CN106998229A - It is a kind of based on variable step without constraint FD LMS mode division multiplexing system Deplexing method - Google Patents
It is a kind of based on variable step without constraint FD LMS mode division multiplexing system Deplexing method Download PDFInfo
- Publication number
- CN106998229A CN106998229A CN201611150443.4A CN201611150443A CN106998229A CN 106998229 A CN106998229 A CN 106998229A CN 201611150443 A CN201611150443 A CN 201611150443A CN 106998229 A CN106998229 A CN 106998229A
- Authority
- CN
- China
- Prior art keywords
- signal
- carried out
- lms
- value
- constraint
- 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
- 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
- H04L25/025—Channel estimation channel estimation algorithms using least-mean-square [LMS] method
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/25—Arrangements specific to fibre transmission
- H04B10/2581—Multimode transmission
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04J—MULTIPLEX COMMUNICATION
- H04J14/00—Optical multiplex systems
- H04J14/04—Mode multiplex systems
-
- 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)
- Power Engineering (AREA)
- Electromagnetism (AREA)
- Mathematical Physics (AREA)
- Radio Transmission System (AREA)
Abstract
The invention discloses a kind of based on mode division multiplexing system Deplexing method of the variable step without constraint FD LMS, belong to communication technical field, the first step, electric-wave filter matrix W is set to realize pre- convergence with one section of training sequence, it regard training sequence as desired signal, second step, output data is obtained with the method directly adjudicated, obtained pre- convergent tap vector value is assigned to the balanced device in S2 steps, and the signal for obtaining judgement is used as desired signal, remaining step is carried out according to the method for the first step, continuous iteration, until the output data bit error rate is no longer reduced.Algorithm used in the present invention can change the step value of each frequency cabinet of each data block, so as to reach the purpose of Fast Convergent.The present invention can make up the problem of adaptive unconstrained frequency domain least-mean-square error algorithm (Unconstrained FD LMS) convergence is slow, with higher demultiplexing efficiency.
Description
Technical field
The invention belongs to communication technical field, it is related to a kind of Deplexing method of mode division multiplexing (MDM) system, is walked using becoming
Long unconstrained frequency domain least mean square algorithm realizes the demultiplexing to MDM systems.
Background technology
With the continuous propulsion of social informatization process, large-scale data center, Internet of Things, HD video etc. " bandwidth consumption "
Type business fast development, is continuously increased to the bandwidth demand of Networks of Fiber Communications.And be time-multiplexed (TDM), wavelength-division multiplex
(WDM), the extensive use of the technology such as palarization multiplexing (PDM) and higher-dimension high order modulation so that the power system capacity of single-mode fiber by
Gradually close to shannon limit.For foreseeable " bandwidth crisis ", a kind of new capacity-enlargement technology of tackling optical communication network ----be based on
The mode division multiplexing technology of less fundamental mode optical fibre (FMF) is arisen at the historic moment.MDM technologies utilize the orthogonality between each pattern in less fundamental mode optical fibre,
Each pattern is subjected to information transmission as independent channel, so as to realize high capacity transmission.However, in the mould based on less fundamental mode optical fibre point
In multiplex system, Mode Coupling and differential mode time delay (DMGD) equivalent damage have had a strong impact on the transmission performance of system.Meanwhile, in mould
Formula is coupled with differential mode time delay collective effect, and the demultiplexing of mode division multiplexing system also becomes relative complex.
Adaptive filter algorithm can effectively realize demultiplexing.Wherein, lowest mean square (LMS) algorithm is because simple and be easy to
Realize and be widely applied.LMS algorithm can both be realized in time domain, can also be realized in frequency domain.Compared to Time-Domain algorithm,
Frequency domain lowest mean square (Variable step-size unconstrained FD-LMS) algorithm passes through Fast Fourier Transform (FFT)
(FFT) a large amount of computings are saved.And FD-LMS algorithms are further divided into Constrained frequency domain minimum according to the presence or absence of time-domain constraints
Side's (Constrained FD-LMS) algorithm and unconstrained frequency domain lowest mean square (Unconstrained FD-LMS) algorithm.Wherein
Time-domain constraints block is cast out without bounding algorithm, so as to further save operand, the advantage of this complexity is more in multi input
It is particularly evident in (MIMO) system of output.But without constraint FD-LMS convergences of algorithm speed relatively slowly, it is necessary to more midamble code
Member realizes convergence, so as to reduce the availability of frequency spectrum.Therefore, further lifting, without constraint FD-LMS convergence of algorithm speed, is entered
And the demultiplexing efficiency of algorithm is improved, it is a urgent problem to be solved.
