CN109510649A - The data transmission method of MU-MIMO network based on CSI accuracy, validity and stability - Google Patents
The data transmission method of MU-MIMO network based on CSI accuracy, validity and stability Download PDFInfo
- Publication number
- CN109510649A CN109510649A CN201811629905.XA CN201811629905A CN109510649A CN 109510649 A CN109510649 A CN 109510649A CN 201811629905 A CN201811629905 A CN 201811629905A CN 109510649 A CN109510649 A CN 109510649A
- Authority
- CN
- China
- Prior art keywords
- csi
- user
- value
- data
- channel
- 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
- H04B7/0452—Multi-user MIMO systems
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L5/00—Arrangements affording multiple use of the transmission path
- H04L5/003—Arrangements for allocating sub-channels of the transmission path
- H04L5/0053—Allocation of signaling, i.e. of overhead other than pilot signals
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/02—Arrangements for optimising operational condition
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W24/00—Supervisory, monitoring or testing arrangements
- H04W24/10—Scheduling measurement reports ; Arrangements for measurement reports
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Mobile Radio Communication Systems (AREA)
- Radio Transmission System (AREA)
Abstract
The invention discloses a kind of data transmission methods of MU-MIMO network based on CSI accuracy, validity and stability, comprising: carries out accuracy evaluation to the CSI value H between user and the end AP using the end the AP bit error rate maximum likelihood estimation algorithm based on planisphere distribution distance；The end AP utilizes the CSI stability quantitative statistics algorithm become based on signal-to-noise ratio, calculates instant CSI valueSignal-to-noise ratio become SNRV；The end AP utilizes the AP end equipment status monitoring algorithm to test the speed based on PDP similarity, ToF estimation and stationary field, to instant CSI valueCarry out the analysis of channel status influence factor；The end AP utilizes the maximum time piece service algorithm based on relevant duration, and the maximum service duration of active user is determined according to the CSI coherence time of active user；The end AP utilizes the CSI correction algorithm based on downlink beamforming technology, re-starts the distribution for sending power, and this method can carry out the transmission of data in MU-MIMO wireless network based on CSI accuracy, validity and stability.
Description
Technical field
The invention belongs to wireless communication network security field, it is related to a kind of based on CSI accuracy, validity and stability
The data transmission method of MU-MIMO network.
Background technique
Newest standard 802.11ac agreement introduces multi-user's multiple-input and multiple-output (Multi-user Multiple
Input Multiple Output, MU-MIMO) technology, to promote the channel performance of wireless network.In MU-MIMO system,
Receiving end (Receiver) can receive the mixed signal sent from multiple transmitting terminals (Transmitter), in order to noiseless
Ground decodes the data of each user, needs by channel state information (Channel State Information, CSI).
CSI is the comprehensive measurement of the energy attenuation and phase delay that generate to different path signals, reflects the difference between different channels
It is anisotropic.When carrying out with multiple users concurrent communication simultaneously under synchronization identical frequency band, wireless access points (Access
Point, AP) why can carry out accurate blended data before downlink (Down Link, DL) sends data and prelist
Code, or accurate blended data decoding is carried out after uplink (Up Link, UL) receives data, it places one's entire reliance upon effective
CSI is calculated.Once there is CSI calculating to go wrong, receiving end will be unable to accurately decode corresponding data, especially in UL, one
The CSI of user, which calculates mistake, will lead to the data decoding errors of all users, so as to cause wireless network data efficiency of transmission
It is more low.Therefore, the accuracy of CSI and procurement cost can all directly influence the transmission performance of MU-MIMO network.
To safeguard MU-MIMO network performance, improves work currently based on the data transmission of CSI and focus primarily upon two sides
Face.On the one hand, before data transmission, according to the feature of the detailed process of CSI feedback and feedback frame, the feedback capacity of CSI is compressed
And Times of Feedback, or according to multiuser channel compete matching mechanisms, adaptively carry out user distribution formula selection adjustment with
The channel contention time is saved, so that the CSI feedback time overhead in data transmission procedure is reduced, such as AFC, OPUS and Signpost
Etc. schemes.On the other hand, after data transmission failure occurs, according to CSI accuracy, validity and its index of correlation, phase is carried out
Ground intelligent data recovery or partial retransmission are answered, or tentatively CSI is restored by changing power distribution, with extensive
Complex channel data transmission capabilities, such as PPR, SOFT and SMR scheme.But unfortunately, both of which is adopted just for particular problem
Part Improving Measurements are taken, comprehensive and systematic anticipation never is carried out to channel accuracy, validity and stability, more could not
Reasonable and long-range data transmission plan is made, is difficult to effectively extensive utilization under complicated data transmission conditions.
If not carrying out Pre-Evaluation, entire data transmission to the accuracy of CSI, validity and stability before data transmission
The throughput of process, the bit error rate even channel effectively whether be all unable to get guarantee.Firstly, when channel ambient noise always very
It greatly and can not eliminate, i.e., when channel SNRs (Signal Noise Ratio, SNR) are very low, the cost carried out data transmission will
It can greatly increase, continually request data retransmission or recovery are extremely unadvisable, and blindly reduce CSI feedback expense also in thing
It is of no help, or even originally intolerable SNR can be further decreased.Secondly, when equipment is in Large Amplitude Motion state, in real time
CSI value will appear significant change, and channel SNR can also change correspondingly, so as to cause CSI validity period (i.e. coherence time,
Coherent Time) substantially shorten, carried out data transmission again using expired CSI at this time undoubtedly unpractical.Furthermore when
When there is strong interference in equipment ambient enviroment, such as other equipment or other object of which movement bring signal interferences and multipath effect
It should will make the CSI of present channel that catastrophic fluctuation occur, so that data transmission be made to include that data are restored and retransmit all to be blocked.
