CN110020695A - K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system - Google Patents
K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system Download PDFInfo
- Publication number
- CN110020695A CN110020695A CN201910319988.0A CN201910319988A CN110020695A CN 110020695 A CN110020695 A CN 110020695A CN 201910319988 A CN201910319988 A CN 201910319988A CN 110020695 A CN110020695 A CN 110020695A
- Authority
- CN
- China
- Prior art keywords
- vector
- quantization
- filter bank
- communication system
- data
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- 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
- H04B10/2589—Bidirectional transmission
- H04B10/25891—Transmission components
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Signal Processing (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Networks & Wireless Communication (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Probability & Statistics with Applications (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Optical Communication System (AREA)
Abstract
The present invention proposes a kind of K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system, include: step 1: the data sequence that filter bank multi-carrier modulation transmitter is sent being input to data preprocessing module, using the vector of generation as the input of K mean quantization module；Step 2: choose K vector as initial quantization rank, calculate the distance between each vector and quantization step vector, according to distance by each vector clusters into the set apart from the smallest quantization step vector, calculating quantized interval and each quantized interval center vector；Step 3: the initial quantization rank that the center vector that step 2 is generated is calculated as distance, iterative step 2 obtain K quantized interval and center vector until convergence；Step 4: exporting quantized interval obtained in step 3 and center vector to obtain quantized result.K-means non-uniform quantizing algorithm of the present invention reduces system quantizing error, improves error rate of system performance, saves cost.
Description
Technical field
The present invention relates to a kind of K-means non-uniform quantizing calculations for filter bank multi-carrier modulation optical communication system
Method.
Background technique
In recent years, with the appearance of the emerging technologies such as machine learning, social networks, cloud computing, to traffic throughput and band
Also more stringent requirements are proposed for the optical transport network in rapid growth to bloom spectrum efficiency, low cost for wide demand.Based on overloading
The intensity modulated Direct Inspection Technology of wave modulation technique is standby in short distance optical fiber link since it realizes that simple, complexity is low
Favored.However, compared to traditional multi-carrier modulation technology such as orthogonal frequency division multiplexing (OFDM), overloading based on wave filter group
Wave modulated signal includes multicarrier based on wave filter group (FBMC), universal filter multicarrier (UFMC) etc., in optical fiber, wireless
It is widely used in communication system.Its advantage is that Sidelobe Suppression, than high, asynchronous transmission characteristic is good, before not needing circulation
Sew, spectrum efficiency is high, improves system to the robustness of inter-carrier interference (ICI), reduces out-of-band power leakage power etc..?
In optical transmission system, the digital analog converter (DAC) of high bit resolution is to convert digital signals into electrical analog signals
Ideal chose.And the use of high-resolution uniform quantization DAC will lead to very high power consumption and system cost, therefore need to use
A kind of performance promoting low resolution DAC based on the method for non-uniform quantizing DAC is to reduce system cost.
Quantization method can be divided into uniform quantization and non-uniform quantizing.It include index quantization in non-uniform quantizing method, secondary
Side's quantization, broken line quantization estimate quantization etc. based on signal distributions.Quantization step based on signal distributions estimation can pass through nonparametric
Estimation method determines.In conventional OFDM systems, for example, Jizong Peng in 2017 in IEEE Photonics
" the SQNR Improvement Enabled by Nonuniform DAC Output Levels for that Journal is delivered
IM-DD OFDM Systems ", the fact that meet Gaussian Profile according to ofdm signal, determines quantizing noise minimum using Matlab
When quantization step, improve the transmission performance of IMDD-OFDM system.But in IMDD filter bank multi-carrier system, waveform warp
After wave filter processing, distributed wave and non-fully in Gaussian Profile is not available based on Gaussian Profile it is assumed that just can not yet
The quantization step optimization algorithm proposed using Jizong Peng, so the method for non-parametric estmation can only be used to signal distributions
Carrying out estimation just can determine that quantization step.
2018, " the Performance that the present inventor delivers in IEEE Photonics Journal
Optimization by Nonparametric Histogram Estimation for Low Resolution in
IMDD-OQAM-OFDM System " use non-parametric estmation histogram method to carry out signal estimation to IMDD-OQAM-OFDM waveform
Determine quantization step.But non-parametric estmation histogram method is larger to sample size demand, and computation complexity and storage complexity all compare
It is larger, and it is only applicable to low-dimensional data, when dimension increases, space needed for histogram will exponentially increase with the increase of dimension
Add.So quantization step optimizes determining algorithm, there are still rooms for improvement.
