Symbol constellation generation device, control method, and program

Abstract

The symbol constellation generating device (2000) acquires specification information (10) of an optical communication system (100). The symbol constellation generating device (2000) generates noise feature information (20) using a first estimation model (50) based on the specification information (10). The symbol constellation generating device (2000) generates a symbol constellation (30) suitable for an optical communication system having noise features represented by the noise feature information (20) using a second estimation model (60).

Description

[Technical Field]

[0001] This disclosure generally relates to a symbol control system for use in optical communication systems. Ta Regarding ration. [Background technology]

[0002] In an optical communication system, an optical transmitter converts a message into a symbol sequence and transmits the symbol sequence to an optical receiver. Ta Symbolic Constraints Ta The symbol assignment represents the correspondence between the bit patterns in the message and the symbols to be transmitted for each bit pattern. Ta Providing such a dispersion effectively improves the performance of optical communication systems.

[0003] Symbol Cons Ta As a method for generating a training signal, Patent Document 1 and Non-Patent Document 1 disclose an end-to-end learning system that includes an autoencoder that models an optical communication system. Ta By repeatedly updating the symbol constraints (including the symbol constraints), a symbol constraint is generated that is customized for the optical communication system modeled by the learning system. Ta You can get ration. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Chinese Patent Application Publication No. 109660297 [Non-patent literature]

[0005] [Non-Patent Document 1] Rasmus T. Jones, Tobias A. Eriksson, Metodi P. Yankov, and Darko Zibar, “Deep Learning of Geometric Constellation Shaping including Fiber Nonlinearities,” Computer Research Repository, arXiv:1805.03785, May 10, 2018. Summary of the Invention [Problem to be solved by the invention]

[0006] Obtaining a symbol constellation using the aforementioned end-to-end training system is a time-consuming task since training the training system involves repeatedly updating trainable parameters. It is an object of the present disclosure to provide techniques that reduce the time required to generate a symbol constellation for an optical communication system. [Means for solving the problem]

[0007] The present disclosure discloses a symbol constellation generation device having at least one processor and at least one memory storing instructions, wherein the at least one processor is configured to execute the instructions to acquire specification information representing one or more elements of a specification of an optical communication system including an optical transmitting device, an optical receiving device, and an optical fiber therebetween, generate noise feature information using the acquired specification information and a first guess model, the noise feature information representing one or more noise features of the optical communication system, the first guess model being predefined and representing a correspondence between the specification of the optical communication system and a noise feature of the optical communication system having the associated specification, generate a symbol constellation using the generated noise feature information and a second guess model, the symbol constellation being used by the optical transmitting device to encode a message to be transmitted to the optical receiving device, and the second guess model being predefined and representing a correspondence between the noise feature of the optical communication system and the symbol constellation suitable for the optical communication system having the corresponding noise feature.

[0008] The present disclosure provides a computer-implemented control method, the control method including: acquiring specification information representing one or more elements of specifications of an optical communication system including an optical transmitting device, an optical receiving device, and an optical fiber therebetween; generating noise feature information using the acquired specification information and a first guess model, the noise feature information representing one or more noise features of the optical communication system, the first guess model being predefined and representing a correspondence between the specifications of the optical communication system and noise features of the optical communication system having the associated specifications; generating a symbol constellation using the generated noise feature information and a second guess model, the symbol constellation being used by the optical transmitting device to encode a message to be transmitted to the optical receiving device; and the second guess model being predefined and representing a correspondence between the noise features of the optical communication system and the symbol constellation suitable for the optical communication system having the corresponding noise feature.

[0009] The present disclosure provides a computer-readable storage medium that stores a program for causing a computer to execute the following steps: acquire specification information representing one or more elements of the specifications of an optical communication system including an optical transmitting device, an optical receiving device, and an optical fiber therebetween; generate noise feature information using the acquired specification information and a first guess model, the noise feature information representing one or more noise features of the optical communication system; the first guess model is predefined and represents a correspondence between the specifications of the optical communication system and noise features of the optical communication system having the associated specifications; generate a symbol constellation using the generated noise feature information and a second guess model, the symbol constellation being used by the optical transmitting device to encode a message to be transmitted to the optical receiving device; and the second guess model is predefined and represents a correspondence between the noise features of the optical communication system and the symbol constellation suitable for the optical communication system having the corresponding noise feature. [Effects of the Invention]

[0010] SUMMARY OF THE DISCLOSURE In accordance with the present disclosure, techniques are provided for reducing the time required to generate a symbol constellation for an optical communication system. [Brief explanation of the drawings]

[0011] [Figure 1] FIG. 1 shows an overview of a symbol constellation generating device according to the first embodiment. [Figure 2] FIG. 2 illustrates an example of the functional configuration of a symbol constellation generating device. [Figure 3] FIG. 3 is a block diagram showing an example of the hardware configuration of a computer that realizes a symbol constellation generating device. [Figure 4] FIG. 4 is a flowchart illustrating an example of the flow of processing executed by the symbol constellation generating device. [Figure 5] FIG. 5 shows a flowchart illustrating an example of the overall flow of management of the optical communication system 100. DETAILED DESCRIPTION OF THE INVENTION

[0012] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. The same elements are designated by the same reference numerals throughout the drawings, and redundant description will be omitted as necessary.

