CN119853814A - Link power control method and system for coherent optical wavelength division multiplexing transmission system - Google Patents
Link power control method and system for coherent optical wavelength division multiplexing transmission system Download PDFInfo
- Publication number
- CN119853814A CN119853814A CN202411926195.2A CN202411926195A CN119853814A CN 119853814 A CN119853814 A CN 119853814A CN 202411926195 A CN202411926195 A CN 202411926195A CN 119853814 A CN119853814 A CN 119853814A
- Authority
- CN
- China
- Prior art keywords
- neural network
- attenuation
- channel
- adjustment value
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- 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
- H04B10/50—Transmitters
- H04B10/564—Power control
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
- H04B10/25—Arrangements specific to fibre transmission
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04J—MULTIPLEX COMMUNICATION
- H04J14/00—Optical multiplex systems
- H04J14/002—Coherencemultiplexing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04J—MULTIPLEX COMMUNICATION
- H04J14/00—Optical multiplex systems
- H04J14/02—Wavelength-division multiplex systems
- H04J14/0221—Power control, e.g. to keep the total optical power constant
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Signal Processing (AREA)
- Computer Networks & Wireless Communication (AREA)
- Theoretical Computer Science (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Electromagnetism (AREA)
- Optical Communication System (AREA)
Abstract
The invention provides a link power control method and a link power control system of a coherent optical wavelength division multiplexing transmission system, which comprise the steps of building the coherent optical wavelength division multiplexing transmission simulation system, configuring the transmitting power of each channel in the wavelength division multiplexing system, taking the initial state of the transmitting power as the input of an executive neural network, obtaining the adjusting values of the transmitting power of different channels, calculating the transmitting power of the channels through the initial state and the transmitting power adjusting values, configuring the attenuation of each channel of a wavelength selective switch in the wavelength division multiplexing system, taking the initial state of the wavelength selective switch as the input of the executive neural network, obtaining the adjusting values of the attenuation of the wavelength selective switch, and calculating the transmitting power of the channels through the initial state and the attenuation adjusting values. The invention is based on a deep reinforcement learning algorithm, and realizes the optimal configuration of power in a link by adjusting the transmitting power of different channels in a coherent optical wavelength division multiplexing system and the attenuation of different channels in a wavelength selective switch, thereby optimizing the communication quality of each channel at a receiving end after transmission.
Description
Technical Field
The invention relates to the technical field of link power control, in particular to a link power control method and a link power control system of a coherent optical wavelength division multiplexing transmission system.
Background
The coherent optical wavelength division multiplexing system is a mainstream system of optical transmission, in the coherent optical wavelength division multiplexing co-system, signals of all channels are affected by nonlinearity of optical fibers, and transmission quality is inconsistent, so that the overall transmission level of the system is affected. Link power control is used to balance the communication quality of the individual channels. Current methods based on deep learning control of link power rely on accurate models of link transmissions.
The patent application document CN116938383A discloses a nonlinear compensation method and a nonlinear compensation system based on a module multiplexing system, wherein the method comprises the steps of selecting any one mode to be set as a second mode, setting other modes to be set as a first mode, applying frequency offset to a wavelength division multiplexing optical signal of the second mode by adopting a first optical frequency shifter, inputting the wavelength division multiplexing optical signal of the first mode and the wavelength division multiplexing optical signal of the second mode after the offset to the same module multiplexer, outputting the same module multiplexer through the same multimode optical fiber, applying reverse frequency offset to the wavelength division multiplexing optical signal of the second mode output by the module multiplexer by adopting the second optical frequency shifter, and performing wavelength division multiplexing on the wavelength division multiplexing optical signal of the first mode output by the module multiplexer and the wavelength division multiplexing optical signal of the second mode applying the reverse frequency offset to complete nonlinear compensation. However, the present patent cannot completely solve the existing technical problems, and cannot meet the needs of the present invention.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a link power control method and a link power control system of a coherent optical wavelength division multiplexing transmission system.
