Broadband policy and metrics: reassessing satellites’ role in meeting key connectivity targets

Note: This article is the third in a series inspired by the European Commission's Digital Decade targets and recent White Paper on network infrastructure. It has been commissioned by CCIA Europe. It reflects my existing well-known analysis and positions about policymakers and regulators choosing appropriate metrics and KPIs, as well as my views on satellite connectivity#.

Introduction: are broadband policy targets too arbitrary?

I’ve written a lot over the last few years[1][2] about the importance of choosing the most appropriate #metrics and #KPIs for #telecoms. Policymakers, telcos and commentators are often tempted to use figures that are easy to obtain, or which have nice PR-friendly round numbers, rather than actually thinking about good metrics or meaningful KPIs.

And they also tend to set metrics in stone for long periods of time, without giving scope for reassessment when market expectations, end-user behaviour or technology capabilities evolve quickly. This can have unintended consequences that run counter to the underlying objectives.

Nowhere is that evident than in many targets for #broadband and #mobile coverage, such as those defined by the EU’s Digital Decade[3] strategy or the UK Government’s broadband policies:

• The term “gigabit connectivity” has a nice ring to it, which means almost nobody thinks to ask “why not 700Mbps or 1.3Gbps?” when setting targets for broadband.
• Fibre or #gigabit coverage is usually measured as “homes passed”, rather than “premises connected” measuring both residential and business sites actually signed up for fast broadband service.
• In the mobile world we still measure 5G coverage in terms of a signal outside of a % of people’s homes, despite the importance of enterprise networks, indoor wireless and IoT devices.

Setting arbitrary thresholds and metrics runs the risk of what is sometimes called Goodhart's Law[4], which notes that "When a measure becomes a target, it ceases to be a good measure".? Another commonly-cited maxim is “the perfect is the enemy of the good”.

In an ideal world, everyone would have access to fibre broadband at home and work, fast 5G while mobile, and good last-metre connections via Wi-Fi, Bluetooth or other mechanisms while indoors or in-vehicle.

However in the real world, we face numerous practical, economical, temporal, security and behavioural constraints. Fibre (#FTTP) networks cannot get to every property, mobile coverage struggles with distance or obstacles like walls, and devices and applications have needs that may require specific connectivity modes. Infrastructure takes time and often multi-party cooperation to deploy. Backup or fast-deployment options are often needed.

For broadband and mobile use this means that alternatives such as satellite connectivity, fixed-wireless access (#FWA) or coverage from drones or “high altitude platforms” have important roles to play. All have seen huge advances in recent years, notably the evolution of:

• Low earth orbit (#LEO) #satellite constellations such as Starlink, OneWeb and various others expected or launching soon
• High-capacity, programmable geostationary satellites (GEOs)
• High-performance FWA based on 5G and various other wireless technologies
• Various stratospheric or lower-altitude aerial platforms

Yet in many cases, regulation and especially broadband targets and metrics have not kept up with these developments. Definitions sometimes exclude them, while often arbitrary thresholds mean they do not qualify for recognition, promotion or funding. Similar criticisms apply to inflexible targets for cloud, mobile coverage and other areas.

For instance, in the EU and US there are constraints on satellite connectivity and some forms of terrestrial FWA from being considered as, respectively a #VHCN (Very High Capacity Network), or eligible for BEAD (Broadband Equity, Access, and Deployment) funding.

In general, there seems to be more regulatory support for 5G FWA, as it tends to be emphasised by the powerful cellular industry, while satellite and non-cellular FWA access tend to be less prominent, except in debates around spectrum. Atmospheric platforms are still at early and mostly pre-commercial stages, but that should change over time.

Satellite broadband for the hardest-to-reach destinations

There are many use-cases emerging for new satellite networks – positioning, earth-observation and imaging, direct-to-device connectivity for smartphones and IoT – but perhaps the most high profile and successful so far has been simple broadband access. While this is not a new concept, it is only with the advent of large LEO constellations that it has become accessible to normal residential users, rather than specialist enterprises or branches of government.

This reflects notable advantages of cost, performance and speed of deployment for certain use-cases, especially rural users who lack fibre or even good super-fast (50-100Mbps) broadband.

In general, the cost of deploying fibre to the last 5% of premises in a country is much higher per-site than the main 95%. And that calculus applies even further for the last 1% or hardest-to-reach 0.1%.