[0013] Embodiment 1 <Summary> 1 shows an overview of a symbol constellation generation apparatus 2000 according to embodiment 1. The overview shown in FIG. 1 shows an example of the operation of the symbol constellation generation apparatus 2000, with the aim of facilitating understanding of the symbol constellation generation apparatus 2000, and is not intended to limit the range of possible operations of the symbol constellation generation apparatus 2000.

[0014] The symbol constellation generating device 2000 generates a symbol constellation 30. The symbol constellation 30 is a symbol constellation suitable for the optical communication system 100. Ta One example of a type of optical communication system 100 is a wavelength division multiplexing (WDM) coherent optical communication system via optical fiber. The optical communication system 100 includes a transmitter 110, a receiver 120, and an optical fiber 130. The transmitter 110 encodes a message using a symbol constellation 30 and transmits the encoded message to the receiver 120 via the optical fiber 130. The receiver 120 receives the encoded message transmitted from the transmitter 110 via the optical fiber 130. Using the WDM method, the optical fiber 130 is divided into multiple communication channels 132.

[0015] The symbol constellation generation apparatus 2000 generates a symbol constellation 30 based on the specifications of the optical communication system 100. Specifically, the symbol constellation generation apparatus 2000 estimates the noise characteristics of the optical communication system 100 based on the specifications of the optical communication system 100, and further generates a symbol constellation 30 suitable for the optical communication system 100 having the estimated noise characteristics based on the estimated noise characteristics of the optical communication system 100.

[0016] To this end, the symbol constellation generation device 2000 acquires specification information 10 representing one or more elements of the specifications of the optical communication system 100, and generates noise feature information 20 representing one or more elements of the noise feature of the optical communication system 100. The noise feature of the optical communication system 100 is estimated using the specification information 10 and a first estimation model 50. The first estimation model 50 is prepared in advance and associates the specifications of the optical communication system 100 with the noise feature of the optical communication system 100 having the corresponding specifications.

[0017] Furthermore, the symbol constellation generation device 2000 generates a symbol constellation 30 using the noise feature information 20 and a second estimation model 60. The second estimation model 60 is prepared in advance and associates the noise feature of the optical communication system 100 with a symbol constellation 30 suitable for the optical communication system 100 and having the corresponding noise feature.

[0018] <Examples of effects> As described above, the symbol constellation generation device 2000 uses the first estimation model 50 to estimate the noise characteristics of the optical communication system 100 having the specifications represented by the specification information 10, and further uses the second estimation model 60 to generate the symbol constellation 30 suitable for the optical communication system 100 having the estimated noise characteristics. Because the first estimation model 50 and the second estimation model 60 are prepared in advance, unlike the end-to-end training systems disclosed in Patent Document 1 and Non-Patent Document 1, the symbol constellation generation device 2000 does not need to update these models when generating the symbol constellation 30. Therefore, the symbol constellation generation device 2000 can generate the symbol constellation 30 in a shorter time than the end-to-end training systems disclosed in Patent Document 1 and Non-Patent Document 1.

[0019] Furthermore, since the symbol constellation generation device 2000 takes into account the noise characteristics of the optical communication system 100, it is possible to generate a symbol constellation 30 suitable for an optical communication system having the estimated noise characteristics. Therefore, the symbol constellation generation device 2000 can generate a symbol constellation 30 that can improve the performance of the optical communication system 100 in a shorter time than the end-to-end training systems disclosed in Patent Document 1 and Non-Patent Document 1.

[0020] A more detailed description of the symbol constellation generating device 2000 is provided below.

[0021] <Example of functional configuration> 2 shows an example of the functional configuration of a symbol constellation generation apparatus 2000. The symbol constellation generation apparatus 2000 includes an acquisition unit 2020, a first generation unit 2040, and a second generation unit 2060. The acquisition unit 2020 acquires specification information 10. The first generation unit 2040 generates noise feature information 20 using noise feature information 20 and a first estimation model 50. The second generation unit 2060 generates a symbol constellation 30 using the noise feature information 20 and a second estimation model 60.

[0022] <Example of hardware configuration> The symbol constellation generation apparatus 2000 may be realized by one or more computers, each of which may be a dedicated computer created for realizing the symbol constellation generation apparatus 2000, or a general-purpose computer such as a personal computer (PC), a server machine, or a mobile device.

[0023] Symbol constellation generation apparatus 2000 can be realized by installing an application on a computer. The application is realized as a program that causes the computer to function as symbol constellation generation apparatus 2000. In other words, the program implements the functional components of symbol constellation generation apparatus 2000.

[0024] 3 is a block diagram showing an example of a hardware configuration of a computer 1000 that realizes a symbol constellation generating apparatus 2000. In FIG. 3, the computer 1000 includes a bus 1020, a processor 1040, a memory 1060, a storage device 1080, an input / output interface 1100, and a network interface 1120.

