Optical communication system global optimization method based on neural network combined genetic algorithm
Through the global optimization method of optical communication system with neural network combined with genetic algorithm, the multi-dimensional parameter optimization problem in high-baud rate optical communication system is solved, and the optimization of FIR tap coefficient, Clipping coefficient and DAC peak-to-peak value is realized, which improves the system performance and Signal-to-noise ratio.
Patent Information
- Application Number
- CN202411118646.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-15
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-08-15
AI Technical Summary
In high baud rate optical communication systems, the transceiver bandwidth limitation and high frequency fading make it difficult to optimize the configuration of multi-dimensional parameters, affecting the system signal-to-noise ratio.
The global optimization method of optical communication system using neural network combined genetic algorithm is used to model offline system, and alternately update multi-dimensional parameters using genetic algorithm and backpropagation algorithm, optimize parameters such as FIR tap coefficient, Clipping coefficient and DAC peak-to-peak value.
It realizes high-precision and rapid global optimization configuration of optical communication systems, improves system performance, and meets the timely parameter optimization requirements of high-speed optical communication systems.
Smart Images

Figure CN119030888A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to optical fiber communication technology, in particular to a global optimization technology for a high baud rate optical communication system. Background Art
[0002] In recent years, with the rapid development of cloud computing, data centers, ultra-high-definition video and other technologies, the global network data traffic has shown explosive growth. This trend has prompted the fiber-optic communication system to continue to develop in the direction of larger capacity and higher transmission rate, especially when the baud rate of the system exceeds 100GBaud, the bandwidth of the electronic devices in the transceiver is increasingly limited.
[0003] In order to overcome the band limitation of the transceiver and improve the system performance, on the one hand, at the transmitter, digital pre-emphasis DPE technology can be used to enhance the high-frequency components of the signal. However, DPE will increase the peak-to-average power ratio PAPR of the signal. In a peak-limited system, it will reduce its average power, which ultimately reduces the signal-to-noise ratio SNR of the system. On the other hand, at the receiver, a feedforward equalizer FFE can be used to compensate for the high-frequency components of the signal, but at the same time it will cause the high-frequency noise to be enhanced, affecting the system performance. Therefore, it is necessary to reasonably allocate the degree of compensation of each transceiver, that is, to optimize the tap coefficients of the transmitter DPE. Considering that the use of DPE or FFE alone has its limitations, partial DPE has been proposed. Among the partial DPE schemes, one is to use a two-tap filter 1+αz -1 , by scanning the tap coefficient α, the value at which the system performance is optimal is determined. Another solution is partial channel compensation, that is, setting the pre-emphasis frequency response to |H -β , after Fourier transformation, the time domain tap coefficient can be obtained, where H is the channel frequency response and β is the compensation degree, and its optimal value needs to be obtained through scanning. The above are all single-variable optimization schemes with limited optimization capabilities. In addition, since the PAPR of the signal after DPE is large, the signal needs to be clipped. The digital signal after Clipping passes through the digital-to-analog converter DAC, and the peak-to-peak value Vpp of the DAC output signal needs to be further controlled. If Vpp is too large, it will cause nonlinearity of subsequent devices. If it is too small, the signal power will be low, affecting the signal-to-noise ratio. Summary of the invention
[0004] The technical problem to be solved by the present invention is to provide a method for optimizing the configuration of multi-dimensional parameters of the system, such as FIR tap coefficients, Clipping coefficients, DAC peak-to-peak values, etc., by performing high-precision and rapid modeling of an offline system in order to solve the problem that in a high-baud rate optical communication system, it is difficult to optimize the configuration of multi-dimensional parameters of the system due to limited transceiver bandwidth and severe high-frequency fading, thereby affecting the signal-to-noise ratio of the system.
[0005] The technical solution adopted by the present invention to solve the above technical problems is a global optimization method of an optical communication system using a neural network combined with a genetic algorithm, comprising the steps of:
[0006] Steps to establish the offline optical communication system model:
[0007] Determine a parameter module to be configured in an optical communication system; use multi-dimensional parameters to represent the parameter to be configured, where N parameters to be configured have N dimensions; train a neural network to simulate a non-parameter module to be configured in the optical communication system;
[0008] The parameter module to be configured is placed before the trained neural network to form an offline optical communication system model;
[0009] Multidimensional parameter alternating update steps:
[0010] A genetic algorithm GA is used to update a round of multidimensional parameters, wherein the received BP-optimized multidimensional parameters are used as the first individual in the initial population; the genetic algorithm GA outputs the multidimensional parameters corresponding to the best individual in the last population as the GA-optimized multidimensional parameters to the back propagation algorithm BP; the GA-optimized multidimensional parameters are used as the BP initial parameter values, and BP is used to update a round of multidimensional parameters, and the BP-optimized multidimensional parameters are output, completing a multidimensional parameter alternation update;
[0011] Determine whether the currently updated multidimensional parameters meet the alternating update end condition. If so, output the latest multidimensional parameters as the multidimensional parameters for global optimization. Otherwise, repeat the alternating update of the multidimensional parameters.
