The DOCSIS Effect: Can the Internet Handle it?

July 28, 2016
By Mattias Fridström, Telia Carrier - Internet traffic is increasing steadily, driving significant changes in network operations - both for service ...

By Mattias Fridström, Telia Carrier

Internet traffic is increasing steadily, driving significant changes in network operations - both for service providers and the Internet's backbone itself.

According to ACG Research, by 2018, 5% of bandwidth will be consumed by Internet-connected large-screen TVs. SImilarly, at the end of 2015, Sandvine reported that more than 70% of all Internet traffic consisted of streaming video and audio. As over-the-top (OTT) 4K video, virtual reality and 5G mobile networks bring rich content into homes, cable and Internet providers are implementing gigabit services to address this demand and increase speeds to their customers.

As a result, Comcast (NASDAQ:CMCSA) is currently deploying DOCSIS 3.1-enabled modems and gateways along with its fiber-based Gigabit Pro fiber-to-the-premises (FTTP) services. The MSO is not alone in this endeavor, as its competitor AT&T (NYSE:T) is also employing copper-enhancing technologies such as, where it can serve multiple dwelling units (MDUs) from single, external locations at gigabit speeds. More expensive but reliable fiber-based technologies like GPON (gigabit passive optical network), which powers AT&T's GigaPower network, drive the need for cost-effective and viable alternatives. Verizon's (NYSE:VZ) FiOS and Google Fiber (NASDAQ:GOOG) also provide bandwidth at gigabit speeds, albeit at higher prices than hybrid fiber-coax (HFC) or copper.

Additionally, cable alternatives to DOCSIS technology such as Distributed Access Architecture (DAA) and integrated DOCSIS/PON networks provide yet another set of alternatives for cable operators to pump more bandwidth to the curb. Coupled with centralized content delivery and cloud architectures that transport traffic in existing networks, these technologies will put more stress on the Internet's backbone (or core) as it tries to keep up with such rapid increases in edge access speeds.

Operator obstacles

MSOs, Internet service providers (ISPs) - access, edge, or otherwise - as well as content providers will face growing pressure to figure out how to deliver content to consumers more cost effectively and how to reduce churn from dips in quality of service with delay-sensitive applications. The needs of these providers usually follow a list of network characteristics listed below:

  • Bandwidth: In the gigabit era, scalability and future compatibility (or future-proofing) are vital as demand for 100G services are becoming more commonplace in business, public sector and industrial environments. With 4K video, Internet of things (IoT) and virtual reality becoming actual reality, communications service providers are betting big by investing in network infrastructure to meet bandwidth demands from end users.
  • Flexibility: The more diverse and dense a network is, the better service providers can offer reliable services with minimal delays (usually in the tens of milliseconds) through mesh topologies and redundant routes.
  • Predictability: Outages will happen, and service providers want to experience the same service even when traffic is rerouted over a different path. Network builders must consider predictability when laying new fiber and designing their networks to be more efficient.
  • On-demand provisioning: Technology advancements in software-based provisioning have transformed the way networks are managed and maintained; service providers can simplify processes, improve operational efficiencies, speed deployment of services and pass on the savings to end users.
  • Low latency: For content providers, gaming companies and cloud service providers, minimizing roundtrip time, or latency, limits wait times for downloads, buffering and load times, thus improving the overall experience for the viewer and reducing churn.

With the global average Internet connection speeds currently at or around 5.6 Mbps (Akamai: State of the Internet 2015), how will the backbone address these needs while access technologies are continuing to widen the pipes to gigabit speeds directly to the curb?

What's under the ground matters most

While the most often deployed fiber-based networks include some combination or hybrid of mesh, star or ring topologies, the fact is that the more fiber in the ground to add redundancy to the backbone, the more reliable and predictable the network will be, regardless of the topology employed. More fiber in the ground designed to emulate the fastest route means that in the event of an outage on route A, the service provider can expect service on route B to be nearly identical to the route it's backing up, therefore minimizing any decrease in service quality and improving the experience for end users. One thing to note is that fiber, although robust and durable for the most part, is vulnerable to breaks, hence the need for backup fiber to be laid in a manner that emulates the original route. With more fiber in the ground, service providers will be able to predict the service quality that comes with a network built smarter.

Given the move towards gigabit services, what will operators and carriers do to address the impending network stress caused by the tremendous amount of bandwidth being accessed in homes throughout the country? Network infrastructure builders, those who lay fiber, sell wholesale services and provide the equipment to light those networks, use modern innovative designs and techniques so that service providers can assure their customers a seamless experience. Service providers should therefore work with the wholesale carriers much more to innovate to minimize outage delays and optimize traffic through redundant routes, mesh topology and more fiber in the ground.

Mattias Fridström is the chief evangelist of Telia Carrier.