See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/326051276
Towards multiband optical systems
Conference Paper · January 2018
DOI: 10.1364/NETWORKS.2018.NeTu3E.1
CITATIONS READS
76 1,494
12 authors, including:
Antonio Napoli Nelson M. S. Costa
Infinera Infinera Portugal
319 PUBLICATIONS 5,205 CITATIONS 153 PUBLICATIONS 1,733 CITATIONS
SEE PROFILE SEE PROFILE
Joao Pedro S. Abrate
Bahiana School of Medicine and Public Health LINKS Foundation
238 PUBLICATIONS 2,235 CITATIONS 152 PUBLICATIONS 1,438 CITATIONS
SEE PROFILE SEE PROFILE
All content following this page was uploaded by Antonio Napoli on 11 March 2019.
The user has requested enhancement of the downloaded file.
Towards multiband optical systems
Antonio Napoli1 , Nelson Costa2 , Johannes K. Fischer3 , João Pedro2,4 , Silvio Abrate5 ,
Nicola Calabretta6 , Wladek Forysiak7 , Erwan Pincemin8 , Juan P. F-P. Gimenez9 ,
Chris Matrakidis10 , Gunther Roelkens11 , Vittorio Curri12
1 Coriant R&D GmbH, Germany; 2 Coriant Portugal, Portugal; 3 Frauhofer HHI, Germany; 4 IT, IST, Portugal;
5 ISMB, Italy; 6 TU/e, The Netherlands; 7 AIPT, Aston University, UK; 8 Orange Lab, France; 9 Telefonica, Spain;
10 University of Peloponnese, Greece; 11 Ghent University-IMEC, Belgium, 12 Politecnico di Torino, Italy
antonio.napoli@coriant.com, nelson.costa@coriant.com, johannes.fischer@hhi.fraunhofer.de
Abstract: Multiband transmission is a valid option to significantly increase fiber capacity
and efficiently utilize the available and deployed optical fiber infrastructure. In this contribu-
tion, we evaluate its challenges and possible implementation.
OCIS codes: (060.2330) Fiber optics communications; (060.1660) Coherent communications.
1. Introduction
The race to increase fiber capacity moved from static systems, employing intensity modulation direct-detection (IM-
DD), to advance elastic optical networks [1, 2] where complex quadrature amplitude modulation (QAM) formats
are transmitted by bandwidth variable transponders [3]. This significantly extended the fiber capacity, moving from
<1 Tb/s/fiber with 10 Gb/s IM-DD per channel to >25 Tb/s/fiber with 200 Gb/s polarization multiplexing 16-QAM
per channel. However, forecasts suggest that this capacity might not be sufficient to cope with the current growth of
IP traffic, as we enter the Zettabyte era [4]. Furthermore, the imminent deployment of 5G networks and high-capacity
access networks will further stress telecommunication infrastructures. To cope with this high demand for capacity,
researchers proposed several solutions, mainly based on space division multiplexing (SDM), by transmitting over
multi-fibers (MF) and/or multi-core (MC) and/or multi-mode (MM) fibers. In this contribution, we discuss a different
approach, which aims at transmitting over the entire low-loss optical spectrum of single mode fibers (SMF), namely
from 1260 nm to 1625 nm. We name this as multiband (MB). The main objective of this article is to compare the
advantages and shortcomings of MB and SDM, taking into account technology availability and possible applications.
Fig. 1: (a): Comparison among the different solutions; (b): Fiber attenuation and dispersion parameters over the multi-
band SMF spectrum; (c): Multiband system and bandwidth requirements of the key-components.
2. Solutions for capacity increase
Large investments have been realized with the goal of extending the total capacity of optical systems. Within SDM,
we can identify the following solutions: Multi{-Fiber, -Mode, -Core} and Fig. 1(a) compares them. Only MF can be
already realized, for example by installing several parallel systems [5] over new fibers or by lighting up the existing
dark fibers. However, this is a costly solution due to the multiplication of all network elements, and possibly requiring
also the deployment or renting of additional optical fiber pairs, potentially entailing prohibitive costs. The MC and MM
strategies have been widely investigated and numerous hero experiments have been reported. For instance, capacity
in excess of 138 Tb/s with MM [6] or up to 1 Pb/s with MC [7] has been achieved. Such strategies would greatly
contribute to address the capacity crunch, but they present several important limitations, as listed in Fig. 1(a). For
example, they require the deployment of novel fibers, which is a very costly operation. Moreover, the transponders
and also the digital signal processing units require a large amount of power because of the high-complexity MIMO
equalizers. This makes them less suitable for the most relevant scenario for these solutions, i.e., short links. The last
solution, MB, aims at exploiting spectral bands beyond the C-band, in order to maximize the return on investment of
already deployed optical infrastructure which, in the case of Europe, consists almost entirely of optical cables without
water peak [8]. This is showed in Fig. 1(b), where the five available bands (O, E, S, C, and L) present an attenuation
coefficient below 0.4 dB/km. Naturally, MB still presents numerous challenges, as discussed in next section.