The link power control method of the coherent optical wavelength division multiplexing transmission system provided by the invention comprises the following steps:
Step 1, constructing a coherent optical wavelength division multiplexing transmission simulation system, wherein the system comprises the steps of setting the length of an optical fiber in a link and the parameters of an optical amplifier, and setting the total transmission distance;
Training a deep neural network by using a simulation environment, and configuring the transmitting power of each channel in the wavelength division multiplexing system;
Step 3, taking the initial state of the transmitting power as input, inputting the initial state into a trained executive neural network to obtain adjustment values of the transmitting power of different channels, and calculating the transmitting power of the channel according to the initial state and the transmitting power adjustment values;
training a deep neural network by using a simulation environment, and configuring attenuation of each channel of a wavelength selective switch in a wavelength division multiplexing system;
And 5, inputting the initial state of the wavelength selective switch into a trained executive neural network to obtain an attenuation adjustment value of the wavelength selective switch, and calculating the channel transmitting power through the initial state and the attenuation adjustment value.
Preferably, the step 2 includes:
Step 2.1, taking an initial value of the transmitting power as an initial state, taking the initial state as an input of an executive neural network, obtaining output as an adjustment value of the transmitting power, calculating the transmitting power of a channel through the initial state and the adjustment value of the transmitting power, then transmitting, and obtaining a result of a receiving end;
Step 2.2, training the criticizing neural network through the data in the data pool, taking the current state and the execution action in the data pair as the input of the criticizing neural network to obtain the output of the data pair, taking the absolute value as the loss function of the criticizing neural network by taking the difference between the output and the reward function, and updating the parameters of the criticizing neural network through a gradient descent method;
And 2.3, training the execution neural network through data in the data pool and the criticizing neural network, taking the current state in the data pair as the input of the execution neural network to obtain the output of the execution neural network, taking the output and the current state as the input of the criticizing neural network to obtain the output of the criticizing neural network, taking the opposite number of the output of the criticizing neural network as a loss function, and updating the parameters of the execution neural network through a gradient descent method.
Preferably, the channel transmit power is calculated from the start state and the transmit power adjustment value by the expression:
Pch＝P0+PΔ
Wherein, P ch is channel transmitting power, P 0 is initial state, and P Δ is power adjusting value;
Calculating corresponding rewards through the signal-to-noise ratio and the target signal-to-noise ratio obtained by the receiving end, wherein the expression is as follows:
Ri＝||SNRi-SNRtarget||2
Where R i is the prize, SNR i is the receiver SNR, and SNR target is the target SNR.
Preferably, the step 4 includes:
Step 4.1, taking an attenuation initial value as an initial state, taking the initial state as an input of an executive neural network, obtaining output as an attenuation adjustment value, calculating the attenuation of each channel through the initial state and the attenuation adjustment value, transmitting and obtaining a result of a receiving end; finally, the signal-to-noise ratio of each channel is obtained at the receiving end, the corresponding rewards are calculated through the signal-to-noise ratio and the target signal-to-noise ratio obtained at the receiving end, and the initial state is updated to the next state;
step 4.2, training and judging the neural network by utilizing the data in the data pool;
and 4.3, training and executing the neural network by utilizing the data in the data pool and the criticizing neural network.
Preferably, the attenuation of each channel is calculated by the initial state and the attenuation adjustment value, expressed as:
Ach＝A0+AΔ
Wherein, A ch is channel attenuation, A 0 is initial state, and A Δ is attenuation adjustment value.
The link power control system of the coherent optical wavelength division multiplexing transmission system provided by the invention comprises:
The method comprises the steps of (1) constructing a coherent optical wavelength division multiplexing transmission simulation system, wherein the system comprises the steps of setting the length of an optical fiber in a link and the parameters of an optical amplifier, and setting the total transmission distance;
the module M2 is used for training a deep neural network by utilizing a simulation environment and configuring the transmitting power of each channel in the wavelength division multiplexing system;
The module M3 takes the initial state of the transmitting power as input, inputs the initial state into a trained executive neural network, obtains adjustment values of the transmitting power of different channels, and calculates the transmitting power of the channel according to the initial state and the transmitting power adjustment values;
The module M4 is used for training the deep neural network by utilizing a simulation environment and configuring attenuation of each channel of the wavelength selective switch in the wavelength division multiplexing system;
And the module M5 takes the initial state of the wavelength selective switch as input, inputs the initial state into the trained executive neural network, obtains the attenuation adjustment value of the wavelength selective switch, and calculates the channel transmitting power through the initial state and the attenuation adjustment value.
