In another example of Nordic networks setting a bar for the rest of the world, Finnish service provider DNA has flipped the switch on a revolutionary smart WiFi system based on technology from machine learning start-up Domos.
Reaching 170,000 fixed network modems, DNA is the latest major operator to roll out machine learning directly on the CPE to improve home WiFi experiences.
DNA Nettipulssi (which translates as online pulse) has been launched after DNA acknowledged that WiFi networks are the root cause of most home network woes, yet detecting these issues has been a thorn in the side of remote workers, home schoolers, online gamers, video streamers, and the DNA network team itself. In other words, DNA Finland is fed up of subscribers directing call center complaints at the fixed Internet connection, when – more often than not – there is nothing wrong with the fixed network infrastructure itself, but the home WiFi signal is at fault.
DNA Nettipulssi optimizes the network and provides room-specific network improvement suggestions. The operator advises users to keep their modem up-to-date with the latest software, with version 17.2 making DNA Nettipulssi available free to all DNA fixed network modem users, as well as crucial security updates.
Domos was not actually credited directly in DNA’s press release, but we spotted the Norwegian vendor publicly celebrating the announcement elsewhere. Domos has not been given clearance by DNA to confirm which specific technologies from its WiFi toolkit have been deployed, so we can only speculate for now.
Domos’ technology is split into three core algorithms. Firstly, Domos App Classification detects apps in real-time based on a library of classified services. Next is the QED Probe, which measures the available networking resources from where it makes network quality predictions based on neural network models. QED means QoE delivered, based on the Broadband Forum’s standardized network quality metric framework (QED, MR-452) – which helps bring a level of clarity to the minefield of definitions for QoE in the WiFi industry.
Lastly, comes the Network Resource Orchestration algorithms, where resources are allocated to each application to optimize QoE for the home.
A fourth area for Domos is WiFi slicing, where machine learning algorithms train WiFi networks to become more adaptive in nature. There are many parallels between Domos’ open WiFi slicing software and Open RAN technologies in cellular networks.
Network slicing has long been proposed as the mechanism enabling multiple use cases to coexist over a common physical network, with each allocated the resources and QoS it needs under distinct SLAs (service level agreements). Network slicing can therefore be seen as the equivalent of virtualization, divorcing higher level processes from underlying hardware infrastructure.
One of the Domos differences is that while other WiFi optimization engineering outfits will claim to prioritize the most pivotal traffic over networks, they do not achieve this on an application-specific basis. One of Domos’ proudest achievements comes from randomly probing and sending data to the cloud for retraining neural networks. For example, if an end user is playing Fortnite and the estimate is predicting this flow, then Domos can prioritize that application’s traffic over everything else. This model is less about capacity and more about orchestrating to achieve business benefits such as gaming-first.
Data collected from network probe projects, such as one funded by Orange Labs at Telenor Sweden, shows that hidden node issues are a major contributor to WiFi performance anomalies.
Domos came from working on optimization techniques like band steering or channel changing, before pivoting to focus on machine learning and microservices with a standards-based approach, which is what the Domos team does best.
DNA admittedly isn’t in the same bracket as Verizon, Vodafone, and British Telecom – which we have been told are running WiFi slicing technology pilots involving Domos and Kaon Broadband.
Operators typically coincide WiFi upgrades with video platform updates, and this week’s news comes only a year after DNA launched its new DNA TV offering in Finland, an upgraded pay-TV platform with IP video features. CDN technology firm Edgeware supports DNA TV with VoD content ingest, nPVR recordings, and just in time packaging, which together meet all origin server requirements for DNA Oy. Newer capabilities include creating nPVR recordings and catch-up on VoD content, which are made possible via Edgeware StreamBuilder’s flexible and scalable technology – delivering HLS and DASH which saves on storage costs since content is only stored once.
Then in November 2020, DNA Finland appeared to become the world’s first operator to bundle Netflix with broadband – offering a Netflix family 4K subscription together with a home broadband package. DNA has been working on the outlandish double play package since the beginning of this year, taking a little longer than it would have hoped to capitalize on 2020’s heightened broadband and SVoD bingeing trends.
As our opening gambit teased, Domos is hopeful that a current case study in the US market study will blossom into a deployment later this year. Could this be Verizon?