Finally, when the CSI of channel is Yi Dan invalid and when can not obtain being effectively improved by certain measure in a short time, in current MU-
Under MIMO wireless network, equipment just no longer has data transmission capabilities, other any type of CSI feedback compressions, channel are adaptive
It should correct and data are restored or retransmit all will be skimble-skamble.Therefore, before the transmission of data next time, gained CSI is carried out
Validity, stabilization and stability carry out it is comprehensive and it is systematic estimate, avoid inefficient and invalid data from transmitting or retransmit,
And reasonable data-transmission mode is selected, it is extremely necessary for providing long-range data transmission plan, therefore is badly in need of wanting a kind of data
Transmission method, the data transmission method can be carried out in MU-MIMO wireless network based on CSI accuracy, validity and stability
The transmission of data.
Summary of the invention
It is an object of the invention to overcome the above-mentioned prior art, provide a kind of based on CSI accuracy, validity
And the data transmission method of the MU-MIMO network of stability, this method can based on CSI accuracy, validity and stability into
The transmission of data in row MU-MIMO wireless network.
In order to achieve the above objectives, the MU-MIMO network of the present invention based on CSI accuracy, validity and stability
Data transmission method the following steps are included:
1) end AP successively broadcasts NDPA and NDP packet, the VHT-LTF field that user passes through leading character in NDP packet to user
Its CSI value H between the end AP is calculated, and CSI value H is encapsulated into data frame field, then feed back in the form of data packet
The end AP；
2) end AP receives the data packet that user feedback is returned, and calculates instant CSI value further according to the leading character of the data packetWhen the end AP be transmitted across NDP/Poll request after receive the data packet when, then go to step 3)；Otherwise, step 5) is gone to；
3) end AP utilizes the instant CSI valueData frame field in the data packet is decoded, when the end AP, decoding should
When data frame field obtains ACK, then go to 5)；The CSI value between user and the end AP is obtained when the end AP decodes the data frame field
H then utilizes the end the AP bit error rate maximum likelihood estimation algorithm based on planisphere distribution distance to the CSI value between user and the end AP
H carries out accuracy evaluation, when the CSI value H between user and the end AP is accurate, then goes back to step 2) and carries out next user feedback
The reception of data packet back；Otherwise, then step 4) is gone to；
4) end AP utilizes the CSI stability quantitative statistics algorithm become based on signal-to-noise ratio, calculates instant CSI valueSignal-to-noise ratio
Become SNRV, when SNRV is less than or equal to preset threshold, then goes to step 5)；Otherwise, step 9) is gone to；
5) end AP utilizes the AP end equipment status monitoring algorithm to test the speed based on PDP similarity, ToF estimation and stationary field, right
Instant CSI valueThe analysis for carrying out channel status influence factor turns if current channel condition is influenced by noise/environmental change
To step 7)；Otherwise, then the CSI coherence time for estimating active user, step 6) is then gone to；
6) end AP utilizes the maximum time piece service algorithm based on relevant duration, according to the CSI coherence time of active user
The maximum service duration for determining active user, the PSP variation further according to adjacent data packet determine whether using instant CSI value
When the maximum service duration for corresponding to user in timer exhausts or PSP variation is more than threshold value, then step 8) is gone to；Otherwise, then
Go to step 2)；
7) end AP utilizes the CSI correction algorithm based on downlink beamforming technology, re-starts point for sending power
Match, then goes to step 8)；
8) end AP sends NDP to user, and request updates the CSI value H between active user and the end AP, then goes to step
2), after the number for going to step 2) is more than or equal to n times (being defaulted as 3 times), still cannot obtain accurate user and the end AP it
Between CSI value H when, then go to step 9)；
9) assert that current channel condition is poor, not can be carried out the transmission of data, abandon transmitting mould using MU-MIMO data
Formula.
Using the end the AP bit error rate maximum likelihood estimation algorithm based on planisphere distribution distance to user and AP in step 3)
CSI value H between end carries out the concrete operations of accuracy evaluation are as follows:
The VHT-LTF frame of leading character calculates channel between user and the end AP in the NDP packet that user is broadcasted by the end AP
CSI value H, and the CSI value H of channel between user and the end AP is encapsulated into data frame field, then is fed back in the form of data packet
The end AP is given, after receiving feedback data packet, the VHT-LTF field for first passing through leading character in the data packet is calculated immediately at the end AP
CSI valueRecycle instant CSI valueThe CSI value H of channel between user in the data packet of feedback and the end AP is carried out extensive
It is multiple, then by the CSI value H of channel between user and the end AP and instant CSI value16QAM planisphere is mapped to as data-signal
In, CSI value H and instant CSI value further according to channel between user and the end APIn 16QAM planisphere between mapping symbols
Distance, the decoded bit error rate of approximate estimationWherein dMIndicate the decision boundary under different modulating mode
Value, when the decoded bit error rate meetsWhereinFor in the error of environmental noise
Boundary, threshold value beta are default upper-bound on BER, then it is assumed that current CSI value H is accurate；Otherwise, then step 4) is gone to.
The end AP utilizes the CSI stability quantitative statistics algorithm become based on signal-to-noise ratio in step 4), calculates instant CSI value's
The concrete operations of signal-to-noise ratio change SNRV are as follows:
The end AP calculates instant CSI value according to the SNR that the SNR and current data packet of historical record are reflectedSignal-to-noise ratio become
SNRV。
The end AP utilizes the AP end equipment status monitoring to test the speed based on PDP similarity, ToF estimation and stationary field to calculate in step 5)
Method, to instant CSI valueCarry out the concrete operations that channel status influence factor is analyzed are as follows:
The channel multipath effect in MU-MIMO network is portrayed at the end AP using PDP similarity, to distinguish locating for active user
Static, rotation/experiencing small oscillating movements and Large Amplitude Motion/environmental change/influence of noise three classes state；By being estimated using ToF
Description the distance between user and the end AP variation is calculated, is in Large Amplitude Motion state to distinguish active user, or become by environment
Change or influence of noise, while according to the whether stable direction of motion increased and decreased to determine user of ToF；By being transported using WiFi signal
Employ the standing wave field-effect generated in family, the speed of approximate estimation active user movement；By analyzing subscriber channel status,
The end AP determines the CSI coherence time of active user, and the service time of user is formulated according to the CSI coherence time of active user.
