplitude wave beam;
(8.4) transmitting terminal is according to index value q*, select one group to have compared with high score in next straton code book of three-dimensional code book The beam forming vector of resolutionQ=1 ..., N, to transmit training signal, so as to further more accurately determine AoD institutes In region;
Wherein bf+1=bf*N-N+q*
For the quad-tree structure in the present embodiment, bfRenewal follow following principle:
If a) q*=LL, then bf+1=4bf-3;
If b) q*=LR, then bf+1=4bf-2;
If c) q*=RL, then bf+1=4bf-1;
If d) q*=RR, then bf+1=4bf
(8.5) repeat step (8.3)~(8.4), the beamforming vectors of highest resolution requirement are met until finding, The vectorial corresponding rectangular domain is AoD regions.
Fig. 6 gives the idiographic flow of three-dimensional training code book design method (BPSA) proposed by the present invention, in addition, figure 6 also give the detailed process trained after code book generation for wave beam.As shown in fig. 7, the figure shows believe when millimetre-wave attenuator The path gain in road is fixed as in the case of 1, and the code book generated using LS methods and the inventive method carries out wave beam training, produces Average wave beam alignment error probability (BAER) with signal to noise ratio (SNR) variation tendency.It can be seen that the inventive method The code book performance of generation is substantially better than LSA code books, can realize more accurate wave beam alignment and channel estimating performance.With SNR Increase, the corresponding BAER of two kinds of code books reduces therewith, but the corresponding error probability decay of BPSA code books is obvious, and LSA code books Corresponding error probability decay is slow.

Claims (8)

1. a kind of three-dimensional training code book design method for millimeter-wave communication system, it is characterised in that comprise the following steps:
(1) according to resolution ratio rs and beam spaceScope to set up depth be the three-dimensional training code book tree construction that S, degree are N;
If resolution ratio rs=ve×va, beam spaceWidth in e directions is We, the width in a directions is Wa, i.e.,For We× WaRectangular area, the degree N=N of tree constructione×Na, depth S and degree N meet condition:And
Each node has N number of child node in 1 to S-1 layer in the three-dimensional training code book tree construction;
(2) by beam spaceN is carried out in e directionseDecile, N is carried out in a directionsaDecile, willIt is divided into N number of rectangular areaN=Ne×Na, it is designated as setEach rectangular areaThere are beamforming vectorsIt is right therewith Should, i=1 ..., N, N number of beamforming vectors constitute beam setI.e.
The root node C of the three-dimensional training code book tree construction1,1Gather for rectangular areaWith beam setCombination
(3) b-th of node C of s layers of the three-dimensional training code book tree construction is determineds,b, Cs,bGather for rectangular areaWith ripple Constriction is closedCombination
WhereinWith s-1 layers of nodeWith following relation:
Wherein s=2 ..., S, b=1 ..., Ns-1,For the computing that rounds up, % is complementation;
It is rightN is carried out in e directionseDecile, N is carried out in a directionsaDecile, obtains N number of sub- rectangular areaEvery sub- rectangle RegionThere are beamforming vectorsCorrespond to therewith, i=1 ..., N, N number of beamforming vectors constitute beam setI.e.
Determine each node in 2 to S layers successively by above-mentioned steps
(4) beamforming vectors that each node is included in the three-dimensional training code book tree construction are solvedObtain three-dimensional training Code book, wherein q=1 ..., N.
2. the three-dimensional training code book design method according to claim 1 for millimeter-wave communication system, it is characterised in that The beamforming vectors that each node is included in the three-dimensional training code book tree constructionHave in main lobe and secondary lobe constant but not Identical amplitude response value, i.e.,
Wherein ψ=(ψea)=(sin (θe),sin(θa)) it is pitching angle thetaeWith azimuth angle thetaaSine value combination,For Antenna-array response vector a in horizontal directionh(2π/λdh sin(θe)cos(θa)) rung with the aerial array in vertical direction Answer vector av(2π/λdv sin(θa)) Kronecker products,It is beamforming vectorsAmplitude response value;Wherein q =1 ..., N.
3. the three-dimensional training code book design method according to claim 2 for millimeter-wave communication system, it is characterised in that Beamforming vectors are solved in step (4)For solving-optimizing problem:
Wherein ε is the ripple of the training main secondary lobe of wave beam, is a minimum arithmetic number,It is main secondary lobe correspondence area respectively Domain.
4. the three-dimensional training code book design method according to claim 3 for millimeter-wave communication system, it is characterised in that It is rightWithCarry out sample variance and turn to discrete beam master, secondary lobe corresponding regionWithRegion.
5. the three-dimensional training code book design method according to claim 4 for millimeter-wave communication system, it is characterised in that Beamforming vectors are solved in step (4)For solving-optimizing problem:
Wherein ε is the ripple of the training main secondary lobe of wave beam, is a minimum arithmetic number;EsFor constant, value is with s increase Reduce;WithRespectively discrete beam master, secondary lobe corresponding region.
6. the three-dimensional training code book design method according to claim 5 for millimeter-wave communication system, it is characterised in that By constraints slacking, using the concavo-convex process iterative algorithm solving-optimizing problem of constraint, step is as follows:
(6.1) iteration initial value f is determined0, build following constrained optimization problem:
Wherein Expression takes the real part of numerical value in bracket,Discretization is represented respectively Main secondary lobe corresponding region;
(6.2) optimization problem in iterative step (6.1), checks whether the r values that current iteration is obtained meet convergence, If meeting, the optimal solution f of current iterationnAs last solution f;If it is not satisfied, according to the optimal solution f of current iterationn, solve again Optimization problem in step (6.1);
(6.3) output last solution f, the three-dimensional training wave beam needed for obtaining.
7. the three-dimensional training code book design method according to claim 6 for millimeter-wave communication system, it is characterised in that Initial value f in step (6.1)0Determined by solving following optimization problem:
8. a kind of beam alignment, it is characterised in that comprise the following steps:
(8.1) parameter configuration and initialization:Sending and receiving end is according to the three-dimensional training of method design any one of claim 1-7 Code book;
If designed three-dimensional training code book contains S training subcode bookS training stage is corresponded to respectively; The degree of designed three-dimensional training code book tree construction is N;Beamforming vectorsIn subscript value be initialized as:S=1, bf= 1;
(8.2) transmitting terminal is continuously using the low resolution training beam vectors in first layer subcode bookTo transmit instruction Practice signal z, and each training wave beam repetition training MsThe secondary received signal to noise ratio to increase system;
(8.3) receiving terminal is according to corresponding reception signalSelection produces highest and connect The beamforming vectors of energy are received, i.e.,
WhereinP is transmit power, and h is channel matrix, hHFor h associate matrix,For additive Gaussian noise;
Receiving terminal is by index value qFeed back to transmitting terminal, thus, transmitting terminal according toAnd its corresponding rectangular areaTentatively It is determined that the corresponding departure angle AoD regions of most high-amplitude wave beam;
(8.4) transmitting terminal is according to index value q, in next straton code book of three-dimensional code book selecting one group has high-resolution Beam forming vectorTo transmit training signal, so as to further more accurately determine AoD locations Domain;
Wherein bf+1=bf*N-N+q
(8.5) repeat step (8.3)~(8.4), the beamforming vectors of highest resolution requirement are met until finding, this to It is AoD regions to measure corresponding rectangular domain.
CN201710293707.XA 2017-04-28 2017-04-28 Three-dimensional training codebook design method and beam alignment method for millimeter wave communication system Active CN107135023B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201710293707.XA CN107135023B (en) 2017-04-28 2017-04-28 Three-dimensional training codebook design method and beam alignment method for millimeter wave communication system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201710293707.XA CN107135023B (en) 2017-04-28 2017-04-28 Three-dimensional training codebook design method and beam alignment method for millimeter wave communication system

