CN110138459A - Sparse underwater sound orthogonal frequency division multiplexing channel estimation methods and device based on base tracking denoising - Google Patents
Sparse underwater sound orthogonal frequency division multiplexing channel estimation methods and device based on base tracking denoising Download PDFInfo
- Publication number
- CN110138459A CN110138459A CN201910271473.8A CN201910271473A CN110138459A CN 110138459 A CN110138459 A CN 110138459A CN 201910271473 A CN201910271473 A CN 201910271473A CN 110138459 A CN110138459 A CN 110138459A
- Authority
- CN
- China
- Prior art keywords
- sparse
- signal
- model
- orthogonal frequency
- 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.)
- Pending
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B13/00—Transmission systems characterised by the medium used for transmission, not provided for in groups H04B3/00 - H04B11/00
- H04B13/02—Transmission systems in which the medium consists of the earth or a large mass of water thereon, e.g. earth telegraphy
-
- 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
-
- 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
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/26—Systems using multi-frequency codes
- H04L27/2601—Multicarrier modulation systems
- H04L27/2602—Signal structure
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Power Engineering (AREA)
- Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Measurement Of Velocity Or Position Using Acoustic Or Ultrasonic Waves (AREA)
Abstract
The invention belongs to OFDM technical field of underwater acoustic communication, in particular to a kind of sparse underwater sound orthogonal frequency division multiplexing channel estimation methods and device based on base tracking denoising, this method includes: constructing sparse underwater sound orthogonal frequency-division multiplex singal model, extracts condition of sparse channel shock response parameter to be estimated in model；Basis signal model, the non-convex optimization problem representation by sparse signal estimation are that base tracks Denoising Problems model, introduce between regularization parameter and signal norm control error and sparsity and balance in base tracking Denoising Problems model；Base tracking Denoising Problems model is solved, in solution procedure, according to noise, when sparse signal matrix is modified regularization parameter to adapt to noise variation, and is solved the obtained sparse estimated result of channel and gone to handle partially, and final channel estimation results are obtained.The present invention solves the problems such as underwater sound communication channel estimating performance is poor under low signal-to-noise ratio, and estimated accuracy is high, is convenient for signal reconstruction, has stronger practical application value and development prospect.
Description
Technical field
The invention belongs to OFDM technical field of underwater acoustic communication, in particular to a kind of sparse underwater sound based on base tracking denoising is just
Hand over frequency division multiplexing channel estimation methods and device.
Background technique
Underwater acoustic channel is a kind of typical time-varying, frequently change and space-variant channel, this is the proposition of steady high-speed underwater sound communication
Challenge.Compared with traditional carrier wave communication system, orthogonal frequency division multiplexing (orthogonal frequency division
Multiplexing, OFDM) because its higher availability of frequency spectrum, stronger ability of anti-multipath and equaliser structure are easily achieved and
As a research hotspot.OFDM mitigates the inter-carrier interference in underwater sound communication by increasing protection interval (cyclic prefix)
(inter-carrier interference, ICI)) and intersymbol interference (inter-symbol interference, ISI).But
Underwater acoustic channel mostly way is usually tens or even several hundred milliseconds, can not overcome Multi-path interference only by protection interval.With it is traditional
Carrier wave communication system is compared, orthogonal frequency division multiplex OFDM because of its preferable ability of anti-multipath and simple equaliser structure thus
It is widely used [1]-[6] under water.Different from conventional wireless channel, underwater acoustic channel multipath effect is obvious, Doppler
Effect is serious, this proposes challenge for steady high rate communication.In order to overcome the problems, such as more ways in underwater sound ofdm communication, accurately
Ground channel estimation method is essential.
Traditional underwater acoustic channel estimation mainly have channel estimation based on least square (Least Square, LS) criterion and
Channel estimation based on minimum mean square error criterion (Minimum Mean Squared Error, MMSE).Underwater acoustic channel when
Domain and frequency domain all have sparse characteristic, with the development of compressed sensing (Compressed Sensing, CS) theory, condition of sparse channel
Estimation has obtained more and more concerns, can be obtained more preferably using the estimation condition of sparse channel estimation of CS theory than conventional channel estimation
Performance.Channel estimation based on CS theory is broadly divided into three classes: (1) greedy tracing algorithm: for example gradient tracks (Gradient
Pursuits, GP), match tracing (Matching Pursuit, MP), compression sampling match tracing (Compressive
Sampling Matching Pursuit, CoSaMP), orthogonal matching pursuit (Orthogonal Matching Pursuit,
OMP), regularization orthogonal matching pursuit (Regularized Orthogonal Matching Pursuit, ROMP), segmentation just
Match tracing (Stagewise Orthogonal Matching Pursuit, StOMP), subspace is handed over to track (Subpuist
Pursuit, SP), degree of rarefication Adaptive matching tracking (Sparsity Adaptive Matching Pursuit, SAMP) etc.；
(2) convex loose class algorithm, such as base tracking denoising (Basis Pursuit De-Nosing, BPDN), gradient projection
(Gradient Projection for Sparse Reconstruction, GPSR) and iterative threshold algorithm；(3) composite class
Algorithm, such as Fourier's sampling algorithm and HHS (Heavy Hitters on Steroids) tracing algorithm.It is traditional based on LS
The channel estimation of criterion is poor in Low SNR lower channel estimated accuracy to noise-sensitive.Channel based on MMSE criterion is estimated
Meter has preferable performance in practical applications, but needs the correlation properties and noise variance of known channel, and calculation amount compared with
Greatly, to hinder its application.Common algorithm has OMP and BPDN in underwater sound OFDM channel estimation method based on CS theory,
OMP algorithm computational efficiency is higher, but needs the prior information (degree of rarefication) of channel as input, however in practical underwater sound communication
In, the degree of rarefication of channel is often to be difficult to obtain.The channel estimation of BPDN has globally optimal solution, and has good robust
Property.Although BPDN algorithm considers the influence of observation noise, estimate that performance is poor when noise is relatively low, serious shadow
Receiving end demodulation performance is rung.
