CN109412996A - Chain circuit self-adaptive method, electronic device and computer readable storage medium - Google Patents
Chain circuit self-adaptive method, electronic device and computer readable storage medium Download PDFInfo
- Publication number
- CN109412996A CN109412996A CN201811510754.6A CN201811510754A CN109412996A CN 109412996 A CN109412996 A CN 109412996A CN 201811510754 A CN201811510754 A CN 201811510754A CN 109412996 A CN109412996 A CN 109412996A
- Authority
- CN
- China
- Prior art keywords
- channel
- coding
- transmission information
- information
- transmitter
- 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
- 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
- H04L27/2647—Arrangements specific to the receiver only
- H04L27/2655—Synchronisation arrangements
- H04L27/2689—Link with other circuits, i.e. special connections between synchronisation arrangements and other circuits for achieving synchronisation
- H04L27/2695—Link with other circuits, i.e. special connections between synchronisation arrangements and other circuits for achieving synchronisation with channel estimation, e.g. determination of delay spread, derivative or peak tracking
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Radio Transmission System (AREA)
Abstract
A kind of chain circuit self-adaptive method, electronic device and computer readable storage medium, wherein the chain circuit self-adaptive method includes: that transmitter carries out message sink coding processing to target information and channel coding is handled, and obtains channel transmission information；Receiver obtains the channel transmission information, and carries out time synchronization, Fast Fourier Transform (FFT) FFT and channel estimation to the channel transmission information, obtains the channel estimation coefficient and noise estimated information of the channel transmission information；Network training unit is trained using the channel estimation coefficient and the noise estimated information as feature input, obtain the adaptive modulation and coding AMC scheme of prediction, the MAC scheme of the prediction is fed back into the transmitter, so that the transmitter is adaptively adjusted according to the AMC scheme that prediction obtains.
Description
Technical field
This application involves electronic technology fields more particularly to a kind of chain circuit self-adaptive method, electronic device and computer can
Read storage medium.
Background technique
In the field of wireless communication, the spectrum efficiency of determining wireless communication link is usually believed with communication on Effectiveness Forecast
The time variation in road is consistent.Specifically, multi-I/O OFDM (Multiple-Input Multiple-
Output Orthogonal Frequency Division Multiplexing MIMO-OFDM, MIMO-OFDM) link need
Suitable adaptive modulation and coding (adaptive modulation and coding, AMC) is wanted, that is, in determining mistake
Under packet rate (packet error rate, PER) constraint, it is based on channel state information (channel state
Information, CSI) select suitable Modulation and Coding Scheme (modulation and coding scheme, MCS) to come most
Improve throughput of system to limits.Each MCS contains quadrature amplitude modulation (Quadrature Amplitude
Modulation, QAM) debugging order and code rate.It is worth noting that, in MIMO-OFDM system, AMC is not
One simple question.Because it may be influenced simultaneously by many complicated factors in actual MIMO-OFDM system, such as
OFDM modulation, mimo channel, convolutional encoding, position interweave, wireless channel effect and circuit it is non-linear etc..Therefore, based on number
Model is learned to assume AMC method as approximate as possible, is not enough to accurately account for had an impact MIMO-OFDM system performance
Factor
AMC algorithm based on machine learning directly learnt from the channel information observed before, training data, is not had
Do any mathematical approach.Therefore they are possible to have reliable performance, but on condition that in complicated MIMO-OFDM system
In, there must be enough abilities using the method for machine learning to capture correct input/output relation.Especially existing
The link circuit self-adapting k-NN sorting algorithm of middle proposition in technology, its each spatial flow sequence based on all OFDM subcarriers
SNR has used a kind of didactic method to reduce the dimension of feature set.It demonstrates and carrys out the use of dimensionality reduction ratio using heuristic
The method performance of average SNR, availability indexes SNR (EESM) mapping is outstanding.However, this heuristic is in reduction process
Need to pay very big calculating cost and reliable data set.In addition, when predicting new data using k-NN, entire training
Collection data be stored in the equipment of user, and need to calculate between the feature of all new datas and sample set feature away from
From, this be it is not realistic, especially when a large amount of training datas.Most of all, feature set dimensionality reduction may be lost
Useful feature information in many original systems is lost so that influencing the accuracy rate of k-NN algorithm prediction.
Summary of the invention
The embodiment of the present application provides a kind of chain circuit self-adaptive method, electronic device and computer readable storage medium, is used for
According to the characteristic for arriving MIMO-OFDM system, suitable debugging encoding scheme is preferably predicted.
The embodiment of the present application first aspect provides a kind of chain circuit self-adaptive method, comprising:
The mode is applied to multi-I/O OFDM MIMO-OFDM system, and the system comprises transmittings
Machine, receiver and network training unit；
The transmitter carries out message sink coding processing to target information and channel coding is handled, and obtains channel transmission information；
The receiver obtains the channel transmission information, and carries out time synchronization to the channel transmission information, quickly
Fourier transformation FFT and channel estimation obtain the channel estimation coefficient and noise estimated information of the channel transmission information；
The channel estimation coefficient and the noise estimated information are inputted as feature and are carried out by the network training unit
Training, obtains the adaptive modulation and coding AMC scheme of prediction, and the AMC scheme of the prediction is fed back to the transmitter,
So that the AMC scheme of the transmitter prediction is adaptively adjusted.