Publications (2)

Publication Number Publication Date
CN107135023A true CN107135023A (en) 2017-09-05
CN107135023B CN107135023B (en) 2020-04-24

Family

ID=59716151

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201710293707.XA Active CN107135023B (en) 2017-04-28 2017-04-28 Three-dimensional training codebook design method and beam alignment method for millimeter wave communication system

Country Status (1)

Country Link
CN (1) CN107135023B (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108809397A (en) * 2018-06-27 2018-11-13 东南大学 High effect numerical model analysis beam-forming method, device and equipment in multiaerial system
CN109379122A (en) * 2018-11-08 2019-02-22 东南大学 A kind of millimetre-wave attenuator multipath channel dynamic beam training method
CN109474549A (en) * 2018-12-04 2019-03-15 上海矽昌通信技术有限公司 A kind of three dimensional channel estimation method based on three-dimensional beam pattern
CN111478724A (en) * 2020-04-15 2020-07-31 南京航空航天大学 Three-dimensional wave beam searching method for millimeter wave platform of unmanned aerial vehicle
WO2020187033A1 (en) * 2019-03-15 2020-09-24 大唐移动通信设备有限公司 Hybrid beam forming architecture-based calibration compensation method and apparatus

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130107977A1 (en) * 2008-03-17 2013-05-02 Qualcomm Incorporated Multi-resolution beamforming based on codebooks in mimo systems
CN104935366A (en) * 2015-04-21 2015-09-23 北京航空航天大学 Wave beam searching method in millimeter wave communication
CN105610478A (en) * 2016-02-15 2016-05-25 东南大学 Millimeter wave MIMO communication multi-subarray cooperative beam alignment method and millimeter wave MIMO communication multi-subarray cooperative beam alignment device
CN105790811A (en) * 2016-04-15 2016-07-20 东南大学 High-resolution analog beam rapid training method and device
CN105959044A (en) * 2016-04-21 2016-09-21 北京航空航天大学 Hierarchical codebook structure design method of joint method