Summary of the invention
For this purpose, the present invention provide it is a kind of based on base tracking denoising sparse underwater sound orthogonal frequency division multiplexing channel estimation methods and
Device solves the problems, such as that underwater sound communication channel estimating performance is poor equal under low signal-to-noise ratio, and estimated accuracy is high, is convenient for signal weight
Structure.
According to design scheme provided by the present invention, a kind of sparse underwater sound orthogonal frequency division multiplexing letter based on base tracking denoising
Channel estimation method includes following content:
Sparse underwater sound orthogonal frequency-division multiplex singal model is constructed, condition of sparse channel shock response ginseng to be estimated in model is extracted
Number, wherein orthogonal frequency-division multiplex singal includes the synchronization signal of sync section and the symbol of OFDM data section, includes in each code element
Cyclic prefix and symbol data；
Basis signal model, the non-convex optimization problem representation by sparse signal estimation are that base tracks Denoising Problems model, should
It introduces between regularization parameter and signal norm control error and sparsity and balances in base tracking Denoising Problems model；
To base tracking Denoising Problems model solve, in solution procedure, according to noise when sparse signal matrix to just
Then change parameter to be modified to adapt to noise variation, and solve the obtained sparse estimated result of channel and gone to handle partially, obtains
Final channel estimation results.
Above-mentioned, it constructs in sparse underwater sound orthogonal frequency-division multiplex singal model, in orthogonal frequency-division multiplex singal code-element period
It is interior, according to complex information symbol transmitted by subcarrier, obtain the bandpass signal transmitted；It is mended by Doppler's estimation and resampling
It repays, obtains underwater sound multipath channel impulse Response Function；Utilize the pilot signal of impulse Response Function, noise vector and transmitting and receiving
Construct signal model.
Preferably, signal model indicates are as follows: Yp=XpFph+Vp, wherein YpAnd XpRespectively received pilot signal and hair
The pilot signal sent, VpFor noise vector, FpFor the Fourier transform matrix containing weight of pilot frequency；H is channel matrix to be estimated,
It is expressed as h=[h (0), h (1) ... h (L-1)]T, h [l] is first of tap coefficient of channel impulse response, and L is channel length.
Above-mentioned, according to compressive sensing theory, the degree of rarefication of sparse signal is defined as nonzero element number, utilizes signal
Norm expression, the non-convex optimization problem representation of sparse signal x estimation are as follows: min | | x | |0, s.t.y=Ax, wherein y is observation square
Battle array, | | x | |0For the degree of rarefication of sparse signal x, A ∈ RN×M, the sub-carrier signal matrix M expression that R expression size is N × M is sub to be carried
Wave number,NIndicate OFDM symbol.
Preferably, consider noisy situation, be that base tracking denoising is asked by non-convex optimization problem representation using lagrange formula
Model is inscribed, is indicated are as follows:Wherein, normλ is regularization parameter.
Preferably, during problem model solves, real-time underwater acoustic channel is reconstructed using SpaRSA algorithm, in the restructuring procedure, by base
Track Denoising Problems model again are as follows:Wherein,
It is iterative solution model by the problem model conversation after rewriting, that is, converts are as follows:
Wherein,αtFor constant,For f (xt) differentiate.
Further, it iteratively solves in model, regularization parameter is modified with adapting to noise variation, correction formula table
It is shown as:SNR indicates signal-to-noise ratio, | | x | |∞Indicate maximum absolute value in amount of orientation x
Element.
Further, it iteratively solves in model, the sparse estimated result of channel obtained to solution goes to handle partially, will weigh first
The item for being zero in the sparse estimated result of structure channel is fixed as zero, carries out minimum solution, target to objective function in residual term
Function minimization, which solves, to be indicated are as follows:Wherein, I is the set of nonzero term in sparse estimated result, xIFor
From in x reject be zero after matrix, AgITo be matrix after zero column from rejecting in A, optimal solution is indicated are as follows: xI=
(AgI TAgI)-1AgI Ty；Final channel estimation results are obtained according to the optimal solution.
A kind of sparse underwater sound orthogonal frequency division multiplexing channel estimating apparatus based on base tracking denoising, includes model construction mould
Block, problem conversion module and problem solver module, wherein
Model construction module is extracted to be estimated dilute in model for constructing sparse underwater sound orthogonal frequency-division multiplex singal model
Dredge channel impulse response parameter, wherein orthogonal frequency-division multiplex singal includes the synchronization signal of sync section and the code of OFDM data section
Member includes cyclic prefix and symbol data in each code element；
Problem conversion module is used for basis signal model, and the non-convex optimization problem representation that sparse signal is estimated is chased after for base
Track Denoising Problems model, which, which tracks, introduces regularization parameter and signal norm control error and sparsity in Denoising Problems model
Between balance；
Problem solver module, for being solved to base tracking Denoising Problems model, in solution procedure, when according to noise
Sparse signal matrix is modified regularization parameter to adapt to noise variation, and solve the obtained sparse estimated result of channel into
Row goes to handle partially, obtains final channel estimation results.
Beneficial effects of the present invention:
The present invention initially sets up sparse underwater sound OFDM channel estimation model, provides the regularization ginseng for being adapted to noise variation
Number promotes the adaptability to the channel estimation under different signal-to-noise ratio；It after the sparse solution reconstructed, is gone to handle partially, be promoted
The precision of channel estimation；Under the premise of not needing channel degree of rarefication, complete to believe the sparse underwater sound OFDM under different signal-to-noise ratio
Road estimation, promotes the precision of channel estimation, provides reliable for subsequent balanced, demodulation as a result, performance is stable, it is efficient to run,
With stronger practical application value and development prospect.
