CN109560841A - Extensive mimo system channel estimation methods based on improved distributed compression perception algorithm - Google Patents
Extensive mimo system channel estimation methods based on improved distributed compression perception algorithm Download PDFInfo
- Publication number
- CN109560841A CN109560841A CN201811528866.4A CN201811528866A CN109560841A CN 109560841 A CN109560841 A CN 109560841A CN 201811528866 A CN201811528866 A CN 201811528866A CN 109560841 A CN109560841 A CN 109560841A
- Authority
- CN
- China
- Prior art keywords
- matrix
- perception
- algorithm
- mimo system
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/309—Measuring or estimating channel quality parameters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/391—Modelling the propagation channel
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/0202—Channel estimation
- H04L25/0224—Channel estimation using sounding signals
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L25/00—Baseband systems
- H04L25/02—Details ; arrangements for supplying electrical power along data transmission lines
- H04L25/0202—Channel estimation
- H04L25/024—Channel estimation channel estimation algorithms
- H04L25/0242—Channel estimation channel estimation algorithms using matrix methods
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Power Engineering (AREA)
- Quality & Reliability (AREA)
- Mathematical Physics (AREA)
- Radio Transmission System (AREA)
Abstract
The invention discloses a kind of extensive mimo system channel estimation methods based on improved distributed compression perception algorithm, and in single subdistrict, base station passes through NTRoot antenna sends pilot frequency information, in NRA single-antenna subscriber end receives, the following steps are included: S1, calculating measure vector y in the pilot frequency information that each user receiving end receives, and according to the sparse consistency of extensive mimo system channel, compressed sensing mathematical model is established to pilot information transmission process, and constructs perception matrix Φ;S2, it converts to obtain block structure perception matrix Ψ by block structure, and passes through restructing algorithm and reconstruct block-sparse signal g;S3, sparse signal h is reconstructed using block structure self-adapting compressing sampling matching pursuit algorithm.The present invention utilizes the sparse consistency of time domain of extensive mimo system channel, reconstructs channel impulse response using block structure self-adapting compressing sampling matching pursuit algorithm, and can be estimated and be can be reduced the use of pilot tone in unknown degree of rarefication.
Description
Technical field
The present invention relates to communications channel estimation, in particular relate to a kind of based on improved distributed compression perception algorithm
Extensive mimo system channel estimation methods.
Background technique
With the fast development of wireless communication transmission technique in recent years and the rapid proliferation of smart phone, extensive MIMO
Technology can construct multiple parallel signal transmission passages in transmission antenna and user terminal by it, make full use of space resources,
The availability of frequency spectrum and the rate of information throughput and capacity for effectively raising communication system, have become the key technology of 5G
One of.The invention belongs to communications channel estimation technical fields, and in particular to based on a kind of based on distributed compression perception algorithm
Extensive mimo system channel estimation technique.
In extensive mimo system, firstly, since there are the scatterings of limited quantity and time delay to expand in extended signal space
Exhibition, so the energy of its channel only concentrates on several main paths, energy very little be can be ignored on other paths,
It can regard channel as sparse in the time domain.Secondly as transmitting antenna and user terminal are in same space, base station
Hold antenna alignment close, the distance between each antenna and the distance between transmitting terminal and the receiving end i.e. propagation distance of signal
Very little for comparing, between different transmitting receiving antennas pair, signal can encounter identical scattering, therefore big rule when transmitting
Mould mimo system has spatial coherence, i.e., the channel between different transmission antenna ends and user terminal can show similar
Path delay.Therefore, the channel sparse mode having the same between different transmitting receiving antennas pair, i.e., extensive MIMO
The channel of system has sparse consistency.Even if finally, actual wireless channel still table in the case where very quickly variation
Emersion time correlation, because the path delay of time changes often to change slowly than path gain, that is to say, that even if several
The path gain of adjacent OFDM symbol changes very significant, but the path delay of these continuous OFDM symbols still keeps
Constant, because the correlation time of the path gain of time varying channel and the carrier frequency of system are inversely proportional, and the path delay of time continues
Time is inversely proportional with system bandwidth.Therefore extensive mimo system has temporal correlation, in the correlation in the path delay of time
In, the path delay of time remains unchanged, therefore channel state information is all having the same sparse in R adjacent OFDM symbols
Position, i.e., they share identical supported collection.When existing compressed sensing algorithm is used for extensive mimo system channel estimation
There is no the bulk properties for considering extensive mimo system, so that the number of pilots used when estimation is still very much.
Summary of the invention
Compressed sensing algorithm is applied to pilot number when extensive mimo system down channel is estimated in view of the prior art
Using excessive, the lower problem of estimated accuracy, the invention aims to provide a kind of channel estimation methods, big rule are utilized
Relevant time domain sparse consistency when the sky of mould mimo system channel reconstructs channel impulse using distributed compression perception algorithm
Response, can improve precision of channel estimation while reducing number of pilots.
