Raising the optical transport goal
A prime goal of optical networking has long been to put more information on a strand of fiber. That’s why technologies like wavelength division multiplexing (WDM) have been so important — they increase capacity of precious optical fiber spans. In time, the technology evolved to include complex techniques for multiplexing many wavelengths, higher modulation rates, faster symbol rates and coherent optical detection- all of which yielded more capacity on that fiber span.
Today, network operators are stretching the bounds of technology, but they are also changing basic requirements. End applications have evolved beyond simple voice and textual communication to sensing and controlling the physical processes of Industry 4.0. The goal is not only for higher line capacity and spectral efficiency, but also improved system scalability, latency and economy.
Nokia’s 1830 PSS photonic switching system is built upon decades of learning and patented invention. The Photonic Service Engine (PSE) is the DSP at the core of Nokia’s optical transport system’s power. PSE-3 implements probabilistic constellation shaping (PCS) to maximize capacity over any distance and on any fiber — from data center interconnect to subsea spans.
A concept pioneered by Nokia Bell Labs, PCS pushes optical performance towards the Shannon limit, the maximum possible information transfer rate. Engineered with the only algorithm proven to approach the Shannon limit, it combines advanced electronics, photonics and software to reach performance not seen before.
PCS allows optical networking systems to adapt to the physical realities of each optical span. By controlling the modulation format to optimize both capacity and reach for each span, network designs can be simplified. In some cases this requires fewer regeneration sites or better utilizes each wavelength. In other cases, better span reach and capacity result in lower latency as shorter end-to-end routes can be engineered to minimize total fiber length.
Earlier this year, Germany’s M-net trialed PCS. It allowed them to exceed 500Gbps over existing optical spans, maximizing the capacity of their network. Similar trials on Netia’s production networks in Poland demonstrated the ability to maximize span reach and capacity with the 600Gbps-capable PSE-3. And perhaps the highest bar was a demonstration with Etisalat in the UAE, where the world’s first 1Tb/s, single carrier transmission was tested over a deployed fiber network. PSE-3 made it possible to operate each optical carrier at 100 Gigabaud for a total fiber capacity of 50.8 terabits per second across multiple wavelengths operating simultaneously on a 93km span.
These results show promise for all network operators, from metro service providers maximizing shorter reach capacity to subsea cable operators seeking to optimize reach and minimize system latency. Research networks can realize truly hyper-scale DCI through the 25 percent shaping gain realized with super-coherent PCS. With PSE-3 built into the Nokia 1830 PSS and 1830 PSI-M solutions, network operators of all types will be able to economically build networks that connect, literally, any end application.
How does PCS work?
Conventional modulation formats result in a performance gap to the Shannon Limit because Shannon’s theorem is based on a Gaussian noise model whereas standard modulation formats are based on square constellation patterns. By focusing energy on lower amplitude symbols as the data rate decreases, probabilistic shaping transforms a square constellation pattern into a more Gaussian-like shape. This enables tailoring the transmission rate to the special optical channel delivering up to 30 percent greater reach while also significantly improving spectral efficiency.
Share your thoughts on this topic by joining the Twitter discussion with @nokianetworks using #optical #opticalnetworking #PSE3 #PCS