Reliability, Not Just Speed
How PNM and PMA are Rewriting the Future of DOCSIS
Experts from Comcast, CableLabs, Technetix, and OpenVault unpack how PNM and PMA are rewriting cable’s reliability playbook.
Reliability Is the New Battleground
For more than three decades working in cable technology, I’ve watched our industry transform from analog to digital, from megabits to gigabits, from “best effort” to near real-time operational intelligence. But if there’s one truth that has become clearer with each generation of DOCSIS, it’s this:
Speed may get the headline, but reliability wins the subscriber.
When a network becomes so stable, so predictable, that the customer forgets it exists, that’s when we’ve truly delivered what broadband is supposed to be.
This year’s TechExpo25 fireside discussion brought together three of the industry’s sharpest minds: Mike O’Dell of Comcast, Dr. Jason Rupe of CableLabs, and John Downey of Technetix. As moderator, I had a front-row seat to a consistent message: Reliability is no longer a byproduct. It is the product.
And with DOCSIS 3.1, DOCSIS 4.0, and the maturation of tools like PNM and PMA, we finally have the means to engineer reliability, not hope for it.
What reliability really means in 2025
When asked to define network reliability in a single sentence, the panelists aligned quickly.
“Availability and consistency,” said Mike O’Dell. “If the subscriber never thinks about the network, that’s success.”
Dr. Jason Rupe put it this way:
“Reliability is when end users can connect and use the network seamlessly, even if throughput momentarily varies.”
And John Downey added nuance:
“Speed tests alone aren’t performance. Latency, jitter, robustness, recovery time, those are the metrics that reflect real reliability.”
This aligns with what we see across the broadband industry: customers value consistency over peaks. They value a network that simply works every time they need it.
DOCSIS reliability: Stronger than perception
Despite its hybrid architecture, DOCSIS continues to outperform expectations. Opensignal’s 2024 broadband reliability index[1] ranked major cable operators well above their fiber telco counterparts. The FCC’s Measuring Broadband America report[2] showed cable operators meeting or exceeding advertised speeds even during peak hours.
Why? As the panelists noted:
- A DOCSIS architecture is measurable and tunable.
- DAA, PMA, better telemetry, and shorter actives improve stability.
- Operational maturity has caught up with technical capability.
As O’Dell put it:
“COVID was proof. Traffic shifted overnight, and cable kept up. That wasn’t luck, it was years of engineering discipline.”
The perception gap: Why customers don’t always “feel” reliable service
Even with strong technical performance, customer perception sometimes lags.
J.D. Power’s 2024 broadband satisfaction study[3] showed fixed wireless subscribers reporting higher satisfaction than many wired subscribers.
The panelists unpacked this quickly:
- Micro-outages disrupt live calls.
- Latency spikes frustrate gamers and remote workers.
- Wi-Fi issues and in-home wiring create “operator blame.”
- Power issues create service drops unrelated to DOCSIS.
- Legacy expectations linger even when networks improve.
“Statistics always include outliers,” Rupe noted. “Some users genuinely get a better experience from alternative tech depending on location.”
The takeaway: Reliability must be engineered and communicated, not assumed.
PNM: Detecting problems before customers do
Proactive network maintenance is the early-warning system DOCSIS has always needed.
“Networks should be stable,” O’Dell explained. “What matters is change; how much and how often. That’s the sign of a defect worth attention.”
PNM identifies impairments days or weeks before customers notice, including:
- Upstream ingress
- Reflections
- Tilt or suckouts
- Power supply degradation
- Water in drops
- Amplitude ripple
- Noise bursts or slow declines masked by adaptive techniques
“These impairments are seeds,” Rupe said. “It’s better to catch seeds than pull weeds.”
Downey added that operators must focus on the signals that matter most: connectivity, packet loss/codeword errors, and network efficiency.
Combined, these metrics form the backbone of reliability triage.
Where operators should start
Rupe’s advice was simple and actionable:
- Use the telemetry you already have.
- Focus on your worst-first areas.
- Fix upstream noise before anything else.
- Address wet drops and plant damage.