Technical scheme is as follows:
A kind of extensive mimo system channel estimation methods based on improved distributed compression perception algorithm, when big rule
When mould mimo system has empty when correlation, in single subdistrict, base station passes through NTRoot antenna sends pilot frequency information, in NRA list
Antenna user end receives, it is characterised in that the following steps are included:
S1, each transmission antenna send R adjacent OFDM symbol, the company of receiving at each user's receiving antenna
Continuous R OFDM pilot frequency information constitutes calculation matrix Y, right according to the sparse consistency of extensive mimo system channel temporal and spatial correlations
Pilot information transmission process establishes distributed compression sensing mathematics model, and constructs perception matrix Φ;
S2, it converts to obtain block structure perception matrix Ψ by block structure;
S3, sparse signal H is reconstructed using distributed compression perception algorithm
The present invention proposes the extensive mimo system channel estimation methods based on improved distributed compression perception algorithm,
The algorithm solves distributed compressions perception using more measurement vector patterns and asks according to correlation when the sky of extensive mimo system
Topic estimates the channel state information of R adjacent OFDM symbol simultaneously in receiving end, and selects it mode of supporting block collection
It improves.It is verified by experiment simulation, mentioned algorithm can increase substantially reconstruct success rate, reduce pilot number, improve
Channel estimating performance.It solves and has pilot tone when compressed sensing algorithm is applied to the estimation of extensive mimo system down channel
Quantity uses the problems such as excessive, estimated accuracy is lower.
Detailed description of the invention
For the clearer technical solution for illustrating the embodiment of the present invention or the prior art, below will to embodiment or
Attached drawing needed to be used in the description of the prior art does one and simply introduces, it should be apparent that, the accompanying drawings in the following description is only
It is some embodiments of the present invention, for those of ordinary skill in the art, without creative efforts,
It is also possible to obtain other drawings based on these drawings.
Fig. 1 is flow diagram of the present invention.
Fig. 2 is the transmission flow of the extensive mimo system in the present invention.
Fig. 3 is the channel impulse response block structure conversion process in the present invention.
Fig. 4 is the present invention and other methods mean square error performance comparison under different signal-to-noise ratio.
Fig. 5 is that the present invention is compared with other methods bit error rate performance under different signal-to-noise ratio.
Fig. 6 is that the present invention and other methods are reconstructed into power-performance comparison under different signal-to-noise ratio.
Fig. 7 is the method for the present invention and other methods mean square error performance comparison under different pilot tone occupation rates.
Fig. 8 is that the method for the present invention and other methods bit error rate performance compare under different pilot tone occupation rates.
Fig. 9 is that the method for the present invention and other methods reconstruct success rate estimation performance comparison under different pilot tone occupation rates.
Specific embodiment
To keep the purposes, technical schemes and advantages of the embodiment of the present invention clearer, below with reference to the embodiment of the present invention
In attached drawing, technical solution in the embodiment of the present invention carries out clear and complete description:
A kind of extensive mimo system channel estimation side based on improved distributed compression perception algorithm as shown in Figure 1:
Method, when extensive mimo system has sky when correlation, in single subdistrict, base station passes through NTRoot antenna sends pilot tone letter
Breath, in NRA single-antenna subscriber end receives, it is characterised in that the following steps are included:
S1, each transmission antenna send R adjacent OFDM symbol, the company of receiving at each user's receiving antenna
Continuous R OFDM pilot frequency information constitutes calculation matrix Y, right according to the sparse consistency of extensive mimo system channel temporal and spatial correlations
Pilot information transmission process establishes distributed compression sensing mathematics model, and constructs perception matrix Φ;It specifically includes:
S101, the OFDM symbol with N number of subcarrier is sent in each antenna of base station, the OFDM symbol passes through IFFT
Cyclic prefix CP is added to weaken channel delay spread before each OFDM symbol of output to realize that OFDM is modulated in transformation
The influence of generation is sent to each user through treated ofdm signal after digital-to-analogue conversion in wireless channel
At the antenna at end;
S102, the operation that cyclic prefix CP and FFT transform are removed at jth root receiving antenna, then receiving end receives
Information be
Wherein, X (i) is the frequency-region signal that i-th antenna is sent, and H (i, j) is channel frequency matrix, and n (i, j) is random
Additive white Gaussian noise;
S103, the M position randomly selected in N number of subcarrier are used for transmission frequency pilot sign, then the pilot tone that receiving end receives
Information is
Wherein, pmFor the pilot frequency information of M position of selection,For NTThe M pilot tone letter that root antenna is sent
The sum of number, For the M of corresponding pilot tone position in leaf transformation matrix F in N point discrete Fourier
Capable and channel length preceding L arranges the submatrix constituted, then extensive mimo system mode is
S104, arrangement merging is carried out to above-mentioned mode, obtains compressed sensing model corresponding with channel model:
ThenVector y is as measured, Φ is perception matrix.
