Feature - Part 2: Optical networking R&D
Predicting the future is a foolhardy endeavour, at best one can make educated guesses.
Ioannis Tomkos is better placed than most to comment on the future course of optical networking. Tomkos, a Fellow of the OSA and the IET at the Athens Information Technology Centre (AIT), is involved in several European research projects that are tackling head-on the challenges set to keep optical engineers busy for the next decade.
“We are reaching the total capacity limit of deployed single-mode, single-core fibre,” says Tomkos. “We can’t just scale capacity because there are limits now to the capacity of point-to-point connections.”
Source: Infinera
The industry consensus is to develop flexible optical networking techniques that make best use of the existing deployed fibre. These techniques include using spectral super-channels, moving to a flexible grid, and introducing ‘sliceable’ transponders whose total capacity can be split and sent to different locations based on the traffic requirements.
Once these flexible networking techniques have exhausted the last Hertz of a fibre’s C-band, additional spectral bands of the fibre will likely be exploited such as the L-band and S-band.
After that, spatial-division multiplexing (SDM) of transmission systems will be used, first using already deployed single-mode fibre and then new types of optical transmission systems that use SDM within the same optical fibre. For this, operators will need to put novel fibre in the ground that have multiple modes and multiple cores.
SDM systems will bring about change not only with the fibre and terminal end points, but also the amplification and optical switching along the transmission path. SDM optical switching will be more complex but it also promises huge capacities and overall dollar-per-bit cost savings.
Tomkos is heading three European research projects - FOX-C, ASTRON & INSPACE.
FOX-C involves adding and dropping all-optically sub-channels from different types of spectral super-channels. ASTRON is undertaing the development of a one terabit transceiver photonic integrated circuit (PIC). The third, INSPACE, will undertake the development of new optical switch architectures for SDM-based networks.
Tomkos’s research group is also a partner in three other EU projects. One of them - dubbed ACINO - involves a consortium developing a software-defined networking (SDN) controller that oversees sliceable transponders.
These projects are now detailed.
FOX-C
Spectral super-channels are used to create high bit-rate signals - 400 Gigabit and greater - by combining a number of sub-channels. Combining sub-channels is necessary since existing electronics can’t create such high bit rates using a single carrier.
Infinera points out that a 1.2 Terabit-per-second (Tbps) signal implemented using a single carrier would require 462.5 GHz of spectrum while the accompanying electronics to achieve the 384 Gigabaud (Gbaud) symbol rate would require a sub-10nm CMOS process, a technology at least five years away.
In contrast, implementing the 1.2 Tbps signal using 12 sub-channels, each at 100 Gigabit-per-second (Gbps), occupies the same 462.5 GHz of spectrum but could be done with existing 32 Gbaud electronics. However, instead of one laser and four modulators for the single-carrier case, 12 lasers and 48 modulators are needed for the 1.2 Tbps super-channel (see diagram, top).
Operators are already deploying super-channels on existing networking routes. For example, certain 400 Gbps links use two sub-channels, each a single carrier modulated using polarisation-multiplexed, 16 quadrature amplitude modulation (PM-16-QAM).
Meanwhile, CenturyLink was the first operator, in the second quarter of 2012, to deploy a 500 Gbps super-channel using Infinera’s PIC. Infinera’s 500 Gigabit uses 10 sub-channels, each carrying a 50 Gbps signal modulated using polarisation-multiplexed, quadrature phase-shift keying (PM-QPSK).
There are two types of super-channels, says Tomkos:
- Those that use non-overlapping sub-channels implemented using what is called Nyquist multiplexing.
- And those with overlapping sub-channels using orthogonal frequency division multiplexing (OFDM).
Existing transport systems from the optical vendors use non-overlapping super-channels and Optical Transport Networking (OTN) at the electrical layer for processing, switching and grooming of the signals, says Tomkos: “With FOX-C, we are developing techniques to add/ drop sub-channels out of the super-channel without going into the electronic domain.”
Accordingly, the FOX-C project is developing transceivers that implement both types of super-channel, using non-overlapping and overlapping sub-channels, to explore their merits. The project is also developing techniques to enable all-optical adding and dropping of sub-channels from these super-channel types.
With Nyquist-WDM super-channels, the sub-channels are adjacent to each other but are non-overlapping such that dropping or adding a sub-channel is straightforward. Today’s 25 GHz wide filters can separate a sub-channel and insert another in the empty slot.
The FOX-C project will use much finer filtering: 12.5GHz, 6.25GHz, 3.125GHz and even finer resolutions, where there is no fixed grid to adhere to. “We are developing ultra-high resolution filtering technology to do this all-optical add/drop for Nyquist multiplexed sub-channels without any performance degradation,” says Tomkos. The FOX-C filter can achieve a record resolution of 0.8GHz.
OFDM is more complicated since each sub-channel interacts with its neighbours. “If you take out one, you disturb the neighbouring ones, and you introduce severe performance degradation,” says Tomkos. To tackle this, the FOX-C project is using an all-optical interferometer.
“Using the all-optical interferometer introduces constructive and destructive interference among the OFDM sub-channels and the sub-channel or channels we want to drop and add,” says Tomkos. “By properly controlling the interferometer, we are able to perform add/ drop functions without performance degradation.”