The end AP utilizes the maximum time piece service algorithm based on relevant duration in step 6), according to the CSI phase of current device
The dry time determines the concrete operations of the maximum service duration of active user are as follows:
If the continuous service duration of maximum that each user can obtainTc,lIt is user current
When the channel inscribed be concerned with duration, TmaxFor maximum channel coherency time in the CSI Matrix List of the end AP, the end AP obtain PDP,
User Status and after service time determines whether that request updates current CSI value according to the most strong path energy PSP at the end AP, works as phase
The PSP variation of adjacent data packet is more than threshold value ThrPOr service time, when exhausting, the end AP then requests to update CSI；Otherwise, continue former
There is data transmission procedure, and carries out corresponding calculate and assess.
The end AP utilizes the CSI correction algorithm based on downlink beamforming technology in step 7), re-starts transmission function
In the specific operation process of the distribution of rate, the end AP passes through CSI accuracy evaluation algorithm, validity period prediction algorithm and stability assessment
Algorithm, which is selected, needs to carry out CSI amendment user collection Λ, if the minimum power that sends of each user isWhen total transmission at the end AP
Power is more than or equal toWhen, then the CSI that user collects all users in Λ is disposably modified at the end AP, otherwise, then right in batches
The user collects the amendment that the user in Λ carries out CSI.
The invention has the following advantages:
The data transmission method of MU-MIMO network of the present invention based on CSI accuracy, validity and stability exists
When specific operation, using the end the AP bit error rate maximum likelihood estimation algorithm based on planisphere distribution distance, current CSI is assessed in real time
Accuracy, avoid because inaccuracy CSI due to cause data transmission failure；In addition, using the CSI stability become based on signal-to-noise ratio
Quantitative statistics algorithm, the stability of real time reaction present channel further clarify more long-range and reasonable data transmission mould
Formula guarantees the reliability of data transmission in a short time；Simultaneously using the equipment to be tested the speed based on PDP similarity, ToF estimation and stationary field
Status monitoring algorithm effectively distinguishes static, rotation/experiencing small oscillating movements locating for current user equipment, big from gained CSI information
The states such as amplitude movement and movement velocity and environment/noise variation influence, determine the CSI coherence time of current device；In addition, adopting
With the maximum time piece service algorithm based on relevant duration, by setting balancing, the use for effectively avoiding CSI life cycle long
Family excessively occupies channel and harmful competition problem that other channels cannot be serviced effectively, in combination with the PSP of adjacent data packet
Variation, further clarifies whether the end AP needs to request the problem of updating CSI；Finally using based on downlink beamforming technology
CSI correction algorithm, the case where influence for CSI by noise/environmental change, a degree of reasonable reparation is carried out to CSI, is
It attempts recovery normal communication and provides possibility；It should be noted that the present invention is based on existing standard 802.11ac agreement, it is only right
The end AP program, which is made, slightly to be modified, and has been carried out comprehensive and systematic judge for the instant CSI of active user and has been estimated,
Consideration is comprehensive, practicability and with strong applicability.
Detailed description of the invention
Fig. 1 is typical case's MU-MIMO wireless network architecture schematic diagram；
Fig. 2 a is the NDP frame format figure of AP broadcast；
Fig. 2 b is the data packet frame format chart for encapsulating CSI vector that user sends；
Fig. 3 is the end AP CSI appraisal procedure, equipment condition monitoring mechanism and data-transmission mode planning flow chart；
Mapping schematic diagram of the CSI that Fig. 4 is obtained by the end AP in 16QAM modulation constellation；
Fig. 5 is the flow diagram of the end AP ToF estimation；
Fig. 6 a is the two-dimentional display diagram of CSI amplitude in the WiFi signal received when the end AP is moved with user equipment；
Fig. 6 b is the three-dimensional display figure of CSI amplitude in the WiFi signal received when the end AP is moved with user equipment；
Fig. 7 is the equipment CSI coherence time contrast schematic diagram under different motion speed.