Remote and mountainous regions, farmhouses far from the nearest road, small islands, holiday homes, houseboats, construction sites, nomadic groups such as Roma and Sami people – or even old inner-city areas with aesthetic limits on construction – are all often unrealistic for deployment of FTTP.

While some of these segments may be accessible by gigabit forms of FWA, that still likely requires fibre to the base station, which may itself be challenging for similar reasons.

In such cases, satellite is therefore the best option, especially given the last few years’ massive growth in LEO constellations and also new high-throughput GEO satellites.

Another angle here is the importance of satellite connectivity for resilience, either if fixed access networks face particular risks, or where 5G mobile is intended for use for critical national infrastructure. Satellite-based backup networks are relevant both for endpoints such as terminals for hospitals, utility sites or banks, or as redundant backhaul from cell towers.????

Overall, satellite connectivity has five main roles:

• Helping the hardest-to-reach properties to obtain (good enough) broadband rapidly, with a longer-term roadmap for gigabit, or close to it.
• Improving back/fronthaul access to 5G or other wireless sites, enabling provision of gigabit FWA, as well as reliable critical services such as public safety networks
• Enhancing connectivity for maritime, offshore and aviation sectors, as well as remote sections of railways.
• Enabling connectivity to be offered rapidly to new-built properties or temporary premises, perhaps months or even years before cell towers or FTTP infrastructure can be deployed.
• Providing much greater reliability and resilience for broadband for both citizens and public services, in the case of natural disasters, hacking, aggression or technical outages.

All this said, it is important not to indulge in overhype it. The aggregate capacity of even the most advanced constellations is, at best, single-digit percentages of terrestrial mobile networks, and a much smaller fraction of a good FTTP network. It cannot replace mainstream broadband for mass-market residential or business users, especially in dense urban areas. Neither can direct-to-device NTN connectivity enable full mobile broadband on smartphones. It is an important complementary technology, not a primary substitute.

European policy on broadband coverage and performance

European policy towards satellite broadband is less than emphatic at present, perhaps because of the legacy of existing targets and programmes for broadband that were created before widespread availability of LEO-based connectivity.

The European Commission Digital Decade initiative has a wide variety of key metrics and KPIs that it aims to attain across the EU, by 2030. Some of these targets are outside the focus here – ICT skills, access to online medical records, and the number of €1bn+ valuation unicorns for instance. But its connected infrastructure goals are different. It aims for network coverage with “gigabit for everyone”, is pushing hard for ubiquitous FTTP (or close equivalents). It also has a mandate for “10,000 climate-neutral highly secure edge nodes” which I have criticised previously.

The EECC defines a VHCN as “either an electronic communications network which consists wholly of optical fibre elements at least up to the distribution point at the serving location, or an electronic communications network which is capable of delivering, under usual peak-time conditions, similar network performance in terms of available downlink and uplink bandwidth, resilience, error-related parameters, and latency and its variation; network performance can be considered similar regardless of whether the end-user experience varies due to the inherently different characteristics of the medium by which the network ultimately connects with the network termination point”

This is a laudable goal, but hits the barriers of the geographic and other constraints discussed earlier. And while the longer-range / wider-area wireless options such as LEOs and HAPS can offer the necessary coverage, they typically have somewhat lower typical performance levels in terms of criteria such as latency and throughput.

The Commission often notes the level of investment required to meet its own broadband coverage goals, without analysis of whether these should be re-evaluated in terms of either timing or performance for the hardest-to-reach premises. This could be one of the trade-offs to be considered.

The recent Commission White Paper noted that:

“satellite broadband can bring broadband services with up to 100 Mbps download speeds to very rural and remote areas, where no very high-capacity networks are available, even if affordability remains crucial to facilitate take-up in these areas. They can also provide resilient emergency services in disaster or crisis situations. However, while satellite services can bridge the digital divide, they cannot currently replace the performance of ground-based networks”

However, there seems little current appetite to recognise satellite broadband as a mechanism for meeting #DigitalDecade targets, or to revise those targets to explicitly include alternative networks where they make most sense. The February 2024 Commission Recommendation on the regulatory promotion of gigabit connectivity and VHCNs does not even mention the word “satellite”.

