Out of Orbit: How Outdated Spectrum Rules Hold Back Satellite Broadband
Executive Summary
The International Telecommunication Union’s (ITU) equivalent power-flux density (EPFD) limits govern how much signal power non-geostationary orbit (NGSO) systems may direct toward geostationary orbit (GSO) receivers. Adopted in 2000, those limits rest on outdated reference links, antenna patterns, propagation models, and worst-case assumptions. They also treat NGSO systems as functionally subordinate even though NGSO and GSO systems generally hold co-primary rights in the relevant spectrum bands. The result is an overly conservative regime that restricts coverage, power, and spectrum reuse while leaving valuable capacity idle.
A modern framework should replace theoretical masks with performance-based standards tied to actual network effects. Existing rules for nearby frequency bands, U.S. sharing frameworks, and the interference GSO systems already tolerate from one another all point toward similar benchmarks. For systems using adaptive coding and modulation (ACM), long-term protection should focus on throughput degradation, while short-term protection should measure absolute increases in link unavailability. For systems that do not use ACM, an interference-to-noise ratio (I/N) can provide a more appropriate measure. Field tests in Romania, Colombia, Nigeria, Botswana, and Jordan show that these standards can permit far more intensive NGSO operations without materially degrading GSO service.
Reform would yield benefits well beyond the satellite industry. More efficient spectrum use would increase capacity, lower deployment costs, strengthen competition, and expand broadband access in rural and remote areas. The United States has already begun implementing performance-based rules domestically, creating a practical model for international reform. Administrations should use that operational record to build consensus before the 2027 World Radiocommunication Conference (WRC-27), reject cosmetic changes that preserve the current regime’s core defects, and modernize Article 22 around verifiable performance rather than obsolete precaution.
I. Introduction
In the final minutes of the 11th plenary meeting concluding the International Telecommunication Union’s (ITU) 2023 World Radiocommunication Conference (WRC-23), delegates issued a clear directive:
[T]o ensure the rational, equitable, efficient and economical use of the radio-frequency spectrum and associated orbit resources, with a focus on non-GSO satellite systems including compatibility among systems. . . . WRC-23 invites ITU-R to conduct technical studies on the EPFD limits in Article 22 . . . in order to ensure the continued protection of GSO FSS and BSS networks, and to inform WRC-27 of the results of the studies[.][1]
That directive followed mounting evidence before both the ITU and national regulators that critical spectrum resources remain underused and inefficiently allocated. The equivalent power-flux density (EPFD) limits exemplify the problem. These limits govern how much signal power a non-geostationary orbit (NGSO) satellite system may deliver to the receiving antennas of a geostationary orbit (GSO) system. They have remained largely unchanged since their adoption in 2000.
The limits originally sought to provide GSO operators with stable expectations about interference protection. Over time, however, they have become perhaps the largest regulatory constraint on next-generation satellite broadband. Launching, operating, and maintaining communications infrastructure hundreds of kilometers above Earth already presents formidable technical and commercial challenges. Outdated EPFD limits add an avoidable one.
Satellite technology has advanced considerably since 2000. The rules have not. The task now is to preserve appropriate protection for GSO infrastructure without unnecessarily constraining newer NGSO systems.
Some national administrations have begun that work. The United States, for example, recently adopted new interference-protection criteria for certain satellite communications within its jurisdiction. At the same time, the ITU Radiocommunication Sector’s Working Party 4A is conducting technical studies and developing recommendations for the 2027 World Radiocommunication Conference (WRC-27).
The immediate question for policymakers, regulators, and satellite operators is what to expect as the ITU completes those studies and prepares for WRC-27. The broader questions are more consequential: How should the EPFD regime change? What level of protection should GSO systems receive? And do special protections designed for an earlier satellite market remain justified as communications increasingly shift toward low-Earth orbit (LEO)?
This issue brief addresses both the immediate process and the longer-term policy stakes. Section II explains how the ITU developed the EPFD limits more than 25 years ago, including the assumptions and methods that shaped them. Section III identifies potential areas for reform, drawing on recently adopted sharing regimes in nearby frequency bands and analogous rules governing NGSO and GSO systems within national jurisdictions. Section IV examines the first real-world tests of modernized protection limits on operating satellite infrastructure and shows that updated limits can protect GSO systems while giving NGSO operators greater flexibility. Section v assesses the economic and social benefits of technically sound interference protections in core satellite bands. Section VI then maps the path to reform, from action by national administrations to the adoption of updated international limits through the ITU.