The content of the invention
For, without the problem of FD-LMS algorithmic statements are slow is constrained, being based on present in prior art the invention provides one kind
Reception signal Deplexing method of the variable step without constraint FD-LMS algorithms, this method has faster convergence rate, relatively low answered
Miscellaneous degree, while ensureing preferable error performance, with preferable demultiplexing effect.
Influenceed by differential mode time delay and Mode Coupling equivalent damage, the output signal of mode division multiplexing system be input signal with
The convolved mixtures of channel transfer matrices.The principle of the present invention is to estimate channel without constraint FD-LMS algorithms by variable step to pass
The inverse matrix W of defeated matrix H, makes W=H-1, wherein W is M × Metzler matrix, by the time-domain signal u (n) of reception=[u1(n),...,uM
(n)]TThe recovery signal y (n) obtained with electric-wave filter matrix W convolution, being expressed as matrix form is：
The present invention is achieved through the following technical solutions：
It is a kind of based on variable step without constraint FD-LMS mode division multiplexing system Deplexing method, comprise the following steps that：
The first step, makes electric-wave filter matrix W realize pre- convergence with one section of training sequence, using training sequence as desired signal,
Wherein, W is M × Metzler matrix, and its detailed process comprises the following steps：
S1, to jth (1≤j≤M) the road time domain signal u to be equalized after two times of over-samplings, dispersion compensationsj(n) odd even is carried out
Branch, obtains parallel strange circuit-switched dataWith even circuit-switched dataPiecemeal is distinguished to two paths of data, and according to Time Delay of Systems
Every piece of he number is obtained for N with the relation of shock response, two successive sub-blocks are put together, is sequentially carried out FFT and is obtained
To kth block odd even road frequency-region signalWherein, diag { } represents diagonal matrix computing, FFT []
Represent Fourier transformation；
S2, to weight vector length be N the sub- balanced device of odd evenInitialize installation is carried out, will when i and j are unequal
Each weights are set to 0；When i is equal with j, the N/2 weights is set to 1, remaining weights is set to 0, and N number of weights thereafter are mended
Zero；Then FFT is carried out again to obtain
S3, calculating i-th (1≤i≤M) road frequency domain outputFourier's contravariant is carried out to it
Change (IFFT) and give up top n value, obtain its kth block time domain output signal yi(k)；
S4, addition Carrier Phase Noise estimation item φi(k) kth block desired signal d, is utilizedi(k) time domain error e is calculatedi
(k)=(di(k)-yi(k))·exp(jφi(k) zero padding), is carried out to N number of value before it and FFT computings are carried out, kth block is obtained
Error of frequency domain signal Ei(k)；
S5, in every piece each frequency cabinet variable step：Wherein α is normal
Number,For kth block input signal power, Se,i,m(k)=λ Se,i,m(k-1)+(1-λ)|
Ei(k)|2For kth block error power,For the mutual work(of kth block signal error
Rate, wherein λ are forgetting factors；
S6, the tap weights vector for updating balanced deviceWherein []HFor
Take conjugate operation,For Step matrix, its occurrence is the step value obtained in step S5；
S7, judge whether convergence, if do not restrained, jump to step S1, if restrain if terminate, obtain pre- convergent tap
Vector
Second step, output data is obtained with the method directly adjudicated, and obtained pre- convergent tap vector value is assigned into S2
Balanced device in step, and obtained signal will be adjudicated as desired signal, remaining step is carried out according to the method for the first step, no
Disconnected iteration, until the output data bit error rate is no longer reduced.
