Noise nonlinear high-precision joint modeling method, device, equipment and medium
Through the time-domain frequency domain joint modeling method, combined with Fourier transform and neural network, unified modeling of noise and nonlinear factors in microwave photonic devices is achieved, solving the problem of difficulty in dealing with noise and nonlinearity simultaneously in the existing technology, and improving simulation accuracy and design capabilities.
Patent Information
- Application Number
- CN202411696258.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-11-25
- Publication Date
- 2025-02-28
AI Technical Summary
The prior art is difficult to accurately handle noise and nonlinear factors in microwave photonic devices simultaneously under a single model framework, resulting in limited simulation accuracy of complex links or systems.
Through the joint modeling method of time domain frequency domain, a behavioral level down-order model is constructed, spectrum components and amplitude components are extracted using Fourier transform, noise characteristics are extracted based on neural networks and equivalent circuits, and nonlinear behavioral level models are multiplied or added with noise characteristics to form a unified output model.
It improves the accuracy of complex links and system simulation, enhances the unified modeling ability of noise and nonlinearity in microwave photonic device models, and improves design capabilities and design success rate.
Smart Images
Figure 00000000_0000_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of microwave photon simulation technology, and in particular to a noise nonlinear high-precision joint modeling method, device, equipment and medium. Background Art
[0002] The behavioral model of microwave photonic devices is the core unit for building microwave photonic components, links and systems, and its accuracy is crucial to support functional simulation and performance evaluation. The important factors affecting the accuracy of microwave photonic device modeling are mainly the effects of noise and nonlinearity on signals. It is worth noting that there are many types of noise and nonlinearity, and the main sources of noise and nonlinearity faced by different types of devices are also different. For example, in terms of noise, there is relative intensity noise and frequency noise inside the laser; the modulator is affected by thermal noise; and the detector may encounter problems such as shot noise. As for nonlinear problems, the performance of the laser is significantly affected by the relaxation oscillation phenomenon; the modulator causes third-order intermodulation distortion due to the third-order nonlinear effect; and the detector will experience second-order nonlinear distortion.
[0003] Obviously, for the simulation of microwave photonic links or systems, the existence of noise and nonlinearity greatly challenges the accuracy of simulation. Specifically, the noise level directly determines the simulation results of the system sensitivity; and the nonlinear characteristics are one of the key factors affecting the simulation effect of the system dynamic range. At present, in most research works on microwave photonic devices, noise and nonlinearity are usually modeled separately. For example, polynomial models, Volterra series, and harmonic distortion models are used to describe nonlinear characteristics, and nonlinear simulation is achieved through frequency domain calculations; for noise, the equivalent circuit method is generally used to analyze noise performance in the time domain. However, so far, there has been no report that comprehensively considers noise and nonlinear factors in the same model.
[0004] When faced with the simulation task of complex microwave photonic links or systems, methods based on pure nonlinear considerations can only be used to evaluate relevant indicators such as the dynamic range of the link; similarly, modeling methods that focus on noise analysis can only provide information about noise-sensitive parameters such as sensitivity. This separate modeling strategy makes it difficult to accurately and comprehensively evaluate the performance indicators of the entire link or system, which in turn limits the guidance for the selection of the optimal design solution in engineering practice. Therefore, in order to achieve the purpose of higher-precision microwave photonic simulation, how to develop a new method that can simultaneously accommodate and properly handle the dual factors of noise and nonlinearity under a single model framework has become one of the research directions of technical personnel in this field. Summary of the invention
[0005] The purpose of this application is to overcome the existing technical defects and provide a noise nonlinear high-precision joint modeling method, device, equipment and medium. By means of joint modeling in the time domain and frequency domain, the nonlinear characteristics and noise characteristics are loaded into the reduced-order model, thereby improving the accuracy of complex link and system simulation, thereby improving the design capability and design success rate of microwave photonic complex links and systems.
[0006] The purpose of this application is achieved through the following technical solutions:
[0007] In a first aspect, the present application proposes a noise nonlinear high-precision joint modeling method, the method comprising:
[0008] Step S1, constructing a behavioral level reduced-order model about time;
[0009] Step S2, performing Fourier transform on the behavioral level reduced-order model to extract the spectral components and corresponding amplitude components of the input port;
[0010] Step S3, obtaining an output signal based on the spectral components and the corresponding amplitude components;
[0011] Step S4, performing inverse Fourier transform on the output signal to obtain a device nonlinear behavior level model;
[0012] Step S5, extracting noise characteristics of microwave photonic devices using neural networks and equivalent circuits based on a device nonlinear behavior level model;
[0013] Step S6, determining whether there is multiplicative noise in the noise characteristics of the microwave photonic device, if yes, multiplying the device nonlinear behavior level model with the multiplicative noise to obtain a multiplicative model, if no, executing step S7;
[0014] Step S7, determine whether additive noise exists, if yes, add the multiplication model and the additive noise to obtain the output model, if no, end.