Specific embodiment
The invention will be described in further detail with reference to the accompanying drawing:
The data transmission method packet of MU-MIMO network of the present invention based on CSI accuracy, validity and stability
Include following steps:
1) end AP successively broadcasts NDPA and NDP packet, the VHT-LTF field that user passes through leading character in NDP packet to user
Its CSI value H between the end AP is calculated, and CSI value H is encapsulated into data frame field, then feed back in the form of data packet
The end AP；
2) end AP receives the data packet that user feedback is returned, and calculates instant CSI value further according to the leading character of the data packetWhen the end AP is then to go to step 3) when receiving the data packet after being transmitted across NDP/Poll request；Otherwise, step 5) is gone to；
3) end AP utilizes the instant CSI valueData frame field in the data packet is decoded, when the end AP, decoding should
When data frame field obtains ACK, then go to 5)；The CSI value between user and the end AP is obtained when the end AP decodes the data frame field
H then utilizes the end the AP bit error rate maximum likelihood estimation algorithm based on planisphere distribution distance to the CSI value between user and the end AP
H carries out accuracy evaluation, when the CSI value H between user and the end AP is accurate, then goes back to step 2) and carries out next user feedback
The reception of data packet back；Otherwise, then step 4) is gone to；
4) end AP utilizes the CSI stability quantitative statistics algorithm become based on signal-to-noise ratio, calculates instant CSI valueSignal-to-noise ratio
Become SNRV, when SNRV is less than or equal to preset threshold, then goes to step 5)；Otherwise, step 9) is gone to；
5) end AP utilizes the AP end equipment status monitoring algorithm to test the speed based on PDP similarity, ToF estimation and stationary field, right
Instant CSI valueThe analysis for carrying out channel status influence factor turns if current channel condition is influenced by noise/environmental change
To step 7)；Otherwise, then the CSI coherence time for estimating active user, step 6) is then gone to；
6) end AP utilizes the maximum time piece service algorithm based on relevant duration, according to the CSI coherence time of active user
The maximum service duration for determining active user, the PSP variation further according to adjacent data packet determine whether using instant CSI value
When the maximum service duration for corresponding to user in timer exhausts or PSP variation is more than threshold value, then step 8) is gone to；Otherwise, then
Go to step 2)；
7) end AP utilizes the CSI correction algorithm based on downlink beamforming technology, re-starts point for sending power
Match, then goes to step 8)；
8) end AP sends NDP to user, and request updates the CSI value H between active user and the end AP, then goes to step
2), after the number for going to step 2) is more than or equal to n times (being defaulted as 3 times), still cannot obtain accurate user and the end AP it
Between CSI value H when, then go to step 9)；
9) assert that current channel condition is poor, not can be carried out the transmission of data, abandon transmitting mould using MU-MIMO data
Formula.
Specifically, the present invention is that have N under the MU-MIMO mode based on MU-MIMO modetThe transmitting terminal of root antenna is sent out
The data sentAnd have NrThe received data in the receiving end of root antennaBetween
Relationship meet
Y=HX+N
Wherein, (0, σ N~CN2) it is Gaussian noise, and H indicates channel state matrix CSI, therefore, is carrying out CSI matrix
When calculating, it then follows
To generate certain estimation variance D (H), wherein
So CSI is influenced obviously by factors such as channel movement, ambient noise and environment transformation, before data transmission, to CSI
It is extremely necessary that accuracy, validity and stability, which carry out assessment,.
The algorithm applied in the present invention will specifically be introduced below
1, a kind of end AP bit error rate maximum likelihood estimation algorithm based on planisphere distribution distance
Fig. 2 a, Fig. 2 b, Fig. 3 and Fig. 4 are please referred to, when carrying out accuracy evaluation to current CSI, user is broadcasted by the end AP
VHT long training frames (Very High in empty data packet (Null Data Packet, NDP, as shown in Figure 2 a)
Throughput-Long Training Field, VHT-LTF) the CSI matrix H of channel between calculating user and the end AP, and will
The CSI value feeds back to the end AP in being encapsulated into after data frame field, as shown in Figure 2 b, due to the leading character of user feedback data packet
In equally include VHT-LTF, the end AP after receiving feedback data frame, can in the same way by leading character calculate work as
The CSI matrix of preceding channelAnd the time interval between the end AP broadcast NDP and user feedback H is very short, therefore can be approximately consideredTherefore the end AP just can utilizeSignaling protein14-3-3 is carried out to the data frame field H in feedback data packet, and the two is regarded
It is mapped in planisphere for data-signal, as shown in figure 4, basis at this timeWith H in planisphere mapping symbols (Symbol) it
Between distance, the end AP can the decoded bit error rate of approximate estimation.As shown in figure 4, all reception symbols are mapped to a 16QAM
In planisphere, the transmission symbol A (0000) original for one, the possibility mapping result after by transmission is A1It arrives
AN.In Fig. 4 rectangular broken line frame indicate decision boundary, if receive symbol mapping result in this border inner, can be accurate
Be mapped as symbol A；On the contrary, an if reception symbol AjThe distance between A is larger, and exceeds decision boundary, then the symbol
Number other symbolic points will be mapped as, and so as to cause decoding error, therefore exceed decision side in planisphere by receiving symbol
A possibility that boundary, can speculate the accuracy for the CSI that channel SNRs and user calculate.A in figureiWith AjAlthough all on decision side
Difference except boundary, but between them is AiPositioned at the variance upper-bound of environmental noiseWithin, it is calculated by certain amendment
Method can make AiIt remaps within decision boundary, but AjIt still may error after correction algorithm；When possessing S day
When the end AP of line receives R symbol in a feedback transmission, then the data Y for the N containing noise that the end AP receivesSFor
Wherein, HlFor the CSI vector of user l, YS=[y1,y2,…,yS]。
Utilize maximum likelihood function L (yi,σ2) to σ2Zero is taken after derivation, is obtained
Wherein, For the real part of plural X, HHFor the hermitian of matrix H
Conjugation；When R value is larger, thenIt will be close to actual σ2；On the contrary, when R value is smaller,σ will be much larger than2；WithIndicate that the two is mapped in the distance in planisphere, when the distance value is less than decision boundary, symbol can be solved correctly
Code, otherwise, will generate a series of error codes, the variance of the distanceIt may be expressed as:
If dMIndicate the decision boundary value under different modulating mode, it is general by the mapping error of Chebyshev inequality call sign
RateAre as follows:
WithInstead of σ2Up to its upper limit value, a settable threshold value beta after many experiments, ifIllustrate that current CSI accuracy is preferable, it can be by the end AP for decoding concurrent data packet；On the contrary, the end AP will be right
The CSI sequence of the user l is analyzed, and judges the reason of influencing channel CSI accuracy, and considers to execute CSI correction algorithm, is used
In the CSI value for calibrating the user.