[0012] Actual configuration steps: Use the globally optimized multi-dimensional parameters to configure the real optical communication system to complete the global optimization of the optical communication system.
[0013] Specifically, the optical communication system is an intensity modulation direct detection IMDD system; the parameters that need to be configured include FIR tap coefficients, Clipping coefficients and DAC peak-to-peak value Vpp; the parameter modules to be configured include FIR filters, Clipping models and digital-to-analog converters DAC; the DAC peak-to-peak value Vpp is input into the neural network as a configurable parameter.
[0014] In order to flexibly configure and optimize parameters, the present invention, on the one hand, models the system through a pre-FIR filter (DPE module) and a Clipping neural network to obtain an offline model with configurable FIR tap coefficients, Clipping coefficients, and DAC peak-to-peak values, so that the parameter optimization process can be completed without affecting the actual system operation. On the other hand, due to the timeliness of communication needs, the system has high requirements on the convergence speed of the algorithm. Genetic algorithm GA is a method for searching for the global optimal solution by simulating the natural evolution process, but the algorithm converges slowly. The back propagation algorithm BP based on neural network guides the parameter optimization direction by calculating the gradient. The algorithm converges quickly, but it is easy to fall into the local optimum. The present invention combines the advantages of the above two algorithms to achieve fast global optimal search capabilities, thereby completing the fast global optimization of various parameters in a severely band-limited optical communication system and improving system performance.
[0015] The beneficial effects of the present invention are as follows: firstly, an offline model of configurable system parameters is established, and then the characteristics of the GA and BP algorithms are combined, and the two are used alternately to accelerate the global optimal search for parameters, and the respective advantages of the two algorithms are fully utilized, while the disadvantages are avoided, so as to realize the high-complexity and multi-dimensional parameter optimization configuration of the entire system, improve the performance of the optical communication system with severe transceiver band limitation, and meet the requirements of timely global optimization of parameters in high-speed optical communication systems. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] Figure 1 This is the GA and BP alternating optimization flow chart of the present invention.
[0017] Figure 2 It is a schematic diagram of the system modeling and parameter global optimization scheme of the present invention. DETAILED DESCRIPTION
[0018] The global optimization method of the optical communication system by combining neural network with genetic algorithm first completes the offline optical communication modeling through the neural network of the module of the parameters to be configured in advance, and obtains the offline optical communication model of the configurable system. Then, based on the offline optical communication model, the optimization of the parameters to be configured is performed alternately in steps by combining the GA and BP algorithms, so that it has the ability of global search and local rapid convergence, and completes the optimization of the parameters to be configured. There can be multiple parameters to be configured, each of which corresponds to a dimension. N parameters to be configured have N dimensions, which are called multidimensional parameters. N is a positive integer greater than or equal to 1.
[0019] The specific steps of combining GA and BP algorithms are as follows: Figure 1 As shown:
[0020] Genetic algorithm GA: Generate an initial population, and the first individual in the population is replaced by the multidimensional parameters output by the BP algorithm. If the GA algorithm has not undergone BP optimization in the first round, it will not be replaced. Each individual in the population is a set of parameters to be optimized, that is, a multidimensional parameter, whose value is randomly generated within the allowable range. Configure each set of parameters in the population to the system model in turn, and calculate the MSE of the model input and output signals under each set of parameters. Select the three individuals with the highest fitness in the population, cross them with each other, and then mutate them. The new individuals obtained form a new population with the original three individuals. Then repeat the above operations of calculating fitness, selecting, crossing, mutating, and generating a new population 5 times, and output the multidimensional parameters corresponding to the best individual in the population for the last time;
[0021] Back propagation algorithm BP: The multidimensional parameters corresponding to the optimal individual output by the GA algorithm are used as the initial parameter values and configured into the system model to calculate the MSE of the model input and output signals under the current multidimensional parameter configuration. The gradient of the MSE function to the multidimensional parameters is obtained through the back propagation method of the neural network, and then the multidimensional parameters are updated using the gradient descent method. The above operations are repeated until the MSE converges and the final multidimensional parameters are output;
[0022] Repeat the GA and BP steps alternately for multiple times until the MSE converges and the performance is stable, that is, the optimal parameter value is output. Figure 1 .