Further, in described step S1, the sub- equaliser operating on signal of odd even is respectively adopted and carries out odd even branch process, can
So that balanced output sum is equal with the down-sampled output at half code element interval, so as to eliminate down-sampled process.
Further, in described step S4, carrier phase estimation item φi(k) obtained using M power algorithm computings, its
Step is：On the one hand, argument is taken to obtain a value time domain error at one moment,；On the other hand, M times of time domain error is taken
Side, enters back into low pass filter, filters out radio-frequency component, then obtained value is carried out taking argument and except M computings, obtains another
Two value additions are obtained carrier phase estimation item by individual value.Carrier phase estimation is carried out in two steps.Wherein the first step is auxiliary in data
The pre- converged state of filter tap is helped, second step is in the decision-feedback stage.
In the step S5, time domain backward error signal is definedWill step
Tap weights vector renewal equation in rapid S6 brings backward error formula into, and carries out Fourier transformation, after can obtaining accordingly
To error frequency domain formAnd then it is rewritten obtain each frequency cabinet
μ value representations：Utilize the amplitude coherent side between input signal and error signal
Journey can be obtained：
By Sv,i,m(k) bring into the step expression tentatively obtained, can be with
Obtain final step expression:
Compared with prior art, the invention has the advantages that：
The present invention is reduced using variable step without constraint FD-LMS algorithms to mixed signal, and it has computation complexity
It is low, the advantages of the availability of frequency spectrum is high, while preferable error performance is ensure that, with preferable demultiplexing effect.
Brief description of the drawings
Fig. 1 is 6 × 6 mode division multiplexing system structure diagrams based on less fundamental mode optical fibre；
Fig. 2 is 6 × 6 mode division multiplexing system receiver structural representations based on less fundamental mode optical fibre；
Fig. 3 is the flow chart of variable step proposed by the invention without constraint FD-LMS algorithms；
Fig. 4 is variable step proposed by the invention without theory diagram of the constraint FD-LMS algorithms for each output；
Fig. 5 is carries out equilibrium with algorithm proposed by the invention, under different OSNRs (OSNR), the mistake of 6 patterns
Code check (BER) change curve；
Fig. 6 be under different OSNRs (OSNR), using Constrained FD-LMS algorithms, without constraint FD-LMS algorithms and
Variable step carries out balanced obtained BER comparison diagrams without constraint FD-LMS algorithms；
Fig. 7 is the convergence graph of 6 patterns in the case that OSNR is 15dB；
Fig. 8 is in the case that OSNR is 15dB, Constrained FD-LMS algorithms, without constraint FD-LMS algorithms and variable step without
The convergence rate comparison diagram of constrained-LMS algorithm.
Embodiment
The present invention is described further below in conjunction with the accompanying drawings.
Implement 16 × 6 mode division multiplexing Transmission systems based on less fundamental mode optical fibre to verify method proposed by the present invention.Fig. 1
Show to verify 6 × 6 mode division multiplexing system schematics based on less fundamental mode optical fibre of the actual use of method proposed by the invention,
Its composition includes：6 data transmission blocks 1, pattern multiplexer 2,80km less fundamental mode optical fibres 3, pattern demultiplexer 4,6 phase stem graftings
Receipts machine 5 and off-line data processing mould 6.Wherein FMF Random Coupling intensity is -30dB/km, and fibre loss is 0.2dB/km,
The chromatic dispersion of LP01 moulds is 20ps/nm/km, and the chromatic dispersion of LP11 moulds is 21ps/km/nm, and differential mode time delay is 27ps/
km.The stiffness of coupling that pattern multiplexer and pattern demultiplexer are introduced is -21dB.
In Fig. 16 56Gbit/s QPSK optical signals by pattern multiplexer be coupled into FMF 3 patterns (LP01,
LP11a and LP11b) and its corresponding polarization mode, after being transmitted through 80km less fundamental mode optical fibres, Dietary behavior demultiplexer is separated.
Obtained signal is received using coherent receiver, and output signal handled by digital signal processing module,
Output signal after being restored.
