CableLabs unveils Point to Point Coherent Optics specifications for access networks

CableLabs, which develops and specifies technology for its cable operator members, released a pair of Point to Point Coherent Optics specifications June 29 that aim to enable coherent transmission capabilities for access networks. The specifications describe coherent optical transceivers that will support 100-Gbps coherent transmission based on dual-polarization differential quadrature phase-shift keying (DP-DQPSK) over at least 40-km unamplified links as well as potential use cases.

The need for coherent-enabled capacity arises as cable operators move to Distributed Access Architectures (DAAs) from their traditional centralized hybrid fiber/coax (HFC) approaches. DAAs see functions removed from the headend and located in remote nodes closer the customer. This architecture change simplifies the headend (and eases power and space requirements there), improves network performance, and offers opportunities for functional virtualization. (For more on DAAs, see this article on Lightwave sister site Broadband Technology Report.) Initially, the fiber connection between the revamped headend and the remote node will use 10 Gigabit Ethernet. However, operators realized that greater capacity eventually will be needed, and multiple wavelengths of 10 Gbps might prove inadequate, too costly, or both.

Enter coherent transmission – at access network prices. CableLabs and working group members began their specification development under the supposition that point-to-point access network links should require less stringent performance than the long-haul and metro networks where coherent transmission currently finds use. Access links should require less optical output power, reduced transmitter wavelength capability, and lower optical signal-to-noise ratio (OSNR), for example. They also should face reduced penalties from chromatic dispersion and polarization mode dispersion (PMD). As a result, less expensive components, as well as less robust digital signal processors (DSPs) and forward error correction (FEC), should do the trick. Curtis Knittle, vice president of wired technologies at CableLabs, pointed out in an interview that the resulting optical transceivers should require less power and cooling than their longer-reach counterparts as well.

Knittle didn’t have an estimate of how much less expensive the upcoming access-focused transceivers will be. But the specifications within the “P2P Coherent Optics Physical Layer 1.0 Specification (P2PCO-SP-PHYv1.0-I01-180629),” many of which derive from the ITU-T G.709 Optical Transport Network family of recommendations (including OTU4 framing and nominal bit rate), offer some hints. Examples include:

• Use of Hard-Decision Staircase FEC as described in ITU-T G.709.2
• Support of a transmitter optical output power of -6 dBm or higher; optical output power greater than or equal to +7 dBm is forbidden
• Support of transmitter OSNR of 35 dB or higher
• At the receive end, the received optical power baseline is a post-FEC BER of ≤10^-15 when the link OSNR is ≥35 dB and the received optical power is ≥-31 dBm.
• Support of a minimum of 2400 ps/nm of chromatic dispersion and at least 10 ps of PMD.

The specifications support both single fiber and dual fiber designs. Single fiber operation will become increasingly important to cable operators, Knittle said. He reported that a recent study indicates that 20% of current headend-to-node connections are single fiber – and operators expect that figure to increase to 60% in five years. Along these lines, the specifications say that the coherent optical transceiver must support transmit and receive via the same frequency. CableLabs has conducted experiments with a technique the organization calls Full Duplex Coherent Optics that leverages circulators to enable transmit and receive simultaneously on a single wavelength.

Coherent in use

Knittle said CableLabs expects to hold coherent transceiver interoperability demonstrations later this year, with the technology moving to field trials by the first quarter of 2019. The transceivers could be used in muxponders, Layer 2 switches, or routers, he said. The system employed likely will depend on the application. The “P2P Coherent Optics Architecture Specification (P2PCO-SP-ARCH-I01-180629)” describes a pair of use cases. The first, aggregation, sees a switch or muxponder in the remote node take in a coherent transmission from the headend and distribute that capacity among a variety of clients (either elsewhere in the node or in the field) via electrical or optical connections. The edge-to-edge use case describes an application in which multiple coherent wavelengths originate at the headend and are demultiplexed at the remote node for distribution to clients, again either elsewhere in the node or further in the field.

In both cases, the physical layer specifications are optimized for 40-km reaches without amplification; such a distance will cover 90% of expected requirements, the specification states. Knittle said that greater distances could be accommodated with the addition of amplifiers or transceiver capabilities beyond those described in the specifications.

Looking ahead, Knittle said that future versions of the specifications will tackle 200 and 400 Gbps transmission. Development of an Operational Support Systems Interface is in its early stages as well, he added.

For related articles, visit the FTTx Topic Center.

For more information on FTTx technology and suppliers, visit the Lightwave Buyer’s Guide.