- Show measurable improvements in 90 days.
And he’s right, PNM is not an all-or-nothing strategy. Small steps generate big wins.
PMA: Reliability in real time
If PNM is the diagnosis, PMA is the reflex.
Downey explained the fundamental shift:
“Before PMA, one noisy modem brought down modulation for the whole group. With PMA, modems are managed individually.”
Rupe offered an elegant analogy:
“PMA is your transmission. It shifts gears to match the terrain.”
PMA dynamically adjusts modulation per subcarrier, per modem, ensuring:
- Maximum performance in good conditions
- Safe fallback in impaired conditions
- No unnecessary impact on the rest of the service group
- PMA shines especially in:
- Downstream OFDM
- Upstream OFDMA
- Mid- and high-split upstream deployments
- Distributed access architectures
But PMA also produces a valuable byproduct: Its profile changes become diagnostic breadcrumbs for PNM analysis.
As Rupe joked:
“PNM and PMA are the cable industry’s peanut butter cup—better together.”
AI’s expanding role in DOCSIS reliability
We’ve hit a tipping point with AI in broadband operations. For years, AI sat at the edge of the NOC; interesting, promising, but not mission-critical.
That time is over.
In 2025, AI has become central to reliability, because no human can watch every subcarrier, every RxMER fluctuation, every equalizer tap, every retry burst across millions of modems—not with the speed and granularity today’s networks demand.
AI does what humans cannot: monitor everything, constantly, and catch the subtle patterns that predict tomorrow’s outage.
PNM gives us visibility. PMA gives us control.
AI gives us foresight.
The most exciting development, in my view, is the rise of agentic workflows. Systems that not only detect issues, but open tickets, attach evidence, propose fixes, and route problems intelligently.
That’s when “self-healing network” becomes more than a buzzword.
It becomes the daily operational reality.
The business case: Reliability pays for itself
Operators using PNM and PMA consistently benefit from:
- Fewer truck rolls
- Fewer repeated service calls
- Lower churn
- Smarter, targeted CAPEX
- Fewer premature node splits
- More stable customer satisfaction
Downey highlighted that traditional metrics like “utilization” can be misleading without proper correlation.
“High utilization doesn’t always equal congestion,” he said. “PNM and PMA help you understand what’s actually happening.”
DOCSIS vs. 5G FWA: Reliability still wins
The panel agreed: DOCSIS remains more consistent than fixed wireless access (FWA) in real-world environments.
DOCSIS excels in:
- Predictable performance under load
- Lower latency variation
- Weather resilience
- Higher aggregate capacity
- Much deeper visibility for operators
But cable must stay vigilant—power issues, Wi-Fi complexity, and in-home networks shape perception as much as plant performance.
Conclusion: The future belongs to the reliable
As we enter 2026, it’s tempting to focus solely on multi-gig speeds, extended spectrum, or new upstream capabilities. Those matter. They move the industry forward.
But they’re not what subscribers will judge us on.
Subscribers judge us on consistency.
On dependability.
On a network that works the same at 8 p.m. as it does at 2 p.m.
The next era of cable won’t be defined by peak speeds—it will be defined by flawless consistency.
PNM, PMA, and AI form the operational stack that will get us there.
If we embrace these tools, technically and culturally, we can finally shift from reactive firefighting to proactive engineering.
- From guessing to knowing.
- From best-effort to nearly always-on.
- From “fast” to “fast and unflappable”.
That’s the broadband future worth building.
References:
[2] https://www.fcc.gov/general/measuring-broadband-america
Brady Volpe
brady.volpe@volpefirm.com
Mr. Volpe, the Chief Product Officer at OpenVault, brings over 30 years of experience in the broadband cable and telecommunications industry. As the founder of The Volpe Firm, Inc., and Nimble This (now OpenVault), he has been instrumental in product development and successful launches. Through his acclaimed blog, podcast, and livestream, he shares expertise in high-speed data, DOCSIS®, PNM, HFC, ML, PMA, and more. Mr. Volpe has an unwavering commitment to broadband innovation.
Image: by author and Sora