Data transmission blocks in Fig. 1, increase Dare to Mach for 28Gbit/s source signal by two independent speed and adjust
Device processed is driven, and produces the QPSK signals that a speed is 56Gbit/s, wherein it is 1550nm, line that light source used, which is wavelength,
A width of 100kHz laser.
The coherent receiver composition structure in Fig. 1 is shown in Fig. 2, and it includes：One local oscillator, one 90 ° mix
Frequency device, two pairs of balanced detectors and two low pass filters.The light of received each pattern is sent into together with local oscillator light
90 ° of frequency mixers, the optical signal after mixing obtains electric signal by balanced detector, then two path signal is sent into
In Bessel low pass filters, two-way output signal is finally obtained.
The signal that receiver is received carries out two times of over-samplings and dispersion compensation processing, then the signal to obtaining first
Carry out equilibrium treatment.The present invention is its equalization algorithm, realizes the demultiplexing to mode division multiplexing system.Fig. 3 gives the present invention's
Particular flow sheet, and Fig. 4 gives the present invention theory diagram for each output.The present invention's comprises the following steps that：
The first step, pre- convergence to filter tap w (n) is realized using data auxiliary law, will originator input data as
Desired signal, takes 1≤i≤6,1≤j≤6, and it specifically includes following steps：
S1, to the signal u to be equalized after two times of over-samplings, dispersion compensationsj(n) odd even branch is carried out, two-way is obtained parallel
Neat circuit-switched dataWith even circuit-switched dataPiecemeal is distinguished to two-way serial data, according to Time Delay of Systems and shock response
It is 64 that relation, which obtains every piece of he number, and successive sub-block is put together, sequentially carries out FFT and obtains kth block frequency-region signalWherein diag { } represents diagonal matrix computing, and FFT [] represents Fourier transformation；
S2, to weight vector length be 64 wave filterInitialize installation is carried out, will be each when i and j are unequal
Weights be set to 0, i it is equal with j when, the 32nd weights are set to 1, remaining weights is set to 0, and to mending 64 null values thereafter, then
FFT is carried out to obtain
S3, calculating frequency domain outputIFFT conversion is carried out to it and gives up first 64
Value, obtains kth block output signal yi(k)；
S4, addition Carrier Phase Noise estimation item φi(k), carrier phase estimation item φi(k) M power algorithm computings are used
Obtain, its step is：On the one hand, argument is taken to obtain a value time domain error at one moment,；On the other hand, time domain is taken to miss
The M powers of difference, enter back into low pass filter, filter out radio-frequency component, then obtained value are carried out taking argument and except M computings,
Another value is obtained, two value additions are obtained into carrier phase estimation item.Utilize kth block desired signal di(k) time domain is calculated to miss
Poor ei(k)=(di(k)-yi(k))·exp(jφi(k) zero padding), is carried out to 64 values before it and FFT computings are carried out, obtained
Kth block error of frequency domain signal Ei(k)；
S5, in every piece each frequency cabinet variable step：α=0.01, k
Block input signal power,For kth block input signal power, SE, i, m(k)=λ SE, i, m
(k-1)+(1-λ)|Ei(k)|2For kth block error power,For kth
Block signal error cross-power, wherein λ=0.6；
S6, the tap weights vector for updating balanced deviceWhereinFor Step matrix；
S7, judge whether convergence, if do not restrained, jump to step S1, terminate if restraining, obtain pre- convergent take out
Head vector
Second step, because the system is modulated using QPSK, then needs data decision after equilibrium to (- 1, -1), (-
1,1) on, (1, -1), (1,1) four points, and obtained signal will be adjudicated as desired signal, by pre- convergent tap vector value
Be assigned to the balanced device in S2 steps, remaining according to the first step method carry out, when iterations be 10 when arrive compared with
For accurate output data.