[0023] Example
[0024] like Figure 2 As shown, for the intensity modulation direct detection IMDD system in actual optical communication, it is composed of a transmitting end including a FIR filter, a Clipping module, a digital-to-analog converter DAC, an electrical amplifier, a laser and a Mach-Zehnder modulator MZM, an optical fiber, and a receiving end consisting of a photodetector PD, an analog-to-digital converter ADC and a feedforward equalizer FFE.
[0025] The pulse amplitude modulation (PAM) signal at the transmitting end needs to be processed by FIR filter and Clipping, and then the analog electrical signal is obtained by DAC and amplified by the electrical amplifier. The electrical signal after the electrical amplifier and the laser output by the laser are input into the MZM for modulation to obtain the optical modulation signal. After that, the optical signal emitted by the MZM reaches the receiving end after being transmitted through the optical fiber. At the receiving end, the optical signal is restored to the original PAM signal after passing through the PD, ADC and FFE in sequence.
[0026] The optimization of the IMDD system in the embodiment is to configure the optimal tap coefficients of the FIR filter at the transmitting end, the Clipping coefficients of the Clipping module and the DAC peak-to-peak value Vpp input to the DAC. Before optimizing the parameters, an IMDD system offline model is first established, such as Figure 2 Only the FIR filter and Clipping module are retained to facilitate the configuration of the parameters to be optimized. The remaining modules in the IMDD system are all modeled using a deep neural network. FIR is a feedforward N-tap filter with a tap coefficient [w 0 ,w 1 ,……,w N-1 ] is a configurable parameter, and the other two configurable parameters are Clipping coefficient and DAC peak-to-peak value Vpp.
[0027] On the established IMDD system offline model, the parameter optimization method of GA and BP is used to complete the fast global optimization of multi-dimensional parameters. The multi-dimensional parameters composed of FIR tap coefficients, Clipping coefficients and Vpp configuration output by GA or BP are used alternately in the IMDD system offline model. The PAM signal x input to the IMDD system offline model is passed through the system to obtain the output PAM signal Finally, the mean square error (MSE) of the input and output signals is calculated. The MSE is returned to the GA or BP, which then updates the multidimensional parameters based on the returned MSE. The above process is repeated to complete parameter optimization. Finally, the optimized final parameters are configured in the actual IMDD system to complete global optimization.
[0028] The specific implementation steps are as follows:
[0029] 1) Collect sample data: Build a Figure 2 The real IMDD system shown in the figure, without adding FIR modules and Clipping to the system, uses random numbers to generate a large number of random signals (limiting the signal values to be within a feasible range) and inputs them into the system. Then, different DAC peak-to-peak values Vpp are configured and input and output test data between DAC and FFE are collected.
[0030] 2) Offline system modeling: Use the data tested in the previous step as training samples for the neural network, and train the parameters of the neural network. After the training is completed, a neural network model corresponding to the system without FIR and Clipping can be obtained. A FIR filter and a Clipping module are cascaded in front of the neural network model to obtain a complete offline IMDD system model. The multi-tap structure of the FIR filter and the structure of the neural network are shown in Figure 2 As shown, the two modules have a multi-stage delay extraction structure before the signal stream input, where D represents the delay of one symbol period. The number of D can be determined according to the symbol length of the crosstalk, and multiple Ds indicate that there is crosstalk between multiple symbols.
[0031] 3) Alternating optimization:
[0032] After establishing the offline IMDD system model, we first generate a population through the GA algorithm. Each individual in the population corresponds to a multidimensional parameter and is configured into the offline system model. Then the PAM signal x is input into the offline IMDD system model. After the offline IMDD system model is completed, the PAM signal is output. The MSE performance of the offline IMDD system model under the multidimensional parameters is calculated and returned to the GA algorithm. According to the returned MSE, the GA algorithm selects the three individuals with the best performance for selection, crossover, mutation and other operations to generate a new population. The GA process is terminated after repeating 5 times, and the multidimensional parameters corresponding to the optimal individuals are output as the multidimensional parameters optimized by GA and passed to the BP process as the initial parameter values.
[0033] In the BP process, the multidimensional parameters determined by GA are first configured into the offline IMDD system model, and then the PAM signal x is input into the offline IMDD system model. After the offline IMDD system model is completed, the PAM signal is output. The BP process determines whether the offline IMDD system model is stable by calculating the MSE. If not, the gradient is calculated by back propagation and the parameters are updated. If not, the BP process is repeated until the MSE performance is stable. If yes, the current multi-dimensional parameters are output as the multi-dimensional parameters optimized by BP, completing a GA-BP alternation.
[0034] After completing the GA-BP alternation once, determine whether the current MSE performance has converged. If there is no obvious improvement in performance, it is converged. If not, the multidimensional parameters optimized by BP are passed to the GA process. Otherwise, it ends and outputs the parameter value as the optimal multidimensional parameter.