Summary of the invention
For the defects in the prior art, the present invention provides a kind of for filter bank multi-carrier modulation optical communication system
K-means non-uniform quantizing algorithm, to improve the quantization performance of DAC.
K mean value (K-means) is a kind of machine learning algorithm based on Statistical Learning Theory.The advantages of K-means is former
Reason is simple, realizes and is easy, fast convergence rate, and Clustering Effect is more excellent.Simultaneously as K-means is a kind of unsupervised cluster point
Algorithm is analysed, the training of early period is not needed, cluster calculation is carried out by the transmitting signal data to filter bank multi-carrier and is somebody's turn to do
K quantized interval of waveform, so that it may determine its quantization step.
Based on this, the present invention adopts the following technical scheme:
For the K-means non-uniform quantizing algorithm of filter bank multi-carrier modulation optical communication system, include the following steps:
Step 1: the data sequence that filter bank multi-carrier modulation transmitter is sent is input to data preprocessing module, it will
Input of the feature vector of generation as K mean quantization module；
Step 2: K feature vector is chosen as initial quantization rank, calculates the distance between each vector and quantization step vector,
According to distance by each vector clusters into the set apart from nearest quantization step vector, quantized interval and each amount is calculated
Change the center vector in section；
Step 3: the initial quantization rank that the center vector that step 2 is generated is calculated as distance, iterative step 2 is until receive
It holds back, obtains K quantized interval and center vector；
Step 4: exporting quantized interval obtained in step 3 and center vector to obtain quantized result.
Preferably, in step 1: the data prediction mode of use includes: the method for 1) vector building are as follows: is prolonged by tap
When device obtain characteristic value of the sampled value as each data of current data and its several front and back data, construction feature to
Amount.2) vector processing method are as follows: assign weight in centre tapped characteristic value, increase its calculate apart from when the influence that generates,
Obtain final feature vector.The weight parameter can take the fixed value between 1~1.5.Wherein, the structure of input data sequence
At method are as follows: split data real and imaginary parts and emit end data, by real part data Re { x1}Re{x2}…Re{xnAnd imaginary data
Im{x1}…Im{xnAccording to former data sequence independent input.
Preferably, in step 2: specific step is as follows: randomly selecting K feature vector as initial quantization rank, calculates each
Euclidean distance between vector and quantization step vector, according to distance by each vector clusters to apart from nearest quantization step vector
In set, the center vector of quantized interval and each quantized interval after a wheel quantization can be obtained by calculating.Wherein,
The choosing method of first group of quantization step are as follows: K vector of random selection is as quantization step, and wherein K is equal to 2n, n is Target quantization position
Number.The distance is Euclidean distance.Cluster operation method are as follows: according to the distance between each vector to quantization step vector,
Vector clusters are formed into a set to apart from nearest quantization step, numerical intervals where the set are quantized interval, are calculated
The center vector of each quantized interval is as quantization step.The new center vector is the mean value of each point vector in quantized interval.
Preferably, in step 3: the convergent center vector that is defined as is no longer changed, in each quantized interval
Point vector no longer changes.
Preferably, in step 4: the mode of the quantized result output are as follows: the center vector of K quantized interval is as each
Included between quantization put vector quantized result, in one-dimensional quantization be data sequence quantized result, by K center
Quantized result can be obtained according to the input sequence output of corresponding feature vector in vector.
Quantization algorithm of the invention, the characteristics of having fully considered high-speed filter bank multi-carrier system, for filter
The characteristics of group multicarrier waveform, finds the quantification manner for being suitble to the waveform, and then greatly improves the performance of DAC quantization.
Compared with prior art, the present invention have it is following the utility model has the advantages that
1, algorithm performance of the invention is compared to conventional uniform quantization method, available better quantization performance, thus
The performance of boostfiltering device group multi-carrier modulation optical communication system.
2, algorithm performance of the invention is available more preferable compared to the Non-Linear Programming quantization method estimated based on signal
Signal estimation effect and quantification effect, and algorithm is simpler, and complexity is lower
3, algorithm of the invention compare with other cluster learning algorithms such as neural network, the principles such as support vector machines are simpler
It is single, it is easy to accomplish.