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20130107977A1 (en) * 2008-03-17 2013-05-02 Qualcomm Incorporated Multi-resolution beamforming based on codebooks in mimo systems
CN104935366A (en) * 2015-04-21 2015-09-23 北京航空航天大学 Wave beam searching method in millimeter wave communication
CN105610478A (en) * 2016-02-15 2016-05-25 东南大学 Millimeter wave MIMO communication multi-subarray cooperative beam alignment method and millimeter wave MIMO communication multi-subarray cooperative beam alignment device
CN105790811A (en) * 2016-04-15 2016-07-20 东南大学 High-resolution analog beam rapid training method and device
CN105959044A (en) * 2016-04-21 2016-09-21 北京航空航天大学 Hierarchical codebook structure design method of joint method

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108809397A (en) * 2018-06-27 2018-11-13 东南大学 High effect numerical model analysis beam-forming method, device and equipment in multiaerial system
CN108809397B (en) * 2018-06-27 2020-06-30 东南大学 High-efficiency digital-analog hybrid beam forming method, device and equipment in multi-antenna system
CN109379122A (en) * 2018-11-08 2019-02-22 东南大学 A kind of millimetre-wave attenuator multipath channel dynamic beam training method
CN109379122B (en) * 2018-11-08 2021-03-16 东南大学 Millimeter wave communication multipath channel dynamic beam training method
CN109474549A (en) * 2018-12-04 2019-03-15 上海矽昌通信技术有限公司 A kind of three dimensional channel estimation method based on three-dimensional beam pattern
CN109474549B (en) * 2018-12-04 2021-08-17 青岛矽昌通信技术有限公司 Three-dimensional channel estimation method based on three-dimensional beam pattern
WO2020187033A1 (en) * 2019-03-15 2020-09-24 大唐移动通信设备有限公司 Hybrid beam forming architecture-based calibration compensation method and apparatus
US11277182B2 (en) 2019-03-15 2022-03-15 Datang Mobile Communications Equipment Co., Ltd. Hybrid beam forming architecture-based calibration compensation method and apparatus
CN111478724A (en) * 2020-04-15 2020-07-31 南京航空航天大学 Three-dimensional wave beam searching method for millimeter wave platform of unmanned aerial vehicle

Also Published As

Publication number Publication date
CN107135023B (en) 2020-04-24

Similar Documents

Publication Publication Date Title
Björnson et al. Massive MIMO is a reality—What is next?: Five promising research directions for antenna arrays
CN107135023A (en) Three-dimensional training code book design method and beam alignment for millimeter-wave communication system
CN111092641B (en) Hybrid precoding design method based on millimeter wave MIMO system deep learning
CN113411110B (en) Millimeter wave communication beam training method based on deep reinforcement learning
CN110401476B (en) Codebook-based millimeter wave communication multi-user parallel beam training method
KR102154481B1 (en) Apparatus for beamforming massive mimo system using deep learning
CN110113088B (en) Intelligent estimation method for wave arrival angle of separated digital-analog hybrid antenna system
CN109946654B (en) Radar array element-pulse coding and processing method
CN107728118B (en) Low sidelobe transmission beam pattern design method without fitting covariance matrix
CN106355245A (en) Method for integrating array antenna directional images on basis of neural network algorithms
CN104935366A (en) Wave beam searching method in millimeter wave communication
CN109714091B (en) Iterative hybrid precoding method based on hierarchical design in millimeter wave MIMO system
CN109922427A (en) Utilize the intelligent radio positioning system and method for large scale array antenna
CN113489521A (en) Intelligent united beam forming method for non-cell large-scale MIMO network assisted by reflecting surface
CN109379122A (en) A kind of millimetre-wave attenuator multipath channel dynamic beam training method
CN115085774A (en) Joint sensation fusion hybrid beam forming method based on Cramer-Lo boundary
CN104994041B (en) Three-dimensional extensive Bo Dabo digression combined estimation method of the antenna system based on pilot tone
CN113922854A (en) Integrated radar sensing and wireless communication method with edge calculation assistance
CN113281732A (en) MIMO radar target positioning method and system based on space-time coding
CN112398513A (en) Beam forming method of massive MIMO system
CN115664471B (en) Millimeter wave MIMO base station cooperative beam selection method based on wide learning
CN107276657A (en) Millimeter wave mixed-beam forming method based on parallel algorithm
CN114844538B (en) Millimeter wave MIMO user increment cooperative beam selection method based on wide learning
CN111245525B (en) Large-scale MIMO underwater acoustic communication method and system
Bing A simplified genetic algorithm in multi-beam combination for mmWave communication system

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