[0035] 4) Actual test: On the actual hardware IMDD system, the optimal multi-dimensional parameter tap coefficients, Clipping coefficients and Vpp are respectively configured to the FIR filter, Clipping and DAC to complete the global optimization of the IMDD system, and the performance of the IMDD system is tested to determine whether it meets the system requirements.
Claims
1. A global optimization method for optical communication systems using a neural network combined with a genetic algorithm, characterized in that: Includes steps: Steps to establish the offline optical communication system model: Determine a parameter module to be configured in an optical communication system; use multi-dimensional parameters to represent the parameter to be configured, where N parameters to be configured have N dimensions; train a neural network to simulate a non-parameter module to be configured in the optical communication system; The parameter module to be configured is placed before the trained neural network to form an offline optical communication system model; Multidimensional parameter alternating update steps: A genetic algorithm GA is used to update a round of multidimensional parameters, wherein the received BP-optimized multidimensional parameters are used as the first individual in the initial population; the genetic algorithm GA outputs the multidimensional parameters corresponding to the best individual in the last population as the GA-optimized multidimensional parameters to the back propagation algorithm BP; the GA-optimized multidimensional parameters are used as the BP initial parameter values, and BP is used to update a round of multidimensional parameters, and the BP-optimized multidimensional parameters are output, completing a multidimensional parameter alternation update; Determine whether the currently updated multidimensional parameters meet the alternating update end condition. If so, output the latest multidimensional parameters as the multidimensional parameters for global optimization. Otherwise, repeat the alternating update of the multidimensional parameters. Actual configuration steps: Use the globally optimized multi-dimensional parameters to configure the real optical communication system to complete the global optimization of the optical communication system.
2. The method according to claim 1, characterized in that: The specific method of using the genetic algorithm GA to update the multidimensional parameters is as follows: generate an initial population, each individual in the population is a multidimensional parameter, and its value is randomly generated within the allowed range, and the first individual in the population is replaced by the multidimensional parameter optimized by BP; The multidimensional parameters corresponding to each individual in the population are configured to the offline optical communication system model in turn, and the mean square error (MSE) of the input and output signals of the offline optical communication system model under each multidimensional parameter is calculated to reflect the fitness; the smaller the MSE, the higher the fitness; the three individuals with the highest fitness in the population are selected, crossover and mutation are performed with each other, and the new individuals obtained form a new population with the original three individuals; then the above operations of calculating fitness, selecting, crossover, mutating, and generating a new population are repeated 5 times; the multidimensional parameters corresponding to the best individual in the population at the last time are output as the multidimensional parameters optimized by GA.
3. The method according to claim 1 or 2, characterized in that: When the genetic algorithm GA is used to update the first round of multidimensional parameters, before receiving the multidimensional parameters optimized by BP, the multidimensional parameters corresponding to the first individual in the initial population are randomly generated within the allowed range.
4. The method according to claim 1, characterized in that: The specific method of using the back propagation algorithm BP to update the multidimensional parameters is: configure the multidimensional parameters optimized by GA into the offline optical communication system model; calculate the mean square error MSE of the input and output signals of the model under the current multidimensional parameter configuration, and then calculate the gradient of the current multidimensional parameters through the back propagation method of the neural network according to the MSE, and then use the gradient descent method to update the multidimensional parameters, and configure the updated multidimensional parameters into the offline optical communication system model, and then calculate the mean square error MSE of the input and output signals of the model under the current multidimensional parameter configuration, repeat the multidimensional parameter update operation until the BP iteration termination condition is reached, and output the current multidimensional parameters as the multidimensional parameters optimized by BP.
5. The method according to claim 1, characterized in that: The optical communication system is an intensity modulation direct detection IMDD system; the parameters that need to be configured include FIR tap coefficients, Clipping coefficients and DAC peak-to-peak value Vpp; the parameter modules to be configured include FIR filters, Clipping models and digital-to-analog converters DAC; the DAC peak-to-peak value Vpp is input into the neural network as a configurable parameter.
6. A computer program product comprising a computer program / instructions, characterized in that When the computer program / instructions are executed by a processor, the steps of the method according to claim 1 are implemented.
Citation Information
Patent Citations
Supercritical fluid heat transfer coefficient prediction method based on interpretable machine learning
CN117150896A
SGP4 model precision improvement method and system based on GA-BP neural network
CN117454963A
Method and system for predicting performance degradation of SOFC (solid oxide fuel cell) stack based on GA-BP (genetic algorithm-back propagation)
CN117725446A
Simulation-based parameter presetting for dynamic driving functions
DE102023203125A1