Further, the transmitter carries out message sink coding processing to target information and channel coding is handled, comprising:
The transmitter successively carries out error control coding, space analysis, quadrature amplitude modulation QAM tune to target information
System, Space-Time Block Coding STBC coding, space reflection, Fast Fourier Transform Inverse IFFT processing.
Further, described to carry out time synchronization to the channel transmission information, Fast Fourier Transform (FFT) FFT and channel are estimated
Meter, comprising:
Short training field (legacy short training field, L-STF) is extracted from the channel transmission information
Coarse frequency offset correction is executed, long training field (legacy long training field, L-LTF) is extracted and carries out fine frequency
Rate offset correction and noise estimation, according to high speed long training field (the high throughput in the channel transmission information
Long training field, HT-LTF) channel estimation is done, estimated matrix is obtained, the estimated matrix includes: fading channel
And space reflection；
The channel estimation coefficient includes the estimated matrix；
The noise estimated information includes the noise estimated result.
Further, described to carry out time synchronization to the channel transmission information, Fast Fourier Transform (FFT) FFT and channel are estimated
After meter, further includes:
Using the matrix parameter in the estimated matrix, and (Spatial Equalization) compartment equalization and
(Channel Equalization) channel equalization carries out multiple-input and multiple-output MIMO detection；
According to MIMO detection as a result, successively carrying out STBC decoding, QAM is demodulated, space analysis and Error Control solution
Code.
Further, the depth convolutional neural networks DCNN in the network training unit includes: convolutional layer, average pond
Layer and full articulamentum are constituted；
Wherein, the first layer of the DCNN and second layer hidden layer separately include 16,32 filters；It and then is average
Pond layer, moving step length 4；Third layer convolutional layer filter quantity increases to 64, is and then second average pond layer, moves
Dynamic step-length is 2；4th layer of convolutional layer filter is reduced to 32；The average pond layer of back to back third and second are unanimously；With
The 5th convolutional layer of subsequent continued access, filter are 16；Wherein, five convolution Chengdu be using the filter of 5*1 size,
It and is all to use RELU as activation primitive；Followed by 2 layers of fully-connected network, first layer includes 100 neurons, activation
Function is also RELU；Second layer neuronal quantity u and Modulation and Coding Scheme (modulation and coding scheme,
MCS quantity) is identical, and activation primitive uses softmax.
The embodiment of the present application second aspect provides link-adaptation system, comprising:
Transmitter obtains channel transmission information for carrying out message sink coding processing and channel coding processing to target information；
Receiver carries out time synchronization for obtaining the channel transmission information, and to the channel transmission information, quickly
Fourier transformation FFT and channel estimation obtain the channel estimation coefficient and noise estimated information of the channel transmission information；
Network training unit is carried out for inputting the channel estimation coefficient and the noise estimated information as feature
Training, obtains the adaptive modulation and coding AMC scheme of prediction, and the AMC scheme of the prediction is fed back to the transmitter,
So that the AMC scheme of the transmitter prediction is adaptively adjusted.
Further, the transmitter is specifically also used to:
Error control coding, space analysis are successively carried out to target information, quadrature amplitude modulation QAM modulation is grouped when empty
Code STBC coding, space reflection, Fast Fourier Transform Inverse IFFT processing.
Further, the receiver is specifically also used to:
L-STF is extracted from the channel transmission information and executes coarse frequency offset correction, and L-LTF carries out fine frequency offset
Correction and noise estimation, do channel estimation according to the HT-LTF in the channel transmission information, obtain estimated matrix, the estimation
Matrix includes: fading channel and space reflection；
The channel estimation coefficient includes the estimated matrix；
The noise estimated information includes the noise estimated result.
The embodiment of the present application third aspect provides a kind of electronic device, comprising: memory, processor and is stored in described deposit
On reservoir and the computer program that can run on the processor, when the processor executes the computer program, realize
The chain circuit self-adaptive method that above-mentioned the embodiment of the present application first aspect provides.
The embodiment of the present application fourth aspect provides a kind of computer readable storage medium, is stored thereon with computer program,
When the computer program is executed by processor, the link circuit self-adapting side that above-mentioned the embodiment of the present application first aspect provides is realized
Method.
Therefore the channel estimation coefficient that is handled using receiver of application scheme and noise estimated information are as net
The feature input of network training unit is trained depth convolutional neural networks, obtains the AMC scheme of prediction, and by the prediction
AMC scheme feed back to the transmitter so that the transmitter prediction AMC scheme adaptively adjusted.Even if depositing
In the case where actual damage, also suitable modulating-coding can be reliably predicted according to the characteristic for arriving MIMO-OFDM system
Scheme.