[0015] In a possible implementation, step S3 includes:
[0016] Step S31, based on the spectral components and the corresponding amplitude components, a nonlinear mapping relationship between the spectral components and the corresponding amplitude components is obtained by extracting the nonlinear characteristics of the device using a neural network;
[0017] Step S32: Obtain an output signal based on the nonlinear mapping relationship.
[0018] In a possible implementation, the step of performing Fourier transform on the behavioral level reduced-order model includes:
[0019] Use continuous Fourier transform on continuous-time signals for behavioral level reduced-order models;
[0020] Use discrete Fourier transform or fast Fourier transform on discrete-time signals of behavioral reduced-order models.
[0021] In a second aspect, the present application proposes a noise nonlinear high-precision joint modeling device, the device comprising:
[0022] A building block for building behavioral reduced-order models with respect to time;
[0023] An extraction module, used for performing Fourier transform on the behavioral level reduction model, and extracting the spectral components and corresponding amplitude components of the input port;
[0024] A first generating module, configured to obtain an output signal based on the spectral components and the corresponding amplitude components;
[0025] The second generation module is used to perform inverse Fourier transform on the output signal to obtain a nonlinear behavior level model of the device;
[0026] An extraction module for extracting noise characteristics of microwave photonic devices using neural networks and equivalent circuits based on a device nonlinear behavior level model;
[0027] A multiplication module is used to determine whether there is multiplicative noise in the noise characteristics of the microwave photonic device. If so, the device nonlinear behavior level model is multiplied with the multiplicative noise to obtain a multiplication model. If not, an addition module is executed;
[0028] The addition module is used to determine whether there is additive noise. If so, the multiplication model and the additive noise are added to obtain the output model. If not, the process ends.
[0029] In a possible implementation, the first generating module is used to:
[0030] Based on the spectral components and the corresponding amplitude components, a neural network is used to extract the nonlinear characteristics of the device to obtain a nonlinear mapping relationship between the spectral components and the corresponding amplitude components;
[0031] An output signal is obtained based on the nonlinear mapping relationship.
[0032] In a possible implementation, the extraction module is used to:
[0033] Use continuous Fourier transform on continuous-time signals for behavioral level reduced-order models;
[0034] Use discrete Fourier transform or fast Fourier transform on discrete-time signals of behavioral reduced-order models.
[0035] In a third aspect, the present application further proposes a computer device, comprising a processor and a memory, wherein the memory stores a computer program, and the computer program is loaded and executed by the processor to implement the noise nonlinear high-precision joint modeling method as described in any one of the first aspects.
[0036] In a fourth aspect, the present application further proposes a computer-readable storage medium, in which a computer program is stored. The computer program is loaded and executed by a processor to implement the noise nonlinear high-precision joint modeling method as described in any one of the first aspects.
[0037] The above-mentioned main scheme of the present application and its further options can be freely combined to form multiple schemes, all of which are schemes that can be adopted and claimed for protection in the present application; and in the present application, (non-conflicting options) options and other options can also be freely combined. After understanding the scheme of the present application, those skilled in the art can understand that there are multiple combinations based on the prior art and common knowledge, all of which are technical schemes to be protected by the present application, and they are not exhaustively listed here.
[0038] The present application discloses a noise nonlinear high-precision joint modeling method, device, equipment and medium, constructs a behavioral-level reduced-order model, performs Fourier transform to extract the spectral components and corresponding amplitude components of the input port, obtains the output signal based on the spectral components and the corresponding amplitude components, performs inverse Fourier transform to obtain the device nonlinear behavioral-level model, uses neural networks and equivalent circuits to extract the noise characteristics of microwave photonic devices, and determines whether there is multiplicative noise in the noise characteristics of microwave photonic devices. If so, the device nonlinear behavioral-level model is multiplied by the multiplicative noise to obtain a multiplicative model. Otherwise, it is determined whether there is additive noise. If so, the multiplicative model is added to the additive noise to obtain the output model. By means of joint modeling in the time domain and frequency domain, the nonlinear characteristics and noise characteristics are loaded into the reduced-order model, the accuracy of complex link and system simulation is improved, and the design capability and design success rate of microwave photonic complex links and systems are thereby improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0040] Figure 1 A flow chart of a noise nonlinear high-precision joint modeling method proposed in an embodiment of the present application is shown.
