CN114722905A - Training method and device for optical communication receiving model - Google Patents
Training method and device for optical communication receiving model Download PDFInfo
- Publication number
- CN114722905A CN114722905A CN202210228714.2A CN202210228714A CN114722905A CN 114722905 A CN114722905 A CN 114722905A CN 202210228714 A CN202210228714 A CN 202210228714A CN 114722905 A CN114722905 A CN 114722905A
- Authority
- CN
- China
- Prior art keywords
- data
- model
- optical fiber
- training
- dimensional
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects
-
- 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
- H04B10/60—Receivers
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Evolutionary Computation (AREA)
- Computer Networks & Wireless Communication (AREA)
- Computational Linguistics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Electromagnetism (AREA)
- Bioinformatics & Computational Biology (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computer Graphics (AREA)
- Geometry (AREA)
- Optical Communication System (AREA)
Abstract
The present disclosure provides a training method for an optical communication receiving model, including: acquiring N strings of information source communication symbols and N optical fiber transmission signals; n is more than or equal to 2, and N optical fiber transmission signals are signals transmitted by each string of information source communication symbols in N strings of information source communication symbols at different optical fiber distances; carrying out normalization processing on the N optical fiber transmission signals to obtain normalized model input signals; constructing a three-dimensional tensor of the normalized model input signal and a two-dimensional tensor of the N strings of information source communication symbols to obtain three-dimensional model input data and two-dimensional label data; and training an optical communication receiving model by using part of the three-dimensional model input data and the two-dimensional label data as training data, wherein the optical communication receiving model is used for restoring the optical fiber transmission signal into a source communication symbol corresponding to the optical fiber transmission signal. The disclosure also provides a restoration method and device of the optical fiber transmission signal, electronic equipment, a storage medium and a computer program product.
Description
Technical Field
The present disclosure relates to the field of optical fiber communication and machine learning technologies, and in particular, to a training method, a recovery method, an apparatus, an electronic device, a storage medium, and a program product for an optical communication receiving model.
Background
The optical receiving algorithm is used as an important research field of optical fiber communication, and plays a crucial role in the final extraction of optical information. In order to recover the transmission signal as correctly as possible, the optical reception algorithm needs to compensate not only the dispersion in the optical fiber transmission but also the nonlinearity thereof. In addition, the optical receiving algorithm also has good noise resistance. Therefore, the traditional light receiving algorithm and the optimized research network thereof take the functional modules as the core, and are researched and optimized in blocks. Although the research mode is beneficial to breaking the problem into parts and optimizing the problem in blocks according to different tasks and scenes, the whole receiver system lacks complete optimization and algorithm integration characteristics, so that various functional module algorithms are respectively operated and are difficult to coordinate. And the narrow-band nature of conventional optical or electrical compensation for dispersion also limits the versatility of the receiver.
Disclosure of Invention
In order to solve the problems in the prior art, embodiments of the present disclosure provide a training method, a recovery method, an apparatus, an electronic device, a storage medium, and a program product for an optical communication receiving model, which are intended to improve the general performance of a receiver and the high recovery rate of an information source signal.
A first aspect of the present disclosure provides a training method for an optical communication reception model, including: acquiring N strings of information source communication symbols and N optical fiber transmission signals; n is more than or equal to 2, and N optical fiber transmission signals are signals transmitted by each string of information source communication symbols in N strings of information source communication symbols at different optical fiber distances; carrying out normalization processing on the N optical fiber transmission signals to obtain normalized model input signals; constructing a three-dimensional tensor of the normalized model input signal and a two-dimensional tensor of the N strings of information source communication symbols to obtain three-dimensional model input data and two-dimensional label data; and training an optical communication receiving model by using part of three-dimensional model input data and two-dimensional label data as training data, wherein the optical communication receiving model is used for restoring the optical fiber transmission signal into an information source communication symbol corresponding to the optical fiber transmission signal.
Further, training an optical communication reception model using a part of the three-dimensional model input data and the two-dimensional tag data as training data includes: setting initial training parameters of an optical communication receiving model, wherein the initial training parameters at least comprise a maximum iteration parameter threshold and a minimum loss function threshold; inputting part of three-dimensional model input data and two-dimensional label data into an optical communication receiving model for training; judging whether the algebra of the current iteration is larger than a maximum iteration parameter threshold or smaller than a minimum loss function threshold; if so, ending the iteration, and fixing the model parameters after the optical communication receiving model training.
Further, the method further comprises: testing the optical communication receiving model by using the other part of three-dimensional model input data and the two-dimensional label data as test data to obtain a tested optical communication receiving model; the test data is data except training data in the three-dimensional model input data and the two-dimensional label data.
Further, the step of testing the optical communication receiving model by using another part of the three-dimensional model input data and the two-dimensional label data as test data to obtain a tested optical communication receiving model includes: testing the optical communication receiving model by using the other part of the three-dimensional model input data and the two-dimensional label data as test data; comparing and analyzing the result obtained by the model training with the other part of the two-dimensional label data to obtain the error rate of the optical communication receiving model training; and adjusting the initial training parameters of the optical communication receiving model according to the error rate so as to optimize the optical communication receiving model.