Detailed description of the invention
Fig. 1-a is the implementation process schematic diagram of chain circuit self-adaptive method provided by the embodiments of the present application；
Fig. 1-b is the workflow schematic diagram of transmitter provided by the embodiments of the present application；
Fig. 1-c is the workflow schematic diagram of receiver provided by the embodiments of the present application；
Fig. 2 is the electronic device construction schematic diagram that one embodiment of the application provides；
Fig. 3 is the Electronic Device Hardware structural schematic diagram that another embodiment of the application provides.
Specific embodiment
To enable present invention purpose, feature, advantage more obvious and understandable, below in conjunction with the application
Attached drawing in embodiment, the technical scheme in the embodiment of the application is clearly and completely described, it is clear that described reality
Applying example is only some embodiments of the present application, and not all embodiments.Based on the embodiment in the application, those skilled in the art
Member's every other embodiment obtained without making creative work, shall fall in the protection scope of this application.
It should be noted that, in this document, the terms "include", "comprise" or its any other variant are intended to non-row
His property includes, so that the process, method, article or the device that include a series of elements not only include those elements, and
And further include other elements that are not explicitly listed, or further include for this process, method, article or device institute it is intrinsic
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including being somebody's turn to do
There is also other identical elements in the process, method of element, article or device.
Herein, using the suffix for indicating such as " module ", " component " or " unit " of element only for advantageous
In explanation of the invention, there is no specific meanings for itself.
Embodiment one
The embodiment of the present application provides a kind of chain circuit self-adaptive method, and the mode is applied to multi-input multi-output orthogonal frequency division
Multiplexed MIMO-ofdm system, it is in practical applications, a set of the system comprises transmitter, receiver and network training unit
MIMO-OFDM system may include NtA transmitting antenna, NrA receiving antenna and NcA OFDM subcarrier.Implement in the present invention
In example, only it is illustrated with a set of transmitter and receiver, please refers to Fig. 1-a, which mainly includes following
Step:
101, transmitter carries out message sink coding processing to target information and channel coding is handled, and obtains channel transmission information；
Illustratively, target information can be the bit information flow or OFDM subcarrier of information source generation.
In embodiments of the present invention, message sink coding processing and channel coding processing include: error control coding, space solution
Analysis, quadrature amplitude modulation QAM modulation, Space-Time Block Coding STBC coding, space reflection, Fast Fourier Transform Inverse (inverse
Fast Fourier transform, IFFT) processing.
Specifically, please referring to Fig. 1-b, the bit information flow that transmitter is generated from information source uses convolutional encoding, coding speed
Rate is c, carries out error control coding.Then the bit information after coding is assigned to different spatial flow L by spatial parsers,
Interweave simultaneously in spatial flow and the enterprising line position of subcarrier.It then carries out QAM modulation and spatial time sharing group encodes (Space time
Block coding, STBC), column vector is used hereThe modulated information of M-QAM is indicated, wherein for m expression
OFDM symbol information, m ∈ { 1,2 ..., Mo, n indicates the carrier wave of each determination, n ∈ { 1,2 ..., Nc, LstRepresentation space stream
Number.Space reflection is carried out in next step, it will be by L by linear predictive codingstThe vector x of dimensionm[n] is converted to NtVector matrixIt finally executes Fast Fourier Transform Inverse (inverse fast Fourier transform, IFFT) and adds
In addition prefix (cyclic prefix, CP) before recycling.It is worth noting that, not all subcarrier all carries out data biography
Defeated, a portion is used to do pilot tone and guard band.The signal issued from transmitter passes through frequency-selective channel, this lining
The channel matrix of carrier wave n is expressed as
102, channel estimation coefficient and noise estimated information are obtained；
Receiver obtains the channel transmission information, and carries out time synchronization to the channel transmission information, in quick Fu
Leaf transformation FFT and channel estimation obtain the channel estimation coefficient and noise estimated information of the channel transmission information；
Specifically, described carry out time synchronization to the channel transmission information, Fast Fourier Transform (FFT) FFT and channel are estimated
Meter, comprising:
L-STF is extracted from the channel transmission information and executes coarse frequency offset correction, and L-LTF carries out fine frequency offset
Correction and σ2Noise estimation, does channel estimation according to the HT-LTF in the channel transmission information, obtains estimated matrixThe estimated matrix includes: fading channel H [n] and space reflection P [n]；The channel estimation system
Number includes the estimated matrix；The noise estimated information includes the noise estimated result.
Further, Fig. 1-c is please referred to, after FFT and channel estimation, receiver utilizes the matrix in the estimated matrix
Parameter and compartment equalization and channel equalization carry out multiple-input and multiple-output MIMO detection；According to the MIMO detection as a result, according to
Secondary progress STBC decoding, QAM demodulation, space analysis and error control decoding.
103, the adaptive modulation and coding AMC scheme of prediction is obtained, and the AMC scheme of the prediction is fed back to described
Transmitter.
Network training unit is trained using the channel estimation coefficient and the noise estimated information as feature input,
The adaptive modulation and coding AMC scheme of prediction is obtained, and the AMC scheme of the prediction is fed back into the transmitter, so that
The AMC scheme of the transmitter prediction is adaptively adjusted.