2, a kind of CSI stability quantitative statistics algorithm become based on signal-to-noise ratio
With reference to Fig. 3, when carrying out stability assessment to current CSI, setting signal-to-noise ratio becomes (SNR variation, SNRV)
Are as follows:
SNRV (t, δ)=| SNR (t+ Δ t, δ)-SNR (t, δ) |
Wherein, SNR (t, δ) is the signal-to-noise ratio of t moment under feedback interval δ, and Δ t is the time of adjacent signal-to-noise ratio computation twice
Interval.Many experiments show that the signal-to-noise ratio of static and rotation equipment becomes the small 1dB or so than substantially moving；When between feedback
When δ is from 1 ms to 100 ms, the signal-to-noise ratio no-load voltage ratio under stationary state substantially moves small 3dB；Further, since ambient enviroment is open
The interference of bring multipath reflection is small, and the fluctuation range of signal-to-noise ratio is smaller than under indoor environment；Stable signal-to-noise ratio means current
Channel status is relatively stable, and channel SNR and current channel SNR difference before is smaller, if being based on MU-MIMO network again at this time
Data are sent and parsed using CSI, just not will cause significantly handling capacity and are glided and the mass data parsing pernicious data such as unsuccessfully
Transmission fault.It is contemplated that setting threshold value ThrV, to determine whether the end AP is passed using MU-MIMO mode under the present conditions
Transmission of data, it is proposed that threshold value ThrV is 5 dB.The signal-to-noise ratio of current CSI becomes SNRVi,l>ThrV, then show current channel condition very
Difference, can not restore in a short time, directly abandon using other data-transmission modes instead using MU-MIMO mode.
3, a kind of equipment condition monitoring algorithm to be tested the speed based on PDP similarity, ToF estimation and stationary field
PDP similarity is utilized when carrying out validity period prediction to current CSI with reference to Fig. 3, Fig. 5, Fig. 6 a, Fig. 6 b and Fig. 7
(Power Delay Profile Similarity) portrays the channel multipath effect in MU-MIMO network, being capable of effective district
Divide static, rotation/experiencing small oscillating movements and Large Amplitude Motion/environmental change/influence of noise three classes state locating for current device；
By describing the distance change of user Yu the end AP using ToF estimation (Time of Flight), can clearly distinguish current
User is in Large Amplitude Motion state, or by environmental change or influence of noise, while can be according to the whether stable increase and decrease of ToF
To determine the direction of motion (separate/close end AP) of user；Further, by being generated in moving user using WiFi signal
Standing wave field-effect, approximate estimation active user movement speed, finally, complete analyze subscriber channel status after, AP
End just can determine the CSI coherence time of active user, thus effectively according to coherence time statistical value of the user under each state
Ground is that user formulates reasonable service time.
4, PDP similarity
PDP similarity is that the time domain of channel frequency domain response converts, and is obtained by the CSI of frequency domain by inverse Fourier transform, if
CSI is H (f), i.e., the CSI value extracted in leading characterThen corresponding PDP similarity h (t) are as follows:
Wherein, K is the item number of multipath, αkAnd τkThe decline and delay of kth paths are respectively indicated, δ () is delta letter
Number, enables fi(k)=| | αkδ(t-τk)||2The signal strength for the kth paths that PDP for i-th of data packet includes, then i-th with
PDP similarity S between (i-1)-th data packetiAre as follows:
Wherein,It is averaged norm for the PDP of i-th of data packet.
In the case where being more than 90% experimental situations, the PDP similarity under user's stationary state is greater than 0.95, and rotation/by a small margin
The value of lower PDP is greater than 0.9, therefore choosing the PDP similarity threshold under static and rotation/experiencing small oscillating movements is respectively 0.95 and 0.9,
It is denoted as Thr respectivelySAnd ThrM, as shown in Figure 3.
5, ToF is estimated
ToF is the round trip propagation time between signal AP and mobile device, and in general, user equipment distance AP is got at end
Far, then round trip propagation time is also longer.At the time of commercial network interface card can estimate that physical signal is sent to aerial when sending data packet,
Referred to as ToD (Time-of-Departure).Movable equipment waits the set time of a SIFS after receiving data packet, so
After start to indicate that current data packet has been successfully received to the feeding back ACK frame response of the end AP.Same commercialization network interface card can be evaluated whether user
The time that the ACK frame of feedback reaches, referred to as ToA (Time-of-Arrival).
With reference to Fig. 5, the difference between ToD and ToA includes ToF, it may be assumed that
tToF=tToA-tToD-tSIFS
Setting regular hour window observes ToF value, when ToF value shows stable rising or downward trend,
It may determine that user is in substantially motion state, otherwise it is assumed that it is since ambient enviroment becomes that the PDP similarity of user, which is less than threshold value,
Caused by change or influence of noise.Further, the direction of the trend user movement of ToF value, ToF value can also be passed through
Stablize increase mean user far from AP move, vice versa.
6, WiFi stationary field tests the speed
With reference to 6a and Fig. 6 b, WiFi stationary field is one kind of standing wave, and WiFi stationary field has both both standing wave and WiFi signal
Property.When static, the CSI information in WiFi signal received can not show standing wave phenomena；To being received in motion process
CSI into WiFi signal is extracted, and can find the WiFi standing wave of uniform stripe shape, indoor the formed standing wave of WiFi signal
Wavelength be its original signal wavelength X half, therefore by estimation standing wave signal frequency fo, available user equipment fortune
Dynamic speed υ are as follows:
Wherein,For the expectation of CSI standing wave frequency under each subcarrier.
Since when user equipment is moved along the rectilinear direction being connected with signal source, the WiFi standing wave observed is relatively
To be obvious, it can be considered to use the end AP of two or more different location come to user carry out standing wave test the speed, to avoid with
Measurement blind spot is generated when the tangential motion of family.