II. An EPFD Regime Built for a Vanished Satellite Market
Nearly 30 years ago, the international community adopted EPFD limits to restrict how much signal power NGSO systems could impose on GSO networks.[2] Because satellite operators share spectrum, some interference safeguards were necessary. The limits aimed to protect incumbent GSO networks in heavily used fixed-satellite service (FSS) bands from NGSO systems that, at the time, remained largely theoretical.
But the rules did more than manage interference. They created a hierarchy. Although NGSO and GSO systems generally hold co-primary status in the relevant bands, the EPFD regime effectively made NGSO operations subordinate to any GSO system using the same frequencies. No express allocation rule required that result. The methodology simply built second-class treatment into the interference limits.
That imbalance has aged poorly. Much of the satellite industry’s recent growth has occurred in LEO. NGSO systems now account for more than 90% of new satellite-broadband capacity, a share that has tripled in five years.[3] Nearly all active satellites launched in 2025 entered LEO,[4] and commercial subscriptions to NGSO broadband services now substantially exceed those to GSO services.[5]
Whatever rationale once supported treating NGSO systems as secondary no longer fits the market.[6] As the United States has observed, EPFD limits “represent the most constraining regulatory restrictions imposed on non-GSO systems” and rest on assumptions that “significantly differ from the modern satellite systems in operation and under development today.”[7]
Three features explain the disconnect. First, the regime relies on obsolete GSO reference links, antenna patterns, propagation models, and worst-case geometries that exaggerate interference risk. Second, its aggregate and single-entry limits emerged from incomplete technical analysis, arbitrary assumptions, and political compromise. Third, those flaws force NGSO operators to avoid large portions of the GSO arc, reduce transmission power, and limit the number of satellite beams that simultaneously serve a location on the same frequencies—sacrificing coverage, capacity, and service quality even where harmful interference is unlikely.
A. Outdated Assumptions Embedded in the EPFD Limits
The EPFD limits rest on a set of reference links, antenna patterns, and operating assumptions intended to approximate GSO systems at the time. More than 25 years later, many of those inputs no longer reflect how modern satellite networks operate. The result is a methodology that often overstates likely interference and unnecessarily constrains NGSO systems.
1. GSO Reference Links
Developing the EPFD limits first required selecting representative GSO links against which to simulate the aggregate interference from NGSO systems.[8] The ITU supplied those reference links through Circular Letter 116.
Subsequent analyses have identified several problems with both the reference links and their use in developing the limits. Most notably, many of the assumed power levels, noise temperatures, and other technical parameters have been obsolete for years. The links reflect an era before modern signal-filtering techniques, when GSO receivers were more vulnerable to interference.
Modern GSO systems operate under different conditions. Operators often use power levels above those assumed in the reference links and deploy advanced filtering and signal-processing tools, including adaptive coding and modulation (ACM),[9] which adjusts a signal in response to changing link conditions. The propagation models used in the current EPFD methodology are also more than 10 generations old.[10] The calculation therefore rests on a baseline that no longer resembles modern GSO operations.
2. Antenna Patterns
The EPFD methodology also relies on theoretical receive-antenna patterns developed for the parabolic antennas common at the time.[11] Those patterns assume greater sensitivity to off-axis interference than modern equipment typically exhibits.
In particular, the average sidelobes of modern antennas—the areas outside the main communications beam where unwanted signal reception may occur—are substantially lower than the theoretical patterns used in the EPFD calculations. The methodology therefore exaggerates the interference that NGSO systems are likely to cause. One estimate finds that the current antenna assumptions overstate the EPFD reaching GSO receivers by 7.7 decibels.[12]
3. Worst-Case Geometry
The ITU methodology further assumes a worst-case geometry in which a GSO system shares a frequency with the NGSO satellite nearest the GSO arc. That approach effectively presumes that an NGSO operator with several satellites in view will select the one most likely to cause harmful interference. The EPFD limits thus treat the least favorable possible alignment as though it were typical. For links that do not present this worst-case geometry, received signal power may be as much as 30 decibels—or 1,000 times—below the current short-term EPFD limits.[13]
B. Arbitrary Aggregate and Single-Entry Limits
After settling on the design assumptions described above, ITU working groups conducted studies from 1997 to 1999 to determine how much interference protection the EPFD regime should provide.[14] The resulting framework never fully resolved the underlying technical questions. Its weaknesses become clearer when the limits are separated into their aggregate and single-entry components.