The Gigabit Infrastructure Act does note the emergence of satellite broadband, but seems cautious on current services constellations. It notes that “Recent technical progress has allowed satellite-based communication constellations to emerge and gradually offer high-speed and relatively low-latency connectivity services, to […] outermost regions and rural, remote and scarcely populated areas”. However, it then references the EU’s own proposed satellite system IRIS2 rather than existing solutions, mentioning “the resources provided by Regulation (EU) 2023/588 of the European Parliament” as a possible enabler for VHCN provision.

The GIA also acknowledges the challenges of “progress towards a fast and extensive deployment of VHCN, in rural, insular and remote areas, such as islands and mountainous and scarcely populated regions” but only requests that the Commission reports on this by May 2028.

In short, there’s no immediate sign that it considers LEO or other non-terrestrial broadband as an alternative option to fibre, or maybe gigabit 5G FWA, in the near term. Possibly, BEREC’s scheduled 2025 work to update the guidelines for meeting VHCN KPIs could resolve this issue. (According to Article 82 of the EECC, “BEREC shall update the guidelines by 31 December 2025, and regularly thereafter”).

Other governments are taking satellite broadband seriously

On the other hand, certain governments – and the major national telcos they influence - are taking satellite connectivity seriously.

A good example is Japan, which has unfortunately had many instances of earthquakes or tsunamis destroying terrestrial network infrastructure, as well as its territory spanning many islands and mountainous areas. These broadband challenges are mentioned in various policy papers such as “ICT Policy Initiatives in the Ministry of Internal Affairs and Communications”[5]. In addition to new use-cases for satellite and space technology, its Space Strategy Fund[6] covers “Accelerating the construction of commercial satellite constellations”.

All the major Japanese operators are now collaborating with satellite providers either on fixed broadband access or mobile direct-to-device capabilities, with a particular focus on LEOs. NTT is working with 亚马逊 Amazon Project Kuiper , KDDI Corporation partners with SpaceX #Starlink and Rakuten has a stake in AST SpaceMobile .

As another example, the UK government’s 2023 Department for Science, Innovation and Technology Wireless Infrastructure Strategy[7] has a full section on “Satellites, drones and fixed wireless access technology for rural coverage” and mentions “an ￡8 million fund to provide capital grants to further promote new satellite connectivity to the most remote 35,000 premises”. It started a trial of LEO-based broadband in December 2022. It also notes the potential to integrate satellite into future #6G networks.

Conclusions

Aiming for equitable targets for broadband is a wise and worthy philosophy, but only as long as those targets are both non-arbitrary and actually attainable. Having arbitrary, long-standing numerical targets – especially those using round numbers purely for political simplicity – is counterproductive, as it can exclude perfectly good-enough solutions to an urgent problem, if they cannot reach that arbitrary threshold.

There also needs to be a pragmatic view of viability, especially when physical constraints such as FTTP deployment are considered. Many jurisdictions have a sensible approach to recognising “very hard to reach” locations or contexts for broadband – either with different timelines, financial support, or relaxation of specific quantitative metrics.

There also needs to be a recognition of trade-offs between different targets and goals. There may be more economic and social value in improving resiliency (for instance by stimulating deployment of backup connections) rather than a narrow focus on reaching maximum theoretical performance of a single connection.

Europe is overdue for a reassessment and update for its Digital Decade targets, especially around VHCNs, to align with the wider objectives for inclusion and economic/social benefits. A recognition of the role of more complex multi-technology networks and converged or hybrid systems would go a long way to resolving Goodhart’s paradox applied to broadband access.

[1] https://www.dhirubhai.net/posts/deanbubley_mobile-telecoms-fwa-activity-7104379043617071105-8RLp/

[2] https://www.dhirubhai.net/pulse/worst-metrics-telecoms-dean-bubley/

[3] https://digital-strategy.ec.europa.eu/en/policies/2024-state-digital-decade-package

[4] https://en.wikipedia.org/wiki/Goodhart’s_law

[5] https://www.soumu.go.jp/johotsusintokei/whitepaper/eng/WP2023/pdf/01-chap5_sec3.pdf

[6] https://www.meti.go.jp/english/press/2024/0426_002.html

[7] https://www.gov.uk/government/publications/uk-wireless-infrastructure-strategy/uk-wireless-infrastructure-strategy