28 SEPTEMBER – 2 OCTOBER 2025

ECOC 2025

Workshops

Workshop 1: What type of optical fibre will be deployed, When and Where?

Sunday 28 September from 9.00-12.30 in Room 1 (break included)

SC 1: Novel fibres, fibre devices and amplifiers.

Short description: A range of transmission technologies and optical fibres have been extensively studied to sustainably meet the growing demand for data capacity. This workshop will explore current and future applications for various optical fibres, including conventional single-mode fibres (such as bundle and reduced diameter types), SDM optical fibres, and hollow-core fibres.

Organizers: Alan McCurdy, Takeshi Hoshida, Pascal Pecci

Data hungry technologies and applications, such as artificial intelligence, machine learning, virtual reality, autonomous driving, optical network sensing, telemedicine and so on, strongly increase the capacity demand. A variety of transmission technologies and corresponding optical fibres have been investigated intensively over the past few decades to support the increasing capacity demand in a sustainable way. For example, wideband and space division multiplexing (SDM) transmission have been major technical directions for increasing the total capacity by considering new degrees of freedom. Fibres which can support wideband and/or SDM transmission include not only one or multiple existing single-mode fibres, but also various SDM optical fibres, such as reduced geometrical diameter, multicore and multimode fibres. Moreover, hollow core fibres may provide the lowest attenuation and minimum latency properties for future large capacity transmission systems. However, there needs to be more discussion on the application areas of these various optical fibres as well as the related system requirements. It may also be difficult to select one unique optical fibre to support all future optical communication systems. This workshop will discuss current and anticipated applications for various optical fibres such as conventional single-mode fibres (including the bundle and reduced diameter types), SDM optical fibres, and hollow core fibres.

Workshop 2: AI-Driven Innovations in Photonic Device Design, Fabrication and Testing.

Sunday 28 September from 9.00-12.30 in Room 2 (break included)

SC 2: Discrete photonic devices and technologies & SC 3: Photonic integrated circuits, assemblies and packaging.

Short description: AI is revolutionizing photonic devices and integrated systems, driving breakthroughs in design, fabrication, data analysis, and integration. This workshop brings together top experts from academia and industry to explore cutting-edge AI applications in photonics, featuring AI tool demos, real-world case studies, and discussions on the future of AI-driven innovation in product design and realization.

Organizers: Selina Farwell, Francesco Da Ros, Stylianos Sygletos, Eric Bernier

Joint workshop between SC2 and SC3.

The integration of artificial intelligence with photonic devices and integrated systems represents a transformative leap in the field, offering unprecedented opportunities for innovation in design, fabrication, packaging, data analysis, and testing. By leveraging AI, researchers and engineers can optimize device performance, streamline manufacturing processes, enhance photonic integrated circuit (PIC) assembly, and unlock new functionalities for both discrete components and large-scale integrated solutions. This workshop aims to explore the cutting-edge intersections of AI and photonics, bringing together thought leaders to discuss breakthroughs, share insights, and shape the future of smarter, more efficient photonic technologies and systems.

Workshop Content Suggestions:
AI in Photonic Device and Integrated Circuit Design:Machine learning techniques for optimizing photonic device structures (e.g., waveguides, modulators, lasers) and photonic integrated circuits (PICs).AI-based generative models for innovative PIC architectures tailored for specific optical communication needs.Co-design strategies for discrete photonic devices and PICs using AI to improve performance and energy efficiency.AI-Enhanced Fabrication and Processing:Real-time AI-driven process control for high-precision fabrication of discrete devices and integrated circuits.AI-guided optimization of packaging processes for PICs, improving thermal management and optical alignment.Leveraging AI to reduce manufacturing defects and enhance yields for both discrete devices and integrated assemblies.AI in Testing and Screening:AI-powered automated testing for discrete photonic devices and PICs, accelerating characterization workflows.Predictive algorithms for screening and diagnosing device and system reliability issues.AI-enabled non-destructive testing methods for integrated photonic assemblies.AI in Photonic Packaging and Assembly:AI-guided optimization of optical coupling and alignment in PIC packaging.Automated assembly processes enhanced by AI to achieve scalable and cost-effective PIC production.Machine learning models for predicting and improving long-term stability and performance in packaged devices.AI for Data Analysis in Photonic Systems:Data-driven approaches to interpret large datasets from discrete device testing and PIC characterization.AI for analyzing optical spectra and system-level performance metrics in integrated circuits.Insights into system-level performance bottlenecks using AI-enhanced analytics.Real-World Applications and Case Studies:Industry case studies showcasing the application of AI in the design and deployment of discrete devices and PICs.Examples of AI-enhanced integration for telecommunications, data centers, and sensing applications.Success stories in scaling up production and integration processes using AI.Future Directions and Challenges:Exploring the potential of AI to enable autonomous photonic systems combining discrete devices and PICs.Ethical considerations and transparency in AI-driven design and testing processes.Addressing the challenges of integrating AI tools into existing photonic R&D workflows.The scope of this workshop is quite big and may need reducing or splitting into 2 parts.
Many speaker suggestions regarding AI for device design (Topic1). To make it more approachable we may chose to invite a couple of commercial vendors of cutting edge AI tools such as Matlab & Olympus etc to inform on AI topics topics 2,3,5,6,7.