Preferably, the module M2 comprises:
The module M2.1 takes the initial value of the transmitting power as an initial state, takes the initial state as the input of the execution neural network, obtains output as an adjustment value of the transmitting power, calculates the transmitting power of a channel through the initial state and the adjustment value of the transmitting power, then transmits the channel and obtains the result of a receiving end;
the module M2.2 is used for training the criticizing neural network through the data in the data pool, taking the current state and the execution action in the data pair as the input of the criticizing neural network to obtain the output of the data pair, taking the absolute value as the loss function of the criticizing neural network through the difference between the output and the reward function, and updating the parameters of the criticizing neural network through a gradient descent method;
The module M2.3 trains the executive neural network through the data in the data pool and the criticizing neural network, takes the current state in the data pair as the input of the executive neural network, obtains the output of the executive neural network, takes the output and the current state as the input of the criticizing neural network, obtains the output of the criticizing neural network, takes the opposite number of the output of the criticizing neural network as a loss function, and updates the parameters of the executive neural network through a gradient descent method.
Preferably, the channel transmit power is calculated from the start state and the transmit power adjustment value by the expression:
Pch＝P0+PΔ
Wherein, P ch is channel transmitting power, P 0 is initial state, and P Δ is power adjusting value;
Calculating corresponding rewards through the signal-to-noise ratio and the target signal-to-noise ratio obtained by the receiving end, wherein the expression is as follows:
Ri＝||SNRi-SNRtarget||2
Where R i is the prize, SNR i is the receiver SNR, and SNR target is the target SNR.
Preferably, the module M4 comprises:
The module M4.1 takes the initial attenuation value as an initial state, takes the initial state as the input of the execution neural network, obtains output as an adjustment value of attenuation, calculates the attenuation of each channel through the initial state and the attenuation adjustment value, and then transmits the attenuation and obtains the result of a receiving end; finally, the signal-to-noise ratio of each channel is obtained at the receiving end, the corresponding rewards are calculated through the signal-to-noise ratio and the target signal-to-noise ratio obtained at the receiving end, and the initial state is updated to the next state;
a module M4.2, training and judging the neural network by utilizing the data in the data pool;
and a module M4.3, training and executing the neural network by utilizing the data in the data pool and the criticizing neural network.
Preferably, the attenuation of each channel is calculated by the initial state and the attenuation adjustment value, expressed as:
Ach＝A0+AΔ
Wherein, A ch is channel attenuation, A 0 is initial state, and A Δ is attenuation adjustment value.
Compared with the prior art, the invention has the following beneficial effects:
(1) The invention is based on a deep reinforcement learning algorithm, and realizes the optimal configuration of power in a link by adjusting the transmitting power of different channels in a coherent optical wavelength division multiplexing system and the attenuation of different channels in a wavelength selective switch, thereby optimizing the communication quality of each channel at a receiving end after transmission;
(2) The invention can train the neural network without depending on accurate micro-models by deep reinforcement learning, thereby realizing the problem of different communication quality of different channels in a coherent optical communication system and enabling different channels to obtain the same signal-to-noise ratio.
Drawings
Other features, objects and advantages of the present invention will become more apparent upon reading of the detailed description of non-limiting embodiments, given with reference to the accompanying drawings in which:
FIG. 1 is a diagram of generation of training data and construction of a data pool;
FIG. 2 is a flowchart of an evaluation network training;
FIG. 3 is a flowchart of performing network training;
FIG. 4 is a flow chart of setting the transmit power and wavelength selective switch;
fig. 5 is a set originating power optimization result.
