Summary

  • Ofcom wants the ITU to update, rather than abandon, the EPFD framework governing interference between geostationary (GSO) and non-geostationary (NGSO) satellite systems. Its near-term proposals target uneven Ka-band limits and modelling assumptions; a fuller review would follow at WRC-31.
  • The issue is how to make room for more NGSO capacity while preserving predictable protection for GSO services. The US has already adopted a different national framework, making Ofcom’s preference for an international route a live coordination choice rather than an abstract defence of harmonisation.

One number exposes the age of a global rule

A satellite sharing rule can look settled until two adjacent bands apply materially different protections to the same class of receiver. Ofcom’s October paper points to the upper and lower downlink Ka bands: for a five-metre reference GSO earth-station antenna, the long-term EPFD limits at the zero-per-cent time mark differ by 14.9 decibels. The upper band’s cited limit is more stringent. Ofcom says the least restrictive limit already used in each sub-band could reasonably be applied across the whole band, given that GSO operators already manage different protection levels in practice.

That is a proposal, not a measured finding that every receiver or service would be unaffected. The number is also not a direct measure of a customer’s speed, dropped video or outage probability. It marks a difference in the regulatory interference envelope. Turning it into a service claim would require evidence about the receiver, link, traffic, operating conditions and aggregation of emissions.

The measure at issue is equivalent power flux density, or EPFD. It represents the combined interference from non-GSO satellites as the equivalent power that a single source would deliver to a GSO receiver, taking account of the receiving antenna’s directionality. The International Telecommunication Union (ITU) uses Article 22 of its Radio Regulations to set limits that let systems share frequencies without negotiating a separate agreement between every GSO and NGSO network worldwide.

That common rule has real value. A satellite operator designs a system for service across borders; a regulator’s assurance of predictable interference conditions can support network planning and investment. The framework also keeps GSO networks—used for broadcasting, enterprise links, government and defence services, and resilience—from having to rely solely on private bargaining with every moving constellation.

But the framework was designed more than 25 years ago. Ofcom says NGSO operators argue that its limits are overly conservative and restrain capacity, efficiency and competition. GSO operators, in general, want existing protections maintained. These are broad descriptions of the two positions, not proof that every operator has the same commercial interest. Hybrid systems may give some companies incentives on both sides.

Ofcom is proposing a sequence, not a reset

Ofcom’s position is deliberately staged. Before or at the 2027 World Radiocommunication Conference (WRC-27), the ITU could examine targeted changes: align the different Ka-band limits; update the inputs used by Recommendation ITU-R S.1503, which supports compliance assessment; and improve the way aggregate interference from multiple NGSO systems is handled.

The proposed modelling work is practical rather than cosmetic. It could revisit worst-case satellite geometry, tracking strategies, traffic and terminal distributions, antenna patterns and propagation assumptions. A model that assumes an unnecessarily adverse configuration may constrain transmissions that current hardware and operating practices could manage. Yet a more realistic model must still be conservative enough to protect receivers under actual conditions, not only in the most convenient simulations.

Aggregation is the harder edge. A constellation may pass a limit when examined on its own while several co-frequency systems together create a different risk. The ITU’s 2023 Resolution 76 already calls for administrations to assess aggregate interference and coordinate necessary measures; its single-entry limits were derived using an assumption of 3.5 effective NGSO systems. Ofcom supports a clearer way to identify actual aggregate effects and take action where breaches have the potential to cause harmful interference. The proposal therefore joins flexibility with a demand for a credible enforcement path.

The second phase would be a more fundamental review at WRC-31. Ofcom says current studies remain incomplete, and there is not yet enough agreement about how measured or modelled interference effects could be translated into an enforceable global alternative. It wants principles for that work: proportionate GSO protection based on realistic service assumptions; measurable and enforceable limits; room for innovation; predictable rules; assumptions that can be updated; and international consistency.

The distinction matters. A technical adjustment can be tested against the current legal architecture. Replacing that architecture raises deeper questions: what counts as acceptable service degradation, which GSO links deserve protection, how should interference from many constellations be allocated, who measures it, and who must act when a shared threshold is exceeded?

The US has made the alternative concrete

The United States has adopted a different approach for specified bands. The Federal Communications Commission replaced EPFD compliance with performance-based protection criteria for GSO systems, extended good-faith coordination and provided technical backstops if operators do not reach agreement. The Government Accountability Office records that the rule took effect on 13 July 2026, except for two provisions whose effective dates were indefinitely delayed.

That is evidence of a national alternative, not a replacement of the ITU rules worldwide. It also does not settle the engineering debate. It shows that a regulator can choose to put more weight on performance criteria and operator coordination, with prescribed backstops. Ofcom instead argues that unilateral departures risk higher compliance costs, systems designed to the most restrictive jurisdiction and reduced predictability for global services.

The trade-off is not “innovation versus protection” in the abstract. It is whether the protection model can adapt fast enough to support new capacity while remaining clear enough for operators, administrations and customers to plan against. A national exception may deliver quicker local flexibility, but each additional regime could complicate equipment settings, filings and service commitments across borders. Staying inside the ITU process may preserve common expectations, but a slow process can harden a system that its participants increasingly regard as out of date.

Ofcom says it has begun measurement and modelling work on the sensitivity of GSO systems to interference and on which S.1503 inputs have the greatest effect on NGSO constraints. It expects to bring evidence into international discussions in the coming months. That work has not been completed. Nor has the ITU adopted Ofcom’s proposed Ka-band alignment or the wider modelling changes.

The decision now is procedural as much as technical: use WRC-27 to make limited changes that are sufficiently supported, while authorising a deeper evidence programme for WRC-31. If the initial studies reveal a clear risk, limits can remain or change narrowly. If they show that a particular assumption is excessively restrictive without materially improving protection, the rule can be updated against a defined engineering record. Neither outcome should be presumed in advance.

For a global satellite market, harmonisation is infrastructure. It makes different administrations’ filings, interference calculations and system designs legible to one another. But harmonisation is valuable only while the shared method remains credible. Ofcom’s staged approach tries to preserve the global rulebook by changing it before national rulebooks become the easier option.

Sources