[0025] The bus 1020 is a data communication path through which the processor 1040, memory 1060, storage device 1080, input / output interface 1100, and network interface 1120 transmit and receive data to and from each other. The processor 1040 is a processor such as a central processing unit (CPU), a graphics processing unit (GPU), or a field-programmable gate array (FPGA). The memory 1060 is a main memory element such as a random access memory (RAM) or a read-only memory (ROM). The storage device 1080 is an auxiliary memory element such as a hard disk, a solid state drive (SSD), or a memory card. The input / output interface 1100 is an interface between the computer 1000 and peripheral devices (such as a keyboard, a mouse, or a display device). The network interface 1120 is an interface between the computer 1000 and a network. The network may be a local area network (LAN) or a wide area network (WAN).

[0026] The storage device 1080 may store the above-mentioned programs. Processor 1040 executes the program to realize each functional component of the symbol constellation generating device 2000.

[0027] The hardware configuration of the computer 1000 is not limited to the configuration shown in FIG. For example, as described above, the symbol constellation generating apparatus 2000 may be realized by a plurality of computers, which may be connected to each other via a network.

[0028] <Processing flow> 4 is a flowchart illustrating an example of the flow of processing executed by the symbol constellation generation apparatus 2000. The acquisition unit 2020 acquires the specification information 10 (S102). The first generation unit 2040 generates noise feature information 20 based on the specification information 10 (S104). The second generation unit 2060 generates a symbol constellation 30 based on the noise feature information 20 (S106). The symbol constellation generation apparatus 2000 outputs the symbol constellation 30 (S108).

[0029] <Acquisition of specification information: S102> The acquiring unit 2020 acquires specification information 10 (S102). The specification information 10 represents one or more elements of the specifications of the optical communication system 100. Each element of the specification may be a parameter related to the conditions of a transmission or communication channel (e.g., fiber specifications of the optical fiber 130, frequency specifications of the optical fiber 130, launch power specifications of the communication channel 132 of interest, the number of communication channels 132 operating adjacent to the communication channel 132 of interest, and the use of an intervening amplification system). Here, the communication channel 132 of interest is the one through which the transmitting device 110 transmits symbols encoded using the symbol constellation 30.

[0030] The fiber specifications of the optical fiber 130 may include fiber coefficients α, β_2, and γ, the number of spans, and its length, where the fiber coefficients α, β_2, and γ represent the fiber loss coefficient, dispersion loss coefficient, and nonlinear coefficient, respectively. The frequency specifications of the optical fiber 130 may include the center frequency, channel spacing, and the number of communication channels 132 contained therein.

[0031] Here, if system specification elements that do not pertain to all of the multiple communication channels 132 can be changed, the symbol constellation generating device 2000 can generate different symbol constellations 30 for each communication channel 132. The transmitting device 110 can use different symbol constellations 30 for the communication channels 132.

[0032] In this case, the specification information 10 may include identification information of the communication channel 132 and specification elements related to the corresponding communication channel 132. The symbol constellation generating device 2000 identifies for which communication channel 132 the symbol constellation 30 is to be generated, based on the identification information of the communication channel 132 indicated in the specification information 10. An example of such a system specification element is the number of communication channels 132 that are active and adjacent to the target communication channel 132.

[0033] There are various methods for acquiring the specification information 10. For example, the acquiring unit 2020 may acquire the specification information 10 from a storage device that is accessible from the symbol constellation generation device 2000 and in which the specification information 10 is stored in advance. Alternatively, for example, the acquiring unit 2020 may receive the specification information 10 transmitted from another computer. Alternatively, for example, the acquiring unit 2020 may provide a user interface that enables the user of the symbol constellation generation device 2000 to manually input the specifications of the optical communication system 100, and acquire the input data as the specification information 10.

[0034] <Generation of noise feature information: S104> The acquisition unit 2020 generates noise characteristic information 20 based on the specification information 10 (S104). The noise characteristic information 20 represents characteristics of noise generated in the optical communication system 100. For example, the noise characteristic of the optical communication system 100 may include amplified spontaneous emission (ASE) noise generated in optical devices. In addition, for example, the noise characteristic of the optical communication system 100 may include nonlinear interference (NLI) noise. The nonlinear interference noise provides a prediction of nonlinear distortion affecting signal transmission. This includes nonlinear effects of fiber such as self-phase modulation (SPM), cross-phase modulation (XPM), and four-wave mixing (FWM).

