Browsing by Author "Lozada Vergara, Astrid Adriana"
Now showing 1 - 1 of 1
- Results Per Page
- Sort Options
Thesis Mathematical modeling of physical layer impairments in spatial division multiplexing (SDM) optical fiber networks based on few-mode fibers (FMF)(Universidad Técnica Federico Santa María, 2025-01) Lozada Vergara, Astrid Adriana; Universidad Técnica Federico Santa María. Departamento de Electrónica; Olivares Véliz, Ricardocapacity limitations of current wavelength division multiplexing (WDM) networks. Due to the increase in data traffic demand, the theoretical capacity limits of standard single-mode fibers are insufficient to meet the requirements of future networks, hence the importance of developing new technologies that increase this capacity. In the context of this thesis, SDM systems are studied addressing aspects from signal propagation and amplification to the resource allocation problem. This research specifically investigates the performance of SDM networks based on Few-Mode Fiber (FMF) with a focus on how physical layer impairments (PLI) impact transmission quality. SDM-FMF networks experience several linear and nonlinear PLI. These impairments were mathematically modeled under weak and intermediate coupling regimes to analyze their impact on the signal-to-noise ratio (SNR) along FMF links. To analyze optical amplification within SDM-FMF networks, two amplification strategies were studied: lumped amplification using Few-Mode Erbium-Doped Fiber Amplifiers (FM-EDFA) and distributed amplification using Few-Mode Distributed Raman Amplifiers (FM-DRA). Genetic algorithms were employed to optimize the pump configurations of these amplifiers, minimizing Differential Modal Gain (DMG) and Differential Spectral Gain (DSG) to improve system performance. A novel Joint DMG-DSG Minimization Methodology was proposed to optimize the performance of two-stage FM-EDFA, ensuring high mean gain and low DMG and minimal DSG. This methodology involves the design of pumping profile and gain flattening filter. Simulation results show that the two-stage FM-EDFA achieves enough gain to compensate for fiber losses in spans of 100 km. Additionally, optimization of FM-DRA was also performed. The optimization process aims to minimize gain ripple and DMG. Performance under three configurations in the pumping direction-backward, forward, and bi-directionalwas compared for signals in the conventional C-band. In terms of the on-off gain, the three configurations achieved a similar value. The main difference between these three configurations is regarding the noise. Finding that, bidirectional pumping provides the optimal balance between on-off gain and noise. To extend operation to the C + L optical bands, FM-DRA optimization was done. In this case, the number of pump wavelengths and the pumping power budget was increased compared to operation in the C-band alone. Resulting in higher gains and greater ASE noise power, at the expense of larger DMG and gain ripple. Within the context of solving the resource allocation problem, a novel Routing, Spatial Mode, and Spectrum Allocation (RSSA) algorithm, named BANG, was developed. BANG’s mode group selection strategy demonstrated a significant reduction in blocking probability compared to existing algorithms. The findings of this work—optimized amplification profiles using FM-EDFA and FM-DRA, and BANG algorithm—are promising solutions for implementation in nextgeneration SDM-FMF systems to address the limited capacity of existing wavelength division multiplexing networks. Finally, the thesis outlines future research directions, including hybrid amplification techniques and the application of elastic optical networks for improved resource assignment. Keywords. Spatial Division Multiplexing; Few-Mode Fiber; Few-Mode Erbium-Doped Fiber Amplifier; Few-Mode Distributed Raman Amplifier; Mode-Group Division Multiplexing; Routing, Spatial mode, and Spectrum Allocation.
