Ciena has unveiled the next-generation of its programmable coherent optical chipset, dubbed WaveLogic Ai. The letters stand for autonomous and intelligent, to underline how the powerful combination of new hardware and software features could unlock network automation.
Like its predecessor WaveLogic 3, the new chipset is designed to maximise bandwidth and reach so that service providers can best match service throughput to line capacity. Compared to the previous product generation, Wavelogic Ai enables twice the capacity per channel, three times the distance at the equivalent capacity, and four times the service density at less than half the power, the company claims.
For metro and data centre interconnect applications, the modem drives 400G single-carrier transmission using an optimised design and higher baud rate. In regional and long-haul networks, WaveLogic Ai establishes 200G and 300G as the new reference line rates for backbone transmission, dramatically lowering cost per bit. And in submarine networks it provides maximum capacity with ultimate reach at distances up to 14,000km.
To better match capacity to system margin, WaveLogic Ai provides unprecedented tunability, and is the industry’s first coherent modem that can tune capacity from single carrier 100G to 400G in 50G increments, according to Helen Xenos, director of portfolio marketing at Ciena.
But programmable modulation formats are only part of the story. The automation features are where Ciena hopes the new technology will stand out.
“Our industry has been talking about the need for intelligent optical networks for many years, and the industry push around SDN and open networks brought us closer to that reality. WaveLogic Ai denotes a move in this area from conversations to implementation,” said Xenos via an article accompanying the release.
Variable-line-rate coherent transceivers provide flexibility. But if operators cannot get accurate, real-time link data from the network to determine the right channel capacity rate at any point in time, they cannot take advantage of the savings associated with the new technology, she says.
To this end, WaveLogic Ai probes the incoming signal, gathering a range of real-time measurements that are updated every 10ms, including: transmit and receive optical powers, polarisation channel characteristics, electronic chromatic dispersion compensation map, error rates and conditions, total and constituent electrical signal-to-noise ratios (SNRs), and latency.
With the embedded instrumentation in WaveLogic Ai, service providers can now understand exactly how much margin is currently present in the network, as well as the optimal capacity they can deploy, explains Xenos. Through the open software interfaces, they can mine the data and use it to accurately engineer their network for optimal capacity and maximum efficiency.
These capabilities are crucial to research and education network operator Esnet. Inder Monga, executive director, ESnet had this to say: “With exponential growth in science traffic from instruments and facilities around the world, our network needs to be able to adjust in real-time to meet the needs of data-intensive applications and scientific research. In the past, decisions were often made using best-guess or anecdotal information and handled through over-provisioning. We now have a need to use real-time network telemetry and analytics to intelligently engineer the underlying infrastructure for optimal capacity.”
Ciena says WaveLogic Ai will be implemented across its portfolio, starting with the Waveserver and 6500 platforms. Customer trials of WaveLogic Ai are expected to begin during the first half of 2017, with global availability in the second quarter of 2017.