[0035] The noise characteristics of the optical communication system 100 can be represented using a noise model, such as Gaussian noise (GN) for a WDM channel. The GN noise model for WDM channels indicates that the nonlinearity corresponding to a dispersion-uncompensated signal follows a Gaussian distribution with a variance obtained from the system specifications. Specifically, the GN noise model for a WDM channel can be defined as follows:

number

[0036] 1st generation part 2040 The first generator estimates parameters of a noise model that represents noise characteristics of the optical communication system 100, and generates noise characteristic information 20 including the estimated parameters of the noise model. When the GN noise model defined by Equation (1) is used, 2040 The first generator estimates the variance n based on the specifications of the optical communication system 100. Here, by changing the data, any existing or future proposed model (that models optical communication) other than the GN model can be generated by the first generator. 2040 It can be used to implement

[0037] 1st generation part 2040 uses a first guess model 50 to estimate the parameters of the noise model. There are various ways to implement such a first guess model 50. For example, the first guess model 50 may be implemented as a predetermined mathematical function. The mathematical function used depends on the type of noise to be calculated. When the ASE noise in the optical communication system 100 is modeled by a GN model to represent the noise characteristics of the optical communication system 100, the variance n of the GN model may be calculated based on the following equation (2):

number

[0038] When the NLI noise in the optical communication system 100 is modeled by a GN model to represent the noise characteristics of the optical communication system 100, the variance n of the GN model can be calculated based on the following equations (3) and (4).

number

number

[0039] Alternatively, for example, the first guess model 50 may be implemented as a machine learning-based model that calculates parameters of a noise model. An example of a machine learning-based model that may be used as the first guess model 50 is a neural network. In this neural network, the input layer receives, as input data, elements of the specifications of the optical communication system 100 indicated in the specification information 10. The hidden layer analyzes the input data. The output layer outputs one or more parameters of the noise model (e.g., the variance n of the GN model) that represent the noise characteristics of the optical communication system 100 having the specifications indicated by the specification information 10.

[0040] The first guess model 50 is pre-trained by repeatedly updating its trainable parameters using multiple training data. The training data may include a combination of input data and ground truth data. Specifically, the input data may include a set of specification elements of the optical communication system 100. Meanwhile, the ground truth data may include noise model parameters that should be output from the first guess model 50 in response to input of corresponding input data.

[0041] To train the first guess model 50, ground truth data of noise features should be prepared for various types of specifications of the optical communication system 100. For example, the ground truth data may be generated based on experimental results. Specifically, the optical communication system 100 is operated under various specifications, and the noise features generated in the optical communication system 100 are measured or calculated for each specification. As a result, pairs of specifications of the optical communication system 100 and noise features are obtained for each of the various specifications. Alternatively, for example, the ground truth data may be generated using the mathematical model described above.

[0042] The noise characteristics do not have to be represented by one or more parameters of a noise model (such as the variance of a GN model). For example, the noise characteristics of the optical communication system 100 can be represented by a nonlinearity distribution (hereinafter, nonlinearity distribution). The nonlinearity distribution represents a nonlinear relationship between a transmitted signal transmitted by the transmitting device 110 and a received signal received by the receiving device 120. In this case, the first guess model 50 can be trained using experimental data to learn the relationship between the nonlinearity distribution and the specifications of the optical communication system 100.

[0043] Specifically, experimental data is generated by varying the specifications to several possible specification states and generating enough received signals corresponding to these states to calculate a nonlinear distribution. The number of specification states used must be sufficient in terms of the amount of data needed to accurately model the first guess model 50. The specification states include possible specification variations (such as variations in the number of transmit channels, variations in transmit power, and other possible variations). For each state, a predetermined number of transmit signals are used to generate received signals. These transmit and receive signal pairs are then used to generate a nonlinear distribution corresponding to the specification states. This generated data is then used to train the first guess model 50.

[0044] <Generation of symbol constellation 30: S106> The second generator 2060 generates a symbol constellation 30 based on the noise feature information 20 (S106). The symbol constellation 30 is a symbol constellation to be used by the transmitter 110 to send a message to the receiver 120. N The symbol constellation 30 represents, for each possible bit pattern in the message, the correspondence between the bit pattern and the symbol (i.e., the coordinate on the symbol constellation map) to be used for the corresponding bit pattern. The transmitter 110 converts each bit pattern in the message into one of the symbols represented by the symbol constellation 30.

[0045] 2nd generation part 2060 generates the symbol constellation 30 using the second guess model 60. The second guess model 60 may be implemented as a machine learning-based model. The model is pre-trained to generate a symbol constellation 30 suitable for the optical communication system 100, whose noise characteristics are represented by the input noise feature information 20. 2060 is the first generation part 2040 The noise feature information 20 generated by the above is input to a second estimation model 60 to generate a symbol constellation 30, and the symbol constellation 30 output from the second estimation model 60 is obtained.

[0046] An example of a machine learning-based model used in the second guess model 60 is a neural network. In this neural network, the input layer receives noise feature information 20 as input data. The hidden layer analyzes the input data (e.g., parameters of a noise model indicated in the noise feature information 20). The output layer outputs a symbol constellation 30 suitable for the optical communication system 100, which has noise features represented by the input noise feature information 20.

[0047] The second guess model 60 is pre-trained by iteratively updating its trainable parameters using multiple training data. The training data may include a combination of input data and ground truth data. The input data includes data representing noise characteristics of the optical communication system 100, such as noise model parameters for one or more types of noise (e.g., the variance n of a GN model) and a nonlinearity distribution indicating a nonlinear relationship between a transmitted signal and a received signal. The ground truth data includes a symbol constellation 30 that should be output from the second guess model 60 in response to corresponding input data being input to the second guess model 60.