Detailed description of the invention:
Fig. 1 is channel estimation methods flow diagram in embodiment；
Fig. 2 is ofdm signal model schematic in embodiment；
Fig. 3 is channel estimating apparatus schematic diagram in embodiment；
Fig. 4 is sound velocity profile in embodiment；
Fig. 5 is that Bellhop normalizes impulse response in embodiment；
Fig. 6 is that lower mean square error performance comparison is arranged in different regularization parameters in embodiment；
Fig. 7 is not go SpaRSA channel estimation in inclined situation in embodiment；
Fig. 8 is to go SpaRSA channel estimation under inclined disposition in embodiment；
Fig. 9 is channel estimation methods mean square error performance curve comparison in embodiment；
Figure 10 is channel estimation methods bit error rate performance curve comparison in embodiment.
Specific embodiment:
The present invention is described in further detail with technical solution with reference to the accompanying drawing, and detailed by preferred embodiment
Describe bright embodiments of the present invention in detail, but embodiments of the present invention are not limited to this.
In recent years, OFDM underwater sound communication has become a research hotspot, compared with traditional carrier wave communication system,
Orthogonal frequency division multiplexing (Orthogonal Frequency Division Multiplexing, OFDM) is because of its preferable anti-multipath
Ability and simple equaliser structure are thus widely used under water.In the embodiment of the present application, provide a kind of based on base
The sparse underwater sound orthogonal frequency division multiplexing channel estimation methods of denoising are tracked, include following content:
Sparse underwater sound orthogonal frequency-division multiplex singal model is constructed, condition of sparse channel shock response ginseng to be estimated in model is extracted
Number, wherein orthogonal frequency-division multiplex singal includes the synchronization signal of sync section and the symbol of OFDM data section, includes in each code element
Cyclic prefix and symbol data；
Basis signal model, the non-convex optimization problem representation by sparse signal estimation are that base tracks Denoising Problems model, should
It introduces between regularization parameter and signal norm control error and sparsity and balances in base tracking Denoising Problems model；
To base tracking Denoising Problems model solve, in solution procedure, according to noise when sparse signal matrix to just
Then change parameter to be modified to adapt to noise variation, and solve the obtained sparse estimated result of channel and gone to handle partially, obtains
Final channel estimation results.
By being adapted to the regularization parameter of noise variation, the adaptability to the channel estimation under different signal-to-noise ratio is promoted；
After the sparse solution reconstructed, is gone to handle partially, promote the precision of channel estimation；In the premise for not needing channel degree of rarefication
Under, it completes to promote the precision of channel estimation to the sparse underwater sound OFDM channel estimation under different signal-to-noise ratio, is subsequent equilibrium,
Demodulation provides reliable result.
It constructs in sparse underwater sound orthogonal frequency-division multiplex singal model, in further embodiment of the present invention, in orthogonal frequency division multiplexing
With in the signal element period, according to complex information symbol transmitted by subcarrier, the bandpass signal transmitted is obtained；Estimated by Doppler
Meter and resampling compensation, obtain underwater sound multipath channel impulse Response Function；It is connect using impulse Response Function, noise vector and transmission
The pilot signal of receipts constructs signal model.
It is shown in Figure 2, it is considered as the OFDM letter of insertion cyclic prefix (Cyclic Prefix, CP) in protection interval
Number, if CP time span is Tg, symbol time length is T, then the time span in OFDM symbol period is T'=T+Tg, subcarrier
Between be divided into Δ f=1/T, the frequency of k-th of subcarrier are as follows:
fk=fc+ k Δ f, k=-K/2 ..., K/2-1 (1)
In formula, fcFor carrier frequency, sub-carrier number K, then bandwidth B=K Δ f.
An OFDM symbol cycle T ' in, indicate complex information symbol transmitted on k-th of subcarrier with d [k], then by
The bandpass signal of transmission are as follows:
Wherein g (t)=1, t ∈ [0, T], otherwise g (t)=0.Consider that the impulse response of underwater sound multipath channel can use following formula table
Show:
In formula: L is tap sum, AL(t) and τL(t) be respectively l-th tap amplitude and time delay.Assuming that all taps
Doppler factor it is identical, it may be assumed that
τL(t)≈τL-at (4)
And in an OFDM symbol cycle T ' interior, AL(t), τL(t) and a perseverance is constant.Then by Doppler's estimation and again
After sampling technique compensation, underwater acoustic channel be can simplify are as follows:
Assuming that then receiving pilot signal using frequency domain dressing pilot tone are as follows:
Yp=XpHp+Vp (6)
Wherein, p ∈ Sp, SpFor the set of pilot sub-carrier, YpAnd XpThe respectively pilot signal that sends and receivees of frequency domain, Vp
For noise vector, HpFor the Fourier transformation of channel impulse response, indicate are as follows:
Wherein, N is Fourier transformation points, and h [l] is the tap coefficient of channel impulse response, and channel length L is indicated
Are as follows:
H=[h (0), h (1) ... h (L-1)]T (8)
HpIt can indicate are as follows:
Hp=Fph (9)
FpFor the Fourier transform matrix containing weight of pilot frequency, formula (9) are substituted into formula (6), can be obtained:
Yp=XpFph+Vp (10)
In formula (10), YpIt can indicate observing matrix, channel matrix h to be estimated, using the channel estimation based on BPDN
Algorithm estimates condition of sparse channel impulse response.