Fig. 5 is given under different OSNR, the BER change curves of 6 patterns；And Fig. 6 is then given under different OSNR, have about
Beam FD-LMS algorithms, without constraint the BER comparison diagrams of FD-LMS algorithms and variable step without constrained-LMS algorithm；Fig. 5 shows 6 letters
Performance of the road after algorithm proposed by the present invention demultiplexing is essentially identical；When error rate of system reaches 10-3,6 channels reach
Minimum OSNR needed for target error rate is 15dB；Fig. 6 then shows the bit error rate of the identical lower 3 kinds of algorithms of OSNR quite, and it is solved
It is multiplexed effect almost consistent；As a result show, algorithm proposed by the invention can be demultiplexed effectively to mode division multiplexing system.
Fig. 7 gives OSNR in the case of 15dB, the convergence curve of 6 patterns；And it is 15dB's that Fig. 8, which gives OSNR,
In the case of, Constrained FD-LMS algorithms, without constraint FD-LMS algorithms and convergence of algorithm velocity contrast proposed by the present invention figure；
The abscissa of two figures is fft block number, and ordinate is normalized mean squared error (NMSE)；Fig. 7 shows that 6 patterns are almost
Realized and restrained with same speed, and finally converge on the NMSE that numerical value is less than -12dB；Fig. 8 is shown, is not up to -12dB's
NMSE, Constrained FD-LMS algorithm, the number without the fft block required for constraint FD-LMS algorithms and algorithm proposed by the present invention
Respectively 65,73 and 45；As a result show, convergence of algorithm speed proposed by the invention, can solve the problem that without constraint FD-LMS
The problem of algorithm the convergence speed is slow.
Mode division multiplexing system based on less fundamental mode optical fibre described in examples detailed above is it is known in the art, being obtained by known approach
.
The term such as step-length, weight vector used describes and explains the present invention just to more convenient in the present invention, can not conduct
Its additional limitation.
Claims (3)
1. it is a kind of based on mode division multiplexing system Deplexing method of the variable step without constraint FD-LMS, it is characterised in that specific steps
It is as follows：
The first step, makes electric-wave filter matrix W realize pre- convergence with one section of training sequence, using training sequence as desired signal, wherein,
W is M × Metzler matrix, and its detailed process comprises the following steps：
S1, to jth (1≤j≤M) the road time domain signal u to be equalized after two times of over-samplings, dispersion compensationsj(n) odd even branch is carried out,
Obtain parallel strange circuit-switched dataWith even circuit-switched dataPiecemeal is distinguished two paths of data, and according to Time Delay of Systems and impact
The relation of response obtains every piece of he number for N, and two successive sub-blocks are put together, sequentially carries out FFT and obtains kth block
Odd even road frequency-region signalWherein, diag { } represents diagonal matrix computing, and FFT [] is represented in Fu
Leaf transformation；
S2, to weight vector length be N the sub- balanced device of odd evenInitialize installation is carried out, will be each when i and j are unequal
Weights are set to 0；When i is equal with j, the N/2 weights is set to 1, remaining weights is set to 0, and to N number of weights zero padding thereafter；
Then FFT is carried out again to obtain
S3, calculating i-th (1≤i≤M) road frequency domain outputFourier inversion is carried out to it
(IFFT) and give up top n value, obtain its kth block time domain output signal yi(k)；
S4, addition Carrier Phase Noise estimation item φi(k) kth block desired signal d, is utilizedi(k) time domain error e is calculatedi(k)=
(di(k)-yi(k))·exp(jφi(k) zero padding), is carried out to N number of value before it and FFT computings are carried out, kth block frequency domain mistake is obtained
Difference signal Ei(k)；
S5, in every piece each frequency cabinet variable step：Wherein α is constant,For kth block input signal power, Se,i,m(k)=λ Se,i,m(k-1)+(1-λ)|Ei
(k)|2For kth block error power,For kth block signal error cross-power,
Wherein λ is forgetting factor；
S6, the tap weights vector for updating balanced deviceWherein []HTo take altogether
Yoke computing,For Step matrix, its occurrence is the step value obtained in step S5；
S7, judge whether convergence, if do not restrained, jump to step S1, if restrain if terminate, obtain pre- convergent tap vector
Second step, output data is obtained with the method directly adjudicated, and obtained pre- convergent tap vector value is assigned into S2 steps
In balanced device, and obtained signal will be adjudicated as desired signal, remaining step is carried out according to the method for the first step, constantly repeatedly
Generation, until the output data bit error rate is no longer reduced.