S2, it converts to obtain block structure perception matrix Ψ by block structure, specifically include:
S201, the sparse consistency being had according to extensive mimo system, to perception matrix Φ progress and block-sparse signal
Corresponding piece of perception matrixing is converted such as formula:
Ψ(:,(l-1)NT+nt)=Φ (:, (nt- 1) L+l),
Obtain the perception matrix Ψ of block compressed sensing algorithm;
S202, block compressed sensing model is obtained according to the perception matrix Ψ of block compressed sensing algorithm:
Y=Ψ g+n,
S203, the temporal correlation being had according to extensive mimo system channel, R adjacent OFDM symbols have phase
Same sparse mode continuously transmits R adjacent OFDM symbols in base station end, then can indicate in the signal that receiving end receives
Are as follows:
Y=Ψ G+N,
Wherein Y is the matrix for the continuous R pilot frequency information composition that receiving end receives, to obtain distributed compression perception
More measurement vector models reconstruct H by restructing algorithm according to receiving matrix Y and block structure perception matrix Ψ.
S3, sparse signal H is reconstructed using distributed compression perception algorithm, specifically included:
S301, M × (N is extractedTL transformed perception matrix Ψ, M × (N) are tieed upTL) the perception square before the transformation of dimension block structure
The observation vector Y of battle array Φ and M × R dimension, and initialization is reconstructed: initial residual error R0=Y, the number of iterations i=1, it is initial to walk
Long s=1, column serial number indexed set
S302, Ψ is calculatedT×Ri-1, obtain NTThe matrix of L × R dimension, by its every NTRow R column are divided into one group, and L is obtained
A NT× R submatrix, and calculate the F norm of each of which submatrix, i.e.,
A={ Al|Al=| | ΨT(1+NT(l-1):NTl,:)×Ri||F, l=1,2 ..., L },
And traversed after being sorted from large to small to vector A, selection meets all A of A [1]-A [i] < 0.05A [1]
[i] corresponding index value charges to indexed set T;
S303, indexed set J is updated according to indexed set Ti=T ∪ Ji-1;
S304, by indexed set JiIt is extended for block structure and obtains perception array-support collection, i.e.,
Ωi={ JiNT-NT+1:JiNT,
And according to perception array-support collection ΩiThe respective column of selection perception matrix Ψ obtains the submatrix of perception matrix
And obtain index ΩiEstimated valueWhereinMatrix is sought in expressionPseudoinverse;
S305, residual error is updated
S306, judge whether to meet iteration stopping condition: as i < K, i=i+1 continues to execute step 302, otherwise stops
Only iteration executes step S307;
S307, selectionIn preceding K maximum value, and
L value, is logged into Ω corresponding to record value;Ω is expanded are as follows:
Ω2={ Ω+L (nt-1)}nt=1,2 ... NT;
S308, calculating pseudoinverse acquire the H finally estimated:
S309, step 301~step 308 is executed to the information that all receiving antennas receive, obtained estimated result is taken
Union obtains final estimation
It is following that scheme of the present invention is described further and is proved with specific embodiment:
Channel estimation methods of the invention are used for the estimation of single subdistrict down channel.Base station configures NTRoot antenna, NR
A single-antenna user terminal.In this specific embodiment, with NT=32, NR=4 specifically describe.
The invention mainly comprises following two contents: 1) distributed compression sense can be used by converting channel estimation problems to
Know that algorithm solves the problems, such as, establishes more measurement vector compression sensor models, obtain measuring vector model compressed sensing algorithms more
Perceive matrix and calculation matrix;2) channel impulse response is reconstructed using distributed compression perception algorithm.Concrete scheme is as follows:
1. establishing compressed sensing model
Channel impulse response between i-th transmission antenna and j-th of user's receiving antenna is
Wherein, hiFor path gain, τiFor path delay, channel length L, hiMiddle non-zero number is K, K < < L,
It is 256 that channel length is taken in the present embodiment, and it is 6 that non-zero number K, which is 6 i.e. channel degree of rarefication, in channel.
The OFDM symbol with 4096 subcarriers is sent in i-th antenna of transmitting terminal, and carries out IFFT transformation and realizes
OFDM modulation, cyclic prefix CP is added before each OFDM symbol of output to weaken the influence of channel delay spread generation, these
Processed ofdm signal is transmitted at the antenna of each user terminal in wireless channel after digital-to-analogue conversion, in jth root
Receiving antenna is removed cyclic prefix CP and FFT transform.In view of the noise n in channel, then what j-th of user received connects
Receiving symbol is
300 positions on 4096 subcarriers are randomly selected later and place frequency pilot sign, and the transmission process of system is as schemed
Shown in 2, then 300 frequency pilot signs that jth root antenna receives are
Obtain perception matrixAnd calculation matrixWherein pmFor M position of selection
The pilot frequency information set,For NTThe sum of the M pilot signal that root antenna is sent.
Then according to the sparse consistency of channel, such as down conversion is carried out to perception matrix Φ:
Ψ(:,(l-1)NT+nt)=Φ (:, (nt- 1) L+l),
Obtain having the signal of block sparse characteristic to perceive matrix Ψ accordingly, channel impulse response h is transformed to dilute with block
The g of structure is dredged, conversion process is as shown in Figure 3.