Therefore the channel estimation coefficient that is handled using receiver of application scheme and noise estimated information are as net
The feature input of network training unit is trained depth convolutional neural networks, obtains the AMC scheme of prediction, and by the prediction
AMC scheme feed back to the transmitter so that the transmitter prediction AMC scheme adaptively adjusted.Even if depositing
In the case where actual damage, also suitable modulating-coding can be reliably predicted according to the characteristic for arriving MIMO-OFDM system
Scheme.
Embodiment two
Deep learning has become a kind of powerful Machine learning tools, and obtains and be widely applied in many fields, such as
Computer vision, speech recognition etc..The embodiment of the present application provides a kind of depth convolutional neural networks (Deep
Convolutional Neural Network, DCNN), inputted using channel estimation coefficient and noise estimated information as feature into
Row training, the AMC scheme for prediction, comprising:
The learning method that the embodiment of the present application proposes a kind of use supervision is solved using DCNN in MIMO-OFDM system
In AMC problem.The adjoint estimation noise standard deviation of the mimo channel coefficient of estimation, all OFDM subcarriers being observed
It is used to train by the input feature vector as DCNN.In MIMO-OFDM system, AMC is considered as a polytypic problem, often
One classification represents a specific MCS.MCS defines the order of modulation M of QAM, ERROR CODING RATE c and spatial flow
Quantity Ls。
Parameter definition
The signal y of m-th of OFDM symbol of n-th of subcarrier that receiver receivesm[n] is indicated,
Here, ESIndicate transmitter signal power, it is assumed that noiseFor Gaussian noise, vm[n]~CN (0, σ2
Ι), mean value 0, variance σ2.Use estimated matrixDesign least mean-square error (minimum mean square
Error, MMSE) balanced device is equalized matrix Q [n],
HereForConjugate transposition.After equilibrium, the qam symbol vector estimatedxm[n]
By the qam symbol vector x of transmitting terminalm[n] is obtained:
Here it usesInstead of Q [n] G [n].
Feature extraction
In order to estimate to obtain disaggregated model, a training sample set is needed, altogether includes D sample, dthWhich instruction indicated
Practice sample, d ∈ { 1,2 ..., D } constitutes the channel parameter of the estimation of all subcarriers.Contain receiver estimation
Noise standard deviation σd。dthCarrier wave n estimation channel parameter matrix it is as follows:
HereThe channel coefficients for indicating estimation include space reflection and frequency selection fading channel, nrExpression connects
Receipts machine antenna serial number, l indicate space-time stream serial number.D expression training sample serial number, matrixEach value take absolute value so
After be converted to vectorForm is as follows:
Particularly, using feature vectorIndicate the channel estimation vector a of all training sample dd[n],
Wherein comprising all subcarrier n ∈ 1,2 ..., NcAnd noise bias σ.It is as follows:
fd=[ad[1],ad[2],...,ad[NC],σ] (6)
Training sample is based on handling capacity and packet loss (PER) determines class categories.The namely classification of determining sample set d
Label idThe corresponding MCS for realizing maximize handling capacity meets PER constraint ε, condition simultaneously.
Depth convolutional neural networks
Depth convolutional neural networks are made of convolutional layer, average pond layer and full articulamentum.First layer, second layer hidden layer
Separately include 16,32 filters.It and then is average pond layer, moving step length 4, zero padding mode is same.Third layer volume
Lamination filter quantity increases to 64, is and then second average pond layer, moving step length 2, zero padding is similarly same.The
Four layers of convolutional layer filter are reduced to 32.Then followed by the average pond layer of third and second unanimously.It continues thereafter with and connects
5th convolutional layer, filter are 16.Wherein five convolution Chengdu are to be using the filter of 5*1 size, and all
Use RELU as activation primitive.Followed by 2 layers of fully-connected network, first layer includes 100 neurons, and activation primitive is also
RELU.The quantity of second layer neuronal quantity u and MCS is identical, and activation primitive uses softmax.Two layers of fully-connected network
Standardization skill is used to mitigate the influence of over-fitting.Final choice adam optimizer, cross entropy loss function are used to instruct
Practice depth convolutional neural networks.Exercise wheel number is 100, and batch size is also 100.
Embodiment three
The AMC method and k-NN, SVM based on DCNN of proposition are assessed the embodiment of the invention provides Simulation results
It compares, comprising:
In experimental evaluation, the SNR feature of the post-processing of aforementioned whole characteristic dimension and sequence has been used[1]
To test DCNN, k-NN and the SVM using linear kernel respectively.
Emulation experiment is carried out in IEEE802.11n 2*2MIMO-OFDM system, wherein 54 carrier waves are for data biography
Defeated, 4 carrier waves do auxiliary transmission.Assuming that there is prolonged protection interval, cyclic prefix 800ns, carrier frequency is
5.25GHz.Analog frequency selects fading channel to use 9 taps, maximum delay 80ns.Between transmitter and receiver away from
From being 10 meters, transmission bandwidth 20MHz, the length of packet is 128 bytes.Consider that packet loss is set as ε=0.1.Assuming that directly into
Row space reflection, i.e. each space-time stream are mapped to a different transmission antenna, therefore the quantity of space-time stream is equal to hair
Penetrate the quantity of antenna, Lst=Nt=2.If space-time stream quantity be greater than spatial flow quantity in this case, only need to pass through
Spatial flow is mapped to space-time stream to execute STBC.The reliability of MCS can be improved using a transmitter spatial flow.
