The amount of money cable operators spend on broadband access equipment will decrease by 2% from 2019 to 2024, research company Dell’Oro Group predicts in its latest report.

Part of the cause, the firm said, is virtualization—operators will spend less on access hardware tech, as software virtualizes components like cable modem termination system (CMTS). In fact, Dell’Oro predicted revenue from virtual CMTS revenue will grow from $90 million last year to $418 million in 2024, but operators will spend less overall on equipment.

Meanwhile, Dell’Oro said that maturing broadband markets in the U.S. and Europe are also a factor, as is the pause operators are taking following wide-scale deployments of DOCSIS 3.1.

“Global operators continue to invest in their broadband access networks, but are increasingly delivering more capacity with lower-cost, virtualized hardware,” said Jeff Heynen, senior research director at Dell’Oro Group, in a statement. “Virtualization, coupled with subscriber saturation in some mature markets will result in gradually declining revenue for broadband access equipment globally.”

The core of every CATV network is the headend, which houses the active electronics behind every connected experience that business and residential subscribers rely on. All of the investments that multisystem operators (MSOs) make to prepare their outside plant infrastructure for emerging services and applications have an impact on the headend’s limited, high-value real estate. So, increasingly, MSOs are turning their attention to headend transformation, and specifically toward leveraging software-defined networks, network function virtualization, and cloud technologies to bring data center simplicity and cloud flexibility to their headend environments.

SDN/NFV Is Cooler

Software-defined networking/network functions virtualization (SDN/NFV) offers promising potential for space conservation and power consumption reduction on a number of fronts. By collapsing bay space currently occupied by power-hungry cable modem termination systems (CMTS) and quadrature amplitude modulation (QAM) equipment, MSOs can move to cost-effective white-box x86 servers that operate more efficiently and do the work of multiple legacy devices in a ninefold footprint reduction, requiring fewer computer room air conditioners (CRACs) to cool. These upgrades offer the chance to structure trunk cabling in crowded headends more cleanly, collapsing older runs of jumpers and copper cabling into denser, smaller-diameter fiber optic trunks that allow smoother, denser airflow and readily accommodate a hot aisle/cool aisle arrangement with short jumpers that are easy to manage during moves, adds, and changes.

The active gear for an SDN/NFV deployment also affords the operator a chance to deploy multifiber MTP connectors that support a migration path to 100GB service and beyond, plugged in to leading-edge quad small-form-factor pluggable transceiver (QSFP) optics that reduce power consumption. An MTP (QSFP+) transceiver is nearly the same size as a duplex LC (SFP) transceiver, but can supply many times the throughput capacity. Evolving the headend to MTP connectivity offers substantial density improvement, a key consideration for most space-reclamation projects. In addition to density, operators also enjoy significant power conservation, as the wattage requirement for one QSFP+ is less than what four SFP ports require. Concomitant with the savings in power is a notable reduction in cooling requirements for the switches running the QSFPs as well. Less obtrusive cabling, lower power requirements, reduced cooling requirements, and substantial space savings gained by way of collapsing bulky, power-hungry legacy access network gear in favor of switches and servers – the move to SDN/NFV sells itself.

The Power of Virtualization

Recently at a SCTE Energy 2020 plenary discussion, a headend operations manager detailed the processes his organization used to analyze where and how often they were experiencing grid power outages, and how much generator backup time they required to offset grid downtime on a location-by-location basis. While carefully evaluating their reliance on generators, and as an adjunct consideration, how much of a power load the affected headends represented – it seemed relevant to them to consider both the cost and reliability of their generator backup systems. To maintain effective, failsafe, and sustainable power delivery, they periodically test generators to make sure they are ready to shoulder the load for their average outage time of 60 to 90 minutes. For just one multi-state region, the operations manager related that they had spent over a million dollars in 2018 just to test their generator backup systems. Merely firing the generators up periodically to ensure their functional status had cost the MSO a seven-digit number through the first eight months of 2018. According to Cisco’s total cost of ownership analysis of remote PHY vs. analog deep fiber, finding ways to reduce the baseload power requirements can greatly impact an operator’s bottom line. By reducing the power required in the headend, the access network would be more survivable during an outage and a substantial operational cost reduction could be realized.