Detailed description of the invention
Upon reading the detailed description of non-limiting embodiments with reference to the following drawings, other feature of the invention,
Objects and advantages will become more apparent upon:
Fig. 1 is the DAC quantization principles schematic diagram based on K-means.
Fig. 2 is IMDD filter bank multi-carrier optical communication applications system schematic.
Fig. 3 is based on the non-homogeneous amount of one-dimensional K-means using not isotopic number for IMDD filter bank multi-carrier optical communication system
Change method and use conventional uniform quantization method SQNR comparison diagram.Kmeans-q is an example of inventive algorithm, and q is quantization
Resolution ratio.
Specific embodiment
The present invention is described in detail combined with specific embodiments below.Following embodiment will be helpful to the technology of this field
Personnel further understand the present invention, but the invention is not limited in any way.It should be pointed out that the ordinary skill of this field
For personnel, without departing from the inventive concept of the premise, several changes and improvements can also be made.These belong to the present invention
Protection scope.
The present invention provides a kind of K-means non-uniform quantizing calculations for filter bank multi-carrier modulation optical communication system
Method.As shown in Figure 1, data preprocessing module, to input data sequence carry out pretreatment and by data sequence form M feature to
Amount.K mean quantization module according between quantization step vector and each vector Euclidean distance carry out cluster operation to center vector not
It changes again, quantization step of the center vector of each quantized interval as the quantized interval, K mean quantization module after clustering
In the quantized interval that generates of last iteration and center vector be last quantization output.Specifically, comprising the following steps:
Step 1: the data sequence that filter bank multi-carrier modulation transmitter is sent is input to data preprocessing module, it will
Input of the feature vector of generation as K mean quantization module；
Step 2: K feature vector is chosen as initial quantization rank, calculates the distance between each vector and quantization step vector,
According to distance by each vector clusters into the set apart from nearest quantization step vector, quantized interval and each amount is calculated
Change the center vector in section；
Step 3: the initial quantization rank that the center vector that step 2 is generated is calculated as distance, iterative step 2 is until receive
It holds back, obtains K quantized interval and center vector；
Step 4: exporting quantized interval obtained in step 3 and center vector to obtain quantized result.
Further, to hereafter need using term be first introduced:
1) data prediction refers to the process for former transmission of data sequences being changed into feature vector, to obtain preferably number
According to sequence signature, so that its upper limit of K mean quantization Model approximation.
2) feature value vector of n-th of symbol based on vector building can be described as:
Xn=[xn-(L-1)/2,…,xn-1,c×xn,xn+1,…,xn+(L-1)/2]
Wherein, xnIndicate current symbol, xn-(L-1)/2,…,xn-1Indicate prior symbol, xn+1,…,xn+(L-1)/2Indicate subsequent
Symbol, the weight of c current symbol can take the fixed value between 1~1.5.
3) distance calculated refers to the Euclidean distance in N-dimensional space, can be described as:
In step 1: the data prediction mode of use includes: the method for 1) vector building are as follows: is obtained by tap delayer
Characteristic value to the sampled value of current data and its several front and back data as each data, construction feature vector.2) to
Measure processing method are as follows: assign weight in centre tapped characteristic value, increase its calculate apart from when the influence that generates, obtain final
Feature vector.The weight parameter can take the fixed value between 1~1.5.Wherein, the constructive method of input data sequence
Are as follows: it splits data real and imaginary parts and emits end data, by real part data Re { x1}Re{x2}…Re{xnAnd imaginary data Im
{x1}…Im{xnAccording to former data sequence independent input.
In step 2: the choosing method of first group of quantization step are as follows: K vector of random selection is as quantization step, and wherein K is equal to
2n, n is Target quantization digit.The distance is Euclidean distance.Cluster operation method are as follows: according to each vector to quantization step
Vector clusters to nearest quantization step are formed a set by the distance between vector, and the numerical intervals of the set are to quantify
Section calculates the center vector of each quantized interval as quantization step.The new center vector is each point vector in quantized interval
Mean value.
In step 3: the convergent center vector that is defined as is no longer changed, and the point vector in each quantized interval is not
It changes again.