A tentative list of speakers regarding AI for device design may include:
1) Prof. G. Karniadakis. They are the creators of Physics-Informed Neural Networks (PINNs), a robust ML framework for inverse design. Since its introduction, PINNs have achieved significant success in various engineering fields, including photonics and microwaves. While they are mathematicians rather than device or optical engineers, their work has revolutionized engineering.
2) Keisuke Kojima (Mitsubishi) – They specialize in inverse design and would be a valuable industrial speaker.
3) Jelena Vucković (Stanford) – Known for her work on ML-assisted photonic device optimization, her group has developed several interesting open-source software libraries for inverse design (JaxWell and spins-b).
4) Ian Williamson at Google X
5) Shanhiu Fan at Stanford
6) Commercial vendors of cutting edge AI tools

Workshop 3: Which modulator technology will dominate in next-generation transceivers?

Sunday 28 September from 9.00-12.30 in Room 3 (break included)

SC 2: Discrete photonic devices and technologies.

Short description: As optical links push toward higher transmission speeds, the modulator performance requirements are becoming increasingly stringent. This workshop will bring together industry experts and academic researchers to discuss the evolving landscape of modulator technologies, from Si photonics to InP and lithium niobate, as well as emerging hybrid approaches. Through system-level insights the suitability of different modulator material technologies for next-generation high-volume transmitters will be explored.

Organizers: Abdul Rahim, Despoina Petousi, Wei Shi

As optical links are moving to higher and higher transmission speeds, the demands on the modulator performance become stricter. Although, silicon photonics is well-positioned to deliver the performance, price, and volume for high-speed transmitters, the pure Si-based modulators show limitations in terms of bandwidth, loss and chirp. On the other hand, InP or lithium niobate modulators have traditionally outperformed the Si ones. Also, recent years have seen a paradigm shift where the integration of various materials on Si/SiN has opened up new applications for Si modulators. These modulators promise to extend the performance beyond the limits set by the physical properties of silicon. Since the next-generation transmitter technology still remains an open topic and no standard solution is found yet, this workshop will draw significant attention and will create fruitful debates. Experts from the industry can first present their demands, while researchers working on different material platforms can discuss their capabilities. More specifically, the first part of the workshop will present system level requirements for different optical communication needs (inter DC, intra DC, chip scale IO). In the second part we will discuss the individual technologies in terms of modulator performance, scalability, supply chain, yield, etc. By comparing the figures of merit of each platform, their suitability and limitations will be explored and the path towards realizing next-generation transmitters for the Terrabaud era will be revealed. The 3 organizers have extensive experience with photonic modulators and they can attract speakers from both the industry and the academia. Some suggestions of presenters are the following: Andrei Kaikonnen (coherent), Peter Ossieur (Ugent – Imec), Molly Peils ( OpenLight), Juerg Leuthold ( ETHz), Amir Ghadimi (Lightium). This workshop will be very important for ECOC 2025 since it will inform the audience about existing technologies, it will set the boundaries required by the industry and it will try to make conclusions in the never-ending debate of which modulator material platform will dominate in future transceivers for 800G and beyond.

Workshop 4: Reconfigurable, Adaptable and Intelligent Access Networks - Do we have real use cases?

Sunday 28 September from 9.00-12.30 in Room 4 (break included)

SC 7: Access, indoor and short-reach systems for data centres and mobile networks.

Short description: This workshop will explore what is driving the evolution of optical access networks and the potential role for more flexible and intelligent systems, with the ultimate aim to identify technologies that could deliver real value. Topics in scope include more adaptable underlying transmission links, artificial intelligence (AI) and monitoring/sensing enabled access networks. Workshop speakers drawn from the network operator domain, along with experts from system vendors and academia, will share their unique insights and contribute to a lively and interactive session.

Organizers: Derek Nesset, Rene Bonk, Paola Parolari

Optical access networks are ubiquitous today and operators are continually exploring ways to extract more value out of their networks, improve the customer experience and increase operational efficiencies. There are many proposals for how these optical networks can become more flexible, dynamic and intelligent. The aim of this workshop is to bring those ideas together and facilitate a lively discussion about which ideas meet real use cases in fibre access network applications.
At the physical layer, it is possible to envisage access networks with wavelength tunable transceivers, optical switching and routing to enable more dynamic and flexible networks and services. Furthermore, these networks could be augmented by intelligent topology discovery and fault location/isolation capabilities.
There are ideas for adaptable line rates, FEC coding and modulation formats that can enable these networks to be more efficient. For example, by exploiting excess link margins to increase capacity or adapting to low traffic periods to reduce power consumption.
With the rapid rise in AI we can expect huge impacts from this on access networks to offer greater automation of operations along with new capabilities in the intelligent control and optimisation of network and transmission resources. Access networks with intelligence can also benefit from network data to aid various of functions such as network planning, operation, optimization and maintenance.
As optical access networks see more and more use cases, in addition to the conventional FTTH, we can expect new requirements and expectations for high network availability. Flexible and adaptable access networks can be used to enable capacity on demand, with low and deterministic latency while minimising service disruptions. For example, for mobile connectivity in the coming 6G networks and future edge compute applications that may be dominated by machine-to-machine communications.
Smart city concepts are a recent trend and the ubiquity of access networks see them being identified as intelligent infrastructures with monitoring and sensing capabilities that can be offered as an additional service.
So, with so many potential ideas, this workshop asks which ones can deliver real value in future access networks? We invite network operators to share their operational pain points, insights and visions for how access networks might involve to include more intelligence and flexibility. Experts from industry and academia will review the range of ideas to realise future reconfigurable, adaptable and flexible access networks. This workshop will also address the standardization efforts in various SDOs (ITU-T, IETF, BBF and ETSI).
A lively debate is expected as there are many different views on what is required and what can realistically be implemented to offer a commercially viable return on investment.

