By Matthew Hayes and David Ririe

Embracing a PON Strategy in a Cable World

Cox first began residential PON based services in 2014. At that time we were looking for more than simply billboard speeds. We were looking to defend our backyard from new entrants. Cox took a competitive position to overbuild PON in a modest brownfield HFC territory. Along this journey Cox has now built over 300,000 HHP of FTTH across brown and greenfield areas. We have learned some valuable lessons along the way. Our intent is to share those lessons as well as promote the value and readiness of PON/FTTH as an access technology ready for prime time in the cable world.

Access Network of the Future

PON and other FTTH technology options are being used widely for greenfield builds in many, if not most, cable operators today. What does that tell us? Is it the network of the future? Almost certainly. Is it the network we desire? Yes. There are a few key challenges to get there. A couple of obvious concerns are the cost of the last mile install and that all services are not readily available to deliver over IP. In order to deliver all services over IP we require client capabilities in the home, a delivery mechanism to stream services in IP, and the backoffice to provision them. And yes, figure out how to pay for the cost of that last mile fiber install while not angering the customer base as we trench through their neighborhoods. Last mile means something quite different depending on if we are referring to brownfield or greenfield. Last mile in greenfield is a forgone conclusion, the only delta is what medium you are dragging with you. In brownfield the investment must be more surgically and strategically targeted toward high business park concentration, high value residential, or competitive builds, while we maintain a ubiquitous offering of gigabit service across our pervasive HFC network.

RFoG

In lieu of having all of the puzzle pieces available, many operators must also deploy RFoG (radio frequency over glass) for legacy video and voice service delivery. RFoG is only required in greenfield scenarios where a shadow HFC/coax network is not available. RFoG can represent a significant tax financially and in complexity of installation. The home requires a micronode to deliver RF over glass services and convert to coax inside the home. A micronode is just that, a small HFC node serving one home from the side of the house. This enables a traditional QAM-based STB to be connected inside the home. Of additional concern in delivering all services using RFoG is a condition known as optical beat interference (OBI). This interference occurs when multiple transmission signals in the upstream are received at the same time. OBI is very difficult to measure as it can only be observed in the optical domain at the access port upstream. OBI is real and is more likely to occur in serving groups that are delivering more bandwidth intense upstream services — for example, Internet, security camera uploads etc. Cox has observed performance degradation as a result of OBI in densities as low as 30 homes passed. In order to mitigate OBI there are some fairly expensive solutions out there that monitor upstream transmission and stagger potential collisions. These solutions add yet another premium to delivering all services over FTTH. Fortunately, we are right at the cusp of delivering all services over IP which will eliminate RFoG/OBI mitigation expenses, and provide a more elegant solution in customers’ homes. All-IP service delivery is the catalyst that will make PON a more mainstream technology in cable access because it puts the technology on a more equal playing field when considering long-term business cases with alternative approaches in DAA, R-PHY, mid-split, full duplex DOCSIS, etc.

ONT Deployment Strategy

An outdoor ONU is an important tool to have — many customers do not want you boring a hole in the house and tugging a new cable (fiber) to an indoor ONT. In many new developments they have prewired houses. Homes are typically prewired with Cat6. In some cases these land in a precanned wiring encasement in the garage. Technicians must be able and ready to support both an indoor ONU and and outdoor ONU. Be ready to work with the builder or homeowner on making that determination. An indoor ONU is a lower capital cost option since it won’t require the temperature hardened unit or a weather encasement. These can cost an additional $50-$100 per CPE.

All-in-One Integrated Gateway or Two-Box Solution

While the option for an all-in-one integrated ONU/Wi-Fi gateway seems attractive at first glance in terms of elegance and overall cost, it comes with some challenges. As described previously, the placement of the fiber termination in the home might not provide an optimal spot for a Wi-Fi gateway. Having a two-box solution allows the most flexibility to decouple ONU placement from the gateway and leverage the Cat6 wiring in the home. Additionally, since Wi-Fi standards are moving at a faster pace than PON technology, it would seem advantageous to decouple PON transport from Wi-Fi. Most broadband service innovation is happening in the gateway function with parental controls, time of day, malware protection, etc. Terminating voice and IPTV services on the gateway feel like a more permanent landing spot and will enable transport transition to 10G PON with little to no impact. As such we have deployed PON ONU simply as a transport function. The only capabilities we ask of the transport ONU is QoS classification and prioritization, IPDR for usage based metering, and lawful intercept.