In step 4: the mode of quantized result output are as follows: the center vector of K quantized interval is middle institute between each quantization
Include puts the quantized result of vector, is the quantized result of input data sequence in one-dimensional quantization, K center vector is pressed
Quantized result can be obtained in input sequence output according to corresponding feature vector.
This example provides a kind of filter bank multi-carrier modulation optical communication system based on K-means cluster, such as Fig. 2 institute
Show, comprising: optical transmitter module, optical receiver module and fiber channel.It will include data information in the optical transmitter module
Digital signal input and based on the K-means quantization modules clustered obtain analog signal, high-speed light is converted by optical modulator
Signal is sent to fiber channel, converts optical signal into corresponding electric signal by the optical receiver module, demodulated to obtain letter
Cease data.
Optical transmitter module include: digital signal module, baseband-modem module, the non-uniform quantizing module based on K-means,
Optical modulator；The digital signal module is connected with baseband-modem module, carries out coding mapping processing to input data sequence, and
The high-speed digital electric signal for needing to transmit is generated, the baseband-modem module output digit signals are to based on the non-of K-means
Digital signal sequences are converted into feature vector by uniform quantization module, and K vector of random selection as quantization step and calculates remaining
The distance between vector and quantization step vector judge remaining vector sum vector rank vector nearest with it for same category and again
Center vector is calculated, interative computation to center vector no longer changes, according to where data sequence input sequence output vector
Quantized interval and center vector finally complete electro-optic conversion by optical modulator as quantization output.
Optical receiver module includes: photodetector, real-time oscilloscope, baseband demodulation module, data outputting unit；Its
In, by the electric signal received after real-time oscilloscope samples, the output signal of the real-time oscilloscope passes through photodetector
The output of baseband demodulation resume module is crossed, realizes the reception of user data.
The one-dimensional K-means non-uniform quantizing method parameter of this case verification are as follows: signal is modulated using OQAM-OFDM, input
Digital signal sequences length is 516096, tap number 1, tap coefficient 1.It is long that uniform quantization method inputs digital signal sequences
Degree is 516096.
Fig. 3 be high-speed light filter bank multi-carrier system pass through respectively 3-5 uniform quantization and 1 dimension K-means it is non-equal
Emitting the SQNR performance comparison figure of signal after even quantization, in figure: horizontal axis is amplitude limit value, and the longitudinal axis is that SQNR expression amount is made an uproar than size,
Unit is dB.It can be seen that algorithm of the invention is achieved than conventional uniform quantization method more in the case where identical quantization digit
Good quantization performance.
In conclusion small magnitude can be better accounted for using the present invention is based on the non-uniform quantizing algorithm of K-means
Influence factor of the signal in signal quantization.It is lower compared to cost required for uniform quantization DAC, and quantification effect is more preferable.
The principle of K-means is simple simultaneously, easy to accomplish.Therefore, algorithm of the invention relatively good can be applied to filter bank multi-carrier
The demand of glistening light of waves communication system.
A kind of K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system of the present invention, packet
It includes: in filter bank multi-carrier modulation optical communication system, the feature vector of input data is generated by data prediction, each
Vector determines the corresponding quantization step of the vector according to the center vector of nearest quantized interval.The result of K-means cluster
The as output of digital analog converter.The present invention solves that two-forty signal transmission process Small Amplitude signal quantization performance is poor to ask
Topic.Quantify compared to conventional uniform, identical quantization performance is also able to maintain while reducing quantization digit, thus is suitble to filter
Wave device group multi-carrier modulation optical communication system.
Specific embodiments of the present invention are described above.It is to be appreciated that the invention is not limited to above-mentioned
Particular implementation, those skilled in the art can make a variety of changes or modify within the scope of the claims, this not shadow
Ring substantive content of the invention.In the absence of conflict, the feature in embodiments herein and embodiment can any phase
Mutually combination.
Claims (8)
1. the K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system, which is characterized in that including
Following steps:
Step 1: the data sequence that filter bank multi-carrier modulation transmitter is sent being input to data preprocessing module, will be generated
Input of the vector as K mean quantization module；
Step 2: choosing K vector as initial quantization rank, the distance between each vector and quantization step vector are calculated, according to distance
By each vector clusters into the set apart from nearest quantization step vector, quantized interval and each quantized interval is calculated
Center vector；
Step 3: the initial quantization rank that the center vector that step 2 is generated is calculated as distance, iterative step 2 obtain until convergence
To K quantized interval and center vector；
Step 4: exporting quantized interval obtained in step 3 and center vector to obtain quantized result.