List of potential speakers:
Dan Kilper, Trinity College Dublin
Junich Kani, NTT
Annachiara Pagano, Fibercop
Jochen Maes, Nokia
Bruno Cornaglia, Vodafone
Nicola Calabretta – Eindhoven University of Technology
Emmanouil Angelakis, Deutsche Telekom
Ye Zhicheng, Huawei
Philippe Chanclou, Orange
Sander Jansen, Adtran
Roberto Gaudino, Politecnico di Torino

Operators:
Fabienne Saliou (Orange)
Andrew Lord (BT)
Gregory Sherrill (Verizon)
Eddy Barker (AT&T)
Junichi Kani (NTT)
Dezhi Zhang (CTC)

Vendors:
Xiang Liu (Huawei)
Michael Kuipers (Adtran)
Andrew Bender (Nokia)

Application / service side:
e.g. speakers from applications side to be identified (e.g. NVIDIA, Omdia, and academia)
Anna Tzanakaki (National and Kapodistrian University of Athens)

Workshop 5: Quantum Key Distribution: Advancements, Challenges and Real-World Implementation.

Sunday 28 September from 9.00-12.30 in Room 5 (break included)

SC 11: Quantum communications and quantum computing.

Short description: As quantum computers advance and pose an increasing threat to modern cryptographic systems, the need for alternative cryptographic approaches becomes more urgent. Quantum key distribution offers a promising solution for building quantum-secure networks. This workshop explores advancements, challenges, and real-world applications related to QKD and the integration of quantum technology into current telecom infrastructure.

By attending this workshop, you will gain valuable insights into the use cases and challenges of QKD. The session will explore both Discrete Variable (DV) and Continuous Variable (CV) QKD, highlighting their key differences and operational mechanisms. You will learn how each approach functions, their advantages in securing communications, and the obstacles they face in practical implementation. This workshop will provide a comprehensive understanding of QKD’s role in building future-proof cryptographic systems.
Additionally, the workshop will emphasize the integration of quantum systems into telecom networks, a key advancement in strengthening security and achieving quantum-safe communications. Utilizing existing fiber-optic infrastructure, QKD and other quantum technologies can be efficiently integrated into contemporary networks.

Organizers: Alessandro Zavatta, Davide Bacco, Tobias Gehring

Detailed motivation of the proposal:
The proposed workshop aligns with the Conference’s technical scope 11: Quantum Communications and Quantum Computing. Specifically, the workshop has the main objective of delving into the topic of Quantum Key Distribution (QKD), considered as the most mature among the quantum technologies and at the forefront of ensuring future-proof security of quantum communications. Although QKD can be considered a mature technology, multiple challenges are still to be solved. Examples are the co-existence of quantum and classical communication in deployed networks, the control and management of a quantum network, and the limited documentation in terms of standardization and policy that is available today. All of these challenges are limiting the implementation of QKD in multiple use-cases. ECOC represents the best framework to present and discuss where we are today and where we are heading. The workshop will be the opportunity to highlight the differences between classical and quantum physical layer security, to discuss the integration of quantum and classical light into existing networks, and show which use-cases have been already tested highlighting pro and cons.
Why is this a good topic for a workshop:
Quantum communications stands as one of the core pillars of quantum technologies. At the forefront of these technologies lies Quantum Key Distribution (QKD), a disruptive innovation marked by significant advancements and simultaneous challenges. While the role of QKD in overcoming the limits of current cryptographic systems to ensure the security of communications against both classical and quantum threats has been widely demonstrated, major steps are still to be taken to guarantee a wide implementation of this technology. The challenges are attributed to both the lack of awareness of the potential benefits of the QKD, as well as to technical challenges for, for example, integrating the technology into existing telecommunication infrastructure or overcome long distance implementations. As the proposed topic encompasses various facets, the organisers felt it necessary to propose a dedicated workshop to comprehensively delve into the topic. This workshop aims to introduce the broader audience to the QKD, shedding light on its role in the revision of current cryptographic systems, address encountered challenges, and highlight the potential opportunities for advancing distributed quantum communication networks.
Possible invited speakers
Name, Affiliation
Chigo Okonkwo, Eindhoven University of Technology, The Netherlands
Cristian Antonelli, University of L´Aquila
Martin Charbonneau, Nokia CA-EU
Daniele Mancuso, Sparkle
Antonio Manzalini, TIM
Marco Lucamarini, York University
Mariella Minder, University of Cyprus
Ivan Derkach Palacky, University Olomouc
Davide Rusca, University of Vigo
Feihu Xu, University of Science and Technology of China

We suggest broadening the scope very slightly to QKD and beyond to include topics like entanglement distribution in networks which is believed to be of enormous interest to the SC11 audience today. Could Tobias invite Mads Obro MAOEB@tdcnet.dk the Danish Telco TDC to help organise – they were interested in being co-organisers of the similar workshop 3

Workshop 6: Coherent optical transceiver for Free-Space Optic links: Commercial-off-the-shelf or custom designed?