It is not lost on the MSOs that legacy service delivery models replete with bay after bay of CMTS and QAM gear requires a hefty utility bill, and they are optimistic that virtualizing some of this functionality will afford them a chance to write a smaller check to the power company. In addition to considering alternative means for supplying baseload power, to include wind and solar installations, and using passive optical network (PON) technologies for greenfield deployments to reduce the number of power-hungry radio frequency (RF) amplifiers in the field, MSOs are starting to look to virtualizing the core functionalities in the headend to chop power consumption and conserve available floorspace in already cramped headend environments.

Tipping the Scale

The move to SDN/NFV is not merely a power conservation play. While the reduced baseload power requirements for virtualized converged cable access platform (CCAP) architectures are extremely attractive, with some estimates showing as much as a 90 percent reduction in power required – a hefty savings for MSOs spending millions on fuel for generators on top of their utility bill – SDN/NFV also entails more flexible access networks, networks that leverage fiber deep architectures to provide scalable services to customers with ever-increasing bandwidth demands. SDN-enabled bandwidth is flexible capacity that can be provisioned remotely on-demand by the customer without a laborious provisioning process. Business customers can add and suspend wide-area networks and other enterprise-class services remotely with the click of a mouse and start using them on demand, and the bandwidth capacity they need for events and challenges can be provided to them on the fly, meaning the MSO sales operations teams can realize new revenues while driving customer satisfaction and creating loyal customers that represent consistent profitable business.

Cable operators face a significant challenge in the coming decade, as driving fiber ever deeper in their networks will require substantial building to stay ahead of demand for reliable, high-bandwidth services. To hold onto their competitive advantages, especially with an industry-wide push to attract and retain enterprise customers, MSOs will need to simultaneously make strategic capital expenditures on fiber distribution and density while aggressively pursuing operational savings. The utility bill will continue to be a substantial operational expense, so the time is right for MSOs to consider how virtualization strategies fit into their business models.

Sebastian Sassi is a market development manager for Corning Optical Communications. He has 20 years of telecommunications industry experience and currently focuses on helping operators transform their inside plant networks to stay ahead of demand for high-bandwidth services and applications. For more information, email SassiSJ@corning.com or visitwww.corning.com/ISP. 

Offering a way for cable operators to migrate to a more efficient, virtualized platform, Calix has introduced a product that can plug into their legacy access networks and utilize software-defined capabilities.

Calix said that product, the AXOS DPx Connector, is a DOCSIS Virtualized Network Function (VNF) that can be woven into existing OSS (operational support systems) for HFC and PON networks deployed by MSOs.

Cable operators are eager to shift to software-defined networking and virtualized apps and systems, but they are challenged with how to develop a framework that can fit that into those legacy OSS environments, Shane Eleniak, vice president of product line leadership at Calix, said.

Calix has already done some work with DOCSIS Provisioning of EPON and DOCSIS Provisioning of GPON, which enable operators to utilize their DOCSIS backoffice systems for PON deployments, but want to move away from a classic, monolithic software approach, he said.

MSOs, Eleniak added, “want to invest in something that is SDN-aware…The tricky part for an operator, is how do I bridge?”

Calix claims that the AXOS DPx Connector can be plugged into the existing architecture and virtualize the applications without altering the underlying backoffice systems. MSOs would also use the same servers, but the new connector would add scale and tie it back to a data center or network operation center (NOC).

To aid the effort, Calix has teamed up with Momentum Telecom, a company that supplies DOCSIS management/provisioning products and hosted voice services to cable operators.

Eleniak said the primary, early targets for the new product are cable operators that have been deploying PON or have it on their roadmap for certain scenarios in a way that complements the HFC side of their business. He said it will also fit in as operators deploy DOCSIS 3.1 and starting to pursue “virtual” Converged Cable Access Platforms (CCAPs) and distributed access architectures.

Calix said it has field trials underway. It will be using a two-pronged approach for trials and deployments – working with Momentum and its base of MSOs as well as direct engagements with the operators, according to Doug Blue, Calix's solutions marketing director for cable. 