2. the K-means non-uniform quantizing according to claim 1 for filter bank multi-carrier modulation optical communication system
Algorithm, which is characterized in that it includes modulating optic communication based on filter bank multi-carrier that filter bank multi-carrier, which modulates optical communication system,
System and universal filter multi-carrier modulation optical communication system.
3. the K-means non-uniform quantizing according to claim 2 for filter bank multi-carrier modulation optical communication system
Algorithm, which is characterized in that in the step 1, the data prediction mode that data preprocessing module uses includes:
1) method of vector building: the sampled value of current data and its front and back data is obtained by tap delayer, by sampled value
As the characteristic value of each data, construction feature vector；
2) vector processing method: assigning weight in centre tapped characteristic value, increase its calculate apart from when the influence that generates, obtain
To final feature vector, weight parameter takes the fixed value between 1~1.5.
4. the K-means non-uniform quantizing according to claim 3 for filter bank multi-carrier modulation optical communication system
Algorithm, which is characterized in that in the step 1:
The constructive method of input data sequence are as follows: split data real and imaginary parts and emit end data, by real part data Re { x1}Re
{x2}…Re{xnAnd imaginary data Im { x1}…Im{xnAccording to former data sequence independent input.
5. the K-means non-uniform quantizing according to claim 4 for filter bank multi-carrier modulation optical communication system
Algorithm, which is characterized in that the step 2 is specific as follows:
K feature vector is randomly selected as initial quantization rank, calculates the Euclidean distance between each vector and quantization step vector, root
According to Euclidean distance by each vector clusters into the set apart from nearest quantization step vector, be calculated by one wheel quantization after
Quantized interval and each quantized interval center vector.
6. the K-means non-uniform quantizing according to claim 5 for filter bank multi-carrier modulation optical communication system
Algorithm, which is characterized in that in the step 2,
The choosing method of first group of quantization step are as follows: K vector of random selection is as quantization step, and wherein K is equal to 2n, n is Target quantization
Digit,
Cluster operation method are as follows: according to the distance between each vector to quantization step vector, recently by vector clusters to distance
Quantization step forms a set, and numerical intervals where the set are quantized interval, and the center vector for calculating each quantized interval is made
For quantization step.
7. the K-means non-uniform quantizing according to claim 6 for filter bank multi-carrier modulation optical communication system
Algorithm, which is characterized in that the step 3 is specific as follows:
The convergent center vector that is defined as is no longer changed, and the point vector in each quantized interval no longer changes.
8. the K-means non-uniform quantizing according to claim 7 for filter bank multi-carrier modulation optical communication system
Algorithm, which is characterized in that in the step 4, the mode of quantized result output are as follows:
The center vector of K quantized interval is the quantized result that vector is put included in each quantized interval, in one-dimensional quantization
In be data sequence quantized result, by K center vector according to corresponding feature vector input sequence export can be obtained
Quantized result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910319988.0A CN110020695A (en) | 2019-04-19 | 2019-04-19 | K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910319988.0A CN110020695A (en) | 2019-04-19 | 2019-04-19 | K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110020695A true CN110020695A (en) | 2019-07-16 |
Family
ID=67191995
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910319988.0A Pending CN110020695A (en) | 2019-04-19 | 2019-04-19 | K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110020695A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111327558A (en) * | 2020-02-28 | 2020-06-23 | 杭州电子科技大学 | Method and system for GMM non-uniform quantization for filter multi-carrier modulation optical communication |