Sunday 28 September from 9.00-12.30 in Room 6 (break included)

SC 9: Free-space optics and optical wireless technologies.

Short description: Commercial Off-The-Shelf (COTS) solutions like Digital Coherent Optical (DCO) systems are increasingly used in space applications, such as by Starlink. However, Free Space Optical (FSO) communications face challenges due to atmospheric turbulence affecting signal processing and synchronization. There is a need to balance the high development costs of specialized modems with the ability to compensate atmospheric effects.

Organizers:, Simon Fabbri, Jeffrey Lee

“Commercial Off-The-Shelf (COTS) solutions like Digital Coherent Optical (DCO) systems are increasingly used in space applications, such as by Starlink. However, Free Space Optical (FSO) communications face challenges due to atmospheric turbulence affecting signal processing and synchronization. There is a need to balance the high development costs of specialized modems with the ability to compensate atmospheric effects.
FSO is promising for space-based earth observation and data relay, facilitating rapid data transfer to earth at low costs. Additionally, FSO could significantly enhance feeder links to satellites in both LEO and GEO orbits, potentially enabling multi-terabit per second capacities at reduced costs. The ongoing issue is whether to adapt existing systems or develop new ones to manage atmospheric conditions and the tough space environment.”

Terrestrial
Daniele Raiteri (Aircision)
Devin Brinkley (Taara)
Space (LEO/GEO)
Mustafa Cardakli (Kuiper)
Ravij Boddeda (Nokia Bell Labs)
Ramon Mata Calvo (European Space Agency)
Anaëlle Maho (Thales Alenia Space)
Curt Schieler (MIT LL)
Jean Christophe Richard (Airbus)
Baris Erkmen (Aalyria)
Xiaoxia Wu (Starlink)
Cyrille Laborde (Thales Alenia Space)
Juraj Poliak (DLR)
Modem
Ravij Boddeda (Nokia Bell Labs)
Mike Dollard (Viasat)
Bryan Tatman (Celestia STS)
Cyrielle Richard (Safran)
Donald Govan (mBryonics)

Workshop 7: Open Optical Networks-as-a-Service for 6G and AI: Vision or Reality?

Sunday 28 September from 9.00-12.30 in Room 7 (break included)

SC 6: Architecture, modelling and performance of optical networks.

Short description: Open Optical Networks-as-a-Service (ONaaS) offers a groundbreaking approach, enabling on-demand, high-capacity, and ultra-low latency connectivity across access, metro, and core segments. This workshop delves into the feasibility of ONaaS, focusing on advancements driven by key industry consortia and standardization organizations. We will explore how open standards, software-defined networking (SDN), and digital twin technologies can enable seamless interoperability and intelligent automation.

Organizers: Mr. Gert Grammel, Dr. Andrea D’Amico, Mr. Sai Kishore Bhyri, Dr. Hideki Nishizawa

The convergence of 6G, data-center exchange and AI data transfer demand will require unprecedented levels of network flexibility, scalability, and performance. Open Optical Networks-as-a-Service (ONaaS) offers a groundbreaking approach, enabling on-demand, high-capacity, and ultra-low latency connectivity across access, metro, and core segments. This workshop delves into the feasibility of ONaaS, focusing on advancements driven by key industry consortia and standardization organizations. We will explore how open standards, software-defined networking (SDN), and digital twin technologies can enable seamless interoperability and intelligent automation. Through expert presentations and discussions, we aim to evaluate whether ONaaS is poised to become a tangible reality or remains an aspirational vision, addressing critical challenges such as orchestration, AI-driven optimization, and multivendor integration.

Workshop 8: Digital signal processing for optical fiber sensing.

Sunday 28 September from 14.00-17.30 in Room 1 (break included)

SC 4: Signal processing for optical communication and computing.

Short description: Digital signal processing has become paramount for modern fiber sensing technologies. It is one of the rare fields of fiber optics which brings together some of the most advanced signal processing techniques. Its signal processing quickly evolved from using simple pulses to employing advanced traditional signal processing techniques such as spread spectrum, handling of laser phase noise, polarization handling, digital backpropagation, etc. On top of that, machine learning techniques are extensively used. This workshop brings together experts in the field to discuss modern fiber sensing technologies from employed pulses, DSP, to feature extraction and event recognition.