Training dataset producing method is as follows, simulates SNR different in decibel range from 8 to 40 in MIMO-OFDM system
Value, step-length are 1 decibel.For each SNR value, transmitted using all MCS, MCS is from 0 to 15, in 1000 different letters
The PER of each MCS is realized while calculated on road.Then, select ideal MCS as label, that is, in determining SNR channel
On can be realized maximum handling capacity while meeting the MCS of ε=0.1.The packet that estimation CSI and receiver are correctly detecting is made an uproar
Sound standard deviation is saved together to form training dataset with corresponding MCS label.In addition, post-processing the SNR of each spatial flow
By being calculated, then sort from low to high.The two characteristic data sets can all be used to train DCNN, k-NN and SVM.
In algorithm evaluation, emulation using the range of SNR value from 8 to 40 decibel, step-length 0.25.For each SNR
Value, has used 1000 channels to realize, it is therefore an objective to test ability of the different algorithms in some Unknown Channels.These algorithms
All employ the channel estimation feature of preceding formula (6) proposition and the post-processing SNR feature of sequence.It can be seen from experimental result
It observes, using the performance of the DCNN of channel estimation feature training and the packet loss of SVM algorithm and handling capacity all than using sequence
It is good to post-process SNR feature.In contrast, channel estimation feature is used using the k-NN algorithm ratio of the post-processing SNR feature of sequence
It performs better than.Under the constraint condition for meeting packet loss ε=0.1, the DCNN algorithm using the training of channel estimation feature of proposition
It is best in the performance of packet loss, in the very close ideal situation of the performance of handling capacity.
Example IV
Referring to Fig. 2, providing a kind of electronic device for the embodiment of the present application.The electronic device can be used for realizing above-mentioned Fig. 1-
The chain circuit self-adaptive method that a illustrated embodiment provides.As shown in Fig. 2, the electronic device specifically includes that
Transmitter 201 obtains transmission letter for carrying out message sink coding processing and channel coding processing to target information
Breath；
Receiver 202 carries out time synchronization for obtaining the channel transmission information, and to the channel transmission information,
Fast Fourier Transform (FFT) FFT and channel estimation obtain the channel estimation coefficient and noise estimated information of the channel transmission information；
Network training unit 203, for being inputted using the channel estimation coefficient and the noise estimated information as feature
It is trained, obtains the adaptive modulation and coding AMC scheme of prediction, and the AMC scheme of the prediction is fed back into the hair
Machine is penetrated, so that the AMC scheme of transmitter prediction is adaptively adjusted.
Further, the transmitter 201 is specifically also used to:
Error control coding, space analysis are successively carried out to target information, quadrature amplitude modulation QAM modulation is grouped when empty
Code STBC coding, space reflection, Fast Fourier Transform Inverse IFFT processing.
Further, the receiver 202 is specifically also used to:
L-STF is extracted from the channel transmission information and executes coarse frequency offset correction, and L-LTF carries out fine frequency offset
Correction and noise estimation, do channel estimation according to the HT-LTF in the channel transmission information, obtain estimated matrix, the estimation
Matrix includes: fading channel and space reflection；
The channel estimation coefficient includes the estimated matrix；
The noise estimated information includes the noise estimated result.
It should be noted that the division of each functional module is only to lift in the embodiment of the exemplary electronic device of figure 2 above
Example explanation, can according to need in practical application, such as the convenient of realization of configuration requirement or software of corresponding hardware considers,
And be completed by different functional modules above-mentioned function distribution, i.e., the internal structure of electronic device is divided into different function moulds
Block, to complete all or part of the functions described above.Moreover, in practical applications, the corresponding function in the present embodiment
Module can be by corresponding hardware realization, can also execute corresponding software by corresponding hardware and complete.This specification provides
Each embodiment all can apply foregoing description principle, repeat no more below.
The detailed process of the respective function of each Implement of Function Module, refers to above-mentioned figure in electronic device provided in this embodiment
Particular content described in 1-a illustrated embodiment, details are not described herein again.
Embodiment five
The embodiment of the present application provides a kind of electronic device, referring to Fig. 3, the electronic device includes:
Memory 301, processor 302 and it is stored in the computer journey that can be run on memory 301 and on processor 302
Sequence when processor 302 executes the computer program, realizes chain circuit self-adaptive method described in earlier figures 1-a illustrated embodiment.
Further, the electronic device further include:
At least one input equipment 303 and at least one output equipment 304.
Above-mentioned memory 301, processor 302, input equipment 303 and output equipment 304, are connected by bus 305.
Wherein, input equipment 303 concretely camera, touch panel, physical button or mouse etc..Output equipment
304 concretely display screens.
Memory 301 can be high random access memory body (RAM, Random Access Memory) memory,
It can be non-labile memory (non-volatile memory), such as magnetic disk storage.Memory 301 is for storing one
Group executable program code, processor 302 are coupled with memory 301.