Further, the optical communication reception model includes: a Transformer encoder comprising: a first residual error structure, a first data processing structure, a second residual error structure and a second data processing structure; the first residual error structure is used for copying the three-dimensional model input data into two identical three-dimensional optical fiber data; the first data processing structure is used for performing feature extraction on one part of three-dimensional model input data and performing data aggregation on the extracted feature data and the other part of three-dimensional model input data to obtain first aggregated data; the first residual error structure is used for copying the first aggregation data into two identical optical fiber data; the second data processing structure is used for performing secondary feature extraction on one part of data output by the first residual error structure and performing secondary data aggregation on the extracted feature data and the other part of data to obtain second aggregated data; the depth full-connection network structure is used for performing information fusion on the second aggregation data to obtain fused data; and the signal judgment layer is used for restoring the fused data into the information source communication symbol corresponding to the fused data.
Further, the method further comprises: outputting a source communication symbol using a source generator; converting the information source communication symbols into electric signals by using a waveform generator; loading the electric signal to an optical carrier by using a modulator to obtain a modulated initial optical fiber signal; and transmitting the modulated initial optical fiber signal by adopting N optical fiber links with different transmission distances to obtain N optical fiber transmission signals.
A second aspect of the present disclosure provides a method for restoring a signal transmitted through an optical fiber, including: acquiring at least one optical fiber transmission signal; the at least one optical fiber transmission signal is restored to the source communication symbol corresponding to the at least one optical fiber transmission signal by using an optical communication receiving model, wherein the optical communication receiving model is trained by using the training method of the optical communication receiving model provided by the first aspect of the disclosure.
Further, the communication symbol recovered by the at least one optical fiber transmission signal is consistent with the corresponding source communication symbol.
A third aspect of the present disclosure provides a training apparatus for an optical communication reception model, including: the signal acquisition module is used for acquiring N strings of information source communication symbols and N optical fiber transmission signals; n is more than or equal to 2, and N optical fiber transmission signals are respectively signals of N strings of information source communication symbols after transmission through N different optical fiber distances; the signal preprocessing module is used for carrying out normalization processing on the N optical fiber transmission signals to obtain normalized model input signals; the data construction module is used for constructing a three-dimensional tensor of the normalized model input signal and a two-dimensional tensor of the N strings of information source communication symbols to obtain three-dimensional model input data and two-dimensional label data; and the model training module is used for training an optical communication receiving model by using part of three-dimensional model input data and two-dimensional label data as training data, and the optical communication receiving model is used for restoring the optical fiber transmission signal into an information source communication symbol corresponding to the optical fiber transmission signal.
A fourth aspect of the present disclosure provides a restoring apparatus for an optical fiber transmission signal, including: the data acquisition module is used for acquiring at least one optical fiber transmission signal; and a data restoring module, configured to restore the at least one optical fiber transmission signal into a source communication symbol corresponding to the at least one optical fiber transmission signal by using an optical communication receiving model, where the optical communication receiving model is trained by using a training method of the optical communication receiving model provided in the first aspect of the present disclosure.
A fifth aspect of the present disclosure provides an electronic device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the method provided by the first aspect of the present disclosure or the method provided by the second aspect of the present disclosure when executing the computer program.
A sixth aspect of the present disclosure provides a computer readable storage medium having stored thereon a computer program which, when executed by a processor, performs the method provided by the first aspect of the present disclosure or the method provided by the second aspect of the present disclosure.
A seventh aspect of the disclosure provides a computer program product comprising a computer program which, when executed by a processor, performs the method provided by the first aspect of the disclosure or the method provided by the second aspect of the disclosure.
It should be understood that the statements in this section do not necessarily identify key or critical features of the embodiments of the present disclosure, nor do they limit the scope of the present disclosure. Other features of the present disclosure will become apparent from the following description.
Drawings
For a more complete understanding of the present disclosure and the advantages thereof, reference is now made to the following descriptions taken in conjunction with the accompanying drawings, in which:
fig. 1 schematically shows a structural diagram of a conventional optical communication system;
fig. 2 schematically illustrates a flow chart of a training method of an optical communication reception model according to an embodiment of the present disclosure;
FIG. 3 schematically shows a schematic diagram of three-dimensional model input data according to an embodiment of the present disclosure;
FIG. 4 schematically illustrates a schematic diagram of two-dimensional tag data according to an embodiment of the disclosure;
fig. 5 schematically shows a structural diagram of an optical communication reception model according to an embodiment of the present disclosure;
FIG. 6 schematically illustrates a training flow diagram of an optical communication reception model according to an embodiment of the present disclosure;
fig. 7 schematically illustrates a block diagram of an apparatus for training an optical communication reception model according to an embodiment of the present disclosure;
FIG. 8 is a block diagram schematically illustrating an apparatus for restoring a signal transmitted through an optical fiber according to an embodiment of the present disclosure;
fig. 9 schematically shows a block diagram of an electronic device adapted to implement the above described method according to an embodiment of the present disclosure.
Detailed Description
Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. It should be understood that the description is illustrative only and is not intended to limit the scope of the present disclosure. In the following detailed description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the embodiments of the disclosure. It may be evident, however, that one or more embodiments may be practiced without these specific details. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure.
The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the disclosure. The terms "comprises," "comprising," and the like, as used herein, specify the presence of stated features, steps, operations, and/or components, but do not preclude the presence or addition of one or more other features, steps, operations, or components.
All terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art unless otherwise defined. It is noted that the terms used herein should be interpreted as having a meaning that is consistent with the context of this specification and should not be interpreted in an idealized or overly formal sense.
Where a convention analogous to "at least one of A, B and C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B and C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.). Where a convention analogous to "A, B or at least one of C, etc." is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., "a system having at least one of A, B or C" would include but not be limited to systems that have a alone, B alone, C alone, a and B together, a and C together, B and C together, and/or A, B, C together, etc.).