It is when signal x can be represented as following form according to block structure compressive sensing theory:
That is x is made of L sub-block, and its nonzero element only occurs in K sub-block, K < < L, then x is block-sparse signal.
According to above-mentioned theory, perception matrix is also divided according to corresponding block structure, it may be assumed that
Wherein,It is the submatrix of M × b, then the compressed sensing mathematical model of block-sparse signal can be expressed as
Higher-dimension block-sparse signal x is projected on a lower dimensional space y by perceiving matrix Φ, then by solving one
A optimization problem can reconstruct original signal from these a small amount of projections with high probability.
Then according to block compressive sensing theory, transmission model can be expressed as
Further according to the temporal correlation that extensive mimo system channel has, R adjacent OFDM symbols are having the same
Sparse mode, therefore continuously transmit R adjacent OFDM symbols in base station end, then it can be indicated in the signal that receiving end receives
Are as follows:
Y=Ψ G+N,
Wherein, Y=[yr,yr+1,…,yr+R-1]∈RM×RFor the signal matrix that receiving end receives,For perception
Matrix,For equivalent channels matrix, N=[nr,nr+1,…,nr+R-1]∈RM×RTo add
Property white Gaussian noise matrix.Y=Ψ G+N is the model that can be typically solved with distributed compression sensing method.For
This model reconstruction G, i.e. solution following formula:
It is using matrix norm expression are as follows:
Wherein, | | X | |0,qNon- zero row number in as X, referred to as l0Pseudonorm.
Solve the problems, such as that this problem is also np hard problem as solution conventional compression sensing reconstructing, reconstruct MMV mode
Compressed sensing reconstruction is also classified into the i.e. convex optimization class algorithm of two classes and greedy class algorithm.For convex optimization class restructing algorithm,
MMV model can be by l unlike SMV model0Norm, which is converted into, solves l1Mould minimization problem, convex optimization problem are relatively difficult to resolve
Certainly, some algorithms are to be translated into Second-order cone programming problem or semi definite programming problem, but their calculation amount is all very
Greatly, it is not suitable for the processing of more data volumes.For greedy class algorithm restructing algorithm, the greedy class restructing algorithm of MMV model be by
What the greedy class algorithm of SMV model was promoted.Therefore the present invention uses MMV model greediness class algorithm reconstruction signal G.
2. reconstructing channel impulse response
Distributed compression perception algorithm is in the progress for the use of two again on the basis of mostly measurement vector compression perception algorithm
It improves: (1) this more measurement compressed sensing algorithm being used to reconstruct block-sparse signal, reduce number of pilots.(2) change atomic block
Selection mode, improve reconstruction accuracy.
Improved restructing algorithm is as follows:
S301, M × (N is extractedTL transformed perception matrix Ψ, M × (N) are tieed upTL) the perception square before the transformation of dimension block structure
The observation vector Y of battle array Φ and M × R dimension, and initialization is reconstructed: initial residual error R0=Y, the number of iterations i=1, it is initial to walk
Long s=1, column serial number indexed set
S302, Ψ is calculatedT×Ri-1, obtain NTThe matrix of L × R dimension, by its every NTRow R column are divided into one group, and L is obtained
A NT× R submatrix, and calculate the F norm of each of which submatrix, i.e.,
A={ Al|Al=| | ΨT(1+NT(l-1):NTl,:)×Ri||F, l=1,2 ..., L },
And traversed after being sorted from large to small to vector A, selection meets all A of A [1]-A [i] < 0.05A [1]
[i] corresponding index value charges to indexed set T;
S303, indexed set J is updated according to indexed set Ti=T ∪ Ji-1;
S304, by indexed set JiIt is extended for block structure and obtains perception array-support collection, i.e.,
Ωi={ JiNT-NT+1:JiNT,
And according to perception array-support collection ΩiThe respective column of selection perception matrix Ψ obtains the submatrix of perception matrix
And obtain index ΩiEstimated valueWhereinMatrix is sought in expressionPseudoinverse;
S305, residual error is updated
S306, judge whether to meet iteration stopping condition: as i < K, i=i+1 continues to execute step 302, otherwise stops
Only iteration executes step S307;
S307, selectionIn preceding K maximum value, and
L value, is logged into Ω corresponding to record value;Ω is expanded are as follows:
Ω2={ Ω+L (nt-1)}nt=1,2 ... NT;
S308, calculating pseudoinverse acquire the H finally estimated:
S309, step 301~step 308 is executed to the information that all receiving antennas receive, obtained estimated result is taken
Union obtains final estimation
3. simulation result
Extensive mimo system is modeled first, according to ITU Vehicular B standard to channel simulator, it is assumed that
Channel is sparse and obeys Rayleigh distribution, NTIt is 32, NRIt is 4, setting channel length is 256, and channel degree of rarefication is 6.
With signal-to-noise ratio (SNR) the i.e. ratio of signal power and noise power, mean square errorIt leads
Frequency utilization rate is as the index for judging estimation performance, and wherein h is ideal former channel impulse response,It is rushed for the channel of estimation
Swash response, n is simulation times.