Further, the embodiment of the present application also provides a kind of computer readable storage medium, the computer-readable storages
Medium can be in the electronic device being set in the various embodiments described above, which can be earlier figures 3
Memory in illustrated embodiment.It is stored with computer program on the computer readable storage medium, which is held by processor
Chain circuit self-adaptive method described in earlier figures 1-a illustrated embodiment is realized when row.Further, which can storage medium
It can also be that USB flash disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), RAM, magnetic or disk etc. are various
It can store the medium of program code.
In several embodiments provided herein, it should be understood that disclosed device and method can pass through it
Its mode is realized.For example, the apparatus embodiments described above are merely exemplary, for example, the division of the module, only
Only a kind of logical function partition, there may be another division manner in actual implementation, such as multiple module or components can be tied
Another system is closed or is desirably integrated into, or some features can be ignored or not executed.Another point, it is shown or discussed
Mutual coupling, direct-coupling or communication connection can be through some interfaces, the INDIRECT COUPLING or logical of device or module
Letter connection can be electrical property, mechanical or other forms.
The module as illustrated by the separation member may or may not be physically separated, aobvious as module
The component shown may or may not be physical module, it can and it is in one place, or may be distributed over multiple
On network module.Some or all of the modules therein can be selected to realize the mesh of this embodiment scheme according to the actual needs
's.
It, can also be in addition, can integrate in a processing module in each functional module in each embodiment of the application
It is that modules physically exist alone, can also be integrated in two or more modules in a module.Above-mentioned integrated mould
Block both can take the form of hardware realization, can also be realized in the form of software function module.
If the integrated module is realized in the form of software function module and sells or use as independent product
When, it can store in a computer readable storage medium.Based on this understanding, the technical solution of the application is substantially
The all or part of the part that contributes to existing technology or the technical solution can be in the form of software products in other words
It embodies, which is stored in a readable storage medium storing program for executing, including some instructions are used so that a meter
It calculates machine equipment (can be personal computer, server or the network equipment etc.) and executes each embodiment the method for the application
All or part of the steps.And readable storage medium storing program for executing above-mentioned includes: USB flash disk, mobile hard disk, ROM, RAM, magnetic or disk etc.
The various media that can store program code.
It should be noted that for the various method embodiments described above, describing for simplicity, therefore, it is stated as a series of
Combination of actions, but those skilled in the art should understand that, the application is not limited by the described action sequence because
According to the application, certain steps can use other sequences or carry out simultaneously.Secondly, those skilled in the art should also know
It knows, the embodiments described in the specification are all preferred embodiments, and related actions and modules might not all be this Shen
It please be necessary.
In the above-described embodiments, it all emphasizes particularly on different fields to the description of each embodiment, there is no the portion being described in detail in some embodiment
Point, it may refer to the associated description of other embodiments.
The above are retouching to chain circuit self-adaptive method provided herein, electronic device and computer readable storage medium
It states, for those skilled in the art, according to the thought of the embodiment of the present application, can in specific embodiments and applications
There is change place, to sum up, the contents of this specification should not be construed as limiting the present application.
Claims (10)
1. a kind of chain circuit self-adaptive method, the mode is applied to multi-I/O OFDM MIMO-OFDM system,
It is characterized in that, the system comprises transmitter, receiver and network training units, comprising:
The transmitter carries out message sink coding processing to target information and channel coding is handled, and obtains channel transmission information；
The receiver obtains the channel transmission information, and carries out time synchronization to the channel transmission information, in quick Fu
Leaf transformation FFT and channel estimation obtain the channel estimation coefficient and noise estimated information of the channel transmission information；
The network training unit is trained using the channel estimation coefficient and the noise estimated information as feature input,
The adaptive modulation and coding AMC scheme of prediction is obtained, and the MAC scheme of the prediction is fed back into the transmitter, so that
The transmitter is adaptively adjusted according to MAC scheme.
2. the method according to claim 1, wherein
The transmitter carries out message sink coding processing to target information and channel coding is handled, comprising:
The transmitter successively carries out error control coding, space analysis, quadrature amplitude modulation QAM modulation, sky to target information
Time block code STBC coding, space reflection, Fast Fourier Transform Inverse IFFT processing.
3. the method according to claim 1, wherein
It is described that time synchronization, Fast Fourier Transform (FFT) FFT and channel estimation are carried out to the channel transmission information, comprising:
Short training field L-STF is extracted from the channel transmission information and executes coarse frequency offset correction, extracts long training field
L-LTF carries out fine frequency offset correction and noise estimation, according to the high speed long training field HT- in the channel transmission information
LTF does channel estimation, obtains estimated matrix, and the estimated matrix includes: fading channel and space reflection；
The channel estimation coefficient includes the estimated matrix；
The noise estimated information includes: the noise estimated result.