[0041] Figure 2A schematic diagram of device input and output spectrum component mapping under nonlinear action proposed in an embodiment of the present application is shown.
[0042] Figure 3 A schematic diagram showing the mapping relationship between the spectral components and the amplitude components output by the device proposed in the embodiment of the present application is shown.
[0043] Figure 4 A schematic diagram of a basic microwave photonic link model proposed in an embodiment of the present application is shown.
[0044] Figure 5 A schematic diagram of the output result of the spectrum analyzer proposed in the embodiment of the present application is shown. DETAILED DESCRIPTION
[0045] The following describes the embodiments of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. The present application can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that the following embodiments and features in the embodiments can be combined with each other without conflict.
[0046] Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making any creative work shall fall within the scope of protection of this application.
[0047] In the prior art, when faced with the simulation task of complex microwave photonic links or systems, methods based on pure nonlinear considerations can only be used to evaluate relevant indicators such as the dynamic range of the link; similarly, modeling methods that focus on noise analysis can only provide information about noise-sensitive parameters such as sensitivity. This separate modeling strategy makes it difficult to accurately and comprehensively evaluate the performance indicators of the entire link or system, which in turn limits the guidance for the selection of the optimal design solution in engineering practice. Therefore, in order to achieve the purpose of higher-precision microwave photonic simulation, how to develop a new method that can simultaneously accommodate and properly handle the dual factors of noise and nonlinearity under a single model framework has become one of the research directions of technicians in this field.
[0048] Therefore, in order to solve the above technical problems, the embodiments of the present application propose a noise nonlinear high-precision joint modeling method, device, equipment and medium, which can be applied to the simulation of broadband array systems, links and devices including microwave models, light wave models and digital models. In view of the difficulty of traditional theoretical models in evaluating the noise coefficient, sensitivity, dynamic range and other index systems of microwave photonic links or systems, the nonlinear characteristics and noise characteristics are loaded into the reduced-order model by means of joint modeling in the time domain and frequency domain, realizing the unified modeling of noise and nonlinearity in the microwave photonic device model, thereby supporting the efficient and high-precision performance evaluation of system-level simulation, which will be described in detail below.
[0049] Please refer to Figure 1 , Figure 1 A schematic diagram of a noise nonlinear high-precision joint modeling method proposed in an embodiment of the present application is shown, and the method includes:
[0050] Step S1: construct a behavioral reduced-order model about time.
[0051] According to the basic theory, a behavioral reduced-order model E1(t) about time t is constructed to characterize the basic characteristics of the model.
[0052] Step S2: Perform Fourier transform on the behavioral level reduced-order model to extract the frequency spectrum component and the corresponding amplitude component of the input port.
[0053] The Fourier transform of the behavioral reduced-order model includes:
[0054] Use continuous Fourier transform on continuous-time signals for behavioral level reduced-order models;
[0055] Use discrete Fourier transform or fast Fourier transform on discrete-time signals of behavioral reduced-order models.
[0056] The spectral component is ω=[ω 1 ,ω 2 ,....,ω n ], the amplitude component is V = [V 1 ,V 2 ,....,V n ], convert the behavioral reduced-order model E1(t) at time t to the frequency domain, using continuous Fourier transform for continuous-time signals and discrete Fourier transform or fast Fourier transform for discrete-time signals. Assuming E1(t) is a continuous-time signal, the result obtained by Fourier transform is a complex function, whose real and imaginary parts correspond to cosine and sine components, respectively. The spectral components refer to the values at different frequency points, and the amplitude at each frequency point can be obtained by calculating the modulus of the complex value at that point.
[0057] The spectral components and corresponding amplitude components of the input port are extracted, and the time domain signal is converted into a frequency domain representation, so as to analyze the components of the input port at different frequencies and their corresponding amplitudes.
[0058] Step S3, obtaining an output signal based on the spectral components and the corresponding amplitude components;
[0059] By using the known spectral components of the input signal and their corresponding amplitudes, the spectrum and amplitude distribution of the output signal after passing through the device are calculated according to the mapping rules. The processed frequency domain information can be converted back to the time domain through means such as inverse Fourier transform to obtain a complete output signal.