ASTRON
The second project, ASTRON, is developing a one terabit super-channel PIC. The hybrid integration platform uses planar lightwave circuit (PLC) technology based on a glass substrate to which are added the actives: modulator arrays and the photo-detectors in indium phosphide. “We have kept the lasers outside the PIC mostly due to budgetary constraints, but there is no problem to include them also in the PIC,” says Tomkos. The one terabit super-channel will use eight sub-channels, occupying a total spectrum of 200 GHz.
The PLC acts as the integration platform onto which the actives are placed. “We use 3D waveguide inscription inside the glass using high-power lasers and flip-chip bonding to couple the actives to the passives inside the PIC,” says Tomkos.
The modulation arrays and the passives have already been made, and the project members have mastered how to create 3D waveguides in the glass to enable the active-passive alignment.
“We are in the process of finalising the technique for doing the hybrid integration and putting everything together,” says Tomkos.
The physical layer PIC is complemented by developments in advanced software-defined digital signal processing (DSP) and forward error correction (FEC) modules implemented on FPGAs to enhance the transmission performance of the transceiver. The working one terabit PIC, expected from October, will then be used for experimentation in transmission testbeds.
INSPACE
Spatial-division multiplexing promises new efficiencies in that instead of individual transponders and amplifiers per fibre, arrays of transponders and amplifiers can be used, spread across all the spatial super-channels. Not only does the approach promise far higher overall capacities but also lower cost.
The introduction of bundled single-mode fibres, as well as new fibers that transmit over several modes and cores within such SDM systems complicates the optical switching. The channels will be less used for point-to-point transmission due to the huge capacities involved, and there will be a need to process and switch spatial sub-channels from the spatial super-channels. “We are developing a wavelength-selective switch that also operates at the spatial dimension,” says Tomkos.
Already it is clear there will be two main SDM switching types.
The first, simpler case involves spatial sub-channels that do not overlap with each other so that individual sub-channels can be dropped and added. This is the case using fibre with a few cores only, sufficiently spaced apart that they are effectively isolated from each other. Existing cable where a bundle of single-mode, single-core fibres are used for SDM also fits this category. The switching for these fibre configurations is dubbed independent switching.
The second SDM switch type, known as joint switching, uses fibre with multiple cores that are closely spaced, and few core multi-mode fibre. In these cases, individual sub-channels cannot be added or dropped and processed independently as their overlap causes crosstalk. “Here you switch the entire spatially-multiplexed super-channel as a whole, and to do so you can use a single wavelength-selective switch making the overall network more cost effective,” says Tomkos.
Only after dropping the entire super-channel can signal processing techniques such as multiple input/multiple output (MIMO), a signal processing technique already used for cellular, be used in the electronic domain to access individual sub-channels.
The goal of the INSPACE project is to develop a new generation of wavelength-selective switches (WSSes) that operate at the spatial dimension.
“The true value of SDM is in its capability to reduce the cost of transport through spatial integration of network elements: fibers, amplifiers, transceivers and nodes,” says Tomkos. If by performing independent switching of several SDM signals using several switches, no cost-per-bit savings result. But by using joint switching for all the SDM signals with the one switch, the hope is for significant cost reductions, he says.
The team has already implemented the first SDM switches one year into the project.
ACINO
The ACINO project is headed by the Italian Centre of Research and Telecommunication Experimentations for Networked communities (Create-net), and also involves Telefonica I+D, ADVA Optical Networking and Tomkos’s group.
The project, which began in February, is developing an SDN controller and use sliceable transponders to deliver different types of application flows over the optical network.
To explain the sliceable transponder concept, Tomkos uses the example of a future 10 terabit transponder implemented using 20 or 40 sub-channels. All these sub-channels can be combined to deliver the total 10 Tbps capacity between two points, but in a flexible network, the likelihood is that flows will be variable. If, for example, demand changes such that only one terabit is needed between the two points, suddenly 90 percent of the overall capacity is wasted. Using a sliceable transponder, the sub-channels can be reconfigured dynamically to form different capacity containers, depending on traffic demand. Using the transponder in combination with WSSes, the different sub-channel groupings can be sent to different end points, as required.
Combining such transponders with the SDN controller, ACINO will enable high-capacity links to be set up and dismantled on demand and according to the different application requirements. One application flow example is large data storage back-ups scheduled at certain times between an enterprise’s sites, another is backhauling wireless traffic from 5G networks.
Tomkos stresses that the key development of ACINO is not sliceable transponders but the SDN controller and the application awareness that the overall solution will offer.
The roadmap
So how does Tomkos expect optical networking to evolve over the next 10-plus years?
The next five years will see further development of flexible optical networking that makes best use of the existing infrastructure using spectral super-channels, a flexible grid and sliceable software-defined flexible transponders.
From 2020-2025, more of the fibre’s spectral bands will be used, coupled with first use of SDM. SDM could start even sooner by using existing single-core, single-mode fibres and combining them to create an SDM fibre bundle.
But for the other versions of SDM, new fibre must be deployed in the network and that is something that operators will find difficult to accept. This may be possible for certain greenfield deployments or for data centre interconnects, he says.
Only after 2025 does Tomkos expect next-generation SDM systems using higher capacity fibre with a high core and mode count, or even hybrid systems that use both low and high core-count fibre with advanced MIMO processing, to become more widely deployed in backbone networks.