4. according to the method described in claim 3, it is characterized in that,
It is described that time synchronization carried out to the channel transmission information, after Fast Fourier Transform (FFT) FFT and channel estimation, also wrap
It includes:
Utilize the matrix parameter and compartment equalization and channel equalization progress multiple-input and multiple-output MIMO inspection in the estimated matrix
It surveys；
According to MIMO detection as a result, successively carrying out STBC decoding, QAM is demodulated, space analysis and error control decoding.
5. the method according to claim 1, wherein
Depth convolutional neural networks DCNN in the network training unit includes: convolutional layer, average pond layer and full articulamentum
It constitutes；
Wherein, the first layer of the DCNN and second layer hidden layer separately include 16,32 filters；It and then is average pond
Layer, moving step length 4；Third layer convolutional layer filter quantity increases to 64, is and then second average pond layer, mobile step
A length of 2；4th layer of convolutional layer filter is reduced to 32；The average pond layer of back to back third and second are unanimously；With subsequent
The 5th convolutional layer of continued access, filter are 16；Wherein, five convolution Chengdu be using the filter of 5*1 size, and
It is all to use RELU as activation primitive；Followed by 2 layers of fully-connected network, first layer includes 100 neurons, activation primitive
It is also RELU；Second layer neuronal quantity u is identical with the quantity of Modulation and Coding Scheme MCS, and activation primitive uses
softmax。
6. a kind of link-adaptation system characterized by comprising
Transmitter obtains channel transmission information for carrying out message sink coding processing and channel coding processing to target information；
Receiver carries out time synchronization for obtaining the channel transmission information, and to the channel transmission information, in quick Fu
Leaf transformation FFT and channel estimation obtain the channel estimation coefficient and noise estimated information of the channel transmission information；
Network training unit, for being instructed using the channel estimation coefficient and the noise estimated information as feature input
Practice, obtains the adaptive modulation and coding AMC scheme of prediction, and the AMC scheme of the prediction is fed back into the transmitter, make
The AMC scheme for obtaining the transmitter prediction is adaptively adjusted.
7. the method according to claim 1, wherein
The transmitter is specifically also used to:
Error control coding, space analysis, quadrature amplitude modulation QAM modulation, Space-Time Block Coding are successively carried out to target information
STBC coding, space reflection, Fast Fourier Transform Inverse IFFT processing.
8. the method according to claim 1, wherein
The receiver is specifically also used to:
L-STF is extracted from the channel transmission information and executes coarse frequency offset correction, and L-LTF carries out fine frequency offset correction
Estimate with noise, channel estimation is done according to the HT-LTF in the channel transmission information, obtains estimated matrix, the estimated matrix
It include: fading channel and space reflection；
The channel estimation coefficient includes the estimated matrix；
The noise estimated information includes the noise estimated result.
9. a kind of electronic device, comprising: memory, processor and be stored on the memory and can transport on the processor
Capable computer program, which is characterized in that when the processor executes the computer program, realize in claim 1 to 5
Any one the method.
10. a kind of computer readable storage medium, is stored thereon with computer program, which is characterized in that the computer program
When being executed by processor, any one the method in claim 1 to 5 is realized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811510754.6A CN109412996A (en) | 2018-12-11 | 2018-12-11 | Chain circuit self-adaptive method, electronic device and computer readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201811510754.6A CN109412996A (en) | 2018-12-11 | 2018-12-11 | Chain circuit self-adaptive method, electronic device and computer readable storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109412996A true CN109412996A (en) | 2019-03-01 |
Family
ID=65458540
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201811510754.6A Pending CN109412996A (en) | 2018-12-11 | 2018-12-11 | Chain circuit self-adaptive method, electronic device and computer readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109412996A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109379120A (en) * | 2018-12-11 | 2019-02-22 | 深圳大学 | Chain circuit self-adaptive method, electronic device and computer readable storage medium |
| CN110417699A (en) * | 2019-05-30 | 2019-11-05 | 北京邮电大学 | A method of the ofdm system timing synchronization based on machine learning |
| CN111510402A (en) * | 2020-03-12 | 2020-08-07 | 西安电子科技大学 | OFDM channel estimation method based on deep learning |
| CN112383496A (en) * | 2020-11-09 | 2021-02-19 | 西安科技大学 | Mine communication method, system, computer equipment and medium based on depth receiver |
| WO2021094119A1 (en) * | 2019-11-15 | 2021-05-20 | Sony Corporation | Communications devices, infrastructure equipment and methods |
| CN113055110A (en) * | 2021-03-10 | 2021-06-29 | 西安电子科技大学 | Multi-carrier system link self-adaption method based on channel identification and channel classification |
