Simulation data of 400 Gbps digital coherent optical communication systems

The dataset is organized around a root directory named 400Gbps_data. From this main directory, the file structure is divided according to the complexity of the simulated optical communication system, resulting in two primary transmission scenarios: Single-channel scenario: accessed through the folder dados_1_canal. Multi-channel (WDM) scenario: accessed through the folder dados_4_canais. Inside each of these two folders, the data flow is further separated according to the impact of the transmission medium on the optical signal, generating two parallel subfolders: dados_ideais: contains the reference symbol sequences transmitted by the transmitter, free from any degradation. dados_distorcidos: stores the corresponding signals after propagation through the optical fiber link, including attenuation, dispersion, and nonlinear impairments. Within the dados_ideais subfolder (applicable to both the single-channel and four-channel scenarios), the researcher will find specific folders corresponding to each physical channel of the simulation, named according to the generic format dados_ref_canalW (where W identifies the channel number, ranging from 1 to 4). In the single-channel scenario (dados_1_canal), only the folder dados_ref_canal1 exists. In the multi-channel scenario (dados_4_canais), four folders coexist, sequentially numbered from dados_ref_canal1 to dados_ref_canal4. At the deepest level of this directory tree, inside each channel reference folder (dados_ref_canalW), there are two .csv files. These files separate the coherent optical signal according to its orthogonal polarization components in the fiber, following the standardized naming convention: dados_Z_canal_ref_X_pol_canalW.csv: reference transmitted data in X polarization. dados_Z_canal_ref_Y_pol_canalW.csv: reference transmitted data in Y polarization. (Note: the character Z assumes the value 1 or 4, reflecting the global scenario to which the channel belongs.) When exploring the dados_distorcidos subfolder within the single-channel scenario (dados_1_canal), the data flow is subdivided according to the simulated optical fiber link length. The researcher will find two transmission-distance folders named dados_Lkm, where the parameter L assumes the values 80 or 120 km. Inside each distance folder, the data are separated by physical channel using the subfolder dados_Lkm_canalW (where W varies from 1 to 3). This division is justified because, although the overall scenario represents a single active channel at a time, the tests were independently performed at each frequency position of the system grid while keeping only the corresponding transmitter active and all others turned off. At the final and deepest level of this structure, each channel folder contains the .csv files corresponding to the propagation-degraded signals. The files simultaneously discriminate the polarization and the optical launch power into the fiber (P), ranging from −6 dBm to 14 dBm in 1 dB steps. The file naming convention strictly follows the pattern: dados_Z_canal_canal_W_Lkm_X_pol_PdBm.csv: distorted signals in X polarization for the specified power P. dados_Z_canal_canal_W_Lkm_Y_pol_PdBm.csv: distorted signals in Y polarization for the specified power P. For the dados_distorcidos branch within the multi-channel scenario (dados_4_canais), the organization follows a logic analogous to that of the single-channel case, but with a larger set of physical variables due to the characteristics of WDM systems. The researcher will find five distance-based subfolders named dados_Lkm, where the parameter L maps transmission distances of 80, 90, 100, 110, and 120 km. Inside each distance folder, the files are divided according to the monitored physical channel within the multiplexed system using the subfolders dados_Lkm_canalW, where W assumes values from 1 to 4 for the 80, 100, and 120 km cases, while for the 90 and 110 km cases only channel 2 is available. In these scenarios, all four optical channels were simultaneously active and propagating through the fiber, enabling the analysis of both linear and nonlinear inter-channel distortions. At the final storage level, each channel folder provides the .csv files corresponding to the degraded signals. For this WDM case, the launch-power sweep (P) was extended from 0 dBm to 20 dBm in 1 dB steps. The files follow the standardized naming convention: dados_4_canais_canal_W_Lkm_X_pol_PdBm.csv: distorted signals in X polarization for the specified channel W, distance L, and power P. dados_4_canais_canal_W_Lkm_Y_pol_PdBm.csv: distorted signals in Y polarization for the specified channel W, distance L, and power P. Content and Internal File Format (.csv)All files are stored as plain-text comma-separated value (.csv) files encoded in UTF-8. Each file contains initial metadata lines beginning with the comment characters // or #. The internal data structure rigorously follows a fixed two-column format. The first column represents the I component, measured in arbitrary units (a.u.), corresponding to the in-phase component of each point in the complex optical signal constellation. The second column represents the Q component, also measured in arbitrary units (a.u.), corresponding to the quadrature component of each constellation point. Therefore, each line in the file after the header represents the coordinates of an optical symbol in the complex plane.