2019 Broadband Access Market Outlook: Part 3 of 4

Very few subscribers will ever praise their service provider’s customer service. Improving customer service has always been a challenge for major operators, regardless of the service. But with advances in AI and tools for network automation, operators can fundamentally change the customer experience. Major operators–including AT&T Inc., Comcast, and Bell Canada–will implement a combination of AI and network automation as part of their overall network virtualization efforts.

2019 will see the roll out of AI capabilities for the express purpose of providing proactive data management as operators move toward full network automation. For operators, the following use cases and applications will see the light of day this year:

• Machine learning and proactive network maintenance, or PNM, to regularly poll active electronics and detect and troubleshoot issues before they impact end subscribers.
• Software-defined networking (SDN) and network functions virtualization (NFV) will provide far more visibility into individual and collective network elements. They will proactively alert network operations centers (NOCs), about potential hiccups.
• Virtualization of customer premises equipment (CPE) to allow for customer self-installation and management without contacting customer service to initiate a new service.
• Monitoring of traffic flows and data usage combined with algorithms to automate speed boosts when customers increase their data consumption.

Comcast, via its X1 service and platforms, is beginning to roll out a number of these capabilities in its effort to reduce truck rolls and improve the overall customer experience. 2019 will see the expansion of these efforts across a much broader range of network operators, including AT&T, Verizon Communications Inc., NTT DOCOMO, British Telecom, and others.

In Part 4, I will examine the ways that operators continue to push further into the home.

Related blogs:

• 2019 Broadband Access Market Outlook: Part 1 of 4 – Marco Trends
• 2019 Broadband Access Market Outlook: Part 2 of 4 – Cable Operators Use Distributed Access as Their Platform for Edge Computing and Virtualization

2019 Broadband Access Market Outlook: Part 2 of 4

2019 is expected to be the year that Comcast, Mediacom, and others in the North American market move from lab and field trials of their remote PHY and remote MACPHY or RPD and RMD platforms to general availability. Equipment vendors are ramping up production of their node units to meet what is expected to be a major year of deployments in 2019. We expect Comcast to pursue its goal of using distributed access architecture, or DAA, to dramatically reduce service group sizes from an average of 300 to 400 homes to less than 100. Other operators, including Cox and Spectrum, will quickly follow suit.

We’ve already seen a number of deployments in Europe, particularly at Com Hem and Stofa, designed to help these cable operators roll out DOCSIS 3.1 services, while simultaneously moving away from their traditional, integrated CCAP platforms. Both operators face significant competition from fiber providers, so they view R-PHY as a stepping stone to either full duplex DOCSIS 3.1 or FTTH.

The move to DAAs sets the stage for cable operators not only to expand the bandwidth they can offer end customers by reducing service group sizes, but also to push more edge computing capabilities closer to subscribers. Current optical nodes are nothing more than layer 1 and layer 2 platforms focused primarily on converting radio frequency signals to optical and vice versa. RPD and RMD nodes introduce layer 3 capabilities, as well as a road map to edge computing for more localized media processing and decision-making for applications beyond high-speed internet. With these more intelligent nodes, cable operators can better deliver IoT and wireless services.

Managing an expanded network of intelligent nodes, however, will introduce new challenges, which cable operators hope to address by virtualizing their existing cable modem termination system, or CMTS, functions. By centralizing service orchestration and control, cable operators can potentially reduce the time to deployment for their new distributed infrastructure, as well as the operational costs associated with these new architectures.

For cable operators, virtualization brings scale. By virtualizing and distributing the data and control planes from previously centralized and self-contained hardware platforms, operators can ultimately rely on more generic equipment, while also preparing their networks for the anticipated deluge of traffic from IoT and 5G services.

In Part 3, I discuss operators’ implementation of early use cases for artificial intelligence and network automation.

Related blog:

• 2019 Broadband Access Market Outlook: Part 1 of 4 – Marco Trends

Typically, in the early part of the year, I survey the broadband landscape and identify what I believe will be the year’s top trends and technology choices that will result in new product introductions and new service rollouts around the world.

In this 2019 broadband access market outlook series, I first discuss the high-level business and technology trends that will have an impact on network operators. Next, I describe the anticipated trends in broadband access and home networking. Finally, I review how two specific trends—edge computing and virtualization–address the larger business challenges, such as managing an expanded network of intelligent nodes. Here are the posts in my series:

• Part 1: Marco Trends
• Part 2: Cable Operators Use Distributed Access as Their Platform for Edge Computing and Virtualization
• Part 3: Operators’ Implementation Early Use Cases for Artificial Intelligence and Network Automation
• Part 4: Operators Continue to Push Further into the Home

---

Part 1 of 4: Macro Trends

The margins provided by the services network operators will continue to be squeezed by increased competition, market saturation, and pressure from OTT and cloud companies, which can often distribute similar services and applications at a reduced price. As a result, network operators will continue to reduce the cost of delivering services by:

• Moving more layer 3 functionality from dedicated access platforms and routers to servers in the cloud.
• Automating the provisioning, management, and troubleshooting of all network services; and using artificial intelligence to anticipate faults and other network-related issues.
• Increasingly moving away from proprietary hardware platforms in their access, edge, and core networks toward COTS hardware and servers with media gateway functionality.

Along those lines, a growing number of operators will divest their network assets and focus solely on content creation and distribution, while maintaining close relationships with underlying network operators to give them a competitive advantage over other OTT providers. Companies such as Crown Castle, American Tower, and other infrastructure providers are potential candidates for purchasing and managing network assets.

Those operators that don’t split out their network assets will continue to focus on the distribution of network platforms, intelligence, and workload processing. They will move away from centralized headends, data centers, and central offices to edge facilities, including hub sites, remote nodes, and modular edge data centers. Minimizing latency will become just as important to users as total broadband throughput. Operators that can deliver on both will secure the most subscribers and the highest revenue.

One ongoing challenge facing operators as they distribute more active electronics into their access networks is the scarcity of qualified labor to install, provision, and manage those platforms. Because of this labor shortage, operators will have to make smart and efficient decisions when it comes to deploying their technicians to install and provision these advanced platforms.

Though this is not an exhaustive list, I believe that these macro trends are under discussion among executives at many network operators. Operating a network is an extremely capital-intensive business. With the continued influx of software-based networking principles from the IT domain, data centers, and CDNs, service margins can certainly improve. Time will tell how quickly network operators embrace these principles throughout their access, edge, and core networks.

In Part 2, I discuss how cable operators use distributed access as their platform for edge computing and virtualization.