 “As an industry, we know that SDN is the way of the future of service providers and this partnership between Calix and Momentum Telecom will allow our customers to more easily move to an SDN architecture as soon as they are ready,” Scott Helms, senior vice president of advanced services at Momentum Telecom, said in a statement. “The service agility enabled by SDN has become a key strategic goal for cable operators, and the combined solution allows our joint customers to immediately begin to reap the benefits of an SDA architecture without the complexity of high costs of integration.“  

Tibit Communications, a startup with technology that’s focused on telcos and cable operator access networks, said it has landed a $13.8 million “A” round led by a mix of strategic investors and network operators.

Tibit didn’t reveal those investors, but Liberty Global praised the Petaluma, Calif.-based company and identified itself as one of the vendor’s partners in Tuesday’s funding announcement.

Founded in 2014, Tibit’s initial focus is on pluggable devices that are focused on helping telcos and cable operators virtualize their networks.

Tibit’s initial products are the MicroPlug OLT and ONT, claiming they provide FTTx PON fiber access from Ethernet switches and other devices equipped with an Ethernet port, including fiber nodes, and that those modules work with some of the market’s leading Software Defined Networking orchestrators.

“PON is not a box anymore,” Edward Boyd, Tibit’s chief technology officer and co-founder, said, noting that the vendor’s pluggable form factor packs in the MAC and PHY layer components.

He said the technology will fit into cable’s strategy as MSOs shift away from centralized access network architectures to those that distribute those components.

Part of the pitch is that operators can plug in Tibit’s components without altering the fiber plant, and that operators can move to generic Ethernet switching hardware and take advantage of those economics without being locked into a specific vendor’s chassis. It also allows operators to put in PON where they want on a port-to-port, selectable basis, Boyd said.

Cable operators are interested, he said, because they can drop a fiber into a neighborhood in a  greenfield scenario without having to add a hub site or headend.

“Partnering with Tibit fits our strategic GIGAWorld ambition of bringing high-speed gigabit connectivity to homes and businesses as early as possible,” Colin Buechner, managing director or access networks as Liberty Global, said in a statement. “The Tibit MicroPlug provides significant flexibility in the systems that we can use to provide FTTx. We can purchase FTTx solutions from a much wider supplier base and get one that best fits the deployment need.”

Tibit is currently enabling partners to evaluate its products with a programmable FPGA version of the product, and expects to reach production on the actual modules, which will run on custom silicon, later this year, Boyd said.

Tibit, which has about 20 employees, all in engineering, is led by a team that hails from the telecom world. Boyd previsuly was CTO and co-founder of Teknovus (now part of Broadcom). Tibi president and CEO and co-founder Richard Stanfield was once CEO of Imagine Communications, and led the sale of the company to Harris Broadcast Corp. in January 2014.

In 2015, Boyd, along with Kevin Noll, then with Time Warner Cable, and Fernando Villarruel of Cisco Systems, and Saifur Rahman & Nagesh Nandiraju of Comcast co-authored a white paper that compared and contrasted several solutions for implementing a PON in a node-based architecture.

“Remote PONs can be very beneficial to MSOs. They allow the operator to run fiber with very high efficiency or connect customers at very long distances. An operator can reach customers without adding hubsites and could potentially consolidate hubsites,” they concluded. 

Though Tibit is announcing the funding today, the company filed an SEC document on Oct. 14, 2016, showing the $13.84 million raise, with a total remaining to be sold of approximately $805,218.

PHILADELPHIA – A panel of some of the industry’s top engineering executives shed some important light on why distributed architectures and the shift toward the software-driven, virtualization of network functions have rapidly become major priorities for MSOs around the globe.

Boiled down: Old models, driven by purpose-built hardware, simply won’t work anymore.

There’s a huge “disconnect” between the data that people are consuming and what cable operators must spend in order to keep up with that demand, Balan Nair, executive vice president and chief technology officer of Liberty Global said in a wide-ranging panel session that was moderated by Bob Stanzione, executive chairman and board chairman of Arris.

As network capacity planning goes, Liberty Global is already allocating enough capacity to deliver an average of about 1 Mbps per sub. With annual capacity growth in the range of 30% to 40% per year, it won’t be long before that hits 2 Mbps per sub.

STATUS ‘NOT SUSTAINABLE’

That would equate to about 25 8MHz-wide channels at Liberty Global, or “like half of our whole plant,” Nair said, noting that this trend causes MSOs to keep “feeding this beast” with capital just to expand capacity to stay ahead.