| CN112737694A (en) * | 2020-12-28 | 2021-04-30 | 杭州电子科技大学 | Non-uniform quantization system of filter multi-carrier modulation optical communication system based on SOM |
| CN112769728A (en) * | 2020-12-31 | 2021-05-07 | 杭州电子科技大学 | KDE non-uniform quantization method of multi-carrier modulation optical communication system based on filter bank |
Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102324075A (en) * | 2011-10-29 | 2012-01-18 | 大连理工大学 | Green building intelligent energy-saving assessment management system based on Web geographic information system (GIS) |
| CN103001825A (en) * | 2012-11-15 | 2013-03-27 | 中国科学院计算机网络信息中心 | Method and system for detecting DNS (domain name system) traffic abnormality |
| CN105844302A (en) * | 2016-04-07 | 2016-08-10 | 南京新与力文化传播有限公司 | Depth-learning-based method for automatically calculating commodity trend indexes |
| US20170250744A1 (en) * | 2016-02-26 | 2017-08-31 | Hobbit Wave, Inc. | Devices and methods using the hermetic transform for transmitting and receiving signals using multi-channel signaling |
| CN108123908A (en) * | 2017-12-14 | 2018-06-05 | 杭州电子科技大学 | A kind of improvement SVM equalization methods and system for NG-PON |
| CN108141598A (en) * | 2015-09-02 | 2018-06-08 | 汤姆逊许可公司 | For the method and apparatus of the quantization in Video coding and decoding |
| CN108205682A (en) * | 2016-12-19 | 2018-06-26 | 同济大学 | It is a kind of for the fusion content of personalized recommendation and the collaborative filtering method of behavior |
| CN108353193A (en) * | 2015-08-19 | 2018-07-31 | Lg 电子株式会社 | Based on it is multiple based on the model of figure by using the method and apparatus of transformation encoding/decoding video signal |
| CN108494488A (en) * | 2018-02-12 | 2018-09-04 | 杭州电子科技大学 | The SVM equalization methods based on DFE for short distance optical communication system |
| CN109150781A (en) * | 2018-09-04 | 2019-01-04 | 哈尔滨工业大学（深圳） | A kind of modulation format recognition methods based on K-K coherent reception |
| CN109308496A (en) * | 2018-10-11 | 2019-02-05 | 南昌工程学院 | A kind of optimal case method of selecting based on fuzzy clustering iterative and projection pursuit |
-
2019
- 2019-04-19 CN CN201910319988.0A patent/CN110020695A/en active Pending
Patent Citations (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102324075A (en) * | 2011-10-29 | 2012-01-18 | 大连理工大学 | Green building intelligent energy-saving assessment management system based on Web geographic information system (GIS) |
| CN103001825A (en) * | 2012-11-15 | 2013-03-27 | 中国科学院计算机网络信息中心 | Method and system for detecting DNS (domain name system) traffic abnormality |
| CN108353193A (en) * | 2015-08-19 | 2018-07-31 | Lg 电子株式会社 | Based on it is multiple based on the model of figure by using the method and apparatus of transformation encoding/decoding video signal |
| CN108141598A (en) * | 2015-09-02 | 2018-06-08 | 汤姆逊许可公司 | For the method and apparatus of the quantization in Video coding and decoding |
| US20170250744A1 (en) * | 2016-02-26 | 2017-08-31 | Hobbit Wave, Inc. | Devices and methods using the hermetic transform for transmitting and receiving signals using multi-channel signaling |
| CN105844302A (en) * | 2016-04-07 | 2016-08-10 | 南京新与力文化传播有限公司 | Depth-learning-based method for automatically calculating commodity trend indexes |
| CN108205682A (en) * | 2016-12-19 | 2018-06-26 | 同济大学 | It is a kind of for the fusion content of personalized recommendation and the collaborative filtering method of behavior |
| CN108123908A (en) * | 2017-12-14 | 2018-06-05 | 杭州电子科技大学 | A kind of improvement SVM equalization methods and system for NG-PON |
| CN108494488A (en) * | 2018-02-12 | 2018-09-04 | 杭州电子科技大学 | The SVM equalization methods based on DFE for short distance optical communication system |
| CN109150781A (en) * | 2018-09-04 | 2019-01-04 | 哈尔滨工业大学（深圳） | A kind of modulation format recognition methods based on K-K coherent reception |
| CN109308496A (en) * | 2018-10-11 | 2019-02-05 | 南昌工程学院 | A kind of optimal case method of selecting based on fuzzy clustering iterative and projection pursuit |
Non-Patent Citations (6)
| Title |
|---|
| DAOJUN XUE ET AL: "Performance Improvement of IM-DD OFDM Systems by Nonuniform DAC Output Levels", 《2018 ASIA COMMUNICATIONS AND PHOTONICS CONFERENCE》 * |
| ELIAS GIACOUMIDIS ET AL: "Comparison of DSP-based nonlinear equalizers for intra-channel nonlinearity compensation in coherent optical OFDM", 《OPTICS LETTERS》 * |