Detailed Description
The present invention will be described in detail with reference to specific examples. The following examples will assist those skilled in the art in further understanding the present invention, but are not intended to limit the invention in any way. It should be noted that variations and modifications could be made by those skilled in the art without departing from the inventive concept. These are all within the scope of the present invention.
Example 1
The invention provides a link power control method of a coherent optical wavelength division multiplexing transmission system based on deep reinforcement learning, which comprises the following steps:
Step 1, constructing a coherent optical wavelength division multiplexing transmission simulation system, wherein the system comprises the steps of setting the length of an optical fiber in a link and the parameters of an optical amplifier, and setting the total transmission distance;
Training the deep neural network by using a simulation environment, so that the execution neural network can configure the transmitting power of each channel in the wavelength division multiplexing system;
and 2.1, selecting a reasonable initial value of the transmitting power as an initial state, taking the initial state as input of the execution neural network, obtaining output as an adjustment value of the transmitting power, calculating channel transmitting power according to a formula (1) through the initial state and the transmitting power adjustment value, transmitting, and obtaining a result of a receiving end.
Pch＝P0+PΔ (1)
Where P ch is the channel transmit power, P 0 is the initial state, and P Δ is the power adjustment value.
And finally, obtaining the signal-to-noise ratio of each channel at the receiving end, calculating corresponding rewards according to the formula (2) through the signal-to-noise ratio and the target signal-to-noise ratio obtained by the receiving end, and updating the initial state to the next state.
Ri＝||SNRi-SNRtarget||2 (2)
Where R i is the prize, SNR i is the receiver SNR, and SNR target is the target SNR.
The initial state, the transmit power adjustment value, the next state and the prize are aggregated into data pairs, which are stored in a data pool, and the data in the data pool is generated by performing a continuous interaction of the neural network and the simulated transmission system, as shown in fig. 1.
And 2.2, training and judging the neural network through data in the data pool, as shown in figure 2. The current state and the execution action in the data pair are used as the input of the criticizing neural network to obtain the output of the criticizing neural network, the difference is made between the output and the rewarding function, the absolute value is taken as the loss function of the criticizing neural network, and the parameters of the criticizing neural network are updated by a gradient descent method.
And 2.3, training the executive neural network through the data in the data pool and the criticizing neural network, as shown in fig. 3. The method comprises the steps of obtaining the current state of a data pair as the input of an executive neural network, obtaining the output of the wholesale neural network by taking the output and the current state as the input of the wholesale neural network, taking the opposite number of the output as a loss function, and updating the parameters of the executive neural network by a gradient descent method.
And 3, inputting the initial state of the transmitting power into a trained executive neural network to obtain the adjustment values of the transmitting power of different channels. And then calculating the channel transmission power according to the formula (1) and the transmission power adjustment value through the initial state.
And 4, training the deep neural network by using a simulation environment, so that the execution neural network can configure attenuation of each channel of the wavelength selective switch in the wavelength division multiplexing system, as shown in fig. 4.
And 4.1, selecting a reasonable attenuation initial value as an initial state, taking the initial state as an input of the execution neural network, and obtaining an output as an adjustment value of attenuation. The attenuation of each channel is calculated according to the formula (3) through the initial state and the attenuation adjustment value, and then the transmission is performed, and the result of the receiving end is obtained, as shown in fig. 5.
Ach＝A0+AΔ (3)
Wherein, A ch is channel attenuation, A 0 is initial state, and A Δ is attenuation adjustment value.
And finally, obtaining the signal to noise ratio of each channel at the receiving end. And calculating corresponding rewards according to a formula (2) through the signal-to-noise ratio and the target signal-to-noise ratio obtained at the receiving end, and updating the initial state to the next state.
The initial state, the attenuation adjustment value, the next state and the rewards are collected into data pairs and stored in a data pool. The data in the data pool is generated by performing a constant interaction of the neural network and the simulated transmission system.