“That’s a problem for all of us,” he said. “That’s not sustainable.”

And it doesn’t add up with a business that grows revenue at 4% to 5% per year. “Eventually, it breaks,” Nair said. “The math will never work.”

The historic practice of buying specialized devices from traditional suppliers that cost a lot to build and buy won’t work anymore. “That model has to change,” Nair said.

The shift is toward cheap, commodity hardware running on very specialized software.

“So, we’re going to ask our vendors to think about the software and to run it on some cheap hardware because, at the current rate, it's not sustainable,” he said.

Vendors appear to have gotten that message loud and clear. Just this week, Nokia, Casa Systems and Harmonic all announced virtualized versions of Converged Cable Access Platforms that can work in tandem with distributed architectures that move away from traditional monolithic access devices.

Fellow panelist Jim Blackley, EVP, engineering and IT at Charter Communications, called Nair’s characterization of the situation “spot on and accurate.”

He added that, when it comes to capital, “you can’t spend it everywhere. … You can't spend it in the home and spend it in the network and spend it on content and all of the other things we need to deal with."

He said it’s important to think about virtualizing the functions of the consumer premises equipment (CPE) and move the compute, storage, encryption, and packaging into cloud-based centralized systems.

Charter is moving in that direction with ActiveVideo (a vendor it owns as part of a joint venture with Arris). Virtualizing the CPE means "you can put out devices that will last longer,” Blackley said. “The reality of our business is, the day you [deploy] something, it's legacy."

He said Charter was a “huge fan” of network virtualization and separating out the MAC and PHY elements.

Zoran Static, EVP and CTO of Shaw Communications, said he’s also keying on these trends and listening to partners on “how the future is going to land.”

But when it comes to a move toward virtualization, “the starting point becomes remote PHY,” he said. “The old model is not sustainable. It’s not a choice really, anymore.”

Nair called the shift to software-driven models as one of his top worries. “It requires a skillset change in all of our businesses,” he said. “The change is not just the skillsets in our employee base, but the skillsets of our vendor base as well."

Vendors, meanwhile, prefer to innovate on their own platforms. Interests of the two sides might not match up and could become a source of tension.

Nair thinks it’s best to solve that problem, “or else you're going to see more and more operators building their own stuff, and that won't be very good for the industry."

The early part of the talk centered on the move toward Gigabit-level broadband – something that has become a table stakes option for consumers who need it, or at least think they need it.

But Nair doesn’t see 1-Gig going “mainstream” until 2018 or 2019. By that, it won’t become a “bullseye” product for Liberty Global that is backed heavily by marketing and promotion. Today, that’s relegated to the operator’s 250 Mbps product, he said.

“We’re looking at increasing it,” Nair said, noting that MSOs have the luxury to press two levers with broadband – price and speed – while competitors with less robust networks can only fiddle with price. “Speed wins.”

Blackley said Charter is already getting positioned for the 1-Gig future. Its all-digital migration is freeing up capacity and Charter is also plunging fiber deeper into the network.

He added: “It’s not just about speeds; it's about latency, and it's not just about in-home; it's out of home.”

Blackley hesitated to say when consumers will actually need 1-Gig speeds, but won’t be surprised to see it happen eventually. "Like your garage, if you build it, it will get filled,” he said.

‘IRRATIONAL’ GOOGLE MOVES

Stakic, meanwhile, warned against “irrational behaviors” occurring in the United States, largely “triggered by Google,” that might cause MSOs to rush prematurely to 1-Gig.  He questioned if adoption rates will be high enough to justify the returns, noting that there aren’t many apps now that require three-digit megabit speeds at this point, let along 1-Gig.

Nair said the challenge of getting gigabit to the home pales versus getting gigabit support inside the home with WiFi. “The cost of WiFi now is actually more expensive on a per subscriber basis than DOCSIS,” he said.

The panel dove deeper into wireless and mobility now that Comcast and Charter Communications are both looking to put their MVNO agreements with Verizon in play.

Blackley said having an MVNO deal won’t prevent Charter from having to bulk up its own network with denser WiFi and backhauling capabilities.

“Wireless is just an extension of the wired network, and our consumers want their products where they are..and we have to have a network that will support that with economics that make sense." 