7, a kind of maximum time piece service algorithm based on relevant duration
For the fairness problem of user's competitive channel, previous in algorithm, user is within CSI validity period only to the end AP
Feeding back ACK frame although feedback overhead can be greatly reduced, but also brings the fairness problem of user service.
With reference to Fig. 7, the coherence time of CSI is corresponded to by the user equipment under statistics different motion speed, it is found that when
When user equipment remains static, CSI coherence time is more than 250ms, and even if movement speed is 1m/s, coherence time
It is far longer than largest packet transmission time as defined in 802.11ac agreement.Theoretically, possess the use of bigger channel coherency time
Family will obtain more transmission times, but the subscriber channel coherence time under static scene can be much larger than the use under mobile context
Family, and then it is seriously unbalanced to cause the throughput between each user to distribute.
With reference to Fig. 3, weigh to be done between user throughput fairness and network overall situation handling capacity, each user l's
The specific service time can modify are as follows:
Wherein, TlIndicate the continuous service duration of maximum that first of user can obtain, Tc,lIndicate first of user current
When the channel inscribed be concerned with duration, TmaxIndicate maximum channel coherency time in the CSI Matrix List of the end AP.Particularly, work as user
When equipment is static, it may be considered that setting Tl=200ms；When user equipment rotation or experiencing small oscillating movements, it may be considered that setting Tl=
100ms。
With reference to Fig. 3, the end AP obtained PDP, status of user equipment and after service time, can calculate most strong path energy
(Power of Strongest Path, PSP).If PiFor the PSP value of i-th of data packet, wherein
Pi=maxfi(k),1≤k≤K
Wherein, fiIt (k) is the signal strength of k-th of multi-path component in i-th of data packet PDP, K is the total of interference contribution
Number；When the PSP variation of adjacent data packet is greater than ThrPOr service time, when exhausting, the end AP will request to update CSI；Otherwise,
Continue legacy data transmission process, and carries out corresponding calculate and assess.Particularly, when continuous 3 times or more more new capital application CSI
It can not solve the problems, such as, then it is assumed that current channel condition is very poor, is not suitable for using MU-MIMO mode, uses other data transmission moulds instead
Formula.
8, a kind of CSI correction algorithm based on downlink beamforming technology
When calibrating to CSI, after AP termination receives the CSI of user feedback, the end AP is calculated by CSI accuracy evaluation
Method, validity period prediction algorithm and stability assessment algorithm, which are selected, needs to carry out the modified user of CSI, it is assumed that the user being selected
Integrate as Λ={ Λ1,Λ2,…,Λn, andThe end AP by way of DL-BF to these users again
The NDP protocol frame containing VHT-LDF is sent, it, can be with since signal energy can be concentrated on user's direction by DL-BF
Signal energy and channel SNR are effectively improved, since channel SNR meets:
The condition for meeting accuracy evaluation requirement in conjunction with the user l CSI calculated, obtains:
Therefore, the end AP is for the minimum power that sends of each userIf once to correct user to collect in Λ
The CSI of all users, then total transmission power needs at the end AP are greater thanIf the general power at the end AP is insufficient, can divide
Batch collects the amendment that the user in Λ carries out CSI to user.
How power distribution is carried out as the end AP, then can turn to linear optimization problem:
Wherein,User's l distribution basis power is given for the end APThe power that can maximumlly reallocate afterwards, should
Linear optimization problem can be solved by water-filling algorithm, the power that AP is finally redistributed at endAre as follows:
Wherein, (x)+=max { x, 0 }, weight wlAnd water filling grade μ meets:
If continuous 3 times or more the more new capital application CSI can not solve the problems, such as after being calibrated by above-mentioned CSI correction algorithm,
Then think that current channel condition is very poor, is not suitable for using MU-MIMO mode, other data-transmission modes are used in consideration instead.
Claims (6)
1. a kind of data transmission method of the MU-MIMO network based on CSI accuracy, validity and stability, which is characterized in that
The following steps are included:
1) NDPA and NDP packet is successively broadcasted at the end AP to user, and user is calculated by the VHT-LTF field in NDP packet leading character
Its CSI value H between the end AP, and CSI value H is encapsulated into data frame field, then feed back to the end AP in the form of data packet；
2) end AP receives the data packet that user feedback is returned, and calculates instant CSI value further according to the leading character of the data packetWhen
The end AP then goes to step 3) when receiving the data packet after being transmitted across NDP/Poll request；Otherwise, step 5) is gone to；
3) end AP utilizes the instant CSI valueData frame field in the data packet is decoded, when the end AP decodes the data
When frame field obtains ACK, then go to 5)；The CSI value H between user and the end AP is obtained when the end AP decodes the data frame field, then
Using the end the AP bit error rate maximum likelihood estimation algorithm based on planisphere distribution distance to the CSI value H between user and the end AP into
Row accuracy evaluation then goes back to step 2) and carries out next user feedback time when the CSI value H between user and the end AP is accurate
The reception for the data packet come；Otherwise, then step 4) is gone to；
4) end AP utilizes the CSI stability quantitative statistics algorithm become based on signal-to-noise ratio, calculates instant CSI valueSignal-to-noise ratio become
SNRV then goes to step 5) when SNRV is less than or equal to preset threshold；Otherwise, step 9) is gone to；
5) end AP utilizes the AP end equipment status monitoring algorithm to test the speed based on PDP similarity, ToF estimation and stationary field, to instant
CSI valueThe analysis for carrying out channel status influence factor goes to step if current channel condition is influenced by noise/environmental change
It is rapid 7)；Otherwise, then the CSI coherence time for estimating active user, step 6) is then gone to；
6) end AP utilizes the maximum time piece service algorithm based on relevant duration, is determined according to the CSI coherence time of active user
The maximum service duration of active user, the PSP variation further according to adjacent data packet determine whether using instant CSI valueWhen fixed
When device in correspond to user maximum service duration exhaust or PSP variation be more than threshold value when, then go to step 8)；Otherwise, then it goes to
Step 2)；
7) end AP utilizes the CSI correction algorithm based on downlink beamforming technology, re-starts the distribution for sending power, so
After go to step 8)；
8) end AP sends NDP to user, and request updates the CSI value H between active user and the end AP, then goes to step 2), when
The number of step 2) is gone to more than or equal to after n times, when still cannot obtain the CSI value H between accurate user and the end AP, is then turned
To step 9)；
9) assert that current channel condition is poor, not can be carried out the transmission of data, abandon using MU-MIMO data-transmission mode.