[0060] Step S3 includes:
[0061] Step S31, based on the spectral components and the corresponding amplitude components, a nonlinear mapping relationship between the spectral components and the corresponding amplitude components is obtained by extracting the nonlinear characteristics of the device using a neural network;
[0062] Step S32: Obtain an output signal based on the nonlinear mapping relationship.
[0063] Select a suitable neural network architecture (such as a deep feedforward network, a recurrent neural network), model the nonlinear relationship between the spectral components and the amplitude components, and train the selected neural network model. During the training process, the weights are adjusted so that the model can learn the complex nonlinear mapping F from the input spectral components to the output amplitude components. The loss function usually measures the difference between the predicted value and the actual measured value, and backpropagates the error to optimize the parameters. After training, the performance of the model needs to be verified using data independent of the training set to ensure that it has good generalization ability, that is, it can make accurate predictions on new data that has not been seen.
[0064] The nonlinear characteristics of the device are extracted using a neural network to obtain a nonlinear mapping relationship between the spectrum component and the amplitude component. At this time, the spectrum component is ω o =F(ω)=[ω 1 ,ω 2 ,ω 2 ±ω 1 ,2ω 2 ±ω 1 ....,ω n ], the amplitude component is V o =F(V)=[V 1 ',V 2 ',V 2 ' 1 ,V 2 ' 1 ,V 2 ' 21 ,V 1 '22 ,....,V n '], the output signal of the device is:
[0065]
[0066] Figure 2 The schematic diagram of the mapping of the input and output spectrum components of the device under the nonlinear action proposed in the embodiment of the present application is shown. Taking ω1 and ω2 as examples, under the nonlinear action of the spectrum components of the two input signals, a series of frequency components related to ω1 and ω1 appear at the output end, and the amplitudes are different. The mapping relationship between each frequency component at the output end and the frequency component at the input end includes: (1) input ω1 is directly mapped to output ω1; (2) input ω1 is mapped to output 2ω1; (3) input ω2 is directly mapped to output ω2; (4) input ω2 is directly mapped to output 2ω2; (5) input ω1 and ω2 are mapped to ω1±ω2 under the joint action; (6) ω1 and 2ω2 are mapped to 2ω2±ω1 under the joint action; (7) ω2 and 2ω1 are mapped to 2ω1±ω2 under the joint action.
[0067] Step S4: Perform inverse Fourier transform on the output signal to obtain a nonlinear behavior model of the device.
[0068] By performing an inverse Fourier transform on the output signal, the complete time domain signal can be recovered from the frequency domain, thereby constructing a behavioral model that can accurately describe the nonlinear response characteristics of a specific electronic component.
[0069] Use the inverse fast Fourier transform to convert the frequency domain data back to the time domain. After IFFT, a new time series signal is obtained, which represents the actual output waveform of the input signal after being processed by a device with nonlinear characteristics. The final time domain signal is actually the simulation result of the nonlinear behavior of the device being sought. By further analyzing the signal (such as comparing with theoretical expectations or experimental measurements), it can be verified whether the modeled nonlinear characteristics accurately reflect the situation in the real world.
[0070] Step S5: Based on the device nonlinear behavior level model, the noise characteristics of the microwave photonic device are extracted using a neural network and an equivalent circuit.
[0071] After establishing a model that can accurately describe the nonlinear response of microwave photonic devices under different input conditions, an equivalent circuit model is used to better understand how the various parts inside the device interact with each other, and to simulate the various noise sources generated by these interactions. Combining neural networks with equivalent circuit models can further improve the understanding of noise characteristics N(t) = {N 1 (t),N 2 (t),…,Nm The joint model is used to analyze the noise characteristics of microwave photonic devices under different working conditions in detail, including but not limited to phase noise, amplitude noise, thermal noise, etc.
[0072] Step S6, determining whether there is multiplicative noise in the noise characteristics of the microwave photonic device, if yes, multiplying the device nonlinear behavior level model with the multiplicative noise to obtain a multiplicative model, if no, executing step S7.
[0073] Check whether the noise characteristics of microwave photonic devices contain multiplicative noise and determine whether there is multiplicative noise N 1 (t), if there is multiplicative noise N 1 (t), then proceed to the next step. If there is no multiplicative noise, then jump to step S7. non (t) and the multiplicative noise N 1 (t) and obtain the multiplication model E N1 (t), the specific formula is: E N1 (t) = E non (t)×N 1 (t).
