摘要
快速、准确的波形级信道建模对于理解光纤信道特性、研发数字信号处理算法,以及优化光网络配置至关重要。传统的波形级信道建模技术——分步傅里叶法(SSFM),其复杂度较高,并且随着带宽增加呈四次方增长趋势,这限制了它在宽带光传输系统中的应用。深度学习技术凭借强大的非线性拟合能力和并行计算优势,在光纤信道波形建模领域取得了突破,其精度与SSFM相当,但计算时间缩短了1~2个数量级。本文综述了近年来基于深度学习的光传输系统信道波形建模技术,根据不同方案的特性,从长距传输模式、深度学习模型架构,以及是否结合物理信息等3个维度进行分类,阐述了它们的原理、性能及应用场景。最后,我们还讨论了深度学习方案面临的挑战及发展前景。
Significance The exponential growth of data traffic has propelled optical networks towards wideband,high rate,and large capacities.Accurate and rapid optical fiber transmission simulation systems are essential for optimizing optical network configurations,developing advanced digital signal processing(DSP)algorithms,and performing end-to-end(E2E)global optimization.The optical fiber channel model plays a crucial role in simulation systems,as it describes the propagation process of optical signals within the optical fiber.The propagation of optical signals in optical fibers is governed by the nonlinear Schrödinger equation(NLSE).Except in some special cases,the NLSE lacks an analytical solution and must be solved through numerical simulations.The Gaussian noise model(GN model)and the enhanced Gaussian noise model(EGN model)provide precise and fast optical fiber channel modeling,primarily focused on power-level simulations.However,they cannot offer detailed waveform information,limiting their application in the design and optimization of DPS algorithms,especially for nonlinear compensation.The split-step Fourier method(SSFM)offers waveform-level channel modeling but requires extensive iterative calculations,with computational complexity scaling to the fourth power of bandwidth,limiting its applicability in wideband systems.Deep learning(DL)technologies,with their remarkable nonlinear fitting capabilities and efficient parallel computing,have demonstrated comparable accuracy to SSFM in optical fiber channel waveform modeling,while reducing computational time by one to two orders of magnitude.Progress This paper reviews recent advances in DL-based optical fiber channel waveform modeling techniques and categorize them from three perspectives:long-haul transmission modes,DL model structures,and incorporation of physical information(Fig.3).We also present the principles and performance metrics of these approaches(Table 1).In terms of long-haul transmission modes, DL schemes are classified into overall and distributed schemes. Overallschemes utilize a single DL model to represent the entire long-haul optical fiber channel, offering lower computationalcomplexity and simplified data collection. However, they face challenges in handling amplified spontaneous emission(ASE) noise and achieving effective convergence. In contrast, distributed schemes employ multiple cascaded DL models toachieve long-haul transmission, each representing a single fiber span. This approach reduces the complexity of the channeleffects the models must fit, improves accuracy, and simplifies the handling of ASE noise between models. By adjusting thenumber of models, distributed schemes allow for flexible distance generalization. Therefore, distributed schemesoutperform overall schemes in handling ASE noise, achieving distance generalization, and improving model accuracy,making them the preferred method for waveform modeling in multi-channel wavelength division multiplexing (WDM)systems.Regarding DL model structures, schemes can be divided into neural networks and neural operators. Neural networks,such as conditional generative adversarial network (CGAN), bi-directional long short-term memory (BiLSTM), and multiheadattention, demonstrate strong nonlinear fitting capabilities. Among them, BiLSTM and multi-head attention, astemporal neural networks, exhibit superior accuracy when modeling time-dependent optical fiber channel characteristicsdue to their recurrent structure and self-attention mechanisms. Neural operators, an emerging DL method, map betweeninfinite-dimensional function spaces rather than discrete vector spaces, offering stronger generalization abilities comparedto traditional neural networks.In terms of incorporation of physical information, DL schemes are categorized pure data-driven and physics-datahybrid-driven methods. Pure data-driven methods require no complex domain-specific knowledge and simpler dataprocessing and training processes. However, they demand larger datasets and longer training times, and may struggle tomaintain high accuracy in multi-channel, high-rate WDM systems. Physics-data hybrid-driven methods combine physicalinformation with data-driven approaches. There are two main strategies to incorporate physical knowledge. First, thephysical constraint of the optical fiber channel, described by the NLSE, can be incorporated into the loss function,enhancing model interpretability and reducing the need for extensive datasets. Second, hybrid models combining physicalmodels and DL models can jointly perform channel modeling, leveraging the interpretability of physical models and thenonlinear fitting capabilities of DL models for improved results. The physics-data hybrid-driven approach shows significantpotential for scaling to multi-channel, high-rate WDM systems.Conclusions and Prospects Over several years of development, DL-based optical fiber channel waveform modeling hasemerged as a powerful technology, offering high accuracy and low complexity. It addresses the limitations of the traditionalSSFM, which is plagued by high computational complexity, and becomes a key technology for future optical fiber channelwaveform modeling. This paper first reviews recent advances in DL-based channel waveform modeling techniques,detailing their principles and performance metrics. We also explore the challenges of applying DL schemes to multichannel,high-rate systems from the perspective of the more complex linear and nonlinear effects, as well as generalizationof various system parameters. Additionally, we discuss potential optimization strategies from the perspective ofincorporating more physical prior information, optimizing the structure of DL models, and improving the generalizationcapability of DL models. With ongoing technological advancements, we believe the challenges faced by DL approacheswill be progressively overcome, ultimately positioning them as the dominant technology for channel waveform modeling innext-generation optical network.
作者
史明辉
牛泽坤
杨航
靳开颜
周欣怡
孙中原
张兆研
胡卫生
义理林
Shi Minghui;Niu Zekun;Yang Hang;Jin Kaiyan;Zhou Xinyi;Sun Zhongyuan;Zhang Zhaoyan;Yi Lilin(State Key Laboratory of Photonics and Communications,School of Electronic Information and Electrical Engineering,Shanghai Jiao Tong University,Shanghai 200240,China)
出处
《光学学报》
北大核心
2025年第14期281-299,共19页
Acta Optica Sinica
基金
国家重点研发计划(2023YFB2905400)
国家自然科学基金(62025503)
上海交通大学2030计划。
关键词
光纤与光通信
深度学习
信道建模
波形级仿真
optical fiber and optical communication
deep learning
channel modeling
waveform-level simulation
作者简介
通信作者:义理林,lilinyi@sjtu.edu.cn;通信作者:牛泽坤,zekunniu@sjtu.edu.cn。