Some block diagrams and/or flow diagrams are shown in the figures. It will be understood that some blocks of the block diagrams and/or flowchart illustrations, or combinations thereof, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the instructions, which execute via the processor, create means for implementing the functions/acts specified in the block diagrams and/or flowchart block or blocks. The techniques of this disclosure may be implemented in hardware and/or software (including firmware, microcode, etc.). In addition, the techniques of this disclosure may take the form of a computer program product on a computer-readable storage medium having instructions stored thereon for use by or in connection with an instruction execution system.
Fig. 1 schematically shows a structural diagram of a conventional optical communication system. As shown in fig. 1, in a conventional optical communication system, a communication symbol transmitted from a source is mapped to a waveform signal by an arbitrary waveform generator and then modulated on an optical carrier by a modulator. After the received modulated optical signal is transmitted in a long distance in an optical fiber link, the modulated optical signal is demodulated by a photoelectric detection module at a receiving end and then input into a signal processing module to recover information source information.
In a conventional signal processing module as identified by a black box in fig. 1, aliasing occurs between signal symbols due to chromatic dispersion and nonlinear equivalent of an optical fiber, so that a specific dispersion compensation algorithm and a nonlinear equalization algorithm for different transmission scenarios are applied to the module to eliminate intersymbol interference of signals. In addition, the module needs to make decisions on the signal after crosstalk cancellation to recover the information transmitted by the source as accurately as possible.
The model/device obtained by the training method of the optical communication receiving model provided by the embodiment of the disclosure aims to compensate the non-ideal factors of the signal, judge the code element and finally recover the information source information.
Fig. 2 schematically shows a flow chart of a training method of an optical communication reception model according to an embodiment of the present disclosure. As shown in fig. 2, the method includes: steps S201 to S204.
In operation S201, N series of source communication symbols and N optical fiber transmission signals are obtained. N is more than or equal to 2, and the N optical fiber transmission signals are signals transmitted by each string of information source communication symbols in the N strings of information source communication symbols through different optical fiber distances.
In the embodiment of the present disclosure, an optical fiber communication system is constructed according to the structure shown in fig. 1, and N different transmission distances are selected as the transmission distances of an optical fiber link, for example: 0.5km, 1km, 1.5km, 2km, …, 99.5km, and 100km, symbol rates may be 10GBaud, 20GBaud, and 40GBaud, respectively, a sampling rate K times the symbol rate, and a modulation format may be OOK, PAM, or the like.
Further, as shown in fig. 1, N series of information source communication symbols may be N series of different or the same communication symbols transmitted by the information source, the N series of information source communication symbols are converted into electrical signals by using a waveform generator, then the electrical signals are loaded onto an optical carrier by using a modulator to obtain modulated initial optical fiber signals, and then the modulated initial optical fiber signals are transmitted by using N optical fiber links with different transmission distances to obtain N optical fiber transmission signals. Following the above embodiment, N series of source communication symbols are processed in a series and then transmitted through optical fiber links of 0.5km, 1km, 1.5km, 2km, …, 99.5km, and 100km, for example: and a first string of information source communication symbols in the N strings of information source communication symbols are transmitted through a 0.5km optical fiber link, a second string of information source communication symbols are transmitted through a 1km optical fiber link, a third string of information source communication symbols are transmitted through a 1.5km optical fiber link, and the rest is done in sequence to obtain N optical fiber transmission signals respectively corresponding to the N strings of information source communication symbols. N optical fiber transmission signals are used as input data sources of model training, and N strings of information source communication symbols are used as label signals for subsequent model training and comparative analysis of training results.
It should be noted that the transmission distance range of the optical fiber link in the above embodiments is 0km to 100km and the different transmission distances are only exemplary illustrations, and it does not limit the embodiments of the disclosure. In a practical application scenario, when training or reducing signals in a long-distance optical communication system, the transmission distance of the optical fiber link may be 500km, 1000km, etc., and the signals obtained by equidistant transmission distance may be signals spaced by 20km, 50km, or 100km, etc.
In operation S202, N optical fiber transmission signals are normalized to obtain a normalized model input signal.
In the embodiment of the disclosure, before the signal is input into the model training, the signal needs to be preprocessed, including normalization processing of the signal, so as to facilitate the subsequent construction of tensor on the signal and generate the model training data set.
Specifically, the N optical fiber transmission signals are complex signals, and the N optical fiber transmission signals are normalized to obtain normalized model input signals, including: and dividing the N optical fiber transmission signals by the maximum value of the corresponding mode, and respectively solving the real part and the imaginary part of the N optical fiber transmission signals to obtain the normalized model input signal.
In operation S203, a three-dimensional tensor of the normalized model input signal and a two-dimensional tensor of the N-string source communication symbols are constructed, and three-dimensional model input data and two-dimensional tag data are obtained.
In the embodiment of the disclosure, for the model input signal, the three-dimensional tensor is constructed to meet the model trainingAnd inputting the data format of the data. As shown in fig. 3, the model input data is a three-dimensional tensor of m × n × 2k, which can be regarded as m × n storing each communication symbol VijM x k two-dimensional matrix of (2), the element V thereinijIs a vector having 2k elements, as shown by a in FIG. 31、a2、..、ak、b1、b2、...、bk. The matrix is a matrix with m rows and n columns, and the number of rows m and the number of columns n are respectively determined by the scale of the model batch calculation and the scale of the neurons, which is not limited in the embodiment of the disclosure.
