2020 has so far seen a significant increase in the number of broadband subscribers being serviced by a virtualized CMTS. More and more cable operators are preparing to make the move from rigid legacy hardware to mature, agile, future-proof virtual solutions. Let's explore some of the reasons for this industry pivot and the expansive opportunities that a virtualized cable access network offers.
The best-known reasons to virtualize the CCAP
- Distributed Access Architectures (DAA): The reasons to move to DAA have now been proven in the field, plenty of times and at volume. The main reasons operators are increasingly adopting DAA are that it drives access network simplification, flexibility and cost-efficiency. In a DAA deployment, you eliminate the entire legacy analog optical network and push power-consuming CMTS PHYs out to smaller hub-sites or into the hybrid fiber-coaxial (HFC) network itself. The obvious financial gains regarding OPEX and CAPEX are more than enticing. These savings then enable operators to make better use of DOCSIS 3.1 capabilities.
Some still argue that you don’t need a virtualized CMTS to deploy DAA. While technically this is true, such an architecture lacks most of the powerful advantages of DAA. Analogous to strapping a battery to an internal combustion engine car, tying DAA modules to a legacy hardware CMTS results in an overall solution that still suffers from slow development cycles, heavy power consumption and massive space inefficiency. You don’t need complex and costly hardware to run a simple software setup: a vCMTS running on state-of-the-art Intel-based servers is clearly a more efficient, affordable and higher-performance solution. vCMTS software requires only commercial off-the-shelf CPU, Ethernet interfaces and power supply. Why not opt for a simplified and streamlined solution that costs less and delivers the full advantages of DAA?
- Service velocity : The ongoing global health crisis and subsequent lockdowns have shown that the current model of cable network infrastructure needs to be virtualized, and fast. Splitting segments will always require new nodes, not much can be done about that. Without DAA however, node splitting requires major headend and hub-site work (adding space, cooling and power) to deploy new CMTS hardware, optical gear and combiner wiring. Forget about the ability to swiftly deploy new capacity.
With this traditional model, you're looking at months before an upgrade can be completed. Even a DAA deployment with a legacy CMTS would still require going through such major hub-site work. To add insult to injury, ordering new CMTS hardware, even without supply chains being disrupted, will likely take at least three months to arrive on-site. This is one of the big issues with legacy hardware deployments. These less-modern approaches towards capacity planning and deployment prevent operators from moving fast and adding new capacity with swift precision, whenever and wherever needed. When deploying a virtual CMTS solution, adding new capacity is as simple as spinning up new CPU instances in a datacenter. We are talking about minutes, compared to months.
- Service continuity: the great thing about a virtualized solution is, that it's virtualized in exactly the same way your other critical IT infrastructure is becoming virtualized! It benefits from all the advantages that state-of-the-art IT solutions provide: speed, flexibility, scalability, simplicity and affordability. Harmonic's virtualized CableOS solution, runs multiple redundant containers across redundant servers. Imagine your legacy hardware CMTS, such as a Cisco cBR8, Casa C100G, or a CommScope 6K that runs with about 60,000 subscribers or more. A minor software or hardware issue occurs by chance, as it may; that means an entire city is suddenly without any internet access.
Yes, issues can still happen with a virtualized CMTS. But fault domains are significantly smaller. We're talking about a few hundred instead of 60,000. That scenario is easily avoided with more modern systems that leverage virtualized cloud-native infrastructures. In addition to a more reliable system, our global services team works hand-in-hand with your network operations center (NOC) to monitor and proactively intervene if ever the need arises. The continuous integration and development (CI/CD) capability that leverages the cloud-native, microservices architecture allows us to do a quick verification of software fixes and revert to previous code using a “canary release” ability. And this support agility is executed without impacting services, at all. With your legacy CMTS you would be looking at a couple of months to get a fix.
Reasons to deploy vCMTS that will impress your C-suite
The reasons in the previous section are relevant and operationally straightforward. However, there are even more strategic reasons to go for a virtual infrastructure for cable access that is maybe lesser-known. Here are some of the more under-referenced benefits that might have you rethink your options when it comes to buying more legacy hardware CMTSs.