(1) when number of pilots is fixed, inventive algorithm and non-innovatory algorithm performance comparison
Pilot frequency locations are randomly choosed in emulation, set pilot tone as random positive negative one sequence, number of pilots 300 is right
OMPMMV, Exact_LS and this paper ABOMPMMV algorithm carry out 100 emulation experiments, and are averaged to obtain to experimental result
Following result.
Fig. 4,5,6 are the mean square error (MSE), the bit error rate (BER), reconstruct success rate of these types of algorithm with different noises
Than the simulated effect figure of (SNR) variation.It can be seen from the figure that interchannel noise is fewer and fewer with the increase of SNR, MSE,
BER also becomes smaller therewith, and reconstruct success rate increases with it.Compare ABOMPMMV algorithm and OMP algorithm of the present invention, it can be seen that
When pilot number is 300, correlation is calculated compared to basic OMP when ABOMPMMV algorithm utilizes the sky of extensive mimo system
Method effectively reduces the mean square error under identical signal-to-noise ratio, and the bit error rate simultaneously improves reconstruct success rate;ABOMPMMV is compared to calculate
Method and OMPMMV algorithm, the MSE of ABOMPMMV algorithm under identical signal-to-noise ratio, BER are significantly less than OMPMMV algorithm, and with
The MSE value of Exact_LS is not much different.And significantly improve reconstruct success rate.The ABOMPMMV that this chapter is mentioned is demonstrated to calculate
The selection mode for block-sparse signal reconfiguring and change atomic block that method carries out on the basis of OMPMMV can actually reduce
Mean square error improves estimation performance;Very big channel estimating performance is brought to be promoted.
(2) when signal-to-noise ratio is fixed, the influence of inventive algorithm and non-innovatory algorithm difference number of pilots to estimation performance
In SNR=25, for pilot tone utilization rate and MSE to inventive algorithm, OMP, OMPMMV, EXACT_LS and
Know that the inventive algorithm of degree of rarefication has done experiment simulation.Define pilot tone utilization rate such as following formula
Fig. 7,8,9 are the MSE of these types of algorithm with the change curve of pilot tone utilization rate, it can be seen from the figure that OMP is calculated
Method is due to being SMV model, and MSE and BER are far longer than ABOMPMMV and OMPMMV algorithm, and the performance of estimation is poor.With
The increase of pilot tone occupation rate, number of pilots be increasing, MSE, BER are gradually reduced, reconstruct success rate gradually increase.When leading
When frequency occupation rate is less than 3%, MSE, BER are declined with faster speed, when pilot tone occupation rate is greater than 4%, ABOMPMMV algorithm
With MSE, BER kept stable of OMPMMV, MSE, the BER for reaching ABOMPMMV algorithm after stablizing are less than OMPMMV calculation
Method.As can be seen from Figure 9 the reconstruct success rate of ABOMPMMV algorithm is higher than other two kinds of algorithms, and in pilot tone occupation rate
Former sparse position can be absolutely reconstructed when reaching 3%, even and if OMP and ABOMPMMV algorithm in pilot tone occupation rate
It still cannot absolutely be reconstructed when being 10%.And it can be seen that MMV model uses 5% or so pilot number from simulation result
Mesh can reach the performance boundary of estimation, need 8% number of pilots to can be only achieved stabilization compared to SMV model, the present invention
Algorithm effectively reduces estimation number of pilots used and improves channel estimating performance.
The foregoing is only a preferred embodiment of the present invention, but protection scope of the present invention be not limited to
This, anyone skilled in the art in the technical scope disclosed by the present invention, according to the technique and scheme of the present invention
And its inventive concept is subject to equivalent substitution or change, should be covered by the protection scope of the present invention.
Claims (4)
1. a kind of extensive mimo system channel estimation methods based on improved distributed compression perception algorithm, when extensive
When mimo system has empty when correlation, in single subdistrict, base station passes through NTRoot antenna sends pilot frequency information, in NRA single antenna
User terminal receives, it is characterised in that the following steps are included:
S1, each transmission antenna send R adjacent OFDM symbol, and continuous R are received at each user's receiving antenna
OFDM pilot frequency information constitutes calculation matrix Y, according to the sparse consistency of extensive mimo system channel temporal and spatial correlations, believes pilot tone
Breath transmission process establishes distributed compression sensing mathematics model, and constructs perception matrix Φ;
S2, it converts to obtain block structure perception matrix Ψ by block structure;
S3, sparse signal H is reconstructed using distributed compression perception algorithm.
