Workshops

tbd

With an annual increase of 40% – 60% in interface bandwidth, yet only a ~13% increase for single lambda transceivers, it is anticipated that future transceivers will utilize one or more types of spatial and/or wavelength parallelization. This workshop aims to have a reality check of where we are on this path.

The discussion includes:

• What level of parallelization do we need in the 5 to 10 years range?
• What needs to be developed (hardware, DSP, packaging, interface solutions) to achieve the desired level of parallelization?
• Innovative transceiver structures for massively parallel transceivers.

The discussions will be arranged into two parts addressing two (very different) applications:

• Ultra-short reach (intra-data center distances and shorter).

Long distances (from inter-data centers to long-haul systems.

Mainstream free-space optical (FSO) communication systems converged on using the 1.55µm band offering data rates up to Tbit/s, e.g. between satellites. Besides, recent years have shown increasing research activities on atmospheric FSO using other wavelength bands reaching tens of Gbit/s in laboratory and outdoor trials over distances up to hundreds of meters. This workshop will examine the best wavelength for FSO communication systems. It will compare near-infrared (NIR, 0.78-1.4 μm), short-wavelength IR (SWIR, 1.4-2 μm), mid-wavelength IR (MIR, 3-5 μm), and long-wavelength IR (LWIR, 8-12 μm). Talks will demonstrate the significant potential for FSO using NIR, MIR, and LWIR, especially in terms of performance, range, and cost under adverse weather conditions (atmospheric scattering, turbulence). Based on the increasing technological maturity of components and sub-systems, speakers will assess the readiness of NIR, MIR, and LWIR for practical FSO applications in commercial deployments. The goal is to advance the use of FSO in future communication infrastructures, including ground-to-ground and ground-to-satellite links.

Key Learning Points:

• How do different wavelengths influence the performance and reliability of FSO systems?
• What are the latest developments of NIR, MIR, and LWIR components?
• How to keep availability high under challenging weather conditions? How to integrate NIR, MIR, and LWIR into existing and future networks?

Target Audience:

This workshop is ideal for network engineers, optical communication professionals, system integrators, and researchers interested in the advancements and application of infrared technologies in free space optics. Participants will leave with a clearer understanding of how different infrared wavelengths can be optimized to improve FSO system performance, especially in less-than-ideal environmental conditions.

Space division multiplexing (SDM) has established itself as one of the most promising technologies to cope with the ever-growing traffic demand in optical networks after more than a decade of fundamental research. Motivated by the higher achievable throughput, SDM recently entered the commercial level in subsea cables, as we experienced the deployment of the first submarine cable based on multi-core fiber technology.

This workshop will explore the benefits, advancements, and challenges of SDM technology and its potential for broader application in future optical networks. The session will feature a dynamic discussion through the interaction between the speakers and the audience, as participants will have the opportunity to engage directly with the speakers, responding to their selected questions to help shape a community consensus about key development matters in SDM. The panel will include academic researchers as well as leading industry technology providers and adopters.

Digital Twin (DT) of optical networks has recently received significant attention from our community as a tool that can use historical data, real-world monitoring data, and expert knowledge to provide a highly reliable simulation environment for modeling, optimization, and predicting the future behavior of the network. DTs may appear in different forms from the DT of a single sub-system, the DT of an entire system, or the DT of a complex meshed network. The advancements in the monitoring capabilities of coherent receivers, versatile telemetry collection protocols for optical line systems and terminals, as well as the fast pace for the realization of self-healing networks, justify the current trends toward the development of a full-fledged DT of optical networks.

Driven by the recent breakthroughs in the development of Large Language Models (LLM) and the success of Generative AI (GenAI) models such as ChatGPT, Bard, and Llama 2, one could see numerous use-cases where GenAI could be valuable for telecom networks, from facilitating device configuration scripts with no limitation on vendor-specific syntaxes, to generating large amount of synthetic, yet realistic, data samples of real-world network failures. This workshop is going to trigger a discussion around the current status of DT development for optical networks and offer a medium to exchange opinions on how GenAI could help DTs become more successful. For instance, will creating a foundational model of various network assets – fiber plants, transceivers, ROAMDs, amplifiers, etc. – result in DTs that can model the real-world phenomenon closer to reality? Can foundation models relying on real-time monitoring data and operational DTs replace engineers to design and operate networks? What are the challenges ahead for curating a multi-purpose, clean dataset for creating such foundational models?

Part 1: Deploying C+L systems: a niche market for “fiber renting” operators, or a future standard in optical networks?

Part2: Extending systems bandwidth beyond C+L through S, E or O bands: cost efficiency of adding additional bands having worse performance?

C-band systems have been widely deployed for over 25 years. Once fiber capacity is saturated on one link, the option of lightning another (possibly wider) C-band system is the one selected by many operators owning their fiber. “Fiber renting“ operators (typical case of OTT) have deployed C+L systems in the past 5 years to reduce network bandwidth cost, even if a C+L system equipment may currently be more expensive than 2 C-band systems. However, these C+L systems may also be saturated given the ceaseless traffic growth.