Organizers: Fatih Yaman, Sjoerd van der Heide

Many exciting applications of fiber sensing has been the focus lately, which has so far overshadowed similarly exciting developments in the fiber sensing signal processing field.
Fiber sensing recently transitioned from mostly analog to mostly digital. With that the signal processing used for sensing quickly evolved from sending a simple pulse to relatively complicated spread-spectrum techniques, to handling and equalization of laser phase noise, polarization management etc. Similarly, in not so far past, resolved measurement used to mean that some sort of backscattering was necessary. However, leveraging digital optical communication’s mature signal processing techniques, especially computation heavy methods like digital back propagation location resolved measurements does not require backscattering anymore. On top of traditional signal processing, fiber sensing is one of the few areas of optics where machine learning is not just a good to have add on but essential for most of its applications. As such it is one of the rare fields of fiber optics which brings some of the most advanced signal processing techniques together.

We aim to advance debates on:
– The effectiveness and limitations of DSP algorithms used for sensing and monitoring;
– The most promising large-scale deployments;
– Comparisons of DSP techniques for forward-based and backscattering-based sensing systems: which techniques are best suited for each monitored parameter?
– Machine Learning for optical fiber sensing;
– DSP allowing for the coexistence of sensing and communication and for seamless integration of sensing in optical networks;
– Handling large amounts of sensed data, feature extraction and event recognition from multiple sensing probes, etc.
– Disambiguation and comparison of spread-spectrum DSP methods for sensing: OFDR, time-gated-OFDR, chirped-pulse, multi-carrier, etc.

Workshop 9: AI Interconnect Dilemma: Which Technology Is Doomed – VCSELs or Silicon Photonics?

Sunday 28 September from 14.00-17.30 in Room 2 (break included)

SC 3: Photonic integrated circuits, assemblies and packaging.

Short description: As AI interconnects push toward 400 Gbps, the industry faces a critical question: Can VCSELs evolve to meet next-generation demands, or will Silicon Photonics emerge as the dominant technology? This workshop brings together experts from system companies, module manufacturers, and chip developers to explore the trade-offs, scalability challenges, and innovation pathways shaping the future of high-speed interconnects. Attendees will gain key insights into technology roadmaps, manufacturability, and the economic viability of these competing solutions.

Organizers: Eric Bernier, Benjamin Lee, Daniel Kuchta

The push for higher data rates in AI interconnects already demands components operating at 224+ Gbps, with the need for 400 Gbps solutions rapidly approaching. Vertical-Cavity Surface-Emitting Lasers (VCSELs) have long been the dominant transmitter technology for short-reach applications due to their cost and energy efficiency. However, Silicon Photonics (SiP) is emerging as a strong contender, offering superior integration and scalability. This workshop will critically evaluate whether VCSELs can adapt to meet future requirements or if SiP will overtake as the leading technology in AI interconnects.

Featuring speakers from leading system companies, module manufacturers, and chip developers, the session will explore the trade-offs between these technologies, their scalability challenges, and the innovation pathways required to support next-generation AI systems. Participants will gain actionable insights into technology roadmaps and strategies to navigate the rapidly evolving interconnect landscape.

System Companies and Analysts

System companies and analysts provide a top-down perspective on AI interconnect trends and requirements. They help define end-user demands, identify bottlenecks, and outline the roadmap for scaling to 400 Gbps and beyond.

Expected Impact:

Identify key drivers shaping AI interconnect requirements.
Provide a roadmap for scaling system-level performance.
Highlight potential bottlenecks and trade-offs in adopting VCSELs versus Silicon Photonics.

Module Companies
Module companies bridge the gap between devices and systems, focusing on transceiver design, packaging, and integration. They offer insights into manufacturability, cost, and the need to evolve form factors for 400 Gbps applications. Additionally, their perspective on cost structures and techno-economic challenges is crucial for understanding market adoption.

Expected Impact:

Share insights on packaging, cost optimization, and energy efficiency of modules.
Highlight challenges in scaling from 224 Gbps to 400 Gbps.
Discuss whether form factors need to evolve to meet future interconnect requirements.
Provide a techno-economic evaluation of module-level trade-offs in VCSELs and Silicon Photonics.

Device/Chip Companies
Device and chip companies drive the core technology behind interconnects, developing innovative VCSELs and Silicon Photonics. They address scalability, power efficiency, and performance at the component level while also factoring in cost-efficiency and manufacturability for large-scale adoption. Their input on cost-performance trade-offs is critical to the viability of both technologies.

Expected Impact:

Present advancements in VCSEL and Silicon Photonics technologies.
Discuss material and design innovations for higher data rates.
Explore trade-offs in cost, performance, and scalability at the device level.
Provide a techno-economic perspective on production costs and scalability of VCSELs versus Silicon Photonics.

Workshop 10: High Symbol-rate Transceivers - how to get to the pinnacle of performance?

Sunday 28 September from 14.00-17.30 in Room 3 (break included)

SC 5: Optical transmission systems.

Short description: Aiming for higher symbol rates is from first sight the most obvious approach to meet the increasing demand for larger data capacity, as this approach minimizes the number of channels. However, it is not clear if CMOS ASICs and opto-electronic components will be able to support symbol rates of 300 GBd and beyond. The question remains which set of electronic and photonic technologies will enable the pinnacle of performance.