2. it is as claimed in claim 1 a kind of based on mode division multiplexing system Deplexing method of the variable step without constraint FD-LMS, its
It is characterised by, in described step S1, the sub- equaliser operating on signal of odd even is respectively adopted and carries out odd even branch process, equilibrium can be made
Sum is exported equal with the down-sampled output at half code element interval, so as to eliminate down-sampled process.
3. it is as claimed in claim 1 a kind of based on mode division multiplexing system Deplexing method of the variable step without constraint FD-LMS, its
It is characterised by, in described step S4, carrier phase estimation item φi(k) obtained using M power algorithm computings, its step is：One
Aspect, takes argument to obtain a value time domain error at one moment,；On the other hand, the M powers of time domain error are taken, are entered back into
Low pass filter, filters out radio-frequency component, then obtained value is carried out taking argument and except M computings, another value is obtained, by two
Individual value addition obtains carrier phase estimation item.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611150443.4A CN106998229B (en) | 2016-12-14 | 2016-12-14 | A kind of mode division multiplexing system Deplexing method based on variable step without constraint FD-LMS |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611150443.4A CN106998229B (en) | 2016-12-14 | 2016-12-14 | A kind of mode division multiplexing system Deplexing method based on variable step without constraint FD-LMS |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106998229A true CN106998229A (en) | 2017-08-01 |
| CN106998229B CN106998229B (en) | 2019-02-15 |
Family
ID=59430875
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611150443.4A Active CN106998229B (en) | 2016-12-14 | 2016-12-14 | A kind of mode division multiplexing system Deplexing method based on variable step without constraint FD-LMS |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106998229B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110474662A (en) * | 2019-07-31 | 2019-11-19 | 同济大学 | A kind of mimo wireless communication demodulation method of reseptance and system |
| CN113067662A (en) * | 2021-03-17 | 2021-07-02 | 吉林大学 | Demultiplexing algorithm of FBMC-OQAM mode division multiplexing system based on lattice reduction |
| CN114337840A (en) * | 2022-01-10 | 2022-04-12 | 天津师范大学 | Dispersion compensation method for modulation signal 6PolSK-QPSK |
| CN115473774A (en) * | 2022-09-13 | 2022-12-13 | 北京理工大学 | Channel equalization method of modular division multiplexing communication system |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110011733B (en) * | 2019-03-25 | 2020-10-16 | 华中科技大学 | Method and system for depolarization multiplexing based on momentum factor |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080175308A1 (en) * | 2007-01-19 | 2008-07-24 | Mark Fimoff | Method for determining the step size for an lms adaptive equalizer for 8vsb |
| CN101252559A (en) * | 2007-10-12 | 2008-08-27 | 电子科技大学中山学院 | Method of training sequence time change step length least mean square |
| CN103095373A (en) * | 2013-01-31 | 2013-05-08 | 华中科技大学 | Self-correlation optical fiber communication system based on mode division multiplexing |
| CN103227623A (en) * | 2013-03-29 | 2013-07-31 | 北京邮电大学 | Step value-variable LMS (Least Mean Square) self-adaptation filtering algorithm and filter |
| CN104283528A (en) * | 2014-09-18 | 2015-01-14 | 河海大学 | Variable-step LMS adaptive filtering method |
-
2016
- 2016-12-14 CN CN201611150443.4A patent/CN106998229B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20080175308A1 (en) * | 2007-01-19 | 2008-07-24 | Mark Fimoff | Method for determining the step size for an lms adaptive equalizer for 8vsb |
| CN101252559A (en) * | 2007-10-12 | 2008-08-27 | 电子科技大学中山学院 | Method of training sequence time change step length least mean square |
| CN103095373A (en) * | 2013-01-31 | 2013-05-08 | 华中科技大学 | Self-correlation optical fiber communication system based on mode division multiplexing |