And 4.2, training and judging the neural network by using the data in the data pool through the same flow as the step 2.2.
And 4.3, training and executing the neural network by utilizing the data in the data pool and the criticizing neural network through the flow similar to the step 2.2.
And 5, inputting the initial state of the wavelength selective switch into a trained executive neural network to obtain the attenuation adjustment value of the wavelength selective switch. And then calculating the channel transmitting power according to the formula (3) through the initial state and the attenuation adjustment value.
Example 2
The present invention also provides a link power control system of a coherent optical wavelength division multiplexing transmission system, which may be implemented by executing the flow steps of the link power control method of the coherent optical wavelength division multiplexing transmission system, i.e. those skilled in the art may understand the link power control method of the coherent optical wavelength division multiplexing transmission system as a preferred embodiment of the link power control system of the coherent optical wavelength division multiplexing transmission system.
The link power control system of the coherent optical wavelength division multiplexing transmission system comprises a module M1, a first module and a second module, wherein the module M1 is used for constructing a coherent optical wavelength division multiplexing transmission simulation system, setting the length of an optical fiber in a link, the parameters of an optical amplifier and setting the total transmission distance; the method comprises the steps of training a deep neural network by using a simulation environment, configuring the transmitting power of each channel in a wavelength division multiplexing system, inputting the initial state of the transmitting power into the trained executing neural network to obtain the adjusting values of the transmitting power of different channels, calculating the transmitting power of the channels by the initial state and the transmitting power adjusting values, training the deep neural network by using the simulation environment, configuring the attenuation of each channel of a wavelength selective switch in the wavelength division multiplexing system, and inputting the initial state of the wavelength selective switch into the trained executing neural network to obtain the adjusting values of the attenuation of the wavelength selective switch by using the initial state and the attenuation adjusting values, and calculating the transmitting power of the channels by using the initial state and the attenuation adjusting values, wherein the module M2 is used for training the deep neural network by using the simulation environment, and the module M5 is used for inputting the initial state of the wavelength selective switch into the trained executing neural network to obtain the attenuation adjusting values of the wavelength selective switch.
The module M2.1 takes an initial value of the transmitting power as an initial state, takes the initial state as an input of an execution neural network, obtains output as an adjustment value of the transmitting power, calculates the transmitting power of a channel through the initial state and the transmitting power adjustment value, then transmits the channel and obtains a result of a receiving end, obtains signal to noise ratios of all channels at the receiving end, calculates corresponding rewards through the signal to noise ratio and a target signal to noise ratio obtained by the receiving end, updates the initial state into a next state, integrates the initial state, the transmitting power adjustment value, the next state and the rewards into a data pair, stores the data pair into a data pool, generates data in the data pool through continuous interaction of an execution neural network and a simulation transmission system, obtains output of the neural network through the data training neural network in the data pool, takes the current state and the execution action of the execution neural network in the data pool as input of the neural network, obtains output of the neural network through the output of the judgment neural network, takes the difference between the output and the reward function as a loss function of the neural network, updates the parameters of the neural network through a gradient descent method, and the data in the judgment pool is obtained through the data in the judgment pool, and the current state is obtained as the output of the neural network, and the output of the neural network is used as the output of the judgment network is compared.
Calculating channel transmitting power through the initial state and the transmitting power adjusting value, wherein the expression is:
Pch＝P0+PΔ
Wherein, P ch is channel transmitting power, P 0 is initial state, and P Δ is power adjusting value;
Calculating corresponding rewards through the signal-to-noise ratio and the target signal-to-noise ratio obtained by the receiving end, wherein the expression is as follows:
Ri＝||SNRi-SNRtarget||2
Where R i is the prize, SNR i is the receiver SNR, and SNR target is the target SNR.