Organizers: Markus Grözing, Georg Rademacher, Qian Hu

There are several approaches like wavelength division multiplex (WDM), spatial division multiplex (SDM) or the use of new spectral bands within the available light spectrum to meet the ever-increasing demand for data bandwidth in the local, regional, and global optical fiber-based information networks. But the most obvious and – from first sight – most cost-effective approach is aiming for higher symbol rates and maintaining high modulation orders, as this minimizes the number of needed electro-optical transceivers. Thus, commercial transceivers strive to operate at higher and higher symbol rates, motivated by cost-per-bit efficiency. Coherent (IQ) dual-polarization (DP) transceivers running at 200 GBd and 8 bit net information per IQ+DP symbol have been announced recently. This corresponds to a net information rate of 2 bit/symbol on each of the four electrical 200 GBd transmit or receive channels. The announced transceiver ASICs are implemented on 3 nm CMOS from TSMC, a technology with limited accessibility for some companies due to high start-up costs and strict confidentiality regulations. Further, it gets more and more difficult to squeeze out more signal bandwidth from the most advanced CMOS technology nodes, as transistor corner frequencies tend to stagnate or even decrease.

As we are approaching the limit of CMOS transistor switching speeds and physical process node sizes, alternative approaches start to become relevant. A question to be answered is which electronic and optic concepts can lead to a further increase of symbol rates to 300 GBaud and beyond without a significant reduction of the spectral efficiency or signal integrity. Future solutions may bring in alternative semiconductor technologies like SiGe- or InP-HBTs for the front-end high-speed symbol interleaving and/or alternative electronic-photonic concepts.

This workshop will bring together experts from the fields of optics and electronics to explore potential paths towards >300 GBaud transceivers.

Workshop 11: Will photonics-enabled THz communication and sensing play a role in 6G?

Sunday 28 September from 14.00-17.30 in Room 4 (break included)

SC 8: Sensing and microwave photonics.

Short description: This workshop explores how photonics-driven THz generation can unlock new horizons for next-generation mobile networks. Industry leaders, standardization bodies, and academic experts will share their insights on the potential of THz frequencies for both communication and sensing applications, highlighting the ongoing efforts to shape the 6G era.

Organizers: Oskars Ozolins, Sebastian Randel, Chris Vagionas

Supported by significant public and industrial funding, extensive research programs like Hexa-X (https://hexa-x.eu/) and 6G-Platform (https://www.6g-platform.com/) are underway with the goal of shaping the 6th generation mobile communication standard. On a hardware level, the (sub-)THz range is promising new spectral resources and research is focusing on efficient ways to generate, detect and process sub-THz waves efficiently and to realize high-speed communication and sensing systems. The IEEE 802.15 WSN SCTHz is working on the standardization of THz communication systems. Photonics-enabled THz generation is a particularly promising approach since it builds upon recent progress in commercial high-speed optoelectronic devices and DSP. However, it remains unclear what role THz systems can play in 6G: whether it will be communications or sensing? This workshop will bring together key players from industry, standardization, and academia to discuss the prospects of photonics-enabled THz communication and sensing in 6G and beyond.

Key learning points:
How can recent advances in photonics enable efficient generation and detection of THz frequencies?
What new opportunities do THz-based technologies present for high-speed communication in 6G networks?
How is IEEE 802.15 WSN SCTHz driving standardization for THz communication systems?
What role do high-speed optoelectronic devices and DSP play in THz signal processing and system performance?
Can THz technologies bridge communication and sensing, and what could that mean for future mobile standards?

Workshop 12: Is the access network ready to host quantum technologies?

Sunday 28 September from 14.00-17.30 in Room 5 (break included)

SC 7: Access, indoor and short-reach systems for data centres and mobile networks.

Short description: The integration of quantum technologies, including quantum key distribution, into access networks presents unique challenges due to architectural constraints, coexistence issues, and high losses in the point-to-multipoint links of passive optical networks. Despite these difficulties, the short-reach nature of access networks may facilitate the support of quantum secure communications. This workshop brings together industry experts, operators, vendors, and academic researchers to explore the challenges and opportunities of deploying quantum technologies in access networks, stimulating an engaging and dynamic discussion.

Organizers: Paola Parolari, Michela Svaluto Moreolo

The adoption of quantum technologies, especially quantum key distribution (QKD), has been mainly investigated in point-to-point metro links. However, limited research has addressed their feasibility in access networks, which present unique architectural and operational challenges. To enhance the network security, beyond niche applications in dedicated networks, and in view of the quantum Internet, quantum technologies must be integrated into existing communication infrastructures, sharing transmission media, network equipment, and management frameworks. However, this poses a threat on the quantum layer deployment due to coexistence issues, particularly in the access segment. The passive optical network (PON) is a very challenging scenario due to the presence of several access standards and to high losses of point-to-multipoint (PtMP) links. Nevertheless, the short-reach nature of access networks may offer advantages for quantum secure communications and may also stimulate innovative PtMP QKD solutions.
This workshop brings together industry experts, operators, system vendors, and academic researchers to discuss the technical and operational challenges, as well as the opportunities, of incorporating quantum technologies into access networks. It extends beyond QKD to explore the broader potential of quantum technologies for new services and applications in access networks.
Key questions to be addressed include:
• How can the access and quantum worlds not struggle?
• Are there network architectures that facilitate the integration of quantum technologies in the access?
• What role does photonic integration play in advancing quantum technologies for access networks?
• Which networking technologies can ease the convergence of quantum and classical optical layers?
• What impact will quantum technologies in access networks have on shaping the future quantum Internet?