2. the data of the MU-MIMO network according to claim 1 based on CSI accuracy, validity and stability are transmitted
Method, which is characterized in that the end the AP bit error rate maximum likelihood estimation algorithm pair based on planisphere distribution distance is utilized in step 3)
CSI value H between user and the end AP carries out the concrete operations of accuracy evaluation are as follows:
The VHT-LTF frame of leading character calculates the CSI value of channel between user and the end AP in the NDP packet that user is broadcasted by the end AP
H, and the CSI value H of channel between user and the end AP is encapsulated into data frame field, then feed back to AP in the form of data packet
End, after receiving feedback data packet, the VHT-LTF field for first passing through leading character in the data packet calculates instant CSI at the end AP
ValueRecycle instant CSI valueThe CSI value H of channel between user in the data packet of feedback and the end AP is restored, so
Afterwards by the CSI value H of channel between user and the end AP and instant CSI valueIt is mapped in 16QAM planisphere as data-signal, then
According to the CSI value H of channel between user and the end AP and instant CSI valueIn the distance in 16QAM planisphere between mapping symbols,
The decoded bit error rate of approximate estimationWherein dMIt indicates the decision boundary value under different modulating mode, works as solution
The bit error rate of code meetsWhereinFor the upper error of environmental noise, threshold value beta
To preset upper-bound on BER, then it is assumed that current CSI value H is accurate；Otherwise, then step 4) is gone to.
3. the data of the MU-MIMO network according to claim 1 based on CSI accuracy, validity and stability are transmitted
Method, which is characterized in that the end AP utilizes the CSI stability quantitative statistics algorithm become based on signal-to-noise ratio in step 4), calculates instant
CSI valueSignal-to-noise ratio become SNRV concrete operations are as follows:
The end AP calculates instant CSI value according to the SNR that the SNR and current data packet of historical record are reflectedSignal-to-noise ratio become
SNRV。
4. the data of the MU-MIMO network according to claim 1 based on CSI accuracy, validity and stability are transmitted
Method, which is characterized in that the end AP utilizes the AP end equipment shape to test the speed based on PDP similarity, ToF estimation and stationary field in step 5)
State monitoring algorithm, to instant CSI valueCarry out the concrete operations that channel status influence factor is analyzed are as follows:
The channel multipath effect in MU-MIMO network is portrayed at the end AP using PDP similarity, quiet locating for active user to distinguish
Only, rotation/experiencing small oscillating movements and Large Amplitude Motion/environmental change/influence of noise three classes state；By being retouched using ToF estimation
State the distance between user and the end AP variation, be in Large Amplitude Motion state to distinguish active user, or by environmental change or
Influence of noise, while according to the whether stable direction of motion increased and decreased to determine user of ToF；By being used using WiFi signal in movement
The standing wave field-effect generated in family, the speed of approximate estimation active user movement；By analyzing subscriber channel status, the end AP
It determines the CSI coherence time of active user, and formulates the service time of user according to the CSI coherence time of active user.
5. the data of the MU-MIMO network according to claim 1 based on CSI accuracy, validity and stability are transmitted
Method, which is characterized in that the end AP utilizes the maximum time piece service algorithm based on relevant duration in step 6), according to current device
CSI coherence time determine active user maximum service duration concrete operations are as follows:
If the continuous service duration of maximum that each user can obtainTc,lIt is user at current time
Under channel be concerned with duration, TmaxFor maximum channel coherency time in the CSI Matrix List of the end AP, PDP, user are obtained at the end AP
State and after service time determines whether that request updates current CSI value according to the most strong path energy PSP at the end AP, works as consecutive number
It is more than threshold value Thr according to the PSP variation of packetPOr service time, when exhausting, the end AP then requests to update CSI；Otherwise, continue original number
According to transmission process, and carries out corresponding calculate and assess.