[0048] To train the second guess model 60, ground truth data of the symbol constellation 30 should be prepared for various noise characteristics of the optical communication system 100. For example, the ground truth data can be generated using a geometric shaping algorithm. An example of an implementation of the geometric shaping algorithm is an end-to-end training system including an autoencoder. By executing the training system for various specifications of the optical communication system, multiple sets of noise characteristics and symbol constellations suitable for the optical communication system 100 with the corresponding noise characteristics can be obtained. Each of the obtained sets can be used as training data. Here, specific methods for using an end-to-end training system to obtain symbol constellations suitable for an optical communication system are disclosed in, for example, Patent Document 1 and Non-Patent Document 1, and are known techniques.

[0049] <Output of 30 symbol constellations> The symbol constellation generation apparatus 2000 may output the symbol constellation 30 in a manner that allows the transmitting apparatus 110 to use the symbol constellation 30 to encode a message. For example, the symbol constellation generation apparatus 2000 may store the symbol constellation 30 in a storage device that is also accessible by the transmitting apparatus 110 (in other words, a storage device shared by the symbol constellation generation apparatus 2000 and the transmitting apparatus 110). Alternatively, for example, the symbol constellation generation apparatus 2000 may transmit the symbol constellation 30 to the transmitting apparatus 110.

[0050] <Example of Operation of Optical Communication System 100> For reference, an example of the operation of the optical communication system 100 will be described with reference to Fig. 5. Fig. 5 shows a flowchart illustrating an example of the overall flow of management of the optical communication system 100. During operation of the optical communication system 100, its specifications may change multiple times. Therefore, in this example, the symbol constellation 30 to be used by the transmitting device 110 is repeatedly updated in response to changes in the specifications of the optical communication system 100. In this way, the optical communication system 100 can be operated efficiently even if the specifications of the optical communication system 100 change during operation.

[0051] Before starting operation of the optical communication system 100, the symbol constellation 30 is initialized based on the initial specifications of the optical communication system 100 (S202). Hereinafter, this stage will be referred to as the "offline stage." In the offline stage, the symbol constellation 30 may be generated by the symbol constellation generating device 2000 or by another device. In the latter case, the symbol constellation 30 may be generated by an end-to-end learning system including an autoencoder, such as those disclosed in Patent Document 1 and Non-Patent Document 1. Using It has been mentioned here that generating a symbol constellation using an end-to-end training system is a time-consuming task. but, It is acceptable to use this system in an offline phase.

[0052] After the optical communication system 100 starts operation, the symbol constellation generating device 2000 is used to update the symbol constellation 30. The phase during which the optical communication system 100 is in operation is called the "online phase." In order to update the symbol constellation 30 in response to changes in the specifications of the optical communication system 100, the optical communication system 100 is monitored during the online phase, and changes in the specifications of the optical communication system 100 are detected. For example, a network management monitoring system may monitor the optical communication system 100 using sensors and online performance to detect or recommend changes in the system specifications of the optical communication system 100.

[0053] The operation of the optical communication system 100 is depicted as loop process A (S204 to S212). Loop process A is executed until a predetermined termination condition, such as "a predetermined time has elapsed," is met. In step S206, it is determined whether the specifications of the optical communication system 100 have changed. If it is determined that the specifications of the optical communication system 100 have changed (S206: YES), the symbol constellation generation apparatus 2000 generates a new symbol constellation 30 (S208) by executing, for example, steps S102 to S108 shown in FIG. 4. Then, the optical communication system 100 is updated to use the symbol constellation 30 newly generated in step S208 (S210).

[0054] Although the present disclosure has been described with reference to the embodiments, the present disclosure is not limited to the above-described embodiments, and various modifications that are understandable to those skilled in the art can be made to the configuration and details of the present disclosure without departing from the scope of the invention.