The degree of rarefication of signal x is defined as the number namely l of nonzero element0Norm, the then degree of rarefication of signal x is defined as:
K=| | x | |0 (11)
CS theory solves the problems, such as to be exactly the estimation to sparse signal x, is indicated with mathematical formulae are as follows:
min||x||0, s.t.y=Ax (12)
Wherein, y ∈ RN×1For observing matrix, x ∈ RM×1For the sparse signal to be estimated, A ∈ RN×MFor sensing matrix, wherein
R is set of real numbers, and N is the dimension of observing matrix, and M is the dimension of signal to be estimated and meets K < N < < M.On.The above problem
The problem of being a non-convex optimization, directly solving this problem is non-deterministic NP problem.L under certain condition1Norm and l0
Norm can be of equal value, therefore formula (12) is converted to formula (13), l1Norm minimum useable linear programming technique solves.
min||x||1, s.t.y=Ax (13)
Wherein,It is defined as the l of signal x1Norm considers in noisy situation, using lagrange formula, on
Stating optimization problem may be expressed as:
Formula (14) is base tracking Denoising Problems, and wherein λ is regularization parameter, is controlled flat between error and sparsity
Weighing apparatus.
Base tracking noise reduction (Basis Pursuit Denoising:BPDN) restructing algorithm in have l1_ls, SpaRSA,
YALL1 scheduling algorithm, wherein SpaRSA algorithm has lower complexity and higher reconstruction accuracy, is suitable for real-time underwater acoustic channel
Estimation.Formula (14) is rewritable are as follows:
WhereinC (x)=| | x | |1, the problem of formula (15), is converted to the iteration of following point:
Wherein, αtFor the t times iteration choose step-length,For f (xt) differentiate,
When c (x) is represented by following form:
Formula (16) may be expressed as:
When c (z)=| | z | |1When, formula (19) minimum value is given by:
Wherein
It is defined as soft-threshold function.
For SpaRSA algorithm, the setting of regularization parameter λ can determine its performance, when white Gaussian noise variance is 10-4When,
Select λ=0.1 | | ATy||∞, which cannot well adapt to the variation of noise.For this problem, another reality of the invention
It applies and a kind of regularization parameter setting method of variation for being adapted to noise is provided in example, by setting the regularization parameter toSNR indicates signal-to-noise ratio, wherein | | x | |∞For the element of maximum absolute value in vector x.
Due to introducing l1Norm, the sparse solution amplitude for causing the sparse solution for obtaining channel by SpaRSA algorithm to estimate is less than normal, this hair
In another bright embodiment, handled partially by going, first by be zero in SpaRSA reconstruction result item be fixed as zero, then surplus
Minimum solution is carried out to objective function in remainder, that is, is exactly optimized-type (22).
Wherein, I is the set of nonzero term in sparse solution, xIFor from x reject be zero after matrix, AgITo be picked from A
Except for the matrix after zero column.The optimal solution of formula (22) may be expressed as:
xI=(AgI TAgI)-1AgI Ty (23)
Final channel estimation results are obtained according to the optimal solution.
Based on above-mentioned channel estimation methods, the embodiment of the present invention also provides a kind of sparse underwater sound based on base tracking denoising
Orthogonal frequency division multiplexing channel estimating apparatus, it is shown in Figure 3, comprising model construction module 101, problem conversion module 102 and ask
Topic solves module 103, wherein
Model construction module 101 is extracted to be estimated in model for constructing sparse underwater sound orthogonal frequency-division multiplex singal model
Condition of sparse channel shock response parameter, wherein orthogonal frequency-division multiplex singal includes the synchronization signal and OFDM data section of sync section
Symbol includes cyclic prefix and symbol data in each code element；
Problem conversion module 102 is used for basis signal model, and the non-convex optimization problem representation by sparse signal estimation is base
Denoising Problems model is tracked, which, which tracks, introduces regularization parameter and signal norm control error and sparse in Denoising Problems model
It is balanced between property；
Problem solver module 103, for being solved to base tracking Denoising Problems model, in solution procedure, according to noise
When sparse signal matrix is modified regularization parameter to adapt to noise variation, and solves the sparse estimation knot of obtained channel
Fruit is gone to handle partially, obtains final channel estimation results.
For the validity for verifying technical solution of the present invention, explanation is further explained below by emulation experiment:
To investigate influence of the regularization parameter for channel estimating performance, emulated at Matlab 2015a.Emulation
Channel is generated by Bellhop ray model, and the horizontal distance of transmitter and receiver is 500m, mean depth 100m, transmitting
Machine and receiver are all placed in underwater 10m, and transmitted frequency of sound wave is 15kHz, and sound ray number is 10.Emulate signal parameter are as follows:
OFDM carrier frequency is 24kHz, bandwidth 12kHz, cyclic prefix 25ms, symbol lengths 85.3ms, shares 1024 sons
Carrier wave, using the convolutional encoding of 1/2 code rate, QPSK constellation mapping.Fig. 4 is the sound velocity profile of Bellhop mode input, the velocity of sound
Section is that the measured data at 9838 station of the Taiwan Straits is passed through obtained by cubic interpolation, as shown, the velocity of sound is in [0m, -40m] interior table
It is now strong negative gradient characteristic.Fig. 5 is normalized Bellhop impulse response, and channel tap distribution has apparent sparse
Characteristic, the channel are a typical condition of sparse channel.