6. the data of the MU-MIMO network according to claim 1 based on CSI accuracy, validity and stability are transmitted
Method, which is characterized in that in step 7) end AP utilize the CSI correction algorithm based on downlink beamforming technology, again into
Row is sent in the specific operation process of the distribution of power, and the end AP passes through CSI accuracy evaluation algorithm, validity period prediction algorithm and steady
Qualitative evaluation algorithm, which is selected, needs to carry out the modified user's collection Λ of CSI, if the minimum power that sends of each user isWork as AP
Total transmission power at end is more than or equal toWhen, then the CSI that user collects all users in Λ is disposably modified at the end AP, otherwise,
The amendment that the user in Λ carries out CSI then is collected to the user in batches.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811629905.XA CN109510649B (en) | 2018-12-28 | 2018-12-28 | Data transmission method of MU-MIMO network based on CSI accuracy, validity and stability |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811629905.XA CN109510649B (en) | 2018-12-28 | 2018-12-28 | Data transmission method of MU-MIMO network based on CSI accuracy, validity and stability |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109510649A true CN109510649A (en) | 2019-03-22 |
| CN109510649B CN109510649B (en) | 2021-02-09 |
Family
ID=65756795
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811629905.XA Active CN109510649B (en) | 2018-12-28 | 2018-12-28 | Data transmission method of MU-MIMO network based on CSI accuracy, validity and stability |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109510649B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111372195A (en) * | 2020-01-23 | 2020-07-03 | 鹏城实验室 | Method, apparatus and storage medium for tracking position of mobile terminal in mobile communication network |
| CN111786917A (en) * | 2020-08-03 | 2020-10-16 | Oppo广东移动通信有限公司 | Channel estimation method, receiver and storage medium |
| CN113411177A (en) * | 2021-08-10 | 2021-09-17 | 广州慧睿思通科技股份有限公司 | Signal type detection method, device, equipment and storage medium |
| CN114268420A (en) * | 2021-12-23 | 2022-04-01 | 西安交通大学 | Method and system for improving throughput of wireless network uplink |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105637772A (en) * | 2013-10-17 | 2016-06-01 | 高通股份有限公司 | Methods and apparatus for channel state information feedback |
| CN106452674A (en) * | 2016-09-30 | 2017-02-22 | 西安交通大学 | Actual data packet restoration method based on 802.11ac |
| CN106712815A (en) * | 2017-01-18 | 2017-05-24 | 西安交通大学 | Validity examination based CSI feedback overhead reduction method |
-
2018
- 2018-12-28 CN CN201811629905.XA patent/CN109510649B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105637772A (en) * | 2013-10-17 | 2016-06-01 | 高通股份有限公司 | Methods and apparatus for channel state information feedback |
| CN106452674A (en) * | 2016-09-30 | 2017-02-22 | 西安交通大学 | Actual data packet restoration method based on 802.11ac |
| CN106712815A (en) * | 2017-01-18 | 2017-05-24 | 西安交通大学 | Validity examination based CSI feedback overhead reduction method |
Non-Patent Citations (2)
| Title |
|---|
| KUN ZHAO ; WEI XI ; JINSONG HAN ; ZHI WANG ; YAOXUAN YUAN ; JIZH: "Efficient CSI Correction with CFO for Uplink MU-MIMO Networks", 《IEEE》 * |
| YUANHANG CAI ; WEI XI ; KUN ZHAO ; JINSONG HAN ; JIZHONG ZHAO: "Smart retransmission for CSI recovery in uplink multi-user MIMO networks", 《IEEE》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111372195A (en) * | 2020-01-23 | 2020-07-03 | 鹏城实验室 | Method, apparatus and storage medium for tracking position of mobile terminal in mobile communication network |
| CN111786917A (en) * | 2020-08-03 | 2020-10-16 | Oppo广东移动通信有限公司 | Channel estimation method, receiver and storage medium |
| CN111786917B (en) * | 2020-08-03 | 2023-07-18 | Oppo广东移动通信有限公司 | Channel estimation method, receiver and storage medium |
| CN113411177A (en) * | 2021-08-10 | 2021-09-17 | 广州慧睿思通科技股份有限公司 | Signal type detection method, device, equipment and storage medium |
| CN113411177B (en) * | 2021-08-10 | 2021-11-16 | 广州慧睿思通科技股份有限公司 | Signal type detection method, device, equipment and storage medium |
| CN114268420A (en) * | 2021-12-23 | 2022-04-01 | 西安交通大学 | Method and system for improving throughput of wireless network uplink |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109510649B (en) | 2021-02-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8644408B2 (en) | Method and apparatus for channel feedback in a wireless communication system | |
| Bejarano et al. | Resilient multi-user beamforming WLANs: Mobility, interference, and imperfect CSI | |
| CN109510649A (en) | The data transmission method of MU-MIMO network based on CSI accuracy, validity and stability | |
| US8036131B2 (en) | Method for transmitting channel quality information based on differential scheme | |
| US8923219B2 (en) | Method and apparatus for supporting adaptive channel state information feedback rate in multi-user communication systems | |
| CN103262458B (en) | A kind of method and apparatus for adjusting channel quality instruction | |
| US7822142B2 (en) | Method and device for cancelling interference in wireless communication system | |
| US7191381B2 (en) | Mode lookup tables for data transmission in wireless communication channels based on statistical parameters | |
| US20150333812A1 (en) | Method and apparatus for supporting adaptive channel state information feedback rate in multi-user communication systems | |
| CN101631355B (en) | Selection method of modulation encoding way for TD-HSPA+multi-user MIMO system | |
| CN101848023B (en) | Multiuser MIMO downlink transmission mode selection method based on rateless code | |
| US20150245360A1 (en) | Staged multi-user grouping for multiple-input, multiple-output wireless communications | |
| US20200136700A1 (en) | Channel Prediction for Adaptive Channel State Information (CSI) Feedback Overhead Reduction | |
| EP3806344A1 (en) | Channel estimation method and apparatus, and communication system | |
| EP4059259A1 (en) | Efficient beam management for wireless networks | |
| US8913515B2 (en) | Measuring and improving multiuser downlink reception quality in wireless local area networks | |
| Bojović et al. | Realistic beamforming design using SRS-based channel estimate for ns-3 5G-LENA module | |
| CN110199485A (en) | Wireless device, serving network node and its corresponding method for communicating with one another | |
| WO2021032009A1 (en) | Automatic gain control for serving cell activation based on two different reference signals | |
| CN101577575B (en) | Method, device and system for realizing switching of multiple antenna modes | |
| WO2022129237A1 (en) | Transceiver method between receiver (rx) and transmitter (tx) in an overloaded communication channel | |
| CN101933269B (en) | Wireless communication system, wireless communication apparatus and communication control method | |
| EP4207649A1 (en) | Interference reporting method and apparatus | |
| EP3301824B1 (en) | Communication network component and method for requesting channel information | |
| WO2024002003A1 (en) | Channel feedback model determination method, terminal, and network side device |
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 |