The module M4.1 takes the initial value of attenuation as an initial state, takes the initial state as the input of the execution neural network, obtains output as an adjustment value of attenuation, calculates the attenuation of each channel through the initial state and the attenuation adjustment value, and then transmits the attenuation and obtains the result of a receiving end; and finally, obtaining the signal-to-noise ratio of each channel at the receiving end, calculating corresponding rewards through the signal-to-noise ratio and the target signal-to-noise ratio obtained at the receiving end, updating the initial state to the next state, collecting the initial state, the attenuation adjustment value, the next state and the rewards into data pairs, storing the data pairs into a data pool, generating data in the data pool through continuous interaction of the execution neural network and the simulation transmission system, performing training on the neural network by using the data in the data pool, and performing training on the execution neural network by using the data in the data pool and the training on the neural network by using the module M4.3.
And calculating the attenuation of each channel through the initial state and the attenuation adjustment value, wherein the expression is as follows:
Ach＝A0+AΔ
Wherein, A ch is channel attenuation, A 0 is initial state, and A Δ is attenuation adjustment value.
Those skilled in the art will appreciate that the systems, apparatus, and their respective modules provided herein may be implemented entirely by logic programming of method steps such that the systems, apparatus, and their respective modules are implemented as logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers, etc., in addition to the systems, apparatus, and their respective modules being implemented as pure computer readable program code. Therefore, the system, the device and the respective modules thereof provided by the invention can be regarded as a hardware component, and the modules for realizing various programs included therein can be regarded as a structure in the hardware component, and the modules for realizing various functions can be regarded as a structure in the hardware component as well as a software program for realizing the method.
The foregoing describes specific embodiments of the present application. It is to be understood that the application is not limited to the particular embodiments described above, and that various changes or modifications may be made by those skilled in the art within the scope of the appended claims without affecting the spirit of the application. The embodiments of the application and the features of the embodiments may be combined with each other arbitrarily without conflict.
Claims (10)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411926195.2A CN119853814B (en) | 2024-12-25 | 2024-12-25 | Link power control method and system for coherent optical wavelength division multiplexing transmission system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411926195.2A CN119853814B (en) | 2024-12-25 | 2024-12-25 | Link power control method and system for coherent optical wavelength division multiplexing transmission system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119853814A true CN119853814A (en) | 2025-04-18 |
| CN119853814B CN119853814B (en) | 2025-10-28 |
Family
ID=95359451
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411926195.2A Active CN119853814B (en) | 2024-12-25 | 2024-12-25 | Link power control method and system for coherent optical wavelength division multiplexing transmission system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119853814B (en) |
Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050225840A1 (en) * | 2004-02-11 | 2005-10-13 | Drasek William A V | Dynamic laser power control for gas species monitoring |
| US20160315729A1 (en) * | 2015-04-23 | 2016-10-27 | Fujitsu Limited | Optical transmission device that transmits wavelength division multiplexed optical signal and optical transmission system |
| CN108880692A (en) * | 2018-06-05 | 2018-11-23 | 西南交通大学 | A kind of identification of coherent optical communication system modulation format and optical signal to noise ratio monitoring method |
| US20210135764A1 (en) * | 2020-11-17 | 2021-05-06 | Intel Corporation | Optical analog matrix multiplier for optical neural networks |