2. channel estimation methods according to claim 1, it is characterised in that the step S1 is specifically included:
S101, send the OFDM symbol with N number of subcarrier in each antenna of base station, the OFDM symbol by IFFT convert with
It realizes OFDM modulation, and cyclic prefix CP is added to weaken the shadow of channel delay spread generation before each OFDM symbol of output
It rings, is sent to the antenna of each user terminal in wireless channel after digital-to-analogue conversion through treated ofdm signal
Place;
S102, the operation that cyclic prefix CP and FFT transform are removed at jth root receiving antenna, the then letter that receiving end receives
Breath is
Wherein, X (i) is the frequency-region signal that i-th antenna is sent, and H (i, j) is channel frequency matrix, and n (i, j) is random additivity
White Gaussian noise;
S103, the M position randomly selected in N number of subcarrier are used for transmission frequency pilot sign, then the pilot frequency information that receiving end receives
For
Wherein, pmFor the pilot frequency information of M position of selection,For NTRoot antenna send M pilot signal it
With, For the M row and letter of corresponding pilot tone position in leaf transformation matrix F in N point discrete Fourier
The preceding L of road length arranges the submatrix constituted, then extensive mimo system mode is
S104, arrangement merging is carried out to above-mentioned mode, obtains compressed sensing model corresponding with channel model:
ThenVector y is as measured, Φ is perception matrix.
3. channel estimation methods according to claim 2, it is characterised in that the step S2 is specifically included:
S201, the sparse consistency being had according to extensive mimo system carry out perception matrix Φ corresponding with block-sparse signal
Block perceive matrixing, convert such as formula:
Ψ(:,(l-1)NT+nt)=Φ (:, (nt- 1) L+l),
Obtain the perception matrix Ψ of block compressed sensing algorithm;
S202, block compressed sensing model is obtained according to the perception matrix Ψ of block compressed sensing algorithm:
Y=Ψ g+n,
S203, the temporal correlation being had according to extensive mimo system channel, R adjacent OFDM symbols are having the same dilute
The mode of dredging continuously transmits R adjacent OFDM symbols in base station end, then can indicate in the signal that receiving end receives are as follows:
Y=Ψ G+N,
Wherein Y is the matrix for the continuous R pilot frequency information composition that receiving end receives, to obtain the more measurements of distributed compression perception
Vector model reconstructs H by restructing algorithm according to receiving matrix Y and block structure perception matrix Ψ.
4. channel estimation methods according to claim 3, it is characterised in that the step S3 is specifically included:
S301, M × (N is extractedTL transformed perception matrix Ψ, M × (N) are tieed upTL) dimension block structure transformation before perception matrix Φ with
And the observation vector Y of M × R dimension, and initialization is reconstructed: initial residual error R0=Y, the number of iterations i=1, initial step length s=1,
Column serial number indexed set
S302, Ψ is calculatedT×Ri-1, obtain NTThe matrix of L × R dimension, by its every NTRow R column are divided into one group, and L N is obtainedT×R
Submatrix, and calculate the F norm of each of which submatrix, i.e.,
A={ Al|Al=| | ΨT(1+NT(l-1):NTl,:)×Ri||F, l=1,2 ..., L },
And traversed after being sorted from large to small to vector A, it is right that selection meets A [1]-A [i] < 0.05A [1] all A [i]
The index value answered charges to indexed set T;
S303, indexed set J is updated according to indexed set Ti=T ∪ Ji-1;
S304, by indexed set JiIt is extended for block structure and obtains perception array-support collection, i.e.,
Ωi={ JiNT-NT+1:JiNT,
And according to perception array-support collection ΩiThe respective column of selection perception matrix Ψ obtains the submatrix of perception matrixAnd it obtains
Take index ΩiEstimated valueWhereinMatrix is sought in expressionPseudoinverse;
S305, residual error is updated
S306, judge whether to meet iteration stopping condition: as i < K, i=i+1 continues to execute step 302, otherwise stops changing
In generation, executes step S307;
S307, selectionIn preceding K maximum value, and record value
Corresponding l value, is logged into Ω;Ω is expanded are as follows:
Ω2={ Ω+L (nt-1)} nt=1,2 ... NT;
S308, calculating pseudoinverse acquire the H finally estimated:
S309, step 301~step 308 is executed to the information that all receiving antennas receive, obtained estimated result is taken into union,
Obtain final estimation
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811528866.4A CN109560841B (en) | 2018-12-13 | 2018-12-13 | Large-scale MIMO system channel estimation method based on improved distributed compressed sensing algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811528866.4A CN109560841B (en) | 2018-12-13 | 2018-12-13 | Large-scale MIMO system channel estimation method based on improved distributed compressed sensing algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109560841A true CN109560841A (en) | 2019-04-02 |