As shown in FIG. 3, for each element V of the matrixijThe waveform of the signal corresponding to the communication symbol represented by the signal is described by a 2 k-element vector, and the length parameter k of the description vector is determined by the sampling rate of the communication signal. In the vector, the elements 1 to k describe the real part of the sampled signal value, such as: a is1、a2、…、ak(ii) a The elements (k +1) to 2k describe the imaginary part of the sampled signal value, e.g. b1、b2、…、bk. Although the model has the functions of processing and judging received signals at different distances, the input data of the model does not contain transmission distance information because the model has good distance generalization performance.
In the embodiment of the disclosure, for the tag data, a two-dimensional tensor is constructed to obtain two-dimensional tag data, and a data format of the two-dimensional tag data is shown in fig. 4. It should be noted that the format of the output data of the three-dimensional tensor data after model training provided by the present disclosure is consistent with the data format shown in fig. 4, so that the output result of the model and the label data are compared and analyzed to obtain the training result or the test result of the model, and further, error analysis and the like can be performed.
Specifically, as shown in fig. 4, the physical meaning of the data format is the communication symbol obtained by decision, which is an m × n matrix corresponding to the element c in the ith row and the jth columnijThe corresponding waveform can be regarded as the element V of the ith row and the jth column of the input dataijThe 2k sampled signal points represented. Under the digital communication system, each of the tag dataThe elements typically take discrete values.
In the above embodiments, m, n, k, i, and j are all any positive integer, i is not more than m, and j is not more than n. In addition, the examples of the real part, the imaginary part, and the like in the foregoing embodiments are only for illustrative purposes, and do not limit the embodiments of the disclosure.
In operation S204, the optical communication reception model is trained using the partial three-dimensional model input data and the two-dimensional tag data as training data. The optical communication receiving model obtained by training restores the optical fiber transmission signal into the information source communication symbol corresponding to the optical fiber transmission signal.
In the embodiment of the present disclosure, the structure of the constructed optical communication receiving model is shown in fig. 5, and specifically includes: the device comprises a Transformer coding module, a depth full-connection network structure and a signal judgment layer.
Specifically, the Transformer encoding module comprises: the device comprises a first residual error structure, a first data processing structure, a second residual error structure and a second data processing structure. The first residual error structure is used for copying the three-dimensional model input data into two identical three-dimensional optical fiber data. A first data processing structure comprising: the multi-head attention mechanism layer is used for carrying out feature extraction on a part of three-dimensional model input data, and specifically extracting multiple aliasing features of the signals after aliasing; and the addition and normalization layer is used for carrying out data aggregation on the extracted characteristic data and the other three-dimensional model input data to obtain first aggregation data. The second residual structure is used to duplicate the first aggregated data into two identical copies of the fiber data. A second data processing structure comprising: the device comprises a forward calculation structure and an addition and normalization layer, wherein the forward calculation structure is used for performing secondary feature extraction on one part of data output by the second residual error structure, and the addition and normalization layer is used for performing secondary data aggregation on the extracted feature data and the other part of data to obtain second aggregated data.
Further, the deep fully-connected network structure is used for compensating signal aliasing caused by various chromatic dispersion and nonlinear effects in a transmission process, and specifically comprises the following steps: the device comprises a first full-connection layer, a first nonlinear activation layer, a second full-connection layer, a second nonlinear activation layer and a third full-connection layer, wherein the three full-connection layers are respectively used for carrying out deep fusion on the primary features extracted by the transform coder by utilizing the connection among network neurons, and the deeper the layer number is, the higher the feature fusion degree is.
And further, the signal decision layer is used for restoring the fused data into the source communication symbol corresponding to the fused data.
In the embodiment of the disclosure, the deep fully-connected network structure performs information fusion so as to better eliminate crosstalk between symbols and recover signals before aliasing to the maximum extent. The arrangement of the residual error network structure can effectively avoid the phenomenon that the gradient disappears in the training process. And the signal decision layer positioned at the extreme end of the network model maps the signals after the de-aliasing to the source communication information. Different from the traditional decision algorithm designed based on the optimal decision probability of matched filtering, the decision network can continuously optimize the decision threshold according to the learning of a large amount of data, so that the performance of optimal decision can be approached or even reached.
In the embodiment of the present disclosure, the three-dimensional model input data obtained in step S203 and the two-dimensional label data are in one-to-one correspondence to establish a total data set for model training, and then part of data is selected from the total data set as test set data to train the established model. Continuing to use the above embodiment, data with transmission distances of 1km, 2km, 3km, … km and 100km in the total data set can be selected as a training data set for model training; and data with transmission distances of 0.5km, 1.5km, 2.5km, 3.5km,. and 99.5km are used as a test data set for model testing.
It should be noted that the division of the data set is only an exemplary illustration, and it may be data selection in other manners, and the embodiment of the disclosure is not limited thereto.
According to an embodiment of the present disclosure, as shown in fig. 6, training an optical communication receiving model by using a part of three-dimensional model input data and two-dimensional label data as training data specifically includes: steps S601 to S606.
In operation S601, initial training parameters of an optical communication reception model are set. The initial training parameters at least comprise a multi-head number and a nonlinear activation function, a model training optimizer, a learning rate, a maximum iteration parameter threshold and a minimum loss function threshold.
In the embodiment of the disclosure, the initial training parameters of the model may be directly configured by default of the model or manually set according to the requirements of the data type and the model precision, so as to meet the requirement of accuracy of the data training result in the actual application requirement process, for example, the maximum iteration parameter threshold may take a value of 100, 500, 1000 or a larger/smaller value, etc., to prevent over-fitting or under-fitting of the model, so that the training result is as close to the real result as possible.