1 difference regularization parameter value MSE of table is compared with BER performance
Table 1 is regularization parameter that will be given under 2dB signal-to-noise ratio, and amplification and MSE and BER when reducing one times, two times
Performance comparison, as it can be seen from table 1 on the one hand, the setting of regularization parameter can determine channel estimating performance and bit error rate performance,
Under the signal-to-noise ratio of 2dB, when the increase of regularization parameter value is arranged, the numerical value of MSE and BER first reduce to be increased afterwards, wherein setting
Setting regularization parameter isWhen can obtain optimal MSE and BER performance, be more than or lower than the ginseng
When number, MSE and BER performance can sharply deteriorate.On the other hand, the MSE performance of channel estimation influences whether BER performance, and channel is estimated
The MSE of meter is smaller, and demodulation performance is better.Therefore it selects suitable regularization parameter to be affected Demodulation Systems performance, adopts
It is arranged with the regularization parameter of this patentIt is carried out pair from the MSE performance of different regularization parameters
Than.λ=0.1 is set by regularization parameter respectively | | ATy||∞, the σ of λ=0.12||ATy||∞With
Wherein σ2It is defined as normalization (when the power of signal is 1) noise variance.Mean square error (Mean Square Error, MSE) is fixed
Justice are as follows:
Wherein H is Bellhop simulated channel frequency domain response,To estimate channel frequency domain response.Using channel estimation in frequency domain
Model uses Comb Pilot when emulation, and pilot sub-carrier is uniformly distributed, and is divided into 6, keeps other conditions identical, carries out channel
Estimation.The mean square error performance curve obtained under different signal-to-noise ratio is as shown in Figure 6.From fig. 6 it can be seen that believing in 0~10dB
It makes an uproar than in range, with the increase of signal-to-noise ratio, the channel estimating performance under three kinds of regularization parameters is increased accordingly.In documents
Parameter setting compared to parameter constant when can MSE performance boost about 1dB, technical solution is compared to existing in the embodiment of the present invention
Scheme can promote MSE performance about 1dB.Regularization parameter is better than regularization parameter with channel estimating performance when signal-to-noise ratio changes
Performance when constant, this is because regularization parameter controls the balance between sparse solution and error, in order to adapt to different letters
It makes an uproar and compares, which should reduce with the increase of signal-to-noise ratio, more accurately estimate performance to obtain.Therefore implemented using the present invention
Regularization parameter setting can effectively promote channel estimating performance after technical solution amendment in example, and then improve the solution of system
Tonality energy.
It goes to handle the influence for channel estimating performance partially to investigate, channel is using the Bellhop channel in emulation one, just
Then changing parameter is the regularization parameter emulated after correcting in one, compares do not go partially and go inclined disposition lower channel to estimate respectively
Performance, in two kinds of processing, pilot sub-carrier is that pectination is uniformly distributed, pilot interval 6, signal-to-noise ratio 2dB, other conditions
Keep identical.Channel result is as shown in FIG. 7 and 8, does not go respectively partially and removes inclined disposition lower channel estimated result, You Tuzhong
It can be seen that do not go under inclined disposition, practical tap is compared, estimating to obtain tap, to show as amplitude less than normal, this is because
L is introduced in SpaRSA channel estimation1Norm causes not going the amplitude for obtaining sparse solution in inclined situation less than normal；It goes to handle partially and go
In the case of, the amplitude for the tap estimated is improved, and has been carried out double optimization to sparse solution this is because going to handle partially, has been made
It is more accurate to obtain sparse solution.In conclusion go it is partially larger for the performance boost of channel estimation, using go partially handle can improve
The performance of channel estimation.
In order to compare the performance of channel estimation method Yu LS, SAMP, OMP, SpaRSA algorithm, pass through Monte Carlo experiment ratio
Compared with the channel estimating performance and demodulation performance of these algorithms, corrected in above-mentioned emulation experiment wherein improved SpaRSA algorithm uses
Regularization parameter and go to handle partially.In emulation experiment, pilot sub-carrier is that pectination is uniformly distributed, pilot interval 3,
His condition is identical.Fig. 9 and 10 is respectively the mean square error performance and bit error rate performance correlation curve of different channels algorithm for estimating, by
Figure can be seen that the increase with signal-to-noise ratio, MSE and BER performance gradually improves, the channel estimation method of base and compressed sensing
It can be substantially better than traditional LS channel estimation method, in compressed sensing based channel estimation method, improved SpaRSA letter
Channel estimation algorithm is better than SpaRSA algorithm；SpaRSA channel estimation method is better than OMP channel estimation method；OMP channel estimation is calculated
Method is better than SAMP channel estimation method.As it can be seen that the performance of improved SpaRSA channel estimation method is substantially better than other algorithms, when
The bit error rate is 10-2When, improved SpaRSA channel estimation method received signal to noise ratio is about 6dB, and SpaRSA channel estimation method
Received signal to noise ratio is about 7dB.In 0~12dB SNR ranges, compare the ber curve of the two as it can be seen that improved SpaRSA
Channel estimation method can promote the BER performance of about 1dB compared to SpaRSA channel estimation method.
In the present invention, it is based on the sparse underwater sound OFDM channel estimation of BPDN, is estimated by initially setting up sparse underwater sound OFDM channel
Model is counted, regularization parameter is analyzed and goes to handle the influence to BPDN channel estimation method performance partially；And it is real by specifically emulation
The performance of method in more existing LS, OMP, BPDN and the embodiment of the present invention is tested, theory analysis and simulation result show the present invention
Scheme has more preferably estimated accuracy in embodiment, compared with base tracks denoising channel estimation, in underwater sound OFDM channel estimation
The mean square error performance gain of about 1dB can be improved, there is biggish practical value and application prospect.
Each embodiment in this specification is described in a progressive manner, the highlights of each of the examples are with other
The difference of embodiment, the same or similar parts in each embodiment may refer to each other.For device disclosed in embodiment
For, since it is corresponded to the methods disclosed in the examples, so being described relatively simple, related place is said referring to method part
It is bright.
The unit and method and step of each example described in conjunction with the examples disclosed in this document, can with electronic hardware,
The combination of computer software or the two is realized, in order to clearly illustrate the interchangeability of hardware and software, in above description
In generally describe each exemplary composition and step according to function.These functions are held with hardware or software mode
Row, specific application and design constraint depending on technical solution.Those of ordinary skill in the art can be to each specific
Using using different methods to achieve the described function, but this realization be not considered as it is beyond the scope of this invention.