Tentative speaker list (to be re-defined)
Operators:
Annachiara Pagano (Fibercop, IT)
Paulette Gavignette (Orange, FR)
Cathy White/Andrew Lord (BT, UK)
Jose Manuel Rivas Moscoso/ Diego Lopez (Telefonica ID, ES)
Vendor/ SME:
James Dynes, (Toshiba UK)
Davide Bacco, (QTI, IT)
Reza Nejabati, (CISCO, US)
Florian Fröwis (ID Quantique, CH)
Academia:
Eleni Diamanti, (Sorbonne Université, FR)
Chigo Okonkwo/ Ozan Çirkinoglu, (TuE, NL)
Rui Wang (University of Bristol, UK)
Go Kato (NICT, Japan)

As we are approaching the limit of CMOS transistor switching speeds and physical process node sizes, alternative approaches start to become relevant. A question to be answered is which electronic and optic concepts can lead to a further increase of symbol rates to 300 GBaud and beyond without a significant reduction of the spectral efficiency or signal integrity. Future solutions may bring in alternative semiconductor technologies like SiGe- or InP-HBTs for the front-end high-speed symbol interleaving and/or alternative electronic-photonic concepts.

This workshop will bring together experts from the fields of optics and electronics to explore potential paths towards >300 GBaud transceivers.

Workshop 13: In-Building Networks: Ways to lower energy and cost per bit.

Sunday 28 September from 14.00-17.30 in Room 6 (break included)

SC 9: Free-space optics and optical wireless technologies.

Short description: The workshop addresses the combination of fiber-to-the-room (FTTR) with Wi-Fi as a promising solution to increase the coverage of high data rates in households and industries. We will discuss recent developments of next-generation in-building networks with a focus on lower cost and carbon footprints.

Organizers: Volker Jungnickel, Christian Bluemm

In-Building Networks: Lower energy and cost per bit for 80% of our data
ORGANIZERS: Volker Jungnickel (Fraunhofer HHI), Christian Bluemm (Huawei)
ABSTRACT: The workshop addresses the increasingly important in-building network segment, where more than 80% of data traffic is consumed and created, while people get used to connect wireless. Until 2030, the average data consumption in European households may well grow by 20% each year [1] and will be only dwarfed by other, even denser communication scenarios in industries, enterprises or sport stadiums. This is when 4G and 5G services become incapable to offer ultra-fast and reliable wireless connectivity into every corner of a building or campus, making fiber to the room (FTTR) combined with Wi-Fi the capacity densifier of choice. While the basic concept attracted significant attention at ECOC 2024, new aspects need to be discussed, among which are cost and energy when increasing data rates. Although total costs increase when densifying a network, recent techno-economic studies indicate that cost per bit are actually lowered [2]. It is anticipated that energy per bit reduces likewise. FTTR facilitates the coordination of multiple Wi-Fi access points deployed in different rooms to overcome attenuation through the walls for unrestricted line-of-sight in many cases. It is still debated if such coordination will be centralized or distributed. The combination of FTTR and Wi-Fi achieves better coverage with typically reduced transmit power, meaning lower energy consumption. Both cost and energy savings complement classical FTTR advantages in terms of latency, capacity and scalability and could stimulate further investments into modern in-building network infrastructures. FTTR and coordinated Wi-Fi are a contribution towards a green and sustainable information and communication technology. Similarly, novel services which only become possible through FTTR and Wi-Fi, such as edge storage and edge computing, can further enhance efficiency.
In the context of surging efforts towards efficiency in the entire telecommunication sector, panelists will discuss recent developments of next-gen in-building networks towards lower cost and carbon footprint. Potential Panelists are (not yet confirmed):
Fraunhofer HHI – Kai Habel
Orange / France – Fabienne Saliou
Nokia / Germany – Rene Bonk
Adtran / Klaus Grobe
Fraunhofer IZM / Lutz Stobbe
University of Bundeswehr / Carmen Mas Machura
Huawei / Tony Zeng & Wu Junhong
Maxlinear / US – Marcos Martinez OR Rainer Strobel
China Unicom / Yue Sun & Hai Ding, OR China Mobile / Ning Wang, OR China Telecom / Qizheng Li
Trinity College Dublin / Marco Ruffini

Workshop 14: Optical Networks and AI: do we need a brand-new infrastructure for AI, and can AI help run it?

Sunday 28 September from 14.00-17.30 in Room 7 (break included)

SC 10: Control and management of optical networks.

Short description: While AI can help with operation in next generation networks (i.e., AI for Optical Networks), how can optical networks be leveraged to help AI model training (Optical Networks for AI)?
AI has been proposed to tackle many design and operation topics in optical networks for the past 10 years or so, and it is now time for a reality check and possibly on a new outlook on the design of optical networks tailored for DCI transport requirements as genAI is becoming ubiquitous.

Organizers: Yvan Pointurier, Raul Muñoz, Behnam Shariati

The relationship between AI and optical networks is two-ways. In one way, the new workloads incurred by the generalized adoption of AI models on different sectors (transportation, education, health, industry, etc.) impacts on the network design and operation to support distributed AI clusters (“optical networks for AI”). On the other way, AI can be leveraged to perform and automate operations that are tedious or too complicated for network operators (“AI for optical networks”). We target a two-part workshop to tackle those two aspects at the same time, to close the AI <-> optical networking design and operation loop.
Training generative AI models, including large (language) models, is a new driving force not only for the short-reach, metro size DCI networks, but also in the transport optical networks, as distributed cross-datacenter model training is garnering pace and the volume of data during training is enormous, outpacing regular consumer traffic growth. The optical transport needs vary depending on the operator – traditional telco operators may be enticed to carry DCI traffic on their current (possibly expanded) infrastructure or on the contrary build a new dedicated generation network for the new traffic, while hyperscalers may simply optimize their network for DCI traffic, which dominates their mix, some of them even advocating replacing ROADMs with IP routers.
Building a new network also brings opportunities in terms of operation. New and innovative operation, administration and management (OAM) processes to meet specific DCI traffic requirements can easily be rolled out.

While optical networks can be leveraged to help AI model training, how can AI help with OAM in next generation networks? AI has been proposed to tackle many design and operation topics in optical networks for the past 10 years or so, and it is now time for a reality check and possibly on a new outlook as genAI is becoming ubiquitous.
Specifically, the planned topics are:

Requirements for the next generation optical network optimized or designed for the AI traffic flood in distributed AI clustersNew multi-fiber architectures (e.g., ROADM vs opaque) and equipment (including novel cables) for such networksNew planning techniques enabled by AINew foundation model dedicated to telecom/optical networksNew operation methods enabled by AI (not limited to genAI), including human/physical layer interactions, sensing, fault management (detection, localization, mitigation e.g. through physical layer reconfiguration)Proposed speakers:
Dave Ward, Lumen, USA (or analyst such as OMDIA) – market trends and requirements for long haul DCI networks
XX, China Telecom, China – Long haul networks in 2030 – the need for separate long-haul networks for 5/6G traffic and DCI traffic
Hideki Nishizawa, NTT – The role of AI in NTT’s IOWN’s requirements, design and operation
Juan Pedro Fernandez-Palacios, Telefonica, Spain – The long-haul network in 2030 will be converged
Arash Vakili or Ligia Maria Moreira Zorello, Meta, USA/UK – The hyperscaler view on the same topic
XX, Alibaba or Tencent – idem for Asia
Lidia Galdino, Corning, UK – New (high density) cables for long haul DCI networks
David Boertjes or other, Ciena – on the use of AI in production control planes (blueplanet talk)
David Charles, Nokia – same (wavesuite talk)
Lilin Yi, SJTU – Foundational models for optical networking
Cen Wang, KDDI – Generative AI and digital twins for optical networks

Proposed organizers: Yvan Pointurier, Huawei; Raul Muñoz, CTTC; Behnam Shariati, HHI

This workshop brings together industry experts, operators, system vendors, and academic researchers to discuss the technical and operational challenges, as well as the opportunities, of incorporating quantum technologies into access networks. It extends beyond QKD to explore the broader potential of quantum technologies for new services and applications in access networks.
Key questions to be addressed include:
• How can the access and quantum worlds not struggle?
• Are there network architectures that facilitate the integration of quantum technologies in the access?
• What role does photonic integration play in advancing quantum technologies for access networks?
• Which networking technologies can ease the convergence of quantum and classical optical layers?
• What impact will quantum technologies in access networks have on shaping the future quantum Internet?

Tentative speaker list (to be re-defined)
Operators:
Annachiara Pagano (Fibercop, IT)
Paulette Gavignette (Orange, FR)
Cathy White/Andrew Lord (BT, UK)
Jose Manuel Rivas Moscoso/ Diego Lopez (Telefonica ID, ES)
Vendor/ SME:
James Dynes, (Toshiba UK)
Davide Bacco, (QTI, IT)
Reza Nejabati, (CISCO, US)
Florian Fröwis (ID Quantique, CH)
Academia:
Eleni Diamanti, (Sorbonne Université, FR)
Chigo Okonkwo/ Ozan Çirkinoglu, (TuE, NL)
Rui Wang (University of Bristol, UK)
Go Kato (NICT, Japan)

As we are approaching the limit of CMOS transistor switching speeds and physical process node sizes, alternative approaches start to become relevant. A question to be answered is which electronic and optic concepts can lead to a further increase of symbol rates to 300 GBaud and beyond without a significant reduction of the spectral efficiency or signal integrity. Future solutions may bring in alternative semiconductor technologies like SiGe- or InP-HBTs for the front-end high-speed symbol interleaving and/or alternative electronic-photonic concepts.

This workshop will bring together experts from the fields of optics and electronics to explore potential paths towards >300 GBaud transceivers.