[0074] Step S7, determine whether additive noise exists, if yes, add the multiplication model and the additive noise to obtain the output model, if no, end.
[0075] Check whether the noise in the system includes additive noise. If additive noise N 2 (t), then the additive noise is added to the existing noise component E N1 (t), the specific formula is: E(t) = E N1 (t)+N 2 (t).E N1 (t) is the noise component (output model) obtained by converting E N1 (t) and additive noise N 2 (t) are added to obtain the final total noise E(t). If there is no additive noise, the modeling process ends directly. The modeling process of noise characteristics is completed. All possible noise types (multiplicative and additive) have been considered. The final E(t) is the total noise model containing all noise components.
[0076] In a possible embodiment, a 1-6 GHz photoelectric detector is taken as an example, and a behavior-level model including noise and nonlinearity is constructed using the noise-nonlinear high-precision joint modeling method proposed in the embodiment of the present application:
[0077] First, based on the basic principles of photodetectors, a reduced-order behavioral model is constructed: where ipd (t) is the photocurrent, is the photodetector responsivity, E in (t) detector input light field, The conjugate of the detector input light field.
[0078] Second, perform FFT transformation on the above formula and extract frequency domain information and amplitude information to obtain ω = [1, 2, 3, 4, 5, 6], V = [1, 1, 1, 1, 1];
[0079] Third, the nonlinear characteristics of the device are extracted using neural networks to obtain the mapping relationship between the spectrum components and amplitude components output by the real device. Figure 3 The schematic diagram of the mapping relationship between the spectral components and the amplitude components of the device output proposed in the embodiment of the present application is shown. ① When the input frequency component is 1 GHz and the normalized amplitude is 1, the frequency components mapped to the output end include 1 GHz, 2 GHz and 3 GHz, and the corresponding amplitude components are 0.3, 0.02, and 0.002, respectively; ② When the input frequency component is 2 GHz, the frequency components mapped to the output end include 2 GHz, 4 GHz and 6 GHz, and the corresponding amplitude components are 0.3, 0.02, and 0.002, respectively; ③ When the input frequency component is 3 GHz, the output frequency components are 3 GHz and 6 GHz. There are only two frequency components here because the 3-fold frequency falls outside the detector bandwidth (2 to 6 GHz) and cannot generate photocurrent at this frequency point; ④ When the input frequency components are 4 to 6 GHz, the output frequency components remain unchanged, and the corresponding amplitude components are 0.3, 0.25, and 0.2, respectively. It can be seen from the above process that the nonlinear effect of the detector is mainly second-order nonlinearity. At the same time, this method not only reflects the nonlinear characteristics, but also presents the inconsistency of the working frequency band, making the model more accurate. Substituting the above characteristics into the frequency domain expression of the detector, we get:
[0080]
[0081] Among them, ω k Indicates the frequency of 1 to 6 GHz, V k '、V k ” and V k ”' respectively represent the amplitudes corresponding to 1 times the frequency, 2 times the frequency and 3 times the frequency.
[0082] Fourth, perform inverse Fourier transform on the above formula to obtain the frequency domain expression:
[0083]
[0084] Fifth, use the equivalent circuit to complete the modeling of the photodetector shot noise:
[0085] N sh=2qI dc R|H pd |
[0086] Where q is the electron charge, Idc is the DC current intensity of the detector, R is the matching resistor, and Hpd is the transfer function of the detector, which can be extracted from the amplitude component data V in nonlinear modeling.
[0087] Sixth, determine whether the noise is multiplicative noise, no;
[0088] Seventh, determine whether the noise is additive noise, yes;
[0089] Eighth, the noise effect is directly loaded into the nonlinear model to obtain:
[0090]
[0091] Ninth, complete the modeling.
[0092] In order to test the model, we encapsulated the above model into a visualization interface and built a basic microwave photonic link. Figure 4 The schematic diagram of the basic microwave photonic link model proposed in the embodiment of the present application is shown, which includes a laser model, a signal source model, a modulator model, a detector model and a spectrum analyzer model. Among them, nonlinear and noise effects are observed through the spectrum analyzer model.