| WO2021259114A1 (en) * | 2020-06-24 | 2021-12-30 | 中兴通讯股份有限公司 | Outer ring value determination method and apparatus, and device and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101340223A (en) * | 2008-08-07 | 2009-01-07 | 中兴通讯股份有限公司 | System resource distribution method and device based on wave beam shaping |
| US20090060076A1 (en) * | 2001-10-17 | 2009-03-05 | Nortel Networks Limited | System access and synchronization methods for mimo ofdm communications systems and physical layer packet and preamble |
| CN105577248A (en) * | 2015-12-25 | 2016-05-11 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | Data demodulating method and system of MIMO-OFDM (Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing) receiver |
-
2018
- 2018-12-11 CN CN201811510754.6A patent/CN109412996A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20090060076A1 (en) * | 2001-10-17 | 2009-03-05 | Nortel Networks Limited | System access and synchronization methods for mimo ofdm communications systems and physical layer packet and preamble |
| CN101340223A (en) * | 2008-08-07 | 2009-01-07 | 中兴通讯股份有限公司 | System resource distribution method and device based on wave beam shaping |
| CN105577248A (en) * | 2015-12-25 | 2016-05-11 | 广东顺德中山大学卡内基梅隆大学国际联合研究院 | Data demodulating method and system of MIMO-OFDM (Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing) receiver |
Non-Patent Citations (1)
| Title |
|---|
| MOHAMED ELWEKEIL等: ""Deep Convolutional Neural Networks for Link Adaptations in MIMO-OFDM Wireless Systems"", 《IEEE WIRELESS COMMUNICATIONS LETTERS》 * |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109379120A (en) * | 2018-12-11 | 2019-02-22 | 深圳大学 | Chain circuit self-adaptive method, electronic device and computer readable storage medium |
| CN110417699A (en) * | 2019-05-30 | 2019-11-05 | 北京邮电大学 | A method of the ofdm system timing synchronization based on machine learning |
| WO2021094119A1 (en) * | 2019-11-15 | 2021-05-20 | Sony Corporation | Communications devices, infrastructure equipment and methods |
| CN111510402A (en) * | 2020-03-12 | 2020-08-07 | 西安电子科技大学 | OFDM channel estimation method based on deep learning |
| CN111510402B (en) * | 2020-03-12 | 2021-08-27 | 西安电子科技大学 | OFDM channel estimation method based on deep learning |
| WO2021259114A1 (en) * | 2020-06-24 | 2021-12-30 | 中兴通讯股份有限公司 | Outer ring value determination method and apparatus, and device and storage medium |
| CN112383496A (en) * | 2020-11-09 | 2021-02-19 | 西安科技大学 | Mine communication method, system, computer equipment and medium based on depth receiver |
| CN113055110A (en) * | 2021-03-10 | 2021-06-29 | 西安电子科技大学 | Multi-carrier system link self-adaption method based on channel identification and channel classification |
| CN113055110B (en) * | 2021-03-10 | 2021-11-30 | 西安电子科技大学 | Multi-carrier system link self-adaption method based on channel identification and channel classification |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109379120A (en) | Chain circuit self-adaptive method, electronic device and computer readable storage medium | |
| CN109412996A (en) | Chain circuit self-adaptive method, electronic device and computer readable storage medium | |
| CN113472706B (en) | MIMO-OFDM system channel estimation method based on deep neural network | |
| Ozpoyraz et al. | Deep learning-aided 6G wireless networks: A comprehensive survey of revolutionary PHY architectures | |
| CN1941659B (en) | Spatial multiplexing detection apparatus and method in MIMO system | |
| Elwekeil et al. | Deep convolutional neural networks for link adaptations in MIMO-OFDM wireless systems | |
| Kim et al. | Towards deep learning-aided wireless channel estimation and channel state information feedback for 6G | |
| Liu et al. | Predictive precoder design for OTFS-enabled URLLC: A deep learning approach | |
| Li et al. | A novel CSI feedback approach for massive MIMO using LSTM-attention CNN | |
| CN113452642A (en) | Reconfigurable intelligent surface enhanced SISO-OFDM downlink transmission method | |
| CN114745233B (en) | Joint channel estimation method and device based on pilot frequency design | |
| Xu et al. | RC-struct: A structure-based neural network approach for MIMO-OFDM detection | |
| Chen et al. | Viewing channel as sequence rather than image: A 2-D Seq2Seq approach for efficient MIMO-OFDM CSI feedback | |
| Dehri et al. | Blind digital modulation classification for STBC‐OFDM system in presence of CFO and channels estimation errors | |
| Ye et al. | Bilinear convolutional auto-encoder based pilot-free end-to-end communication systems | |
| Raviv et al. | Data augmentation for deep receivers | |
| CN101136721A (en) | Mixing decision feedback layered detection method based on suboptimal sorting | |
| CN102025459B (en) | Nonparametric estimation ICA-based MIMO-OFDM system blind deconvolution method | |
| CN101621306B (en) | Mapping method and device for multiple-input multiple-output system precoding matrix | |
| Khosraviyani et al. | Higher order statistics for modulation and STBC recognition in MIMO systems | |
| Li et al. | Model-driven deep learning scheme for adaptive transmission in MIMO-SCFDE system | |
| Yıldırım et al. | Deep receiver design for multi-carrier waveforms using cnns | |
| CN116383656A (en) | Semi-supervised characterization contrast learning method for large-scale MIMO positioning | |
| US20100074376A1 (en) | Sphere decoding method applied to multi-input multi-output (mimo) channel | |
| CN106301493A (en) | A kind of beam-forming method based on MIMO-OFDMA system down link and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190301 |
|
| RJ01 | Rejection of invention patent application after publication |