| MUTSAM A. JARAJREH ET AL: "Artificial Neural Network Nonlinear Equalizer for Coherent Optical OFDM", 《IEEE PHOTONICS TECHNOLOGY LETTERS》 * |
| TIANCHENG HUANG ET AL: "A novel FABP quantization scheme for improving performance in low-bit IMDD-FBMC system", 《OPTICS COMMUNICATIONS》 * |
| 俞嘉生等: "基于非均匀量化ADC/DAC的IMMDD-UMFC系统性能提升", 《聊城大学学报（自然科学版）》 * |
| 虞淼: "高速相关光通信系统中的调制与解调技术研究", 《中国博士学位论文全文数据库 信息科技辑》 * |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111327558A (en) * | 2020-02-28 | 2020-06-23 | 杭州电子科技大学 | Method and system for GMM non-uniform quantization for filter multi-carrier modulation optical communication |
| CN111327558B (en) * | 2020-02-28 | 2022-06-21 | 杭州电子科技大学 | Method and system for GMM non-uniform quantization for filter multi-carrier modulation optical communication |
| CN112737694A (en) * | 2020-12-28 | 2021-04-30 | 杭州电子科技大学 | Non-uniform quantization system of filter multi-carrier modulation optical communication system based on SOM |
| CN112737694B (en) * | 2020-12-28 | 2021-11-16 | 杭州电子科技大学 | Non-uniform quantization system of filter multi-carrier modulation optical communication system based on SOM |
| CN112769728A (en) * | 2020-12-31 | 2021-05-07 | 杭州电子科技大学 | KDE non-uniform quantization method of multi-carrier modulation optical communication system based on filter bank |
| CN112769728B (en) * | 2020-12-31 | 2022-06-21 | 杭州电子科技大学 | KDE non-uniform quantization method of multi-carrier modulation optical communication system based on filter bank |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110020695A (en) | K-means non-uniform quantizing algorithm for filter bank multi-carrier modulation optical communication system | |
| CN109104248B (en) | DRoF front-end signal quantization method based on SOFM neural network clustering | |
| CN102724761B (en) | Dynamic resource allocation method for orthogonal frequency division multiple access (OFDMA) system | |
| CN103283200A (en) | Adaptive bit or power loading in optical orthogonal frequency division multiplexing transceivers | |
| CN107949065B (en) | A kind of adaptive bit power distribution method of NOMA | |
| CN110233755B (en) | Computing resource and frequency spectrum resource allocation method for fog computing in Internet of things | |
| CN101902747A (en) | Spectrum allocation method based on fuzzy logic genetic algorithm | |
| CN109104250B (en) | Star 24QAM mapping-based optical probability forming coding method | |
| CN111083078B (en) | Probability shaping quadrature amplitude modulation format blind identification method and system | |
| CN108600999A (en) | FD-D2D is based on channel distribution and power control combined optimization method | |
| CN103124250A (en) | Receiving device, signal processing device, and signal processing method | |
| CN109981223A (en) | Multicarrier FTN transmission/method of reseptance and relevant device based on FRFT | |
| CN115296737A (en) | ADM-GSMPPM constellation construction method and device based on free space optical communication | |
| CN109951412B (en) | Method for suppressing signal cubic metric by deep neural network | |
| CN106899347A (en) | Based on the system and method that 2D TCM PAM8 realize high speed transmission of signals | |
| CN113507712B (en) | Resource allocation and calculation task unloading method based on alternate direction multiplier | |
| CN112543056B (en) | Power and grouping combined optimization method for PD-NOMA-VLC system | |
| CN111327558B (en) | Method and system for GMM non-uniform quantization for filter multi-carrier modulation optical communication | |
| CN102724670A (en) | Dynamic resource distribution method used in single-carrier frequency-division multiple access (SC-FDMA) system | |
| CN111884980A (en) | Generalized index modulation optical OFDM modulation method for wireless optical communication | |
| US20230214667A1 (en) | Methods for long-range federated edge learning with chirp-based over-the-air computation | |
| CN108111226B (en) | A kind of system and method for polar coordinates coded modulation | |
| CN113938209B (en) | Method and system for shaping pulse amplitude modulation signals for probability shaping | |
| CN113271280B (en) | Modulation method, device, equipment and storage medium | |
| Kadhum et al. | Dynamic bit loading with the OGFDM waveform maximises bit-rate of future mobile communications |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190716 |