| CN113301637A (en) * | 2021-05-20 | 2021-08-24 | 东南大学 | D2D communication power control algorithm based on Q learning and neural network |
| CN113595649A (en) * | 2021-07-19 | 2021-11-02 | 联合微电子中心有限责任公司 | Optical nonlinear activation unit, method and photonic neural network |
| CN114337849A (en) * | 2021-12-21 | 2022-04-12 | 上海交通大学 | Physical layer confidentiality method and system based on mutual information quantity estimation neural network |
| CN114499723A (en) * | 2022-01-25 | 2022-05-13 | 西南交通大学 | Optical fiber channel rapid modeling method based on Fourier neural operator |
| WO2022191523A1 (en) * | 2021-03-08 | 2022-09-15 | 삼성전자 주식회사 | Electronic device, and method for controlling power of transmission signal in transformable electronic device |
| US11606160B1 (en) * | 2022-01-10 | 2023-03-14 | X Development Llc | Power control loop for stabilization of link power |
| US20230079797A1 (en) * | 2021-09-14 | 2023-03-16 | Infinera Corp. | Computation of optimum fiber input power |
| CN116318320A (en) * | 2022-09-06 | 2023-06-23 | 哈尔滨工业大学 | A Power Control Allocation Method for Satellite-Earth Link Based on Local Regression Neural Network |
| CN116449499A (en) * | 2023-03-07 | 2023-07-18 | 浙江大学 | Wavelength selective switching device based on micro-ring resonator array |
| US20230402942A1 (en) * | 2020-11-20 | 2023-12-14 | Mitsubishi Electric Corporation | Power conversion device |
| CN117938264A (en) * | 2024-01-31 | 2024-04-26 | 北京邮电大学 | A GRU neural network nonlinear equalization method based on regression coupling value |
| US20240339801A1 (en) * | 2023-04-04 | 2024-10-10 | Guangdong University Of Technology | Intelligent phase control method for laser coherent combination |
| CN118804249A (en) * | 2024-06-18 | 2024-10-18 | 福州大学 | A cellular network power control method and device based on deep reinforcement learning |
| US20240422693A1 (en) * | 2022-03-31 | 2024-12-19 | Industry-University Cooperation Foundation Hanyang University Erica Campus | Full-duplex non-orthogonal multiple access-based transmit power control device employing deep reinforcement learning |
| CN119168881A (en) * | 2024-08-28 | 2024-12-20 | 广东工业大学 | A laser coherent synthesis method based on dual-stream network and reinforcement learning framework |
-
2024
- 2024-12-25 CN CN202411926195.2A patent/CN119853814B/en active Active
Patent Citations (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20050225840A1 (en) * | 2004-02-11 | 2005-10-13 | Drasek William A V | Dynamic laser power control for gas species monitoring |
| US20160315729A1 (en) * | 2015-04-23 | 2016-10-27 | Fujitsu Limited | Optical transmission device that transmits wavelength division multiplexed optical signal and optical transmission system |
| CN108880692A (en) * | 2018-06-05 | 2018-11-23 | 西南交通大学 | A kind of identification of coherent optical communication system modulation format and optical signal to noise ratio monitoring method |
| US20210135764A1 (en) * | 2020-11-17 | 2021-05-06 | Intel Corporation | Optical analog matrix multiplier for optical neural networks |
| US20230402942A1 (en) * | 2020-11-20 | 2023-12-14 | Mitsubishi Electric Corporation | Power conversion device |
| WO2022191523A1 (en) * | 2021-03-08 | 2022-09-15 | 삼성전자 주식회사 | Electronic device, and method for controlling power of transmission signal in transformable electronic device |
| CN113301637A (en) * | 2021-05-20 | 2021-08-24 | 东南大学 | D2D communication power control algorithm based on Q learning and neural network |
| CN113595649A (en) * | 2021-07-19 | 2021-11-02 | 联合微电子中心有限责任公司 | Optical nonlinear activation unit, method and photonic neural network |
| US20230079797A1 (en) * | 2021-09-14 | 2023-03-16 | Infinera Corp. | Computation of optimum fiber input power |
| CN114337849A (en) * | 2021-12-21 | 2022-04-12 | 上海交通大学 | Physical layer confidentiality method and system based on mutual information quantity estimation neural network |
| US11606160B1 (en) * | 2022-01-10 | 2023-03-14 | X Development Llc | Power control loop for stabilization of link power |
| CN114499723A (en) * | 2022-01-25 | 2022-05-13 | 西南交通大学 | Optical fiber channel rapid modeling method based on Fourier neural operator |
| US20240422693A1 (en) * | 2022-03-31 | 2024-12-19 | Industry-University Cooperation Foundation Hanyang University Erica Campus | Full-duplex non-orthogonal multiple access-based transmit power control device employing deep reinforcement learning |