The questions that will be addressed during the workshop are the following:

• What capacity growth expected over the next 10 years? x10, x5, x2?
• Will “fiber owing” operators keep C-band system and increase number of fibers used? Given WSS have a (physically) limited number of ports, which no longer increases, what is the risk of saturating WSS ports?
• What are the main challenges for deploying C+L systems: physical effects (SRS), cost/supply of L-band assets (lasers, WSS, OA)?
• Is C+L mostly for Long Haul networks? Or could that also be interesting for DCI or metro in some regions of the world?
• Will C+L system become the new standard, even for “fiber-owing” operators?
• Will operators renting fiber (mostly OTT) continue to deploy standard C+L (4.8THz+4.8THz) systems, or evolve toward wider bandwidth C+L (6+6THz) system?
• C+L deployment only started ~20 years after C-band deployment. Should we wait 15 more years to see S+C+L? or could S+C+L appears in the next 5 years?
• What could trigger the development of S+C+L systems (up to 18THz, so almost doubling bandwidth of current 4.8+4.8THz C+L system)?
• Technologies providing OA for additional bands is progressing and O/E/S/C/L/U bands research experiments have been reported. While worse performance is observed for additional bands, what could drive the adoption of such systems? Could “renting fiber” operators see interest despite potential higher cost per bit?

Could new generations of fibers (MCF, HCF) become a game changer?

tbd

Fibre Optical Sensing (FOS) and monitoring have been developed to oversee, predict and survey critical field parameters from the environment around us, acquiring, processing, and interpreting vast amounts of rapid changing data. Optical networks today represent one of the most expensive and pervasive distribution infrastructures. Is it possible to employ them as a mean for extensive optical sensing and monitoring? Which FOS applications will drive the development of new business opportunities or will be able to meet societal needs?

Invited speakers will detail the main applications for residential, industry 4.0 and mobile markets, providing solutions for temperature and humidity metering, intrusion detection, seismology, energy saving, road traffic monitoring, fault prediction, fire prevention, and more. In particular, the current and future FOS solutions will be analysed with regard to business development around carrier networks, debating where it is worth to use sensing technologies based on scattering (Brillouin, Rayleigh, Raman) or state of polarization processing.

• What are the main applications that will drive fibre-optic sensing to productization?
• How well suited are the different sensing technologies for the applications and is there a universal technology that addresses all applications?
• How will Artificial Intelligence and machine learning techniques support those applications and how will such amount of data be handled?
• Are there new technologies on the horizon that will revolutionize the fibre-optic industry?

The workshop will consist of operators, system providers and academics, who will provide their insights on the questions above in the form of an invited presentation. Afterwards, a panel will be held for discussion.

With the completion of 50G-PON (Higher Speed PON) standardisation in the ITU-T, the announcement of 50G-PON products and operator deployment plans, the ITU-T has initiated a study into the requirements and candidate technologies for the next generation of Very High Speed PON (VHSP) expected ~ 2030 timeframe.

In parallel, the development of 6G and the evolution of mobile towards smaller cells will position PON as a primary candidate technology for deep fibre connectivity. Any convergence of residential and wireless access services on the same PON architecture will bring new challenges due to requirements for very low and deterministic latency, higher reliability and high data rates for continuous learning, fuelled by edge-AI computing and by a future network of intelligent machines.

With the expectation from network operators on very high service capacity (e.g. 200G) in combination with demands for longer reach and higher loss budgets, it seems like we could be at a technology inflection point for PON. Such requirements could be met by coherent technology but questions remain concerning the economics in this highly cost sensitive network segment.

This workshop will review the different technologies in scope for VHSP and address the question as to whether IMDD is still a candidate technology for this beyond 50G generation of ITU-T PON or whether coherent technology is needed to meet the high capacity and loss budget expectations.

Network operators will be invited to give their view on the key requirements for VHSP and the potential applications. Experts from industry and academia will review the different technology options and potential challenges to meet the operator requirements.

A lively debate is expected as PON pushes the performance limits of conventional IMDD PON and challenges the costs of the alternatives. Such a debate will be very timely and a great chance to contribute research ideas that could impact the ITU-T standardisation effort.

The traffic demands of emerging AI systems appear to be imposing fundamentally new demands on optical networks. Many varied systems are being deployed while utilizing massive amounts of data; both for training models and distributing model parameters. In this workshop, we will discuss how AI workloads will affect the design of optical transmission systems, and the components and subsystems that comprise them.

• How will AI affect traffic growth in the core, metro and edge?
  • Will this lead to increasing decoupling of rates required by Telecom and ICPs?
• What is the impact of AI on coherent transceivers?
  • Will it drive coherent inside the datacenter?
  • Do we need more options for coherent pluggable transceivers?
• How will regulations impact network designs?
• Will AI drive widespread adoption of hollow core fiber?

What is the main advantage of HCF? TOF, thermal stability, dispersion etc.?