Proactive network maintenance (PNM) continues to make inroads on its goals of reducing troubleshooting time and cost while reducing the impact of network impairments on customer service. The broadband industry achieves this through grassroots efforts developed and shared by participants in two working groups. This global ecosystem of vendors and operators share their ideas, efforts and challenges to help move the community further on the path to efficient operations and improved services.
In a recent face-to-face meeting, we continued to make progress on several of the working groups’ workstreams, sprint toward some near term goals and set some long-term goals. And there is more to come!
Working Groups
CableLabs’ PNM Working Group (PNM WG) meets bi-weekly for all current work, with a focus on some specific workstreams on the off-weeks. SCTE’s Network Operations Subcommittee Working Group 7 (NOS WG7) also focuses on proactive network maintenance and related workstreams. The CableLabs group focuses on developing the engineering and science to enable better PNM while NOS WG7 focuses on the field implementation of that engineering and science.
In addition, the chairs of these working groups meet often to manage the workstreams and ensure progress on key developments that benefit the industry most. Many members overlap between the two groups, which ensures an effective pipeline of research, development and implementation.
Workstreams
The proactive network maintenance working groups maintain progress on several developing workstreams, and a few of note have recently concluded.
Our Methodology for Intelligent Network Discovery (MIND) effort is an important workstream where we target repair efficiency through topology discovery and automation. Participants share developments and ideas on Thursdays, then bring the best results to Tuesday meetings, where we contribute canonical methods and software code for reference. This high cadence and open contribution are leading to well-developed ways to utilize channel estimation data for identifying features in the radio frequency (RF) network that aid in determining the ordinality and cardinality of network components. We accomplish this through an initial process of data cleanup to reveal the features in the signal, then through clustering and pattern matching methods we identify the features in the network and reveal the network’s topology. As with all research, the initial process demonstrates functionality, which we have done; now we are working to make it reliable for all the diverse network designs and modify it for various PNM use cases.
The PNM WG has been working hard to develop GAI for Network Operations (NetOps) — or, more simply, AIOps. This group develops and shares the software and tooling to enable the output from NOS WG7 and PNM WG to be automated into network operations through retrieval-augmented generation (RAG) models and agentic AI solutions. In parallel, industry experts are peer reviewing the output from these GAI tools to ensure quality and engineering precision can be reliably delivered, improving on our own peer review process as we go.
Meanwhile, the PNM WG is developing standard ways to quantify impairments in the network as they are revealed in PNM data, referred to as the measurements work, to enable GAI applications and streamline PNM automation.
The PNM WG looks forward to publishing an update to our “Galactic Guide” at the beginning of 2026. The current version is available to the public. In the new version, you can look forward to receiving updates on the development of the measurements work and MIND work, in addition to initial methods to use Upstream Data Analysis (UDA), for those cable modems that can reveal the upstream spectrum.
New workstreams start when the groups complete earlier workstreams and queue up the next priority effort. The initial work on UDA was completed early this year, and more work on it is planned for the future. Likewise, NOS WG7 has recently completed outlining some new PNM training to come from SCTE, which updates the training to utilize new learning methods, and includes our newest knowledge about maintenance efficiency and proactivity.
PNM Face-to-Face
About once a year, CableLabs hosts a face-to-face meeting for these PNM groups and their members. We hosted a hybrid event in October with more than 30 participants representing many operators and vendors at our Colorado headquarters and through virtual attendance. Our time during the face-to-face focused on providing updates about our workstream progress, discussing potential future workstreams and developing content for the upcoming Galactic Guide update.
Near-Term Goals
The working groups are hyper-focused on a few key near-term results:
- Completion of the initial use case for channel estimation data to identify network features and determine the network topology. This is the MIND work.
- Unification on impairment quantification — and, specifically, the ability to monitor resiliency in DOCSIS® networks to transform PNM into managing capacity. Our measurements work will help operators better assess urgency with proactive repairs to prioritize work based on risk of customer impact.
- Identifying new UDA opportunities as new methods are published and further validated.
- Completing updates to the Galactic Guide, including incorporating new, more precise treatments of several important RF concepts published in our technical reports (such as several monographs published recently by SCTE’s NOS WG1).
Long-Term Goals
Accomplishments on our near-term goals open up new opportunities:
- The MIND work is not done; we still can develop better, automated localization for troubleshooting proactive and reactive network faults.
- One topic that will be important for our MIND goals is standardizing GIS information. This would allow the automation we created to work across tools and platforms for greater ease of use and broader adoption.
- We look forward to a forthcoming release of the new PNM training material from SCTE, which our expert participants intend to peer review further.
- As our work on GAI use for network operations continues, we intend to develop better input and better testing methods of these new tools to ensure their reliable application.
- The face-to-face meeting also revealed that we should make better use of our knowledge of noise, distortion and interference (NDI). By looking more closely and improving the categorization of these various signal impairments, we expect to identify their sources and locations through automation — which will help form additional efficient PNM opportunities.
Beyond Proactive Network Maintenance
While there is a substantial community working on RF and DOCSIS-related PNM, CableLabs doesn’t want to leave fiber out of the fun!
We are currently identifying near- and far-term opportunities in the passive optical networking (PON) world as well. As many operators push fiber deeper — all the way to the premises — the Optical Operations and Maintenance Working Group (OOM WG) works toward identifying use cases and aligning them to the telemetry that network elements provide. By better aligning PON and DOCSIS technologies, the group helps streamline network operations and maintenance further.
While the fun continues with DOCSIS technology and there is much more to develop, we are applying our experience with RF over coax to RF over fiber.
Whether your concerns are with maintaining DOCSIS or PON deployments, or ensuring network or service reliability, there is a community ready to work with you. Join a working group, and join the fun!