[0093] When the signal source model input frequency is set to 1GHz~6GHz, Figure 5 The schematic diagram of the spectrum analyzer output results proposed in the embodiment of the present application is shown. The peak values of each harmonic are the simulation results of the nonlinear effect, and the bottom of each figure is the noise simulation result. It can be seen from the figure that the simulation results of the model include both nonlinear effects and noise effects. At the same time, the results are consistent with the theoretical analysis, proving the effectiveness of this solution. Among them, (a) to (f) respectively represent the output spectrum characteristics when the input signal frequency is 1 to 6 GHz.
[0094] The present invention provides a noise and nonlinear high-precision joint modeling method for microwave photonic devices, which can solve the problem that it is difficult to simultaneously simulate and evaluate noise and nonlinear effects in traditional microwave photonic link and system simulations. By means of joint modeling in the time domain and frequency domain, nonlinear characteristics and noise characteristics are loaded into a reduced-order model, thereby realizing unified modeling of noise and nonlinearity in a behavioral-level model of microwave photonic devices, and can improve the accuracy of complex link and system simulations, thereby improving the design capability and design success rate of complex microwave photonic links and systems.
[0095] Compared with the prior art, the embodiments of the present application have the following beneficial effects:
[0096] First, by jointly solving in the time domain and frequency domain, nonlinear characteristics and noise characteristics are loaded into the reduced-order model, which can more comprehensively capture the behavior of the device under different conditions and improve the accuracy and reliability of the model.
[0097] Second, the unified modeling of noise (such as multiplicative noise and additive noise) and nonlinearity in microwave photonic device models is realized. By considering both noise and nonlinear effects at the same time, the model can more accurately reflect the behavior of the actual device, thereby improving the accuracy of the simulation results. This is crucial for high-precision simulation of complex systems.
[0098] Third, this method can evaluate key performance indicators such as link, system noise figure, dynamic range and sensitivity, which are important parameters for measuring system performance. By comprehensively evaluating these indicators, we can fully understand the performance of the system and provide reliable data support for design and optimization.
[0099] Fourth, it is not only applicable to the field of microwave photonics, but also can be applied to device modeling in the microwave and optical fields. This wide applicability enables this method to play a role in multiple related fields and improve the level of overall design and simulation. Whether it is the microwave, optics or microwave photonics field, it can benefit from it.
[0100] A possible implementation of a noise nonlinear high-precision joint modeling device is provided below, which is used to execute the various execution steps and corresponding technical effects of the noise nonlinear high-precision joint modeling method shown in the above embodiment and possible implementation, and the device includes:
[0101] A building block for building behavioral reduced-order models with respect to time;
[0102] An extraction module, used for performing Fourier transform on the behavioral level reduction model, and extracting the spectral components and corresponding amplitude components of the input port;
[0103] A first generating module, configured to obtain an output signal based on the spectral components and the corresponding amplitude components;
[0104] The second generation module is used to perform inverse Fourier transform on the output signal to obtain a nonlinear behavior level model of the device;
[0105] An extraction module for extracting noise characteristics of microwave photonic devices using neural networks and equivalent circuits based on a device nonlinear behavior level model;
[0106] A multiplication module is used to determine whether there is multiplicative noise in the noise characteristics of the microwave photonic device. If so, the device nonlinear behavior level model is multiplied with the multiplicative noise to obtain a multiplication model. If not, an addition module is executed;
[0107] The addition module is used to determine whether there is additive noise. If so, the multiplication model and the additive noise are added to obtain the output model. If not, the process ends.
[0108] In a possible implementation, the first generating module is used to:
[0109] Based on the spectral components and the corresponding amplitude components, a neural network is used to extract the nonlinear characteristics of the device to obtain a nonlinear mapping relationship between the spectral components and the corresponding amplitude components;
[0110] An output signal is obtained based on the nonlinear mapping relationship.
[0111] In a possible implementation, the extraction module is used to:
[0112] Use continuous Fourier transform on continuous-time signals for behavioral level reduced-order models;
[0113] Use discrete Fourier transform or fast Fourier transform on discrete-time signals of behavioral reduced-order models.
[0114] This preferred embodiment provides a computer device, which can implement the steps in any embodiment of the noise nonlinear high-precision joint modeling method provided in the embodiments of the present application. Therefore, the beneficial effects of the noise nonlinear high-precision joint modeling method provided in the embodiments of the present application can be achieved. Please see the previous embodiments for details and will not be repeated here.