In operation S602, part of the three-dimensional model input data and the two-dimensional tag data are input to the optical communication receiving model for training.
In operation S603, it is determined whether the generation number of the current iteration is greater than the maximum iteration parameter threshold or less than the minimum loss function threshold. If not, go to step S604; if yes, go to step S606.
In operation S604, a part of data from the training data set is randomly extracted and input into the model for training, and a loss function value and a label point-by-point mean square error are calculated.
In the embodiment of the disclosure, part of data is randomly extracted from a training data set and input into a model for training again, the loss function value after model training and the label point-by-point calculation mean square error are calculated, at the moment, the result after model training and the selected label data are compared and analyzed, and the mean square error is calculated so as to judge the accuracy of the model calculation result, thereby facilitating model optimization.
In operation S605, the gradient of the trainable network parameters is calculated according to the loss function value, and parameters such as Q, K, V matrix in the transform coding structure and weights between the other structures in the model are updated according to the training optimizer algorithm, the number of iterative parameters is increased by one, and the process returns to step S603.
In operation S606, the model training is ended, the iteration is ended, and the model parameters after the optical communication reception model training are fixed.
In the embodiment of the disclosure, the method includes testing the trained model, determining the error rate of the trained model, and testing the model. As shown in fig. 6, in operation S607: after the model training is finished, data in the test data are input into the model for testing, errors obtained by the final model and the traditional method are calculated, and error results are output. The test data is data other than the training data in the three-dimensional model input data and the two-dimensional label data, and the step S607 is used for evaluating the training effect and generalization performance of the optical communication receiving model.
According to the embodiment of the present disclosure, testing the optical communication receiving model by using another part of the three-dimensional model input data and the two-dimensional label data as the test data to obtain the tested optical communication receiving model specifically includes: testing the optical communication receiving model by using the other part of the three-dimensional model input data and the two-dimensional label data as test data; comparing and analyzing the result obtained by the model training with the two-dimensional label data corresponding to the other part of test data to obtain the error rate of the optical communication receiving model training; and adjusting initial training parameters of the optical communication receiving model according to the error rate so as to optimize the optical communication receiving model and further improve the model performance.
It should be noted that, the foregoing embodiments are merely exemplary, and should not be construed as limiting the embodiments of the present disclosure.
According to the training method of the optical communication receiving model provided by the embodiment of the disclosure, under the training of data and a gradient descent algorithm, the rules of signal de-aliasing and optimal judgment in different transmission scenes can be progressively mastered. Thereafter, the model can process transmission signals with different transmission lengths and signal-to-noise ratios, and the source information can be recovered with a low error rate.
Another embodiment of the present disclosure provides a method for restoring an optical fiber transmission signal, including: acquiring at least one optical fiber transmission signal; and restoring at least one optical fiber transmission signal into a source communication symbol corresponding to the optical fiber transmission signal by using an optical communication receiving model, wherein the optical communication receiving model is obtained by training by using the method provided by the embodiment.
In the embodiment of the present disclosure, the model obtained by training the method provided in the above embodiment is applied to the optical communication system shown in fig. 1 instead of the signal processing module, so that information transmitted by the information source can be quickly and accurately recovered from the optical fiber transmission signal.
In the embodiment of the disclosure, the restoration method of the optical fiber transmission signal provided by the disclosure can judge and restore the optical fiber transmission signals of different scenes to the information source information before the optical fiber transmission signals are transmitted by the restoration method with higher precision. The restoration method effectively utilizes the generalization capability of the neural network to improve the general performance of the traditional receiver, so that the traditional receiver can be used for different transmission scenes.
Fig. 7 schematically illustrates a block diagram of an apparatus for training an optical communication reception model according to an embodiment of the present disclosure.
As shown in fig. 7, the training apparatus for an optical communication reception model includes: a signal acquisition module 710, a signal preprocessing module 720, a data construction module 730, and a model training module 740. The apparatus 700 may be used to implement the training method of the optical communication reception model described with reference to fig. 2.
A signal obtaining module 710, configured to obtain N strings of information source communication symbols and N optical fiber transmission signals; n is more than or equal to 2, and N optical fiber transmission signals are respectively signals of N strings of information source communication symbols after transmission through N different optical fiber distances. The signal obtaining module 710 may be configured to perform the step S201 described above with reference to fig. 2, for example, and is not described herein again.
And the signal preprocessing module 720 is configured to perform normalization processing on the N optical fiber transmission signals to obtain a normalized model input signal. The signal preprocessing module 720 can be used to perform the step S202 described above with reference to fig. 2, for example, and is not described herein again.
The data construction module 730 is configured to construct a three-dimensional tensor of the normalized model input signal and a two-dimensional tensor of the N-string source communication symbols, and obtain three-dimensional model input data and two-dimensional tag data. The data building module 730 can be used to perform the step S203 described above with reference to fig. 2, for example, and is not described herein again.
And a model training module 740, configured to train an optical communication receiving model using part of the three-dimensional model input data and the two-dimensional label data as training data, where the optical communication receiving model is configured to restore the optical fiber transmission signal to a source communication symbol corresponding to the optical fiber transmission signal. The model training module 740 may be used to perform the step S204 described above with reference to fig. 2, for example, and is not described herein again.
Fig. 8 schematically shows a block diagram of a restoration apparatus for an optical fiber transmission signal according to an embodiment of the present disclosure.