[0055] A part or all of the above-described embodiments can be described as, but not limited to, the following supplementary notes. <Additional Notes> (Appendix 1) at least one processor; at least one memory storing instructions; The at least one processor executes the instructions to: obtaining specification information representing one or more elements of a specification of an optical communication system including an optical transmitter, an optical receiver, and an optical fiber therebetween; generating noise feature information using the acquired specification information and a first guess model, the noise feature information representing one or more noise features of the optical communication system, the first guess model being predefined and representing a correspondence between the specifications of the optical communication system and the noise features of the optical communication system having the corresponding specifications; and generating a symbol constellation using the generated noise feature information and a second guess model, the symbol constellation being used by the optical transmitting device to encode a message to be transmitted to the optical receiving device, the second guess model being predefined and representing a correspondence between the noise feature of the optical communication system and the symbol constellation suitable for the optical communication system having the corresponding noise feature. (Appendix 2) 2. The symbol constellation generating apparatus of claim 1, wherein the first guess model is a machine learning-based model that receives as input data the one or more elements of the specifications of the optical communication system and is pre-trained to output the noise characteristics of the optical communication system having specifications represented by the input data. (Appendix 3) 3. The symbol constellation generation device according to claim 1, wherein the second estimation model is a machine learning-based model that acquires the noise characteristics of the optical communication system as input data and is pre-trained to output the symbol constellation suitable for the optical communication system having the noise characteristics represented by the input data. (Appendix 4) 4. The symbol constellation generating apparatus of claim 1, wherein the at least one processor is further configured to calculate one or more parameters of a noise model as the noise characteristic of the optical communication system. (Appendix 5) 5. The symbol constellation generating apparatus according to claim 4, wherein the noise model represents amplified spontaneous emission noise or nonlinear interference noise. (Appendix 6) The at least one processor: repeatedly acquiring the specification information while the optical communication system is operating; generating the symbol constellation in response to the specification information being acquired; 6. The symbol constellation generation apparatus of claim 1, configured to provide the generated symbol constellation to an optical communication system so that the optical communication system is updated to use the generated symbol constellation. (Appendix 7) obtaining specification information representing one or more elements of a specification of an optical communication system including an optical transmitter, an optical receiver, and an optical fiber therebetween; generating noise feature information using the acquired specification information and a first guess model, the noise feature information representing one or more noise features of the optical communication system, the first guess model being predefined and representing a correspondence between the specifications of the optical communication system and the noise features of the optical communication system having the corresponding specifications; generating a symbol constellation using the generated noise feature information and a second guess model, the symbol constellation being used by the optical transmitting device to encode a message to be transmitted to the optical receiving device, the second guess model being predefined and representing a correspondence between the noise feature of the optical communication system and the symbol constellation suitable for the optical communication system having the corresponding noise feature. (Appendix 8) The control method described in Appendix 7, wherein the first guess model is a machine learning-based model that takes as input data the one or more elements of the specifications of the optical communication system and is pre-trained to output the noise characteristics of the optical communication system having specifications represented by the input data. (Appendix 9) the second estimation model is a machine learning-based model that acquires the noise characteristics of the optical communication system as input data; 9. The control method of claim 7 or 8, which is pre-trained to output the symbol constellation suitable for the optical communication system having noise characteristics represented by the input data. (Appendix 10) 10. The control method of any one of appendixes 7 to 9, further comprising calculating one or more parameters of a noise model as the noise characteristics of the optical communication system. (Appendix 11) 11. The control method of claim 10, wherein the noise model represents amplified spontaneous emission noise or nonlinear interference noise. (Appendix 12) repeatedly acquiring the specification information while the optical communication system is operating; generating the symbol constellation in response to the specification information being acquired; 12. The control method of claim 7, further comprising providing the generated symbol constellation to the optical communication system so that the optical communication system is updated to use the generated symbol constellation. (Appendix 13) obtaining specification information representing one or more elements of a specification of an optical communication system including an optical transmitter, an optical receiver, and an optical fiber therebetween; generating noise feature information using the acquired specification information and a first guess model, the noise feature information representing one or more noise features of the optical communication system, the first guess model being predefined and representing a correspondence between the specifications of the optical communication system and the noise features of the optical communication system having the corresponding specifications; a second guess model that is predefined and represents a correspondence between the noise characteristics of the optical communication system and the symbol constellation suitable for the optical communication system having the corresponding noise characteristics; (Appendix 14) The medium of claim 13, wherein the first guess model is a machine learning-based model that takes as input data the one or more elements of the specifications of the optical communication system and is pre-trained to output the noise characteristics of the optical communication system having specifications represented by the input data. (Appendix 15) The medium described in Appendix 13 or 14, wherein the second estimation model is a machine learning-based model that acquires the noise characteristics of the optical communication system as input data and is pre-trained to output the symbol constellation suitable for the optical communication system having the noise characteristics represented by the input data. (Appendix 16) 16. The medium of any one of appendixes 13 to 15, wherein the program further causes the computer to calculate one or more parameters of a noise model as the noise characteristics of the optical communication system. (Appendix 17) 17. The medium of claim 16, wherein the noise model represents amplified spontaneous emission noise or nonlinear interference noise. (Appendix 18) The program further comprises: repeatedly acquiring the specification information while the optical communication system is operating; generating the symbol constellation in response to the specification information being acquired; 18. The medium of any one of Clauses 13 to 17, causing the computer to: provide the generated symbol constellation to the optical communication system, so that the optical communication system is updated to use the generated symbol constellation. [Explanation of symbols]

[0056] 10. Specifications 20 Noise feature information 30 symbol constellation 50 First guess model 60 Second guess model 1000 computers 1020 Bus 1040 processor 1060 memory 1080 storage device 1100 Input / Output Interface 1120 Network Interface 2000 Symbol Constellation Generator 2020 Acquisition Department 2040 1st generation part 2060 Second generation part