CN109560841B CN109560841B (en) | 2021-06-15 |
Family
ID=65869834
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811528866.4A Expired - Fee Related CN109560841B (en) | 2018-12-13 | 2018-12-13 | Large-scale MIMO system channel estimation method based on improved distributed compressed sensing algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109560841B (en) |
Cited By (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109995376A (en) * | 2019-04-28 | 2019-07-09 | 哈尔滨工业大学 | Signal reconfiguring method based on joint block sparse model |
CN110519189A (en) * | 2019-08-30 | 2019-11-29 | 东南大学 | Compressed sensing based millimeter wave channel estimation methods under highly mobile scene |
CN111107023A (en) * | 2019-09-23 | 2020-05-05 | 南京邮电大学 | Compressed sensing channel estimation method based on smooth norm in large-scale MIMO |
CN111343730A (en) * | 2020-04-15 | 2020-06-26 | 上海交通大学 | Large-scale MIMO passive random access method under space correlation channel |
CN111698182A (en) * | 2020-05-26 | 2020-09-22 | 武汉大学 | Time-frequency blocking sparse channel estimation method based on compressed sensing |
CN111786703A (en) * | 2020-06-16 | 2020-10-16 | 杭州电子科技大学 | Self-adaptive dual-threshold downlink channel estimation method for large-scale MIMO |
CN111865842A (en) * | 2020-02-11 | 2020-10-30 | 北京邮电大学 | Two-stage low-complexity Massive MIMO channel estimation method, device and equipment |
CN111970216A (en) * | 2020-06-24 | 2020-11-20 | 天津工业大学 | Design method for sparse channel estimation of MIMO-OFDM system based on VAMP algorithm |
CN112383492A (en) * | 2020-11-11 | 2021-02-19 | 中国人民解放军陆军工程大学 | Recursive compressed sensing method and system applied to short-wave OFDM double-selection sky wave channel estimation |
CN112910807A (en) * | 2021-02-04 | 2021-06-04 | 华中科技大学 | Intelligent super-surface channel estimation method and system based on space random sampling |
CN113193895A (en) * | 2021-04-19 | 2021-07-30 | 中国人民解放军陆军工程大学 | Method, system and computer storage medium for acquiring massive MIMO channel state information |
CN113242042A (en) * | 2021-04-08 | 2021-08-10 | 浙江大学 | Sparse channel estimation method based on block parallelization of segmented column correlation matrix |
CN113259278A (en) * | 2021-05-20 | 2021-08-13 | 北京邮电大学 | Millimeter wave multi-antenna channel estimation method and device |
WO2023004563A1 (en) * | 2021-07-27 | 2023-02-02 | Oppo广东移动通信有限公司 | Method for obtaining reference signal and communication devices |
CN118174994A (en) * | 2024-05-14 | 2024-06-11 | 南京信息工程大学 | Ultra-large-scale MIMO communication channel estimation method |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104702390A (en) * | 2015-02-04 | 2015-06-10 | 南京邮电大学 | Pilot frequency distribution method in distributed compressive sensing (DCS) channel estimation |
CN108599820A (en) * | 2018-05-07 | 2018-09-28 | 东北大学 | The extensive mimo system channel estimation methods of matching pursuit algorithm are sampled based on block structure self-adapting compressing |
-
2018
- 2018-12-13 CN CN201811528866.4A patent/CN109560841B/en not_active Expired - Fee Related
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104702390A (en) * | 2015-02-04 | 2015-06-10 | 南京邮电大学 | Pilot frequency distribution method in distributed compressive sensing (DCS) channel estimation |
CN108599820A (en) * | 2018-05-07 | 2018-09-28 | 东北大学 | The extensive mimo system channel estimation methods of matching pursuit algorithm are sampled based on block structure self-adapting compressing |
Non-Patent Citations (2)
Title |
---|
何雪云: "基于压缩感知的无线OFDM信道估计及导频优化研究", 《中国博士学位论文全文数据库信息科技辑》 * |
李在林: "MIMO OFDM无线信道估计算法的研究", 《中国硕士学位论文全文数据库信息科技辑》 * |
Cited By (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109995376B (en) * | 2019-04-28 | 2023-02-03 | 哈尔滨工业大学 | Signal reconstruction method based on joint block sparse model |
CN109995376A (en) * | 2019-04-28 | 2019-07-09 | 哈尔滨工业大学 | Signal reconfiguring method based on joint block sparse model |
CN110519189A (en) * | 2019-08-30 | 2019-11-29 | 东南大学 | Compressed sensing based millimeter wave channel estimation methods under highly mobile scene |
CN110519189B (en) * | 2019-08-30 | 2022-12-09 | 东南大学 | Millimeter wave channel estimation method based on compressed sensing in highly mobile scene |
CN111107023A (en) * | 2019-09-23 | 2020-05-05 | 南京邮电大学 | Compressed sensing channel estimation method based on smooth norm in large-scale MIMO |
CN111107023B (en) * | 2019-09-23 | 2024-02-02 | 南京邮电大学 | Compressed sensing channel estimation method based on smooth norm in large-scale MIMO |
CN111865842A (en) * | 2020-02-11 | 2020-10-30 | 北京邮电大学 | Two-stage low-complexity Massive MIMO channel estimation method, device and equipment |
CN111343730A (en) * | 2020-04-15 | 2020-06-26 | 上海交通大学 | Large-scale MIMO passive random access method under space correlation channel |
CN111343730B (en) * | 2020-04-15 | 2023-08-08 | 上海交通大学 | Large-scale MIMO passive random access method under space correlation channel |
CN111698182A (en) * | 2020-05-26 | 2020-09-22 | 武汉大学 | Time-frequency blocking sparse channel estimation method based on compressed sensing |