In today’s environment of cord-cutters and streaming services, cable MSOs are under pressure to offer competitively priced services while providing an enhanced quality of experience.

The pressure is not unfounded – there is a lot at stake for cable companies facing what can be described as an "over-the-top (OTT) double-edged sword.” Internet service has never been more important to cable companies as they watch linear video give way to on demand programming, but meeting customer demand for fast and reliable internet service is ironically what led to the decline of their video revenue stream as it paved the way for OTT service providers such as Netflix, Hulu and Amazon Prime. 

With OTT not going away, cable companies must look to provide a greater experience in order to stay ahead of what will certainly be the future of television to remain viable. But what steps can MSOs take to provide a greater level of service? 

For one, adoption of virtualization technologies and trends like Head Ends Re-Architected as a Data Center (HERD), which is the adaptation of the telco Central Office Re-Architected as a Data Center (CORD) initiative for the MSO market, will be important. Another differentiator is the continued drive to push the network closer to subscribers with what is described as a Fiber Deep evolution—pushing the fiber closer and closer to the end user to provide better service. 

While cable companies may look different by 2020, whether they begin decoupling programming packages or even aggregating OTT packages as Comcast is planning to do with Netflix, they need to focus on their architecture to support a better experience now and allow them to offer differentiating features in the future. 

Importance of Highly Responsive Internet Services

As user demands around Internet speed and availability have increased, the need to store content locally to provide high quality, responsive access to the content that customers want is required in order to address the competitive landscape.

The Internet relies on locally stored or “cached” content to serve up a great user experience. Cable has an advantage here, since they are the network down the block—not miles away in a large data center. That distance will impact responsiveness and, especially in peak hours, quality of service. The cable headend has for many years served up on demand content, but will need to make one more transformation to bring content even closer to consumers.

The first generation of streaming video might take customers to a data center out-of-state providing poor performance while tarnishing a cable company’s internet services reputation. By leveraging the HERD approach and utilizing data center technologies like virtualization, caching and software defined networking in the headend, those organizations can keep customers satisfied, lower the abandonment rate and make sure the advertisers get compensated.

Competition is already robust and to stay ahead, the customer experience is best served from the headend down the road rather than from a random data center in the cloud.

Content Delivery, Mobile Edge Computing and the Redefined Network Edge

Netflix is a great example of a company that innovated its networks out of necessity. While a fledgling streaming media company a few years ago, the company realized that the legacy Internet was not enough for the new mobile generation. Netflix used the cloud and the internet but found that video would not work unless it was delivered by a local server. When streaming from a far-away data center, the user often became dissatisfied and cancelled their service. Content Delivery Networks solved the problem to a degree, but still lacked the ability to scale with demand and would still limp along slowly.

To overcome this challenge, Netflix reimagined content distribution and placed servers close to the Internet’s edge, providing near real time responsiveness. Now picture making the performance even better by having the network respond with more bandwidth by effectively optimizing caching technology.

By moving and virtualizing these servers and sharing precious network resources, cable MSOs can effectively make all OTT content closer to the end user, improving latency and providing a platform for video aggregation that can support live and real time events such as the Super Bowl or a heavyweight boxing match.

And by driving the fiber deeper, they can further improve performance: deep fiber pushes the optical-to-electrical conversion of signals closer to subscribers, which increases potential bandwidth to homes (allowing for support of newer and more numerous services), as well as cutting down on operational costs related to power and maintenance.

Marrying Transport and Caching

Let’s face it, video follows a hockey stick curve, where there may have been slow growth initially but we have seen that growth rate quickly increase to a much faster rate. With the ability today to interact with content and binge watching becoming the new normal, cable MSOs must be able to quickly react and provide a differentiated user experience for their customers. To achieve this, cable MSOs will need to throw a combination of compression, caching, and CDN technologies at the problem.

But what if there was a new option, with the ability to get a file from just a few miles away—near instantly—and the user wouldn’t realize the difference. The future of the cloud is a combination of network, compute and storage architectures to create the best experience. By offering the right mix of bandwidth and orchestration, cable MSOs will be able to take their businesses to the next level.

This will keep the OTT consumers happy while cable MSOs avoid becoming the weak link in the video delivery value chain.

-Elias Cagiannos is MSO Practice Leader at Ciena