[0115] Those skilled in the art will appreciate that all or part of the steps in the various methods of the above embodiments can be completed by instructions, or by controlling related hardware through instructions, and the instructions can be stored in a computer-readable storage medium and loaded and executed by a processor. To this end, an embodiment of the present application provides a storage medium in which a plurality of instructions are stored, and the instructions can be loaded by a processor to execute the steps of any embodiment of the noise nonlinear high-precision joint modeling method provided in the embodiment of the present application.
[0116] The storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.
[0117] Since the instructions stored in the storage medium can execute the steps in any noise nonlinear high-precision joint modeling method embodiment provided in the embodiments of the present application, the beneficial effects that can be achieved by any noise nonlinear high-precision joint modeling method provided in the embodiments of the present application can be achieved. Please see the previous embodiments for details and will not be repeated here.
[0118] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present application should be included in the protection scope of the present application.
Claims
1. A noise nonlinear high-precision joint modeling method, characterized in that: The method comprises: Step S1, constructing a behavioral level reduced-order model about time; Step S2, performing Fourier transform on the behavioral level reduced-order model to extract the spectral components and corresponding amplitude components of the input port; Step S3, obtaining an output signal based on the spectral components and the corresponding amplitude components; Step S4, performing inverse Fourier transform on the output signal to obtain a device nonlinear behavior level model; Step S5, extracting noise characteristics of microwave photonic devices using neural networks and equivalent circuits based on a device nonlinear behavior level model; Step S6, determining whether there is multiplicative noise in the noise characteristics of the microwave photonic device, if yes, multiplying the device nonlinear behavior level model with the multiplicative noise to obtain a multiplicative model, if no, executing step S7; Step S7, determine whether additive noise exists, if yes, add the multiplication model and the additive noise to obtain the output model, if no, end.
2. The noise nonlinear high-precision joint modeling method according to claim 1, characterized in that: Step S3 includes: Step S31, based on the spectral components and the corresponding amplitude components, a nonlinear mapping relationship between the spectral components and the corresponding amplitude components is obtained by extracting the nonlinear characteristics of the device using a neural network; Step S32: Obtain an output signal based on the nonlinear mapping relationship.
3. The noise nonlinear high-precision joint modeling method according to claim 1, characterized in that: The steps of Fourier transforming the behavioral reduced-order model include: Use continuous Fourier transform on continuous-time signals for behavioral level reduced-order models; Use discrete Fourier transform or fast Fourier transform on discrete-time signals of behavioral reduced-order models.
4. A noise nonlinear high-precision joint modeling device, characterized in that: The device comprises: A building block for building behavioral reduced-order models with respect to time; An extraction module, used for performing Fourier transform on the behavioral level reduction model, and extracting the spectral components and corresponding amplitude components of the input port; A first generating module, configured to obtain an output signal based on the spectral components and the corresponding amplitude components; The second generation module is used to perform inverse Fourier transform on the output signal to obtain a nonlinear behavior level model of the device; An extraction module for extracting noise characteristics of microwave photonic devices using neural networks and equivalent circuits based on a device nonlinear behavior level model; A multiplication module is used to determine whether there is multiplicative noise in the noise characteristics of the microwave photonic device. If so, the device nonlinear behavior level model is multiplied with the multiplicative noise to obtain a multiplication model. If not, an addition module is executed; The addition module is used to determine whether there is additive noise. If so, the multiplication model and the additive noise are added to obtain the output model. If not, the process ends.
5. The noise nonlinear high-precision joint modeling device according to claim 4, characterized in that: The first generation module is used to: Based on the spectral components and the corresponding amplitude components, a neural network is used to extract the nonlinear characteristics of the device to obtain a nonlinear mapping relationship between the spectral components and the corresponding amplitude components; An output signal is obtained based on the nonlinear mapping relationship.
6. The noise nonlinear high-precision joint modeling device according to claim 4, characterized in that: Extraction module for: Use continuous Fourier transform on continuous-time signals for behavioral level reduced-order models; Use discrete Fourier transform or fast Fourier transform on discrete-time signals of behavioral reduced-order models.
7. A computer device, characterized in that: The computer device includes a processor and a memory, wherein a computer program is stored in the memory, and the computer program is loaded and executed by the processor to implement the noise nonlinear high-precision joint modeling method as described in any one of claims 1 to 3.
8. A computer-readable storage medium, characterized in that: The storage medium stores a computer program, which is loaded and executed by a processor to implement the noise nonlinear high-precision joint modeling method as described in any one of claims 1 to 3.