| CN103227623A (en) * | 2013-03-29 | 2013-07-31 | 北京邮电大学 | Step value-variable LMS (Least Mean Square) self-adaptation filtering algorithm and filter |
| CN104283528A (en) * | 2014-09-18 | 2015-01-14 | 河海大学 | Variable-step LMS adaptive filtering method |
Non-Patent Citations (1)
| Title |
|---|
| KUN SHI 等: "A Frequency Domain Step-Size Control Method for LMS Algorithms", 《IEEE SIGNAL PROCESSING LETTERS》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110474662A (en) * | 2019-07-31 | 2019-11-19 | 同济大学 | A kind of mimo wireless communication demodulation method of reseptance and system |
| CN113067662A (en) * | 2021-03-17 | 2021-07-02 | 吉林大学 | Demultiplexing algorithm of FBMC-OQAM mode division multiplexing system based on lattice reduction |
| CN114337840A (en) * | 2022-01-10 | 2022-04-12 | 天津师范大学 | Dispersion compensation method for modulation signal 6PolSK-QPSK |
| CN114337840B (en) * | 2022-01-10 | 2023-10-31 | 天津师范大学 | Dispersion compensation method for modulated signal 6PolSK-QPSK |
| CN115473774A (en) * | 2022-09-13 | 2022-12-13 | 北京理工大学 | Channel equalization method of modular division multiplexing communication system |
| CN115473774B (en) * | 2022-09-13 | 2023-07-04 | 北京理工大学 | Channel equalization method for mode division multiplexing communication system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106998229B (en) | 2019-02-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106998229B (en) | A kind of mode division multiplexing system Deplexing method based on variable step without constraint FD-LMS | |
| Bai et al. | Adaptive frequency-domain equalization for mode-division multiplexed transmission | |
| CN101631095B (en) | Adaptive non-linearity compensation in coherent receiver | |
| EP2660984B1 (en) | Optical signal compensation device | |
| CN107171735A (en) | A kind of big line width CO OFDM phase noise compensation methods of time-frequency domain Kalman filtering | |
| CN109347562A (en) | CO-OFDM system phase noise optimization compensation method | |
| CN111934783B (en) | Optical comb system master-slave phase estimation performance improving method based on time delay module | |
| CN104081737A (en) | System and method for polarization de-multiplexing in a coherent optical receiver | |
| Xu et al. | Joint scheme of dynamic polarization demultiplexing and PMD compensation up to second order for flexible receivers | |
| JP6378654B2 (en) | Spatial multiplexed optical transmission system and spatial multiplexed optical transmission method | |
| Barbosa et al. | Scaling spatial multiplexing with principal modes | |
| CN105024763A (en) | A twice fixed point real number de-multiplexing algorithm used for polarization division multiplexing system de-multiplexing | |
| He et al. | Fast convergent frequency-domain MIMO equalizer for few-mode fiber communication systems | |
| CN112713942B (en) | MC-DBP algorithm-based method for jointly equalizing optical fiber signal damage | |
| CN108631904A (en) | A kind of mode division multiplexing system injury compensation method based on lattice reduction | |
| He et al. | Noise power directed adaptive frequency domain least mean square algorithm with fast convergence for DMGD compensation in few-mode fiber transmission systems | |
| Barbosa et al. | On a scalable path for multimode SDM transmission | |
| CN107346991B (en) | A kind of multichannel mpsk signal regeneration method based on Phase sensitive amplification | |
| Li et al. | Hybrid time–frequency domain equalization based on sign–sign joint decision multimodulus algorithm for 6× 6 mode division multiplexing system | |
| Arik et al. | Adaptive MIMO signal processing in mode-division multiplexing | |
| Li et al. | Single-carrier transmission scheme for extracting characteristic parameters of 32-Point 6PolSK-QPSK | |
| Li et al. | An improved LMS algorithm for mode demultiplexing in frequency domain | |
| Zhang et al. | Machine learning for long-haul optical systems | |
| Xing et al. | Study of efficient photonic chromatic dispersion equalization using MZI-based coherent optical matrix multiplication | |
| Kaur et al. | Design of multichannel optical OFDM system using advanced modulation techniques |
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 |