When someone makes a phone call or joins a video chat, streams a favorite show, orders products online or collaborates with co-workers remotely, the information that enables these interactions is comprised of digital data – zeros and ones. It may originate anywhere on the globe and connects people over telecommunications networks.
A fundamental infrastructure that enables data to move around is optical networks. There are primarily three physical ways to transmit data: copper cabling, which debuted for the telegraph in the 1820s and is still used today in more advanced forms; radio transmission, first used by Marconi in 1901; and fiber optics, utilized since 1977, when General Telephone and Electronics (GTE) sent the first live telephone traffic through fiber optics in Long Beach, California.
Each has its benefits: Copper cabling is low cost and thus pervasive in homes and small offices. Radio enables recipients to be untethered from a physical, hard-wired connection, providing locational freedom and mobility. The benefit of fiber optics is its ability to transmit massive amounts of data over practically any distance. As a result, the services and applications upon which people rely are, at some point, carried over an optical network.
Optical Networks Connect The World
Optical networks are based on the use of fiber optic glass strands, each no thicker than a human hair, that can transmit light pulses and thus information. These fibers are clad for physical protection from the elements and to facilitate being handled. They have been buried underground or mounted on aerial lines across neighborhoods, along highways and under oceans for more than four decades, forming an increasingly interconnected web of connectivity.
No matter which access methods connect end users – radio or microwave signal, copper or coax cable or fiber to the home – optical networks provide the most scalable and economical means of transferring massive amounts of user data over distance: within and between data centers; across neighborhoods, cities and countries; and between continents.
Initially, the transmission of data over fiber optics was done using lasers that turned “on” and “off” to send pulses of light through the glass, which is called on-off keying (OOK). Each pulse represented either a one or a zero – the binary data bits that encode all modern communications, applications and connectedness. The practical speed limit for OOK transmission was achieved at a data rate of 10 gigabits per second (Gbps) per wavelength in the late 1990s.
During the same time frame, wavelength division multiplexing (WDM) was developed to allow multiple different colors, or wavelengths, of light to travel through the same fiber. Thus, a single fiber, instead of carrying one channel on one color, could carry multiple channels on multiple, different colors. This allowed network operators to keep scaling the total information-carrying capacity of their optical networks beyond 10 Gbps, achieving capacities of 400 Gbps using 40 wavelengths, each carrying its stream of data.
The use of WDM was complemented by the development of optical amplifiers, which significantly extended the distance that the light pulses could travel before they needed to be converted back into the electrical domain to be retransmitted, if necessary, which saved both cost and power.
This brings up the fundamental physics of transmitting data along an optical fiber and just how much is possible. The ultimate scaling performance of any media is defined by the Shannon Limit – the maximum rate of error-free data transmitted across a given communications channel. It was developed by Claude Shannon at Bell Labs in 1948.
Coherent Transmission Enables Scaling
During the last decade, a new technology called coherent transmission made enormous scaling in optical network transmission possible. Combining the technology of digital signal processors (DSPs) with the application of phase modulation techniques (also common in mobile networks) to optical signals, coherent transmission technology enabled a further 100-fold increase in wavelength speed over what was possible with OOK technology since it was first introduced a decade ago.
Coherent transmission has provided the groundbreaking leap to what became the de facto optical transmission technology. It enables today’s optical networks to operate with a capacity of one terabit per second (Tbps) per wavelength and within a hair’s breadth of the Shannon Limit.
New developments in coherent optics continue to provide more scaling and lower total cost of ownership (TCO) of optical networks. Some advances include miniaturizing coherent optics into highly integrated digital coherent optic (DCO) transceivers, and a push to design and optimize coherent transceivers for different network applications.
The evolution to DCO transceivers is a result of the ever-increasing operating speeds of coherent optics, and the miniaturization of the optical components and integrated circuits used in their design. As optical transmission technologies progress, the optical components that transmit and detect the light, such as lasers and detectors, continually miniaturize to the point that they can fit into ever-smaller, lower-power, pluggable packages.
This miniaturization and integration of all constituent parts used in coherent optics – the DSP, the high-speed optoelectronics and the driver and amplification electronics – is now necessary for high-speed operation because connecting distinct devices mounted across a circuit board is no longer possible. The entire coherent engine must now be carefully co-designed and miniaturized into a DCO transceiver to maintain signal integrity at ever-higher wavelength speeds.
Optimizing Coherent Solutions
The second industry trend recognizes that there is no one-size-fits-all optimization for coherent transceivers. High-performance DCOs are optimized to deliver the next generational advance of faster baud rates, ever-higher wavelength speeds that will soon exceed 1 Tbps, and the maximum capacity-reach performance over distances of thousands of kilometers needed for long-haul and subsea networks.
The other development in the evolution of coherent optics is the introduction of DCOs optimized for low power and built into industry-standardized pluggable transceiver form factors that can be used in routers, switches and access platforms. The advantage of pluggable transceivers is that only a simple plug replacement is required if a transceiver fails, not a restoration of the whole line card. This allows the platforms they are used in to interchangeably use a range of optics with different characteristics for different capacity and distance applications without having to design specific line cards for each.
Examples of new pluggable DCO optics include router-pluggable coherent modules operating at 400 Gbps used for metro data center interconnection (DCI) and metro aggregation, and metro core networks. These will soon be complemented by pluggable DCOs operating at 100 Gbps and 200 Gbps optimized for access/edge applications such as PON, 5G fronthaul, cable DAA backhaul and campus WANs. These optimizations enable the cost-effective, modular deployment of coherent optics across an ever-broader range of use cases in access and metro applications.
These latest trends in coherent optics recognize network operators’ need to simultaneously scale the capacity and performance of their optical networks and to optimize for the needed capacity and lowest TCO across the wide range of network use cases from access and metro networks to the most challenging long-haul and subsea links.
Scaling Optical Networks
With today’s coherent optics now enabling operation within a decibel (dB) of the Shannon Limit, further evolution of optical networks will require further innovations to help networks scale, simplify operations and continue meeting growing bandwidth needs into the future.
One example of this ongoing evolution is to utilize more of the spectrum available in optical fibers, akin to transmitting across ever more radio channels. In the case of optical networks, this means expanding the range across which WDM technology can operate in the fiber spectrum to transmit more wavelengths. Until recently, WDM allowed up to 80 wavelengths to be transmitted in the C-Band of the fiber spectrum. This is now being doubled by extending the usable fiber spectrum into the L-Band, adding another 80 wavelengths of capacity. The potential exists to extend that even further, into what is called the S- and O-Bands of the optical fiber spectrum.
Other promising areas being explored to unlock even more capacity are new designs for optical fibers and cables so when network operators build out new fiber routes, the infrastructure they deploy will be scalable for decades into the future. Examples include optical fibers with narrower diameters, which allows more of them to be packed into a given cable size.
Complementing that is research into spatial division multiplexing (SDM) and multicore fibers, which will enable even denser cables capable of supporting many more fiber paths that can be deployed in simpler, more cost-effective ways. Organizations such as Bell Labs are at the forefront of driving research into these new, exciting areas, and the advances that result will continue to find their way into new products and solutions that will enable the networks of the future.
Serge Melle is the director of IP-optical product marketing for Nokia.
Serge Melle
Comments
Read what others have to say, and share your own thoughts with the community.