580 California St., Suite 400
San Francisco, CA, 94104
Key research themes
1. How do wavelength division multiplexing (WDM) techniques and related hardware advances improve capacity and flexibility in optical networks?
This theme explores the evolution, deployment, and optimization of Wavelength Division Multiplexing (WDM), including Dense WDM (DWDM), as a method to increase data transmission capacity over optical fibers. It covers the hardware enablers such as fiber types, optical amplifiers, transceivers, and wavelength converters, the challenges in wavelength assignment, and emerging architectures that leverage WDM. The theme is crucial as WDM forms the backbone of expanding bandwidth capabilities, enabling scaling from gigabits to terabits per second, and is fundamental for meeting growing network traffic demands.
Key finding: This comprehensive roadmap article systematically presents current state-of-the-art and future prospects in optical communications, highlighting that advances in optical fibers, amplification, spatial division multiplexing,... Read more
Key finding: The paper outlines the maturation of WDM network studies over a decade, emphasizing WDM’s role in exploiting fiber's vast bandwidth potential (up to nearly 50 Tbps per fiber) to handle the surge in data traffic. It clarifies... Read more
Key finding: This work introduces selective transparency—combining wavelength conversion selectively within the optical network as a trade-off between opaque and fully transparent designs. It demonstrates that deploying wavelength... Read more
Key finding: This study rigorously analyzes wavelength allocation and routing in WDM ring networks without assuming traffic arrival process knowledge. It proves that shortest-path routing achieves at most twice the optimal maximum link... Read more
Key finding: This paper proposes and experimentally validates a novel point-to-multipoint (P2MP) optical network architecture utilizing digital subcarrier multiplexing (DSCM) to share a single high-capacity coherent transceiver among... Read more
2. What emerging network architectures and control paradigms enable integrated, flexible, and automated optical networking for 5G and beyond?
This theme examines novel optical network architectures, infrastructure convergence with wireless domains, and advanced control and orchestration mechanisms (including software-defined networking, NFV, and management-enablement) critical for meeting 5G/6G transport requirements. It investigates how optical transport segments adapt to low latency, high bandwidth, and virtualization demands through centralized/distributed network control, flexible functional splits, and programmable photonic devices, enabling automated, converged, and intelligent optical networks.
Key finding: This paper identifies the evolution of optical access transport to meet 5G/6G requirements of low latency, high bandwidth, and enhanced automation. It details silicon photonics as a key enabler for integrated, small... Read more
Key finding: This tutorial reviews control plane (CP), management plane (MP), and orchestration architectures and protocols in optical networks crucial for dynamic on-demand provisioning, QoS/QoT assurance, multilayer/multidomain... Read more
Key finding: The authors propose converged wireless-optical 5G infrastructure that addresses fronthaul and backhaul constraints by flexible functional splits, combining centralized and cloud-based radio access network components with... Read more
Key finding: This work develops a genetic algorithm to optimize virtual network function (VNF) placement and service chain allocation in optical backhaul/fronthaul networks supporting NFV and MEC paradigms, minimizing service blocking and... Read more
Key finding: The paper presents a novel centralized optical line terminal (OLT) leveraging an optical multicarrier (OMC) generation technique integrating cascaded modulators without amplifiers, enabling high aggregate data rates (1.2 Tbps... Read more
3. How can machine learning and advanced optimization methods enhance operational efficiency, quality of transmission (QoT) assessment, and resource allocation in optical networks?
This theme focuses on the adoption of artificial intelligence (AI), supervised machine learning (ML), and optimization algorithms in optical network operations. It addresses enhancing QoT estimation accuracy and speed, optimizing routing, wavelength, and spectrum assignment under static and dynamic traffic, and orchestrating virtualized network functions leveraging evolutionary algorithms. Integration of AI techniques aims to make networks more adaptive, predictive, and resilient, contributing to capacity utilization and service reliability improvements.
Key finding: By evaluating multiple supervised ML models (including support vector machines, random forests, and bagging trees), this study achieves accurate (>99.9%) and fast classification of lightpath quality (QoT) in impairment-aware... Read more
Key finding: This paper critically evaluates offline network dimensioning approaches based on routing and spectrum assignment (RSA), comparing optimization objectives targeting path length minimization versus spectrum occupancy reduction.... Read more
Key finding: The study proposes algorithms for selecting pairs of sub-trees in multicast optical light-tree reconfiguration to minimize optical flow interruptions. By categorizing sub-tree pairs based on shared links and employing precise... Read more
Key finding: This survey extensively catalogs AI techniques applied within optical networks, covering optimization, machine learning, planning, control, and management domains. It contextualizes how search algorithms, metaheuristics,... Read more
Key finding: Focusing on space division multiplexing (SDM) via multi-core and few-mode fibers, this work develops allocation and protection schemes in elastic optical networks supporting dynamic traffic. It formulates optimal routing and... Read more