Those of ordinary skill in the art will appreciate that all or part of the steps in the above method can be instructed by program
Related hardware is completed, and described program can store in computer readable storage medium, such as: read-only memory, disk or CD
Deng.Optionally, one or more integrated circuits also can be used to realize, accordingly in all or part of the steps of above-described embodiment
Ground, each module/unit in above-described embodiment can take the form of hardware realization, can also use the shape of software function module
Formula is realized.The present invention is not limited to the combinations of the hardware and software of any particular form.
The foregoing description of the disclosed embodiments makes professional and technical personnel in the field can be realized or use the application.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the application.Therefore, the application
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (10)
1. a kind of sparse underwater sound orthogonal frequency division multiplexing channel estimation methods based on base tracking denoising, which is characterized in that comprising such as
Lower content:
Sparse underwater sound orthogonal frequency-division multiplex singal model is constructed, condition of sparse channel shock response parameter to be estimated in model is extracted,
In, orthogonal frequency-division multiplex singal includes the synchronization signal of sync section and the symbol of OFDM data section, includes circulation in each code element
Prefix and symbol data；
Basis signal model, the non-convex optimization problem representation by sparse signal estimation are that base tracks Denoising Problems model, which chases after
It introduces between regularization parameter and signal norm control error and sparsity and balances in track Denoising Problems model；
To base tracking Denoising Problems model solve, in solution procedure, according to noise when sparse signal matrix to regularization
Parameter is modified to adapt to noise variation, and is solved the obtained sparse estimated result of channel and gone to handle partially, is obtained final
Channel estimation results.
2. the sparse underwater sound orthogonal frequency division multiplexing channel estimation methods according to claim 1 based on base tracking denoising,
It is characterized in that, constructs in sparse underwater sound orthogonal frequency-division multiplex singal model, in orthogonal frequency-division multiplex singal code-element period, foundation
Complex information symbol transmitted by subcarrier obtains the bandpass signal transmitted；By Doppler's estimation and resampling compensation, water is obtained
Sound multipath channel impulse Response Function；Signal is constructed using the pilot signal of impulse Response Function, noise vector and transmitting and receiving
Model.
3. the sparse underwater sound orthogonal frequency division multiplexing channel estimation methods according to claim 2 based on base tracking denoising,
It is characterized in that, signal model indicates are as follows: Yp=XpFph+Vp, wherein YpAnd XpRespectively received pilot signal is led with what is sent
Frequency signal, VpFor noise vector, FpFor the Fourier transform matrix containing weight of pilot frequency；H is channel matrix to be estimated, is expressed as h
=[h (0), h (1) ... h (L-1)]T, h [l] is first of tap coefficient of channel impulse response, and L is channel length.
4. the sparse underwater sound orthogonal frequency division multiplexing channel estimation methods according to claim 1 based on base tracking denoising,
It is characterized in that, according to compressive sensing theory, the degree of rarefication of sparse signal is defined as nonzero element number, utilizes signal norm table
Show, the non-convex optimization problem representation of sparse signal x estimation are as follows: min | | x | |0, s.t.y=Ax, wherein and y is observing matrix, | | x
||0For the degree of rarefication of sparse signal x, A ∈ RN×M, the sub-carrier signal matrix M expression sub-carrier number that R expression size is N × M, N
Indicate OFDM symbol.
5. the sparse underwater sound orthogonal frequency division multiplexing channel estimation methods according to claim 4 based on base tracking denoising,
It is characterized in that, considers noisy situation, be that base tracks Denoising Problems mould by non-convex optimization problem representation using lagrange formula
Type indicates are as follows:Wherein, normλ is regularization parameter.
6. the sparse underwater sound orthogonal frequency division multiplexing channel estimation methods according to claim 4 based on base tracking denoising,
It is characterized in that, during problem model solves, reconstructs real-time underwater acoustic channel using SpaRSA algorithm, in the restructuring procedure, base is tracked
Denoising Problems model is again are as follows:Wherein,
It is iterative solution model by the problem model conversation after rewriting, that is, converts are as follows:
Wherein,αtFor constant, ▽ f (xt) it is f (xt) differentiate.
7. the sparse underwater sound orthogonal frequency division multiplexing channel estimation methods according to claim 6 based on base tracking denoising,
It is characterized in that, iteratively solves in model, regularization parameter is modified with adapting to noise variation, correction formula indicates are as follows:SNR indicates signal-to-noise ratio, | | x | |∞Indicate the element of maximum absolute value in amount of orientation x.
8. the sparse underwater sound orthogonal frequency division multiplexing channel estimation methods according to claim 6 based on base tracking denoising,
It is characterized in that, iteratively solve in model, the sparse estimated result of channel obtained to solution goes to handle partially, and it is dilute to reconstruct channel first
It dredges the item for being zero in estimated result and is fixed as zero, minimum solution is carried out to objective function in residual term, objective function is minimum
Changing to solve indicates are as follows:Wherein, I is the set of nonzero term in sparse estimated result, xITo be rejected from x
It is the matrix after zero, AgITo be matrix after zero column from rejecting in A, optimal solution is indicated are as follows: xI=(AgITAgI)-
1AgI Ty；Final channel estimation results are obtained according to the optimal solution.
9. a kind of sparse underwater sound orthogonal frequency division multiplexing channel estimating apparatus based on base tracking denoising, which is characterized in that include mould
Type constructs module, problem conversion module and problem solver module, wherein
Model construction module extracts sparse letter to be estimated in model for constructing sparse underwater sound orthogonal frequency-division multiplex singal model
Road shock response parameter, wherein the symbol of the orthogonal frequency-division multiplex singal synchronization signal comprising sync section and OFDM data section, often
It include cyclic prefix and symbol data in a symbol；
Problem conversion module is used for basis signal model, and the non-convex optimization problem representation by sparse signal estimation is that base tracking is gone
It makes an uproar problem model, which, which tracks, introduces in Denoising Problems model between regularization parameter and signal norm control error and sparsity
Balance；
Problem solver module is when sparse according to noise in solution procedure for solving to base tracking Denoising Problems model
Signal matrix is modified regularization parameter to adapt to noise variation, and solves the obtained sparse estimated result of channel and gone
Processing partially, obtains final channel estimation results.