As shown in fig. 8, the apparatus for restoring the optical fiber transmission signal includes: a data acquisition module 810 and a data restoration module 820. The apparatus 800 may be used to implement the restoration method of the optical fiber transmission signal described with reference to the above embodiments.
A data acquisition module 810 configured to acquire at least one optical fiber transmission signal.
A data restoring module 820, configured to restore the at least one optical fiber transmission signal into a source communication symbol corresponding to the at least one optical fiber transmission signal by using an optical communication receiving model, where the optical communication receiving model is trained by using the method according to the foregoing embodiment.
Any number of modules, sub-modules, units, sub-units, or at least part of the functionality of any number thereof according to embodiments of the present disclosure may be implemented in one module. Any one or more of the modules, sub-modules, units, and sub-units according to the embodiments of the present disclosure may be implemented by being split into a plurality of modules. Any one or more of the modules, sub-modules, units, sub-units according to embodiments of the present disclosure may be implemented at least partially as a hardware circuit, e.g., a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), an on-chip device, a device on a substrate, a device on a package, an Application Specific Integrated Circuit (ASIC), or by any other reasonable means of hardware or firmware for integrating or packaging a circuit, or by any one of or a suitable combination of any of software, hardware, and firmware. Alternatively, one or more of the modules, sub-modules, units, sub-units according to embodiments of the disclosure may be at least partially implemented as a computer program module, which when executed may perform the corresponding functions.
For example, any number of the signal acquisition module 710, the signal preprocessing module 720, the data construction module 730, and the model training module 740, or the data acquisition module 810 and the data restoration module 820 may be combined and implemented in one module, or any one of them may be split into a plurality of modules. Alternatively, at least part of the functionality of one or more of these modules may be combined with at least part of the functionality of the other modules and implemented in one module. According to an embodiment of the present disclosure, at least one of the signal obtaining module 710, the signal preprocessing module 720, the data constructing module 730, and the model training module 740 or the data obtaining module 810 and the data restoring module 820 may be at least partially implemented as a hardware circuit, such as a Field Programmable Gate Array (FPGA), a Programmable Logic Array (PLA), a device on a chip, a device on a substrate, a device on a package, an Application Specific Integrated Circuit (ASIC), or may be implemented by hardware or firmware in any other reasonable manner of integrating or packaging a circuit, or implemented by any one of three implementation manners of software, hardware, and firmware, or by a suitable combination of any several of them. Alternatively, at least one of the signal acquisition module 710, the signal pre-processing module 720, the data construction module 730 and the model training module 740 or the data acquisition module 810 and the data restoration module 820 may be at least partially implemented as a computer program module that, when executed, may perform a corresponding function.
Fig. 9 schematically shows a block diagram of an electronic device adapted to implement the above described method according to an embodiment of the present disclosure. The electronic device shown in fig. 9 is only an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present disclosure.
As shown in fig. 9, the electronic device 900 described in this embodiment includes: a processor 901 which can perform various appropriate actions and processes in accordance with a program stored in a Read Only Memory (ROM)902 or a program loaded from a storage section 908 into a Random Access Memory (RAM) 903. Processor 901 can include, for example, a general purpose microprocessor (e.g., a CPU), an instruction set processor and/or related chip sets and/or a special purpose microprocessor (e.g., an Application Specific Integrated Circuit (ASIC)), among others. The processor 901 may also include on-board memory for caching purposes. The processor 901 may comprise a single processing unit or a plurality of processing units for performing the different actions of the method flows according to embodiments of the present disclosure.
In the RAM 903, various programs and data necessary for the operation of the electronic apparatus 900 are stored. The processor 901, the ROM 902, and the RAM 903 are connected to each other through a bus 904. The processor 901 performs various operations of the method flows according to the embodiments of the present disclosure by executing programs in the ROM 902 and/or the RAM 903. Note that the programs may also be stored in one or more memories other than the ROM 902 and the RAM 903. The processor 901 may also perform various operations of the method flows according to embodiments of the present disclosure by executing programs stored in the one or more memories.
According to an embodiment of the present disclosure, the method flow according to an embodiment of the present disclosure may be implemented as a computer software program. For example, embodiments of the present disclosure include a computer program product comprising a computer program embodied on a computer readable storage medium, the computer program containing program code for performing the method illustrated by the flow chart. In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 909, and/or installed from the removable medium 911. The computer program, when executed by the processor 901, performs the above-described functions defined in the apparatus of the embodiments of the present disclosure. According to an embodiment of the present disclosure, the above-described apparatuses, devices, apparatuses, modules, units, and the like may be realized by computer program modules.
An embodiment of the present invention further provides a computer-readable storage medium, which may be included in the apparatus/device/apparatus described in the foregoing embodiment; or may exist alone without being assembled into the apparatus/device/arrangement. The computer-readable storage medium carries one or more programs which, when executed, implement a method for training an optical communication reception model or a method for restoring an optical fiber transmission signal according to an embodiment of the present disclosure.
According to embodiments of the present disclosure, the computer-readable storage medium may be a non-volatile computer-readable storage medium, which may include, for example but is not limited to: a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing. In embodiments of the disclosure, a computer readable storage medium may be any tangible medium that can contain, or store a program for use by or in connection with an instruction execution apparatus, device, or device. For example, according to embodiments of the present disclosure, a computer-readable storage medium may include the ROM 902 and/or RAM 903 described above and/or one or more memories other than the ROM 902 and RAM 903.
Embodiments of the present disclosure also include a computer program product comprising a computer program containing program code for performing the method illustrated in the flow chart. When the computer program product runs in a computer device, the program code is used for causing the computer device to implement the training method of the optical communication receiving model or the restoring method of the optical fiber transmission signal provided by the embodiment of the disclosure.