CN111786703B (en) * | 2020-06-16 | 2021-09-03 | 杭州电子科技大学 | Self-adaptive dual-threshold downlink channel estimation method for large-scale MIMO |
CN111786703A (en) * | 2020-06-16 | 2020-10-16 | 杭州电子科技大学 | Self-adaptive dual-threshold downlink channel estimation method for large-scale MIMO |
CN111970216A (en) * | 2020-06-24 | 2020-11-20 | 天津工业大学 | Design method for sparse channel estimation of MIMO-OFDM system based on VAMP algorithm |
CN112383492A (en) * | 2020-11-11 | 2021-02-19 | 中国人民解放军陆军工程大学 | Recursive compressed sensing method and system applied to short-wave OFDM double-selection sky wave channel estimation |
CN112383492B (en) * | 2020-11-11 | 2022-07-26 | 中国人民解放军陆军工程大学 | Recursive compressed sensing method and system applied to short-wave OFDM double-selection sky wave channel estimation |
CN112910807A (en) * | 2021-02-04 | 2021-06-04 | 华中科技大学 | Intelligent super-surface channel estimation method and system based on space random sampling |
CN112910807B (en) * | 2021-02-04 | 2022-03-29 | 华中科技大学 | Intelligent super-surface channel estimation method and system based on space random sampling |
CN113242042B (en) * | 2021-04-08 | 2022-07-05 | 浙江大学 | Sparse channel estimation method based on block parallelization of segmented column correlation matrix |
CN113242042A (en) * | 2021-04-08 | 2021-08-10 | 浙江大学 | Sparse channel estimation method based on block parallelization of segmented column correlation matrix |
CN113193895A (en) * | 2021-04-19 | 2021-07-30 | 中国人民解放军陆军工程大学 | Method, system and computer storage medium for acquiring massive MIMO channel state information |
CN113193895B (en) * | 2021-04-19 | 2023-03-10 | 中国人民解放军陆军工程大学 | Method, system and computer storage medium for acquiring massive MIMO channel state information |
CN113259278A (en) * | 2021-05-20 | 2021-08-13 | 北京邮电大学 | Millimeter wave multi-antenna channel estimation method and device |
WO2023004563A1 (en) * | 2021-07-27 | 2023-02-02 | Oppo广东移动通信有限公司 | Method for obtaining reference signal and communication devices |
CN118174994A (en) * | 2024-05-14 | 2024-06-11 | 南京信息工程大学 | Ultra-large-scale MIMO communication channel estimation method |
Also Published As
Publication number | Publication date |
---|---|
CN109560841B (en) | 2021-06-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109560841A (en) | Extensive mimo system channel estimation methods based on improved distributed compression perception algorithm | |
CN108599820A (en) | The extensive mimo system channel estimation methods of matching pursuit algorithm are sampled based on block structure self-adapting compressing | |
CN105656819B (en) | A kind of adaptive channel estimation method based on compressed sensing and extensive MIMO | |
CN105119853B (en) | A kind of extensive mimo channel method of estimation of multi-user based on bayes method | |
KR20070038552A (en) | Iterative channel and interference estimation with dedicated pilot tones for ofdma | |
CN102223327B (en) | SAGE (Space-alternating Generalized Expectation-maximization) based channel estimation method in CoMP multi-user system | |
CN105681232B (en) | A kind of extensive mimo channel method of estimation based on shared channel and compressed sensing | |
CN103685088A (en) | Pilot frequency optimization method of sparse channel, device and channel estimation method | |
CN106341216B (en) | A kind of wireless communication link adaptive approach and uplink downlink adaptive approach | |
CN106911622A (en) | ACO ofdm system channel estimation methods based on compressed sensing | |
CN104348763A (en) | Channel measuring method for large-scale antennas, and user terminal | |
Qiao et al. | Unsourced massive access-based digital over-the-air computation for efficient federated edge learning | |
CN105812299A (en) | Channel estimation algorithm and system of wireless sensor network based on joint block sparse reconstruction | |
CN107181705A (en) | A kind of half-blind channel estimating method and system | |
CN110059401B (en) | OFDM system underwater sound channel impulse response reconstruction method | |
CN103368717A (en) | Method and device for multiple access communication system | |
CN104253639A (en) | Channel quality indicator acquisition method and device | |
Vijayalakshmi et al. | A novel intelligent channel estimation strategy for the 5G wireless communication systems | |
CN103368702B (en) | Fluxion adaptive approach in the distributed interference alignment of WLAN | |
CN101098557A (en) | Signal-to-noise ratio (snr) determination in the frequency domain | |
Kaplan et al. | Message Transmission ThroughUnderspread Time-Varying Linear Channels | |
CN113556159A (en) | Channel feedback method of large-scale MIMO multi-user system | |
CN108418615B (en) | User scheduling method based on MU-MIMO limited feedback system | |
CN103873125A (en) | Downlink signal sending method and downlink signal sending equipment in multiaerial system | |
WO2016095092A1 (en) | Method, apparatus and system for sending and receiving information in wireless local area network |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210615 Termination date: 20211213 |