10. the sparse underwater sound orthogonal frequency division multiplexing channel estimating apparatus according to claim 9 based on base tracking denoising,
It is characterized in that, in model construction module, in orthogonal frequency-division multiplex singal code-element period, according to the symbol of complex information transmitted by subcarrier
Number, obtain the bandpass signal transmitted；By Doppler's estimation and resampling compensation, underwater sound multipath channel shock response letter is obtained
Number；Signal model is constructed using the pilot signal of impulse Response Function, noise vector and transmitting and receiving.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910271473.8A CN110138459A (en) | 2019-04-04 | 2019-04-04 | Sparse underwater sound orthogonal frequency division multiplexing channel estimation methods and device based on base tracking denoising |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910271473.8A CN110138459A (en) | 2019-04-04 | 2019-04-04 | Sparse underwater sound orthogonal frequency division multiplexing channel estimation methods and device based on base tracking denoising |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110138459A true CN110138459A (en) | 2019-08-16 |
Family
ID=67569431
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910271473.8A Pending CN110138459A (en) | 2019-04-04 | 2019-04-04 | Sparse underwater sound orthogonal frequency division multiplexing channel estimation methods and device based on base tracking denoising |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110138459A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110661735A (en) * | 2019-10-17 | 2020-01-07 | 哈尔滨工程大学 | Time delay estimation method based on orthogonal matching pursuit sparse channel estimation |
| CN111427046A (en) * | 2020-03-27 | 2020-07-17 | 西安交通大学 | Terahertz pulse echo positioning method for improving detection precision |
| CN113271196A (en) * | 2021-04-07 | 2021-08-17 | 北京邮电大学 | Short guard interval orthogonal frequency division multiplexing wireless transmission method |
| CN114124246A (en) * | 2021-11-12 | 2022-03-01 | 厦门大学 | Underwater acoustic multiple access communication method and system based on multi-frequency airspace sparse reconstruction |
| CN114584433A (en) * | 2022-02-24 | 2022-06-03 | 哈尔滨工程大学 | Method for detecting synchronous signal in multi-path channel under impulse noise environment |
| CN115865574A (en) * | 2022-11-24 | 2023-03-28 | 山东大学 | Sparse channel estimation method and system based on structure detection |
| CN117395104A (en) * | 2023-12-13 | 2024-01-12 | 为准（北京）电子科技有限公司 | Channel estimation method and device in orthogonal frequency division multiplexing system |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110261905A1 (en) * | 2010-04-27 | 2011-10-27 | Qualcomm Incorporated | Compressed sensing channel estimation in ofdm communication systems |
| CN102624666A (en) * | 2012-03-05 | 2012-08-01 | 哈尔滨工程大学 | Cyclic coding method of multi-channel transceiving orthogonal multi-carrier underwater acoustic communication in sparse channel model |
| CN103792000A (en) * | 2014-02-20 | 2014-05-14 | 苏州大学 | Method and device for detecting transient components in signal based on sparse representation |
| CN104780128A (en) * | 2015-04-14 | 2015-07-15 | 哈尔滨工程大学 | Sparse channel estimation and pilot optimization method for underwater sound OFDMA uplink communication |
| US20150382237A1 (en) * | 2014-06-27 | 2015-12-31 | Electronics And Telecommunications Research Institute | Network apparatus based on orthogonal frequency-division multiplexing (ofdm) and data compression and data recovery method thereof using compressed sensing |
| CN106027445A (en) * | 2016-05-13 | 2016-10-12 | 江苏科技大学 | Channel estimation method for sparse property of underwater acoustic block structure |
| CN106199524A (en) * | 2016-06-23 | 2016-12-07 | 中国科学院电子学研究所 | Far field WB-RCS data acquisition and the compression method of denoising is followed the trail of based on base |
| CN106302298A (en) * | 2016-09-27 | 2017-01-04 | 哈尔滨工程大学 | A kind of method eliminating OFDM underwater sound communication system clipped noise |
| CN109361631A (en) * | 2018-10-15 | 2019-02-19 | 中国人民解放军战略支援部队信息工程大学 | Degree of rarefication unknown underwater sound orthogonal frequency division multiplexing channel estimation methods and device |
-
2019
- 2019-04-04 CN CN201910271473.8A patent/CN110138459A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20110261905A1 (en) * | 2010-04-27 | 2011-10-27 | Qualcomm Incorporated | Compressed sensing channel estimation in ofdm communication systems |
| CN102624666A (en) * | 2012-03-05 | 2012-08-01 | 哈尔滨工程大学 | Cyclic coding method of multi-channel transceiving orthogonal multi-carrier underwater acoustic communication in sparse channel model |
| CN103792000A (en) * | 2014-02-20 | 2014-05-14 | 苏州大学 | Method and device for detecting transient components in signal based on sparse representation |
| US20150382237A1 (en) * | 2014-06-27 | 2015-12-31 | Electronics And Telecommunications Research Institute | Network apparatus based on orthogonal frequency-division multiplexing (ofdm) and data compression and data recovery method thereof using compressed sensing |
| CN104780128A (en) * | 2015-04-14 | 2015-07-15 | 哈尔滨工程大学 | Sparse channel estimation and pilot optimization method for underwater sound OFDMA uplink communication |
| CN106027445A (en) * | 2016-05-13 | 2016-10-12 | 江苏科技大学 | Channel estimation method for sparse property of underwater acoustic block structure |
| CN106199524A (en) * | 2016-06-23 | 2016-12-07 | 中国科学院电子学研究所 | Far field WB-RCS data acquisition and the compression method of denoising is followed the trail of based on base |
