At the start of 2019, the cable industry announced its vision for delivering 10 gigabit networks, ramping up from 1 Gbps service offerings to symmetrical speeds of 10 Gbps and beyond while enhancing the customer experience and achieving operational efficiencies. Industry bodies, cable MSOs, and vendors are working together to address industry-wide challenges associated with moving to next generation networks. Moving forward, even more interaction may be necessary if we want to maximize the potential of these new networks - particularly around convergence.
Recently, I've had the opportunity to meet with many MSOs in North America and other regions to talk about one of my favorite topics, the Converged Interconnect Network, otherwise known as CIN. Some MSOs plan multi-service convergence in the CIN from the beginning, while others reserve the idea for future contemplation. For those considering service convergence "out of the gate," it must be capable of providing (or delivering) different revenue services such as residential, mobile backhaul (small cell and macro-cell) and enterprise connections - and this is independent of delivery systems such as R-PHY, R-MACPHY, Flexible MAC Architecture (FMA), and even PON. In many cases, MSOs outside of the United States also have telco services (e.g. mobile networks - LTE, 4G, moving to 5G) and are interested in creating an environment where the last tentacles of the network - the access network - can fully participate in the convergence of services to maximize operational efficiencies.
The CIN and Distributed Access Architecture(s) by design are not proprietary or single vendor solutions, and even less so as we move to achieve convergence of services. However, putting structure around the CIN is a new concept. Unlike analog hybrid fiber/coax (HFC) solutions, there is now a vast array of digital networking options to choose from, which could lead to simplicity and a reduced total cost of ownership - navigating toward best options is done collectively. For example, CableLabs has helped spearhead industry initiatives pertaining to the CIN, such as developing specifications for point-to-point coherent optics, which at 100 or 200 Gbps, over 80km DWDM can be a great general purpose DAA backhaul optical vehicle. CableLabs is also working on the framework for DOCSIS to be the backhaul and fronthaul solution for LTE and eventually 5G small-cell traffic. This initiative toward multi-service convergence on the CIN and can reduce complexity as mobile traffic is included in the DOCSIS transmission. Simultaneously, the SCTE is also working on the Generic Access Platform project with the goal to provide common housing for multi-service aggregation devices in the outside plant, which can simplify the physical integration of multiple services deeper in the fiber plant.
So far the effect of standards for the CIN is clear. They keep us on target - and as vendors collaborate well together, the rapid development of emerging technologies becomes a reality. Current CIN initiatives have shown great promise, but for the sake of service convergence there are additional opportunities MSOs and vendors can work together to address, including:
- Timing - the synchronicity of the network must track its most stringent participant - which will be the delay and jitter expectations for 5G mobile networks and latency critical applications. This is important in bringing the vendor community together because it will have hardware implications that ideally are addressed in initial product cycles of CIN deployments.
- Backhaul capacity modeling for a multi-service CIN infrastructure - there are well-understood models for the CIN capacity evolution in the case of a single service, such as residential services, but we also need capacity models for when the CIN will support multiple services. The industry needs to understand the additive effect on the network when carrying residential, mobile (in various forms) and business services (beyond static connections). This is important because it impacts the planning of fabric and optical scale, ultimately affecting cost and thermal footprint, both delicate items to CIN elements.
- Aggregation in the outside plant - as part of planning for multi-service convergence on the CIN, there are increasingly frequent discussions regarding outdoor aggregation devices such as aggregating traffic in the outside plant. Note, there might be no obvious fiber shortage or throughput capacity trigger today, but some quick back-of-the-envelope calculations of lambda usage, business services and mobile growth capacity can make the case for aggregation in the outside plant a more pressing consideration.
- Management of fiber assets - is a top of mind topic in many circles - what fiber is out there, what is on it, and how fiber ages. This is extremely important to the CIN as fiber is foundational to the capability. There is an opportunity for more discussion on this topic as well as a need to develop a description of the tools which can catalogue fiber assets in a consistent way, particularly as different sectors within an MSO rely on this asset.
- Finally, there is room to have a proactive discussion about best practices in protocol level aggregation of these distinct services through the access plant. Perhaps this is the opportunity to add the dimension of MPLS and segment routing to the access network, where the uniqueness of paths and flows could prove quite useful in a multi-service environment.
As the cable industry grapples with ways to come up with efficient uses of existing cable infrastructure in order to meet demand, there will be an increasing need for continued collaboration among industry bodies, the vendor community and providers. Working together in areas like the CIN is a proactive start toward maximizing the use of these new DAA cable networks.
Fernando Villarruel is chief architect, MSO Practice, at Ciena.