3. Multiband approach, technology requirements and challenges
Fig. 1(b) reports the attenuation and dispersion over wavelength of two exemplary fiber types: ITU-T G.652A (red)
and G.652D (blue). The low-loss of G.652D across all bands (i.e. 1260 - 1625 nm), could be exploited to increase the
transmission bandwidth by several tens of THz up to 365 nm compared to C-band only systems [9]. Such a solution
would be highly competitive against MF, MC and MM, in particular for metro and regional distances, where fiber losses
(even in O-band) should be tolerable at the receiver. Fig. 1(c) displays a simplified block diagram of a MB system for
a point-to-point link, such as the ones for data-center interconnection. In this scenario, we envision that all network
components must be developed. More precisely, the following elements must be covered: (I) the transceiver (laser,
modulator, front-end, etc.) must be realized (in Indium Phosphide - InP or Silicon Photonics - SiPh [10]) possibly with
the capability of being tunable over the entire spectrum. The DSP would likely remain (at least for short reach) similar
to the existing ones; (II) the amplifiers and filters must be designed by utilizing the best doped material for each band.
For instance, in Fig. 1(b) we assume a possible choice for the rare earth materials used within the amplification stage.
Other possibilities, such as Bismuth-DFA, Raman and semiconductor optical amplifiers (SOA) can also be considered.
MB enables a pay-as-you-grow upgrade strategy, so that operators can install the additional bands only when required.
Finally, MB transmission will need a sophisticated software-defined network (SDN) platform to dynamically and
efficiently manage the enormous bandwidth. Such SDN platform requires a physical layer controller that provides
an accurate yet quick quality-of-transmission estimator (QoT-E) that enables exploiting the differently performing
bands [9] for transmitting, e.g., different classes of services. Such a QoT-E will need a feasible transmission model
generalizing the ones that have been developed for the C-band. A first step in this direction has been presented in [11].
4. Conclusions
We discussed the great potential of MB transmission. We compared this solution against the ones commonly proposed,
by evaluating the pros and cons. We conclude that MB transmission is a viable and realistic solution for the imminent
deployment of next generation wireless infrastructure, although specific developments are still required.
Acknowledgment
A. Napoli would like to thank the German BMBF and grant no. 16KIS0487K (Celtic project SENDATE-FICUS) for partially funding this work. N.
Costa, J. Pedro, J. K. Fischer and N. Calabretta would like to thank the EU 5G-PPP project METRO-HAUL grant no. 761727.
References
1. A. Napoli et al. “Next generation elastic optical networks: The vision of the European research project IDEALIST,” IEEE Comm. Mag.,
vol. 53, no. 2, pp. 152-162, 2015.
2. V. López and L. Velasco. “Elastic Optical Networks,” Springer International Publishing:, 2016.
3. N. Sambo et al. “Next generation sliceable bandwidth variable transponders,” IEEE Comm. Mag., vol. 53, no. 2, pp. 163-171, 2015.
4. https://www.cisco.com/c/en/us/solutions/collateral/service-provider/visual-networking-index-
vni/complete-white-paper-c11-481360.html
5. http://www.lightwaveonline.com/articles/2017/04/verizon-signs-1-05b-fiber-deal-with-corning-
to-support-wireless-broadband.html
6. J. van Weerdenburg et al. “138-Tb/s mode- and wavelength-multiplexed transmission over six-mode graded-index fiber,” JLT, 2018.
7. T. Kobayashi et al., ”1-Pb/s (32 SDM/46 WDM/768 Gb/s) C-band dense SDM transmission over 205.6-km...,” paper Th5B.1, OFC, 2017.
8. https://www.fibre-systems.com/viewpoint/getting-most-currently-deployed-optical-fiber-
infrastructure
9. S. Okamoto et al., ”5-band (O, E, S, C, and L) WDM transmission with wavelength adaptive modulation format allocation,” ECOC, 2016.
10. C. Doerr et al., ”O, E, S, C, and L band silicon photonics coherent modulator/receiver,” paper Th5C.4, OFC, 2016.
11. M. Cantono et al. ”Modelling the impact of SRS on NLI generation in commercial equipment...,” Paper M1D.2, OFC, 2018.
View publication stats