| CN106302298A (en) * | 2016-09-27 | 2017-01-04 | 哈尔滨工程大学 | A kind of method eliminating OFDM underwater sound communication system clipped noise |
| CN109361631A (en) * | 2018-10-15 | 2019-02-19 | 中国人民解放军战略支援部队信息工程大学 | Degree of rarefication unknown underwater sound orthogonal frequency division multiplexing channel estimation methods and device |
Non-Patent Citations (3)
| Title |
|---|
| 尹艳玲、乔钢、刘凇佐、周锋: "《基于基追踪去噪的水声正交频分复用稀疏信道估计》", 《物理学报》 * |
| 李娜: "《OFDM水声信道估计与均衡技术研究》", 《中国优秀硕士学位论文全文数据库》 * |
| 李雪: "《水声MIMO-OFDM迭代稀疏信道估计与均衡技术研究》", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110661735B (en) * | 2019-10-17 | 2021-11-09 | 哈尔滨工程大学 | Time delay estimation method based on orthogonal matching pursuit sparse channel estimation |
| CN110661735A (en) * | 2019-10-17 | 2020-01-07 | 哈尔滨工程大学 | Time delay estimation method based on orthogonal matching pursuit sparse channel estimation |
| CN111427046A (en) * | 2020-03-27 | 2020-07-17 | 西安交通大学 | Terahertz pulse echo positioning method for improving detection precision |
| CN111427046B (en) * | 2020-03-27 | 2022-05-06 | 西安交通大学 | Terahertz pulse echo positioning method for improving detection precision |
| CN113271196A (en) * | 2021-04-07 | 2021-08-17 | 北京邮电大学 | Short guard interval orthogonal frequency division multiplexing wireless transmission method |
| CN114124246B (en) * | 2021-11-12 | 2022-11-25 | 厦门大学 | Underwater acoustic multiple access communication method and system based on multi-frequency airspace sparse reconstruction |
| CN114124246A (en) * | 2021-11-12 | 2022-03-01 | 厦门大学 | Underwater acoustic multiple access communication method and system based on multi-frequency airspace sparse reconstruction |
| CN114584433A (en) * | 2022-02-24 | 2022-06-03 | 哈尔滨工程大学 | Method for detecting synchronous signal in multi-path channel under impulse noise environment |
| CN114584433B (en) * | 2022-02-24 | 2024-05-28 | 哈尔滨工程大学 | Synchronous signal detection method in multi-path channel under impulse noise environment |
| CN115865574A (en) * | 2022-11-24 | 2023-03-28 | 山东大学 | Sparse channel estimation method and system based on structure detection |
| CN115865574B (en) * | 2022-11-24 | 2024-05-03 | 山东大学 | Sparse channel estimation method and system based on structure detection |
| CN117395104A (en) * | 2023-12-13 | 2024-01-12 | 为准（北京）电子科技有限公司 | Channel estimation method and device in orthogonal frequency division multiplexing system |
| CN117395104B (en) * | 2023-12-13 | 2024-02-09 | 为准（北京）电子科技有限公司 | Channel estimation method and device in orthogonal frequency division multiplexing system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110138459A (en) | Sparse underwater sound orthogonal frequency division multiplexing channel estimation methods and device based on base tracking denoising | |
| CN111884685B (en) | Synchronous demodulation method and device for digital communication signal | |
| CN103095615B (en) | The method and device that a kind of phase noise is estimated, compensated | |
| CN107682297A (en) | A kind of mobile underwater sound communication method | |
| CN112733811B (en) | Method for identifying underwater sound signal modulation modes based on improved dense neural network | |
| CN108540419A (en) | A kind of OFDM detection methods of the anti-inter-sub-carrier interference based on deep learning | |
| CN110113288A (en) | A kind of design and demodulation method of the ofdm demodulator based on machine learning | |
| CN108712353A (en) | Soft iterative channel estimation method | |
| CN113572708B (en) | DFT channel estimation improvement method | |
| CN111404634B (en) | Large-scale MIMO detection method, system and application based on variable step length iteration | |
| CN110311876A (en) | The implementation method of underwater sound OFDM receiver based on deep neural network | |
| CN109729035A (en) | A kind of impulse noise suppression method in orthogonal frequency division multiplexi communication system | |
| CN110278170A (en) | Short wave communication frequency deviation estimating method based on maximum likelihood | |
| CN109756930A (en) | Ofdm signal compression transmission and reconstructing method based on signal extrapolation and comb filtering | |
| CN107094043B (en) | Improved MMSE low-complexity signal detection method based on block iteration method | |
| CN108696466A (en) | The blind equalization detector of underwater sound communication regulatable view window mouth length | |
| CN108650203B (en) | Modulation mode identification method based on reconnaissance receiver | |
| CN110059401B (en) | OFDM system underwater sound channel impulse response reconstruction method | |
| CN110944002B (en) | Physical layer authentication method based on exponential average data enhancement | |
| CN104270328B (en) | A kind of signal to noise ratio real-time estimation method | |
| CN103117969B (en) | A kind of fractional lower-order statistics mould value transform wavelet frequency domain multimode blind balance method | |
| CN107231323B (en) | Channel estimation methods based on reliable decision feedback in visible light communication system | |
| CN109378837A (en) | A kind of ultra-low frequency oscillation discrimination method based on Matrix Pencil | |
| CN109361631A (en) | Degree of rarefication unknown underwater sound orthogonal frequency division multiplexing channel estimation methods and device | |
| CN105827556B (en) | Two-way turbo equalization methods and system, underwater sound communication system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190816 |
|
| RJ01 | Rejection of invention patent application after publication |