Content area
The growing demand for high-capacity, low-latency services has placed significant pressure on the design and operation of optical transport networks. Multi-layer optical network design—which coordinates the physical layer with higher-layer protocols—has emerged as a critical strategy to enhance resource efficiency, service flexibility, and fault resilience. Enabled by advancements in software-defined networking (SDN) and network softwarization, intelligent multi-layer architectures allow for adaptive, cross-layer control of routing, grooming, and protection mechanisms, ultimately reducing both capital and operational expenditures.
This dissertation investigates the intelligent design and simulation of multi-layer optical networks through the integration of SDN, machine learning, and high-fidelity physical-layer modeling. We first introduce a novel service mesh architecture that supports dynamic, cross-layer service provisioning, improving both Quality of Service (QoS) and resource utilization. Second, we enhance the SimEON optical network simulator by integrating Deep Reinforcement Learning (DRL) to address the Routing, Modulation, and Spectrum Assignment (RMSA) problem in Elastic Optical Networks (EONs). Our results show that DRL-based approaches significantly outperform traditional heuristics under dynamic traffic conditions. To bridge control-plane decisions with physical-layer constraints, we develop a unified simulation framework by integrating SIMON, an optical network simulator, with GNPy, the Optical Route Planning Library. This hybrid approach enables accurate modeling of impairments such as non-linearities and dispersion, while supporting SNR-constrained dynamic routing. Finally, in collaboration with IIT-Madras, we propose a novel node architecture that incorporates optical phase conjugators (OPCs) and regenerators. Our coordinated OPC-regenerator placement strategy mitigates physical-layer impairments, reduces Digital Signal Processing (DSP) complexity, and enhances energy efficiency in long-haul Wavelength Division Multiplexing (WDM) networks.
By combining multi-layer intelligence, machine learning, and realistic simulation, this work contributes a scalable, adaptive, and performance-aware foundation for the next generation of optical transport networks.