- Shared datacenter infrastructure: Even if you virtualize only the cable access infrastructure, it would make a lot of sense from a business and operational standpoint, in relation to the savings and all the other benefits outlined above. But what if you are looking at this from a network-wide and company-wide perspective? Imagine having a significant number of applications all sharing the same docker farm, using containers, that are all orchestrated with the same Kubernetes infrastructure. Your efficiency gains are going to be like nothing you and your management team have ever seen with legacy technology. Then, for an even more attractive, optimized approach, you could add a virtual PON OLT and BNG in different docker containers that are also managed by the same Kubernetes infrastructure. Since virtualization is so flexible thanks to these microservices, why not add peripheral applications into the mix?
The number of applications that can be added on top of the general data center compute is unlimited. That's right. Unlimited. At Harmonic we are currently working on some really exciting new applications. The days of seeing racks and racks and more racks with different pieces of hardware in a hub-site, that each have their own power, cooling and network management requirements, are gone. A shared infrastructure also has an impact on the previously mentioned service velocity. Deploying new services and applications can be done a lot faster, and simultaneously.
- Big data in cable access: With the current legacy CMTS we are only accessing a tiny part of the data that is available in the network. These legacy CMTSs tend to struggle if you try pulling too much data. A virtualized solution like CableOS, integrated with our CableOS Central solution for network monitoring and management, creates a data gathering powerhouse. It's able to collect and leverage all of that network data so much more efficiently thanks to user-friendly analytics, dashboards, and reports. It takes proactive network management to a whole new level of excellence.
These new solutions also integrate AI capabilities to predict network issues and proactively add capacity where traffic is expected to increase. We have empirical data showing a 30-minute reduction in root cause analysis and, as an example, we have one customer that has been able to reduce calls from around 250 to only about 100 calls per day. At this point, you don't even need to quantify the OPEX savings. The AI capability can also be tied to other Cable network-specific applications, such as low latency DOCSIS. The improvements for your customers just keep going from there.
- New business models: Virtualization opens the door for exciting new approaches to the cable business. More and more operators are asking their equipment providers to “lean in” and take on more ownership and responsibility of their products and solutions. The days of shipping a product, and just moving on to the next customer are becoming a thing of the past. Operators deserve more from their product and service providers. Resources are tight, travel may be more restricted, and the complexities of solutions increase. It can be inconvenient for operators to have to increasingly rely on providers that may not be available every time you need them to be. As these previous ways of working are being challenged, virtualization and DAA offer opportunities.
Operators can “syndicate” their offering or offer DOCSIS as a service (DaaS). One of the things Harmonic has done is to reinforce our service and support efforts on behalf of our commitment to quality is to work even more closely with customers. We developed centralized, user-friendly engagement tools and created the ability to embed our expert teams with our customer's operational team. We have created our own Harmonic Network Operations Center that can act as an integral part of your operations team. We have subject matter experts that know the ins and outs of our products, DOCSIS and datacenter technology. We have proven with our dedication to our customer's success that we are capable of proactively managing and operating customer networks.
The industry is rallying around virtualization
Massive investments and innovation have started to grow virtualization adoption, and precipitate a significant slowdown in new legacy CMTS hardware investments. It has become clear that many legacy hardware providers are no longer going to focus on hardware development. Many legacy hardware providers have also decided to cease any major software feature development for that equipment.
We can say a friendly and grateful goodbye to “big iron.” The market can no longer sustain legacy solutions that have failed to make the transition and are falling behind the curve of innovation. Virtualization and DAA hold the key to your transition to the next generation of cable access and more compelling services. Faster broadband has never been so simple.
Harmonic can help you evolve your network to meet tomorrow's demands today. Let us know how we can help.
About Daniel Etman
Daniel is a long time cable industry executive living in Belgium with his wife and four kids. At Harmonic, Daniel is leading the Liberty Global team. He has previously worked in different roles at some of the major vendors in the industry such as Casa Systems and Cisco. Daniel has been involved in DOCSIS as of the start of the journey in 1998.