The computer program performs the above-described functions defined in the apparatus/devices of the embodiments of the present disclosure when executed by the processor 901. According to an embodiment of the present disclosure, the above-described apparatuses, devices, modules, units, and the like may be realized by computer program modules.
In one embodiment, the computer program may be hosted on a tangible storage medium such as an optical storage device, a magnetic storage device, or the like. In another embodiment, the computer program may also be transmitted, distributed in the form of a signal on a network medium, and downloaded and installed through the communication section 909 and/or installed from the removable medium 911. The computer program containing program code may be transmitted using any suitable network medium, including but not limited to: wireless, wired, etc., or any suitable combination of the foregoing.
In such an embodiment, the computer program may be downloaded and installed from a network through the communication section 909, and/or installed from the removable medium 911. The computer program, when executed by the processor 901, performs the above-described functions defined in the apparatus of the embodiments of the present disclosure. According to an embodiment of the present disclosure, the above-described apparatuses, devices, apparatuses, modules, units, and the like may be realized by computer program modules.
In accordance with embodiments of the present disclosure, program code for executing computer programs provided by embodiments of the present disclosure may be written in any combination of one or more programming languages, and in particular, these computer programs may be implemented using high level procedural and/or object oriented programming languages, and/or assembly/machine languages. The programming language includes, but is not limited to, programming languages such as Java, C + +, python, the "C" language, or the like. The program code may execute entirely on the user computing device, partly on the user device, partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).
It should be noted that each functional module in each embodiment of the present disclosure may be integrated into one processing module, or each module may exist alone physically, or two or more modules are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be substantially or partially embodied in the form of a software product, or all or part of the technical solution that contributes to the prior art.
The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that, in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams or flowchart illustration, and combinations of blocks in the block diagrams or flowchart illustration, can be implemented by special purpose hardware-based apparatus that perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
Those skilled in the art will appreciate that various combinations and/or combinations of features recited in the various embodiments and/or claims of the present disclosure can be made, even if such combinations or combinations are not expressly recited in the present disclosure. In particular, various combinations and/or combinations of the features recited in the various embodiments and/or claims of the present disclosure may be made without departing from the spirit or teaching of the present disclosure. All such combinations and/or associations are within the scope of the present disclosure.
While the disclosure has been shown and described with reference to certain exemplary embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the disclosure as defined by the appended claims and their equivalents. Accordingly, the scope of the present disclosure should not be limited to the above-described embodiments, but should be defined not only by the appended claims, but also by equivalents thereof.
Claims (12)
1. A training method for an optical communication receiving model is characterized by comprising the following steps:
acquiring N strings of information source communication symbols and N optical fiber transmission signals; n is more than or equal to 2, and the N optical fiber transmission signals are signals of each information source communication symbol in the N strings of information source communication symbols after transmission by different optical fiber distances;
carrying out normalization processing on the N optical fiber transmission signals to obtain normalized model input signals;
constructing a three-dimensional tensor of the normalized model input signal and a two-dimensional tensor of the N strings of information source communication symbols to obtain three-dimensional model input data and two-dimensional label data; and
and training an optical communication receiving model by using part of the three-dimensional model input data and the two-dimensional label data as training data, wherein the optical communication receiving model is used for restoring the optical fiber transmission signal into a source communication symbol corresponding to the optical fiber transmission signal.
2. The method for training an optical communication reception model according to claim 1, wherein the training an optical communication reception model using a part of the three-dimensional model input data and the two-dimensional label data as training data includes:
setting initial training parameters of the optical communication receiving model, wherein the initial training parameters at least comprise a maximum iteration parameter threshold and a minimum loss function threshold;
inputting part of the three-dimensional model input data and the two-dimensional label data into the optical communication receiving model for training;
judging whether the algebra of the current iteration is larger than the maximum iteration parameter threshold or smaller than the minimum loss function threshold;
if so, ending the iteration, and fixing the model parameters after the optical communication receiving model is trained.
3. The method for training an optical communication receiving model according to claim 1, further comprising:
testing an optical communication receiving model by using the other part of the three-dimensional model input data and the two-dimensional label data as test data to obtain the tested optical communication receiving model; the test data is data except training data in the three-dimensional model input data and the two-dimensional label data.
4. The method for training an optical communication receiving model according to claim 3, wherein the testing an optical communication receiving model using another part of the three-dimensional model input data and the two-dimensional label data as test data to obtain the tested optical communication receiving model comprises:
testing an optical communication reception model using another part of the three-dimensional model input data and the two-dimensional tag data as test data;
comparing and analyzing the result obtained by model training with the other part of two-dimensional label data to obtain the error rate of the optical communication receiving model training;
and adjusting the initial training parameters of the optical communication receiving model according to the error rate so as to optimize the optical communication receiving model.
5. The method for training an optical communication receiving model according to claim 1, wherein the optical communication receiving model comprises:
a Transformer encoder comprising: a first residual error structure, a first data processing structure, a second residual error structure and a second data processing structure; the first residual error structure is used for copying the three-dimensional model input data into two identical three-dimensional optical fiber data; the first data processing structure is used for performing feature extraction on one part of three-dimensional model input data and performing data aggregation on the extracted feature data and the other part of three-dimensional model input data to obtain first aggregated data; the first residual error structure is used for copying the first aggregation data into two identical optical fiber data; the second data processing structure is used for performing secondary feature extraction on one part of data output by the first residual error structure, and performing secondary data aggregation on the extracted feature data and the other part of data to obtain second aggregated data;
the deep full-connection network structure is used for performing information fusion on the second aggregation data to obtain fused data;
and the signal judgment layer is used for restoring the fused data into the information source communication symbol corresponding to the fused data.