OLT Deployment Strategy

One of the fundamental design considerations is placement of the OLTs. There are three primary deployment scenarios 1) OLT out of the headend 2) OLT in an outdoor cabinet 3) Remote OLT (strand mounted in clam shell). We began our deployment with Options 1 and 2. OLT out of the headend requires a great deal of bundled fiber homed back to the headend but works well for reach within 5,000 feet of the hub site. Depending on split ratios PON has a reach of approximately 20 km. Deploying an OLT in an outdoor cabinet relieves this burden by enabling you to place the OLT more strategically closer to customers and limit your fiber build to where it matters most. However, deploying outdoor cabinets comes with additional operating expenses for active units (power, cooling, filter maintenance, etc.). Deploying an OLT in a strand mounted clam shell proved to be the most cost effective and efficient deployment model. This approach enabled a more surgical placement of PON OLTs while avoiding the operational expense attached to temperature conditioned outdoor cabinets. There is a fourth deployment scenario that involves deploying an OLT embedded in an optic. So the deployment would simply go in some pre-existing Ethernet switch (hub site, cabinet, or strand mount). While this approach is interesting, it seems to fit more a niche commercial where serviceability and automated provisioning might not be table stakes.

PON Technology

There has been much debate and conjecture about PON technology and which technology has the greatest benefit, market adoption and long-term viability. This debate seems to get more airtime in the cable community. There are more similarities than differences in the fundamental function of the technology and roadmaps for speed. They are, however, completely incompatible. There have been proclamations in the cable community in support of DPoE and 10G EPON, however, those commitments have not resulted in many substantial deployments. Meanwhile, the rest of the world deploys GPON for any meaningful PON deployments. And clearly, China drives any global trends in PON. Even if we see substantial North American cable commitments toward 10G  EPON it will not overcome a direction China sets. It is our belief that GPON will remain the dominant PON technology deployed through 2021. A crossover point could be reached as early as 2022 where some combination of 10G EPON and XGS (10G GPON) will overtake GPON in new deployments. The bang for the buck is not lost in GPON today. And a GPON deployment should feel totally non-regrettable. That said, there are a couple of hurdles to overcome if you are a cable provider looking to deploy GPON. A few features native in DPoE based EPON are IPDR and lawful intercept (CALEA). As you get ready for PON prime time make sure you explore with your PON supplier support of these features. And ensure you understand what a 10G PON transition would look like on the platform. Do you have flexibility toward 10G EPON or XGS? Can you overlay a business services PON wave?

As cable begins deep fiber deployments for DAA consider FTTH and PON in targeted applications. The real gamechanger is ALL IP based services and what that means for total cost per HH. In greenfield you are paying for a drop regardless… may as well be FTTH. In brownfield, D3.1 + DAA/R-PHY have an immediate edge in ubiquity of service offering and lower upfront cost. Don’t delay, PON is ready for prime time in cable access. Learn how to execute and operationalize FTTH/PON deployment in greenfield because one day soon it will make up most of your network.

Matthew Hayes
Executive Director, Network Engineering
Cox Communications, Inc.
matt.hayes@cox.com

Matt began his career with Cox Communications as a Network Engineer in 2000. He has held various roles in network engineering leadership. Matt’s greatest satisfaction is growing engineers into next level leaders. Promoting leadership at every level. Matt’s secondary passion is influencing the network evolution toward ALL IP. His most current focus is leading the Access Engineering team toward DAA, R-PHY and ALL-IP services over PON. Matt holds a Masters of Science in Telecommunications from University of Colorado-Boulder.

David Ririe
Sr. Director, Access Engineering
Cox Communications, Inc.
david.ririe@cox.com

David Ririe serves as a Senior Director in the Access Engineering group for Cox Communications in Atlanta, GA. He has worked in various engineering roles for Cox Communications starting as a Network Engineer in the Omaha, NE market 14 years ago. He has lead the team through a number of transitions and upgrades on both the DOCSIS and PON technology platforms in that time. He began his telecom career in the U.S. Air Force working on various communications and data networking platforms in the air and on the ground.

Editorial credit: meunierd / Shutterstock.com