Testing and Measurement Panel 3: TEST AND MEASUREMENT IN NETWORK DELIVERY AND BEYOND scte PRESENTS
create a logistics and process nightmare making sure engineers “are plugging the right box into the right OLT [Optical Line Terminal].” He also warned that chasing the cheapest ONT can backfire. In one example, pushing costs down too far led to inadequate power supplies – and expensive truck rolls. Gannon argued modern simulation changes everything: you can replicate operational conditions end-to-end without touching live subscribers. And instead of warehouses full of powered devices, you can simulate “tens of thousands, hundreds of thousands of devices” to stress throughput and back-office interactions (DHCP/DNS/registration storms), cutting test cycles from days to hours.
Steve Kingdom, CTO, Fixed Networks at Xantaro, argued testing is now a baseline requirement because
A recurring theme of the Test and Measurement and Beyond session was that, as networks become multi- vendor, software-driven and operationally complex, test and measurement has to evolve from point checks into continuous,
modern networks are inherently complex. He listed multiple layers of testing: physical fibre tests, performance, systems/OSS/BSS, qualification and regression, monitoring/ telemetry, plus multi-vendor and multi- operator interoperability. The discussion quickly centred on what mergers, acquisitions and phased buildouts mean for altnets. Moderator Paddy Paddison, CEO of INCA, warned about the risk of being told “those ONTs [Optical Network Terminals] work, don’t worry about it,” only to discover later “it’s a lie”. Shrivastava’s answer was that the hardest part is operating two networks – and multiple vendor stacks – at once, often on different software versions, with “legacy” and “modern” environments co-existing, alongside the acquired company’s tools and processes. He described an AIOps [Artificial Intelligence for IT Operations] “federated architecture” that can ingest data from “100 systems”, integrate signals, reduce noise, and output a simpler operational procedure for NOC and field teams. Kingdom said the sharpest pain is often at installation and fulfilment: if you don’t have proven interoperability (for example, one ONT that works across the estate), you
end-to-end assurance that drives decisions – not just dashboards.
Vecima’s Mike Gannon set the tone with a push for a “better-than- source” video quality enhancement that would deliver dramatic improvements in perceived stream quality, up to 80% bitrate reduction, and low- latency (around 30 milliseconds) real-time processing for live streaming and VOD.
Manish Shrivastava described VIAVI’s role as turning measurement into action across the full lifecycle (build, turn-up, verify, operate, optimise).
Digital Twins
A lively exchange followed on digital twins, simulation risk, and over-reliance on software layers. One audience member highlighted the promise – and the anxiety: “The problem in having software layers upon software layers upon software layers upon software I can continue, you know, is that when you start changing not 1, but 2, but 4, but 10, but 20 different parameters at the same time.” He argued digital twins work well in construction- like scenarios, but in networks – where change is continuous – “a single source of truth resides in the actual world, not in a digital simulation.”
His focus was multi-domain service assurance: combining active tests, passive monitoring, telemetry, KPIs/KQIs, and fault/event context, then automating “evidence packs”, repeatable diagnostics, and post-fix verification.
20
MARCH 2026 Volume 48 No.1
Made with FlippingBook - Online magazine maker