AT&T is set trial 10-gigabit symmetric passive optical network technology (XGS-PON), tapping its growing virtualization and software expertise to drive down the cost of next-generation PON deployments. <\/p>\n
The carrier said it plans to later this year conduct the XGS-PON trial as part of its plan to virtualize access functions within the last mile network. Testing is expected to show support for multi-gigabit per second Internet speeds and allow for merging of services onto a single network. Services to be supported include broadband and backhaul of wired and 5G wireless services. <\/p>\n
Eddy Barker, assistant vice president for access architecture and design at AT&T, said the carriers goal was to develop a more cost-efficient network platform to support growing demand. <\/p>\n
In working on next-generation PON, we have focused on trying to get the economics to where we are with GPON, Barker said. A big aspect is just the equipment costs and more significantly the silicon and optics costs. <\/p>\n
Barker explained the XGS-PON technology is a fixed wavelength symmetrical 10 Gb\/s platform. Compared with traditional gigabit PON (GPON), Barker said XGS-PON provides up to four-times greater downlink bandwidth and up to eight-times greater uplink bandwidth capacity. <\/p>\n
The trial is to take place in at least two locations, with Barker stating a likely different focus for each location. <\/p>\n
Some will have greenfield and some will have brownfield elements, Barker said. This will allow for strict XGS trials and then some interoperability trials with current GPON. <\/p>\n
The carrier is looking to expand the XGS-PON deployment into the cloud with software. <\/p>\n
AT&T said it has worked with ON.Lab to develop and test Open Network Operating System (ONOS) and Virtual Optical Line Terminator Hardware Abstraction (VOLTHA) software to hide the lower level details of the silicon. <\/p>\n
AT&T said it was waiting approval on submissions of open white box XGS optical line terminal (OLT) designs to the Open Compute Project (OCP). <\/p>\n
Barker said his team has had a big focus on using software-defined networking (SDN) and network functions virtualization (NFV). However, a bigger challenge was in building a community of suppliers and service providers that are part of ON.Lab and building open source code mirroring what has traditionally been proprietary access micro services that have run on supplier access equipment. <\/p>\n
We are now in the latter part of the development with ON.Lab, Barker said, noting some of that work has included the organizations Central Office Re-Architected as a Data Center (CORD) initiative.We have been trying to bundle up the access components of CORD, he said. Its not that we plan to do it in a turnkey manner as in ON.Lab, but so we can disaggregate it and use parts with what we have already done within AT&T independently of CORD. <\/p>\n
As an example of those plans, Barker said AT&T would use the Linux Foundations Open Network Automation Platform (ONAP) instead of the CORD XOS operating system. This was done because CORD was missing the emphasis on SDN control and virtualization in the access piece, Barker said. <\/p>\n
ONAP sprung from the merger of AT&Ts ECOMP and the Linux Foundations Open-O project. <\/p>\n
Barker said the OCP submissions are part of the carriers open access strategy, and led to the creation and sharing of the OpenOMCI specification. That spec is designed to provide an interoperable interface between the OLT and home devices, and has been distributed to the open source community. <\/p>\n
We are shooting for a common control system that is reusable across multiple access technologies, Barker said. A few things are specific, but we want to reuse as much as possible. <\/p>\n
AT&Ts XGS-PON plans are in contrast to competitor Verizon, which last month commissioned Calix to demonstrate channel bonding using next-generation passive-optical network two (NG-PON2) technology. <\/p>\n
The trial used a software platform sitting in a SDN environment to combine transport channels over a single fiber strand to support speeds up to 80 Gb\/s. <\/p>\n
We believe channel bonding holds the potential to more than double the bandwidth to individual subscribers or network locations and anticipate it could be a means of moving from 10 Gb\/s to 20 Gb\/s and beyond without deploying new technologies, said Vincent OByrne, director of access technology at Verizon, connected with the Calix trial. <\/p>\n
Verizon earlier this year said it completed an interoperability trial of NG-PON technology at its lab in Massachusetts. Vendors involved in the trial includedAdtran,Broadcom, Cortina Access, andEricssonin partnership with Calix. Verizon used its open optical network terminal management and control interface specifications for the trial, which the carrier said it plans to share with the industry. <\/p>\n
AT&Ts Barker said the carrier has looked at NG-PON2, but felt the financial model around XGS-PON was more compelling. Barker added that AT&T may ultimately decide it needs more capacity and go with NG-PON2, though the ongoing evolution of PON could override that need. <\/p>\n
We can see going forward if we need to go there or if we jump over NG-PON2 with the next iteration, Barker said. Standards bodies are working today on next-generation plans that can support up to 100 GB\/s, so we will continue to see where we need to go. <\/p>\n
Dan Meyer is a Senior Editor at SDxCentral, with a focus on containers, lifecycle service orchestration, cloud automation and DevOps. Dan has been covering the telecommunications space for more than 17 years. Prior to SDxCentral, Dan was Editor-In-Chief at RCR Wireless News. <\/p>\n
<\/p>\n