| CN116318320A (en) * | 2022-09-06 | 2023-06-23 | 哈尔滨工业大学 | A Power Control Allocation Method for Satellite-Earth Link Based on Local Regression Neural Network |
| CN116449499A (en) * | 2023-03-07 | 2023-07-18 | 浙江大学 | Wavelength selective switching device based on micro-ring resonator array |
| US20240339801A1 (en) * | 2023-04-04 | 2024-10-10 | Guangdong University Of Technology | Intelligent phase control method for laser coherent combination |
| CN117938264A (en) * | 2024-01-31 | 2024-04-26 | 北京邮电大学 | A GRU neural network nonlinear equalization method based on regression coupling value |
| CN118804249A (en) * | 2024-06-18 | 2024-10-18 | 福州大学 | A cellular network power control method and device based on deep reinforcement learning |
| CN119168881A (en) * | 2024-08-28 | 2024-12-20 | 广东工业大学 | A laser coherent synthesis method based on dual-stream network and reinforcement learning framework |
Non-Patent Citations (3)
| Title |
|---|
| YI-WEI SHEN, RUI-QIAN LI, GUAN-TING LIU, JINGYI YU, XUMING HE, LILIN YI, AND CHENG WANG: "Deep photonic reservoir computing recurrent network", OPTICA, vol. 10, no. 12, 19 December 2023 (2023-12-19) * |
| 杨云朋，柯俊翔，义理林: "基于神经网络同步的全光混沌通信", 光通信研究, vol. 12, no. 06, 10 December 2019 (2019-12-10) * |
| 陈垚至;: "混沌激光通信网络波分复用传输系统设计", 激光杂志, no. 10, 25 October 2018 (2018-10-25) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119853814B (en) | 2025-10-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114793305B (en) | Method, device, apparatus and medium for optical communication | |
| US20230057479A1 (en) | Digital twin with machine learning waveform generation including parameter control for device under test emulation | |
| CN114139324B (en) | Fast and accurate modeling method and system for distributed optical fiber channels based on feature decoupling | |
| CN117997380B (en) | Performance test method and device for dual-mode communication, electronic equipment and medium | |
| CN114626505B (en) | A deep reinforcement learning control method for mobile robots | |
| JP7629247B1 (en) | A ship wake prediction method with self-adaptive time interval based on mirror gate network | |
| Song et al. | Lifecycle management of optical networks with dynamic-updating digital twin: a hybrid data-driven and physics-informed approach | |
| CN119853814B (en) | Link power control method and system for coherent optical wavelength division multiplexing transmission system | |
| JP2020067984A (en) | Information processing system and information processing method | |
| Akinrintoyo et al. | Reconfigurable topology testbeds: A new approach to optical system experiments | |
| Liu et al. | Digital twin modeling and controlling of optical power evolution enabling autonomous-driving optical networks: a Bayesian approach | |
| JP2023073761A (en) | Information processing device, control system, information processing method, and program | |
| JP2010081284A (en) | Wavelength dispersion compensation design system and wavelength dispersion compensation design method | |
| CN117235919A (en) | A parameter optimization method for few-mode optical fiber based on neural network reverse design | |
| CN120072135B (en) | Cement-based grouting material proportion optimization method and system based on deep reinforcement learning | |
| EP3002893B1 (en) | A method for producing a quality of transmission estimator for optical transmissions | |
| CN120567699A (en) | Construction method, device and equipment of digital twin model based on industrial Internet | |
| Minakhmetov et al. | Accurate prediction via artificial neural network of OSNR penalty induced by non-uniform WSS filtering | |
| CN105610496B (en) | The computational methods of the built in advance transmission network OSNR analogues value | |
| CN115695211A (en) | Data-driven end-to-end modeling method and system for optical fiber communication experimental system | |
| CN118740657A (en) | A method and device for predicting optical fiber link gain | |
| CN116437004B (en) | Parameter adjustment method, device, computer readable storage medium and electronic device | |
| CN115549776A (en) | Raman amplification system simulation method and device | |
| CN118101399B (en) | A Chaotic Communication Method Based on Reserve Pool Computation | |
| EP3002892A1 (en) | A method for determining an optical transmission system description |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |