home

Insight top Prysmian Group Insight bottom Prysmian Group

Chigo Okonkwo and Prysmian: making optical transmission fibres fit for the future

optical transmission fibres

“Eventually, over 99% of all data generated will go through fibres”, explains Chigo Okonkwo, Associate Professor at the High-Capacity Optical Transmission Laboratory at Eindhoven Hendrik Casimir Institute (EHCI), Eindhoven University of Technology. “However, optical transmission systems are running out of capacity. A paradigm shift in system design and testing is required to develop next-generation low-energy, robust and high-capacity photonic networks.”

“My group is especially interested in single-mode, few-mode, multimode, and multicore fibres, which is also the area our research with Prysmian focuses on. Single mode is found in many inter-data centre connections, access/metro/edge networks, and long-haul transmission systems. We’re looking at maximizing single mode fibre capacity by optimizing amplifier operation and the way in which signals are sent into the fibre. If we move to a new non-linear regime, we can see what happens to signals and how we can recover them. That means you could potentially operate fibre in a higher power regime, allowing recovery of impacted signals. We’ve realized single mode optimisations that are being used in current products such as transceivers. These use advanced DSP algorithms (‘Constellation Shaping’) on the transmitter and receiver side ASICs.”

Chigo Okonkwo

Associate Professor at the High-Capacity Optical Transmission Laboratory at Eindhoven Hendrik Casimir Institute (EHCI)

“In single mode networks, you can boost capacity to some extent by reducing loss: this is generally 0.15 dB/km and you can try to reduce it to 0.1 dB/km. Most buried fibres measure at around 0.2-0.3 dB/km so over 100km that could easily mean 20-30 dB loss. Reducing the loss of the fibre means you don’t need to deploy amplifiers and active equipment as often, helping reduce fibre plant CAPEX and potentially also OPEX. However, the best way to expand a single-mode based transmission system is by deploying more fibres - room for increasing capacity in a single fibre link is very limited. The maximum transmission value for one fibre is approximately 100 Tb/s. That requires transmitting significant power across many channels simultaneously, which can be affected by limitations related to the amplifier, power, or system operation and reach might be limited. These single-channel constraints don’t apply in the multimode area - in which Prysmian is a leader – that we are also researching. By transmitting orthagonal spatial channels in the fibre and encoding information on each channel a fibre with three modes could have three times the capacity of a ‘regular’ single mode cable - and ten modes might offer ten times the capacity. We’re looking into how we can use fibres as made by Prysmian to extend and expand the capacity of optical communications systems. If fibre interfaces and components are developed correctly, you have plenty of room to do this.”

“The Cost Per Bit is much lower when using multimode fibres, compared to single mode. However, we still need to iron out some signal processing kinks, including those occurring in interfaces and in-line devices. We’re making good progress. We’ve developed prototypes and written several papers with Prysmian, as well as institutes such as Japan’s National Institute for Communications Technology and Nokia Bell Labs in New Jersey USA. Our colleagues in Japan have managed to transmit 10 Pb/s on a single fibre with a large cladding diameter of 312µm, and 1.5Pb/s on a Prysmian few-mode fibre with a standard cladding diameter of 125µm. Imagine the capacity you could reach by using 100 spatial channels within one fibre! Space division multiplexing, enabled by Prysmian fibres, is a very attractive proposition. Combining multiple fibres in one cable and using multimode transmission offers very significant transmission increases. Potentially, you can reach speeds in excess of 100 Pb/s.”

“We’re currently moving what we’ve developed in the lab into the field for testing over a shorter link. This gives us insight into how amplifiers and fibre links work, and the best ways of getting signals into and out of the fibre. There are quite a few complicating factors outside the lab setting. Fibres we receive from Prysmian on a reel haven’t been treated or coated for the environment in which they will be deployed. We test transmission over this spooled fibre and work out how easy it is to splice and connectorise it. Once you provide the fibre with a jacket, you can get very different test readings, for example due to polarisation effects, mode-dependent loss, temperature, or deployment method. Is the cable compressed in a duct, or swaying in an aerial installation? Is it exposed to lightning? Installed in a saline marine environment? In a nuclear environment?”

“A new development Prysmian is supporting is placing multimode fibres in the ground. This is currently being done in a research setting - fibres that can support up to 15 modes.

When we look at multimode fibres, we definitely need to work on maturing the components. For example, there is no off-the-shelf optical amplifier you can use to amplify multiple modes. With Prysmian’s support, we’re looking into creating efficient amplifiers that can do this. These amplifiers need to amplify all modes equally. Ideally one amplifier should be required instead of a hundred identical amplifiers which need to be separately aligned and synchronized to realise the multimode’s capacity potential. We also need to improve multiplexing: how can we realise access to all modes, each carrying independent data, in a small form factor so we easily connect fibres with low loss? The technology underlying WDM has matured over the last 20 to 30 years. Whether you’re using 10 or 100 channels, loss remains the same. When scaling from ten modes to a hundred modes with Space Division Multiplexing we can make sure multiplexer loss is fixed so it remains consistent, even when we increase the number of modes. Coupler loss is fixed but losses can differ per channel. We are working to solve the wavelength and mode-dependent loss. If we can keep the losses across wavelengths the same as the loss across modes, we will really start to reap the benefits of a mode-based system.”

“To date, we’ve mainly been working in long haul transmission systems, where you might want to propagate a signal along a 20,000 km transoceanic distances. In the meanwhile, however, we’ve also been following new application areas, led by data centres. For example. Primarily related to greenfield connections, initiated by large tech companies with deep pockets that want to deploy their own fibre networks. At the relatively short distances involved - typically less than 120 km between DCs - the 24dB loss on fibre is still reasonable and doesn’t require an amplifier. That in turn means lower network cost. In this scenario, you can deploy novel fibres, such as few-mode or multimode fibres. Inter-DC links require very high capacity, possibly at Pb scale, and low latency. So how can we pack and unpack signals without introducing latency? When it comes to fibre, the best choice always depends on distance, type of usage, environment, and so on. There’s no one size fits all! Multimode amplifier technology is still not mature, but this technology might be ideal for deployment across short reaches where you don’t need amplifiers. I think space division multiplexing based on few-mode or multimode fibres can be a good solution for providing the required high capacity. A good example of our take on research and testing: the key goals is to creatively develop links that can not only compete with currently utilized systems, but which offer brand new or alternative benefits for widely diverging applications!

Since 2010, Chigo and his team have been building up a world-class high capacity optical transmission laboratory and collaborates with Prysmian and various partners.

TU/e and Prysmian collaboration: how it started

“After doing my PhD at the University of Essex, I moved to Eindhoven in 2009, where I took up a Postdoctoral research position at the Eindhoven University of Technology. Within two months, I met Adrian Amezcua Correa, currently Director Technical Sales Support at Prysmian Group. We decided to do some low-level collaboration. For example, researching bandwidth limitations and usability in different network setting of Prysmian multimode Fibres.” Within a year or two I met Pierre SIllard, New Fiber products R&D Manager at Prysmian Group, who was in charge of few-mode fibres being designed, was interested in going beyond what competitors were doing in this field. We started characterizing six-mode Prysmian fibres, and identifying interesting parameters, such as mode dependent loss. With state-of-the-art experiments we could uncover the fibre’s potential benefits. We worked on this for some years leading to some well-regarded papers.”

We also worked on a new type of fibre few mode multicore with the University of Central Florida, with Prysmian making the introduction. That showed the benefits of capacity increase. A single-mode fibre typically has one core – what happens when in a multicore fibre you multimode each core? If you have three modes in each core of a seven core fibre, you can have 21 times the capacity of a single-mode fibre. In Florida they manufactured a 1km fibre and analysed its performance. We showed we could transmit a quarter of a Pb over that distance. Still working on this.

Chigo Okonkwo is an Associate Professor leading the high capacity optical transmission laboratory of the Electro-Optical Communications Group at the Institute for Photonics Integration, TU/e. In 2002 he graduated with a Masters Honours degree in Telecommunications and Information systems from the University of Essex, where he also completed his PhD in Optical signal processing at the University of Essex. He joined COBRA TU/e as a Post-doctoral researcher in 2010.

Chigo’s research covers several fields:

  • High Capacity Optical transmission

“How can we maximise the capacity of a channel to carry the largest number of bits without any issues?”

  • Free space optical technology

“This looks at transmitting optical signal using light – so without a conduit such as a fibre – which theoretically makes it possible to cover extremely large distances and significantly reduce latency. Transmission of light over air is potentially 30% faster than via fibre – but there are several limitations to be considered, such as turbulence.”

  • Digital Signal Processing (DSP)

“DSP can play an important role in keeping transmission error-free. For example, a channel consisting of fibres and amplifiers will create impairments in the signal and with DSP technology the signal can be reformatted in a way that allows you to recover your bits accurately.”

In the field of Digital Signal Processing, Chigo is currently working on several specific topics:

  • Maximising single-mode fibre system capacity by employing advanced-coded modulation schemes and Probabilistic/Geometrically shaped signals and low complexity digital signal processing.
  • Future-proof transmission systems based on Space Division Multiplexing
  • Developing new optical components required to scale the capacity of the optical transmission systems towards Petabit/s transmission.

Chigo Okonkwos work has led to a provisional patent and several critical publications in Nature Photonics, JLT and Optics Express. As a Senior Member of IEEE, he serves as a reviewer and has (co)authored more than 200 publications in conferences and journals. Since 2018, he has been a member of the Center for Quantum Materials and Technology (QT/e) working to advance the National agenda on Quantum secure communications. Since 2014, he has been a Technical program committee member for ECOC. In 2018, he was the Sub Committee Chair for Digital Signal Processing track at ECOC and the General Chair for OSA Advanced photonics congress conference on signal processing for photonics.

Read more about optical transmission fibre research

 

255Tb/s

 

R van Uden., et al. Ultra-high-density spatial division multiplexing with a few-mode multicore fibre. Nature Photon 8, 865–870 (2014).

 

15 mode fibre from Prysmian Draka

 

P. Sillard et al., "Low-Differential-Mode-Group-Delay 9-LP-Mode Fiber," in Journal of Lightwave Technology, vol. 34, no. 2, pp. 425-430, 15 Jan.15, 2016, doi: 10.1109/JLT.2015.2463715.

Read more