Sélectionner une langue
In this post:
An overview of DOCSIS and what’s coming in version 4.0
How software-defined networking and cloud-native platforms can drive successful DOCSIS 4.0 implementation
With cable operators looking for ways to simplify their infrastructure for services, including fast data and voice transfer, critical transformation in the cable access network is taking shape. Multiple service operators (MSOs) are moving to a new generation of access network as they shift to digital optical solutions. MSOs are gradually migrating from centralized access architecture to distributed access architecture (DAA).
Fortunately, Data Over Cable Service Interface Specifications (DOCSIS) 3.1 technology is available—and DOCSIS 4.0 is on the horizon—to support the growth of hybrid fiber-coaxial (HFC) network symmetry both upstream and downstream (upload speed and download speed, respectively).
What is DOCSIS?
DOCSIS is a standard for data transfer over coaxial cable TV connections developed by CableLabs, a research and development lab for cable operators. Globally, cable operators have adopted DOCSIS to provide voice, video services, and internet data services using their existing cable TV systems.
The DOCSIS architecture consists of two primary components: a cable modem located with the customer, and the cable modem termination system (CMTS) operated by cable service providers.
When DOCSIS 1.0 emerged in 1997, the specifications enabled 50Mbps downstream and 10Mbps upstream data speeds. This revolutionized the landscape, and coax providers increased their competitive relevance.
New technology being deployed now is pushing broadband speeds into the 10G realm, both downstream and upstream. An illustration of the progression in capabilities and bandwidths is provided below.
DOCSIS 4.0 is an emerging specification that integrates the features of Full Duplex DOCSIS (FDX) and Extended Spectrum DOCSIS (ESD), methods that can supply rejuvenation for today’s HFC networks.
FDX uses the existing spectrum agnostically for both upstream and downstream bandwidth.
ESD advocates for extending the available spectrum and dedicating certain parts of it for upstream or downstream traffic.
This combination offers cable operators two complementary, but differing, options for extending their bandwidth capacity and distributing symmetrical multi-gigabit speeds to their broadband subscribers.
The industry today is running DOCSIS 3.1 with some passive optical network (PON) added. PON is designed to allow a single fiber from a service provider the ability to maintain an efficient broadband connection for multiple end users.
The industry is also focused on mobile. DOCSIS 4.0 is planned to have much more emphasis on PON, mobility and wireless, and push to the cloud, all wrapped together with service and network convergence.
Exploring software-defined technologies and cloud-native platforms with DOCSIS 4.0
Distributed access architecture (DAA), a prerequisite for DOCSIS 4.0 networks, distributes key electronics toward the edges of the network that typically are contained in legacy, centralized Converged Cable Access Platform (CCAP) chassis. DAA enables the evolution of cable networks by decentralizing and virtualizing headend and network functions.
We have seen disaggregation in the traditional HFC network with the development of DOCSIS DAA. These disaggregated components can operate at several locations in the network.
A number of cable operators are in the early stages of upgrading to DOCSIS 3.1, but the high-bandwidth capacities of the upcoming DOCSIS 4.0 standard may be even more appealing for providers with bandwidth-hungry applications.
MSOs have already embraced cloud-native architectures starting with DOCSIS 3.1, and most major cable vendors have introduced virtualized versions of their Converged Cable Access Platforms (vCCAPs). Some MSOs offer both a HFC network and a 4G or 5G mobile network chosen to share a common infrastructure platform that supports the convergence of these services.
When moving to virtualized or cloud-native architectures, two key factors are critical to consider: elasticity and quality of service. Reducing operating domains is one requirement for improving service and reliability.
Artificial intelligence and machine learning (AI/ML) for predictive analysis, modern streaming telemetry, and cloud-native architecture may help companies quickly react to operational issues with predictive tools to identify future issues before they become service impacting.
However, to best maximize advantages of cloud-native economics, the applications need to be modernized and containerized in parallel with building out the horizontal, scalable, distributed virtual architecture. MSOs can consider microservices, small pieces of code that can be scaled independently, to accelerate deployment and minimize outages during service upgrades—along with other advantages like elasticity and intelligent distribution of workloads.
DOCSIS 4.0 may add further value through integrating legacy and new system resources and practices by incorporating cloud infrastructure specifications to address customers’ needs. A hybrid approach—moving services to cloud-native functions when possible and using legacy functions when needed—can help MSOs integrate modern applications and solutions.
How do we continue on this path of success?
While cloud-native application platforms allow developers to innovate, using common development, deployment, automation, and management tools— simplifying creation, use and control of applications across the entire environment—hybrid cloud infrastructures offer scalable and managed foundations for traditional and cloud workloads, across all environments.
As CableLabs and MSOs explore solutions to move towards cloud-native functions for DOCSIS, Red Hat can help.
For example, CableLabs is working to improve edge compute performance and reduce the overall cost of the edge deployment by utilizing Kubernetes-enabled reference architectures in conjunction with hardware components, such as field programmable gate array (FPGA) and graphics processing unit (GPU) accelerators, which accommodate the differences between workloads. With our contributions in upstream open source communities,like Kubernetes, OpenStack, and Fedora, and experience working with GPU hardware partners, Red Hat is well-positioned to identify how certain solutions may be adapted from existing open source functionality, laying the foundation for further innovation.
The cable industry will continue to push technology boundaries of its partners toward real-time services and increased reliability. Software-defined technologies and cloud-native platforms can help cable operators deliver faster speeds, connect to different access networks, and offer a bridge to the future.
To experience more on this topic, we’d welcome you to join us in Red Hat’s sponsored upcoming Light Reading webinar on September 23, 2021.
About the author
Amol Chobe currently leads and delivers the strategy for Red Hat in the North America telecom vertical. He is responsible for the direction of the Red Hat portfolio focusing on cable, media and ecosystem partners.