Claims

1. an acquisition unit that acquires specification information that represents a plurality of elements of the specifications of an optical communication system that includes an optical transmitting device, an optical receiving device, and an optical fiber therebetween; a first generation unit that generates noise feature information representing noise features of the optical communication system using the acquired specification information and a first estimation model; a second generator configured to generate a symbol constellation using the generated noise feature information and a second estimation model; the first estimation model is configured such that a correspondence relationship between each specification of the optical communication system and a noise feature of the optical communication system corresponding to each specification of the optical communication system is defined in advance; the symbol constellation is used by the optical transmitting device to encode a message to be transmitted to the optical receiving device; the second guess model is predefined and represents a correspondence between the noise characteristics of the optical communication system and the symbol constellations suitable for the optical communication system having the corresponding noise characteristics; the first generation unit calculates parameters of a noise model that represents amplified spontaneous emission noise as the noise characteristics of a WDM (Wavelength Division Multiplexing) channel of the optical communication system, which is the specification of the optical communication system; The Gaussian noise model for the WDM channel is given by: [Equation 1] X represents the noise value and follows a normal distribution with mean 0 and variance n, The variance n of the Gaussian noise model representing the amplified spontaneous emission optical noise ASE is expressed by the following formula: [Equation 2] NF represents the noise figure of the optical amplifier, h represents Planck's constant, v represents the channel carrier optical frequency, R represents the symbol rate, and A k represents the transmission loss in one span from the k-1 segment to the k segment, Symbol constellation generator.

2. an acquisition unit that acquires specification information that represents a plurality of elements of the specifications of an optical communication system that includes an optical transmitting device, an optical receiving device, and an optical fiber therebetween; a first generation unit that generates noise feature information representing a noise feature of the optical communication system using the acquired specification information and a first estimation model, a second generator configured to generate a symbol constellation using the generated noise feature information and a second estimation model; the first estimation model is predefined and represents a correspondence between the specifications of the optical communication system and noise characteristics of the optical communication system having the corresponding specifications; the symbol constellation is used by the optical transmitting device to encode a message to be transmitted to the optical receiving device; the second guess model is predefined and represents a correspondence between the noise characteristics of the optical communication system and the symbol constellations suitable for the optical communication system having the corresponding noise characteristics; the first generator calculates parameters of a noise model representing nonlinear interference noise as the noise feature of a WDM (Wavelength Division Multiplexing) channel of the optical communication system; The Gaussian noise model for the WDM channel is given by: [Equation 3] X represents the noise value and follows a normal distribution with mean 0 and variance n, The variance of the Gaussian noise model representing the nonlinear interference noise is expressed as: [Equation 4] [Equation 5] In equation (4), identifiers i and j represent the WDM channels, and the state of the jth communication channel affects the ith communication channel, and X i,j represents the nonlinear interference of the efficiency factor, and α, β 2 , and γ are the fiber coefficients of the optical fiber, L represents the length of the optical fiber, and N s represents the number of spans of the optical fiber, f and Δf represent the center frequency and channel spacing of the communication channel, respectively, and φ represents the time domain of X i,j is α, β 2 , γ, f, and Δf. In formula (5), n NLIi represents the variance n of the Gaussian noise model representing the nonlinear interference noise of the i-th communication channel, and P i represents the transmission power of the i-th communication channel, Symbol constellation generator.

3. 3. The symbol constellation generation device according to claim 1, wherein the second estimation model is a machine learning-based model that acquires the noise characteristics of the optical communication system as input data and is pre-trained to output the symbol constellation suitable for the optical communication system having the noise characteristics represented by the input data.

4. repeatedly acquiring the specification information while the optical communication system is operating; generating the symbol constellation in response to the specification information being acquired; 4. The symbol constellation generating device according to claim 1, further comprising: a first optical communication system configured to receive the generated symbol constellation from the first optical communication system; a second optical communication system configured to receive the generated symbol constellation from the first optical communication system;

5. obtaining specification information representing a plurality of elements of the specification of an optical communication system including an optical transmitter, an optical receiver, and an optical fiber therebetween; generating noise feature information representing a noise feature of the optical communication system using the acquired specification information and a first estimation model; the first estimation model is predefined and represents a correspondence between the specifications of the optical communication system and noise characteristics of the optical communication system having the corresponding specifications; generating a symbol constellation using the generated noise feature information and a second guess model; the symbol constellation is used by the optical transmitting device to encode a message to be transmitted to the optical receiving device; the second guess model is predefined and represents a correspondence between the noise characteristics of the optical communication system and the symbol constellations suitable for the optical communication system having the corresponding noise characteristics; The step of generating noise characteristic information includes calculating parameters of a noise model representing amplified spontaneous emission noise as the noise characteristic of a WDM (Wavelength Division Multiplexing) channel of the optical communication system; The Gaussian noise model for the WDM channel is given by: [Equation 6] X represents the noise value and follows a normal distribution with mean 0 and variance n, The variance n of the Gaussian noise model representing the amplified spontaneous emission optical noise ASE is expressed by the following formula: [Equation 7] NF represents the noise figure of the optical amplifier, h represents Planck's constant, v represents the channel carrier optical frequency, R represents the symbol rate, and A k represents the transmission loss in one span from the k-1 segment to the k segment, A computer-implemented control method.