6. The method for training an optical communication receiving model according to claim 1, further comprising:
outputting a source communication symbol by using a source generator;
converting the source communication symbols into electrical signals using a waveform generator;
loading the electric signal to an optical carrier by using a modulator to obtain a modulated initial optical fiber signal;
and transmitting the modulated initial optical fiber signals by adopting N optical fiber links with different transmission distances to obtain N optical fiber transmission signals.
7. A method for restoring a signal transmitted through an optical fiber, comprising:
acquiring at least one optical fiber transmission signal;
recovering the at least one optical fiber transmission signal into a source communication symbol corresponding thereto using an optical communication reception model, wherein the optical communication reception model is trained using the method according to any one of claims 1 to 6.
8. The method according to claim 7, wherein the recovered communication symbol of the at least one optical fiber transmission signal is identical to the corresponding source communication symbol.
9. An apparatus for training an optical communication reception model, comprising:
the signal acquisition module is used for acquiring N strings of information source communication symbols and N optical fiber transmission signals; n is more than or equal to 2, and the N optical fiber transmission signals are respectively signals of the N strings of information source communication symbols after transmission through N different optical fiber distances;
the signal preprocessing module is used for carrying out normalization processing on the N optical fiber transmission signals to obtain normalized model input signals;
the data construction module is used for constructing the three-dimensional tensor of the normalized model input signal and the two-dimensional tensor of the N strings of information source communication symbols to obtain three-dimensional model input data and two-dimensional label data; and
and the model training module is used for training an optical communication receiving model by using part of the three-dimensional model input data and the two-dimensional label data as training data, and the optical communication receiving model is used for restoring the optical fiber transmission signal into a corresponding information source communication symbol.
10. A device for restoring a signal transmitted through an optical fiber, comprising:
the data acquisition module is used for acquiring at least one optical fiber transmission signal;
a data recovery module for recovering the at least one optical fiber transmission signal into a source communication symbol corresponding thereto using an optical communication reception model, wherein the optical communication reception model is trained using the method according to any one of claims 1 to 6.
11. An electronic device, comprising:
at least one processor; and
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to perform the method of any one of claims 1 to 7.
12. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210228714.2A CN114722905A (en) | 2022-03-09 | 2022-03-09 | Training method and device for optical communication receiving model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210228714.2A CN114722905A (en) | 2022-03-09 | 2022-03-09 | Training method and device for optical communication receiving model |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114722905A true CN114722905A (en) | 2022-07-08 |
Family
ID=82236769
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210228714.2A Pending CN114722905A (en) | 2022-03-09 | 2022-03-09 | Training method and device for optical communication receiving model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114722905A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115632742A (en) * | 2022-08-31 | 2023-01-20 | 深圳市中易腾达科技股份有限公司 | Signal processing method, signal processing device, electronic equipment and storage medium |
-
2022
- 2022-03-09 CN CN202210228714.2A patent/CN114722905A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115632742A (en) * | 2022-08-31 | 2023-01-20 | 深圳市中易腾达科技股份有限公司 | Signal processing method, signal processing device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20230394671A1 (en) | Image segmentation method and apparatus, and device, and storage medium | |
| CN111160171B (en) | Radiation source signal identification method combining two-domain multi-features | |
| CN110932809B (en) | Fiber channel model simulation method, device, electronic equipment and storage medium | |
| CN110570433B (en) | Image semantic segmentation model construction method and device based on generation countermeasure network | |
| CN111526119B (en) | Abnormal flow detection method and device, electronic equipment and computer readable medium | |
| CN112712069B (en) | Question judging method and device, electronic equipment and storage medium | |
| WO2022252881A1 (en) | Image processing method and apparatus, and readable medium and electronic device | |
| CN111428817A (en) | Defense method for resisting attack by radio signal identification | |
| CN112883968B (en) | Image character recognition method, device, medium and electronic equipment | |
| CN109995449A (en) | A kind of millimeter-wave signal detection method based on deep learning | |
| CN112597999A (en) | Question identification method and device, electronic equipment and computer storage medium | |
| WO2023035877A1 (en) | Video recognition method and apparatus, readable medium, and electronic device | |
| CN110633718B (en) | Method and device for determining a driving area in an environment image | |
| CN114722905A (en) | Training method and device for optical communication receiving model | |
| CN113762454A (en) | Track abnormity detection method and device | |
| CN116166271A (en) | Code generation method and device, storage medium and electronic equipment | |
| CN114553315A (en) | Optical fiber nonlinear equalization method and system based on CNN-birNN | |
| CN114647977A (en) | Training method and device for optical fiber transmission signal prediction model | |
| CN113592033A (en) | Oil tank image recognition model training method, oil tank image recognition method and oil tank image recognition device | |
| CN111459780B (en) | User identification method and device, readable medium and electronic equipment | |
| CN115270002B (en) | Knowledge distillation-based item recommendation method, electronic equipment and storage medium | |
| CN113474795A (en) | Answering cognitive queries from sensor input signals | |
| US11507803B2 (en) | System for generating synthetic digital data for data multiplication | |
| CN114647976A (en) | Training method and device for multi-span optical fiber transmission signal prediction system | |
| US20220004817A1 (en) | Data analysis system, learning device, method, and program |
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 |