6. obtaining specification information representing a plurality of elements of the specification of an optical communication system including an optical transmitter, an optical receiver, and an optical fiber therebetween; generating noise feature information representing noise features of the optical communication system using the acquired specification information and a first estimation model; the first estimation model is predefined and represents a correspondence between the specifications of the optical communication system and noise characteristics of the optical communication system having the corresponding specifications; causing the computer to perform a step of generating a symbol constellation using the generated noise feature information and a second guess model; the symbol constellation is used by the optical transmitting device to encode a message to be transmitted to the optical receiving device; the second guess model is predefined and represents a correspondence between the noise characteristics of the optical communication system and the symbol constellations suitable for the optical communication system having the corresponding noise characteristics; The step of generating noise characteristic information includes calculating parameters of a noise model representing amplified spontaneous emission noise as the noise characteristic of a WDM (Wavelength Division Multiplexing) channel of the optical communication system; The Gaussian noise model for the WDM channel is given by: [Equation 8] X represents the noise value and follows a normal distribution with mean 0 and variance n, The variance n of the Gaussian noise model representing the amplified spontaneous emission optical noise ASE is expressed by the following formula: [Equation 9] NF represents the noise figure of the optical amplifier, h represents Planck's constant, v represents the channel carrier optical frequency, R represents the symbol rate, and A k represents the transmission loss in one span from the k-1 segment to the k segment, program.

7. obtaining specification information representing a plurality of elements of the specification of an optical communication system including an optical transmitter, an optical receiver, and an optical fiber therebetween; generating noise feature information representing a noise feature of the optical communication system using the acquired specification information and a first estimation model; the first estimation model is predefined and represents a correspondence between the specifications of the optical communication system and noise characteristics of the optical communication system having the corresponding specifications; generating a symbol constellation using the generated noise feature information and a second guess model; the symbol constellation is used by the optical transmitting device to encode a message to be transmitted to the optical receiving device; the second guess model is predefined and represents a correspondence between the noise characteristics of the optical communication system and the symbol constellations suitable for the optical communication system having the corresponding noise characteristics; The step of generating noise characteristic information includes calculating parameters of a noise model representing nonlinear interference noise as the noise characteristic of a WDM (Wavelength Division Multiplexing) channel of the optical communication system; The Gaussian noise model for the WDM channel is given by: [Equation 10] X represents the noise value and follows a normal distribution with mean 0 and variance n, The variance of the Gaussian noise model representing the nonlinear interference noise is expressed as: [0011] [0012] In equation (11), identifiers i and j represent the WDM channels, and the i-th communication channel is affected by the state of the j-th communication channel, and X i,j represents the nonlinear interference of the efficiency factor, and α, β 2 , and γ are the fiber coefficients of the optical fiber, L represents the length of the optical fiber, and N s represents the number of spans of the optical fiber, f and Δf represent the center frequency and channel spacing of the communication channel, respectively, and φ represents the time domain of X i,j is α, β 2 , γ, f, and Δf. In formula (12), n NLIi represents the variance n of the Gaussian noise model representing the nonlinear interference noise of the i-th communication channel, and P i represents the transmission power of the i-th communication channel, A computer-implemented control method.

8. obtaining specification information representing a plurality of elements of the specification of an optical communication system including an optical transmitter, an optical receiver, and an optical fiber therebetween; generating noise feature information representing noise features of the optical communication system using the acquired specification information and a first estimation model; the first estimation model is predefined and represents a correspondence between the specifications of the optical communication system and noise characteristics of the optical communication system having the corresponding specifications; generating a symbol constellation using the generated noise feature information and a second guess model; the symbol constellation is used by the optical transmitting device to encode a message to be transmitted to the optical receiving device; the second guess model is predefined and represents a correspondence between the noise characteristics of the optical communication system and the symbol constellations suitable for the optical communication system having the corresponding noise characteristics; The step of generating noise characteristic information includes calculating parameters of a noise model representing nonlinear interference noise as the noise characteristic of a WDM (Wavelength Division Multiplexing) channel of the optical communication system; The Gaussian noise model for the WDM channel is given by: [0013] X represents the noise value and follows a normal distribution with mean 0 and variance n, The variance of the Gaussian noise model representing the nonlinear interference noise is expressed as: [0014] [Equation 15] In equation (14), identifiers i and j represent the WDM channels, and the i-th communication channel is affected by the state of the j-th communication channel, and X i,j represents the nonlinear interference of the efficiency factor, and α, β 2 , and γ are the fiber coefficients of the optical fiber, L represents the length of the optical fiber, and N s represents the number of spans of the optical fiber, f and Δf represent the center frequency and channel spacing of the communication channel, respectively, and φ represents the time domain of X i,j is α, β 2 , γ, f, and Δf. In formula (15), n NLIi represents the variance n of the Gaussian noise model representing the nonlinear interference noise of the i-th communication channel, and P i represents the transmission power of the i-th communication channel, program.

Images