1. Aggregate Interference Metrics
Interference-protection frameworks for FSS systems traditionally rely on two measures. A short-term criterion addresses sporadic, high-intensity interference that may temporarily make a communications link unavailable. A long-term criterion addresses persistent interference that raises the background noise floor and reduces data-transmission rates.[15]
Put more simply, the short-term measure limits how often a link may fall below the minimum performance needed to carry data. The long-term measure limits how much continuing interference a bandwidth-constrained communications channel must tolerate.
When the ITU developed the EPFD limits, it reached consensus only on the short-term criterion. Recommendation S.1323 had recently assigned interference from NGSO systems 10% of the aggregate unavailability allowance for FSS networks operating below 30 gigahertz.[16] The EPFD regime adopted the same figure, concluding that “[i]nterference from Non-GSO systems to GSO networks should be responsible for 10% (aggregate) of the unavailability time in a GSO network.”[17]
Later analysis has shown that the 10% relative measure—and relative interference measures more generally—lacks a sound technical basis.[18] Current ITU studies acknowledge that the record does not explain why delegates selected 10% or how they derived it.
The long-term criterion presented an even greater problem. Existing recommendations offered no benchmark, and administrations could not agree on an appropriate protection level. Yet some long-term limit was necessary to permit co-frequency operations while giving GSO operators predictable protection. Without a uniform, technically validated standard, administrations attempted to extrapolate long-term limits from the agreed short-term criterion.
That approach conflated two distinct forms of interference. Short-term and long-term criteria address different phenomena, which is precisely why both exist. Deriving one from the other required administrations to construct operating limits without a reliable technical foundation.
The resulting proposals varied dramatically. For Ku-band downlinks, the United States proposed a long-term EPFD limit 20 times more restrictive than France’s proposal.[19] In the Ka-band, France proposed a limit 30 times more restrictive than the U.S. proposal.[20] These disparities reflected the arbitrary method used to derive the limits, not meaningful differences in interference risk.
The international community nevertheless adopted both short-term and long-term EPFD limits at WRC-2000. With little technical evidence to guide the long-term standard, political bargaining filled the void. The Radio Regulations imposed different EPFD masks across frequency bands, but those distinctions did not consistently track propagation characteristics or other technical conditions. They instead reflected which GSO operators occupied each band and how much influence their administrations wielded at the ITU.
In parts of the Ku-band, delegates adopted a compromise near the median of the competing long-term proposals.[21] In the upper Ka-band, they selected the most restrictive proposal on the table.[22] Those aggregate limits remain in force. Because they emerged from uncertain calculations and political compromise, they often provide GSO networks with more protection than technical conditions warrant. The result is an overly conservative regime that constrains NGSO operations and leaves valuable satellite spectrum underused.[23]
2. Single-Entry Metrics
Aggregate limits define the total interference that a GSO system must tolerate, but they do not tell an individual NGSO operator how much interference its system may generate. The ITU therefore had to divide the aggregate allowance into single-entry limits for each constellation.
Here again, the final measure emerged from negotiations among administrations rather than a clearly demonstrated technical standard.[24] The ITU drew from Recommendation S.1323 and assumed that 3.5 homogeneous NGSO systems would operate simultaneously.[25] It then apportioned the aggregate allowance among those systems, producing an individual long-term interference-to-noise ratio (I/N) of roughly -20 decibels for 80% of the time. That threshold would be considered conservative for most secondary uses of a spectrum band.
The calculation also assumed that every NGSO system would resemble SkyBridge, the proposed constellation used as the reference model.[26] SkyBridge never became commercially viable and never launched a commercial satellite. Today’s EPFD limits therefore rest partly on the architecture of a system that never operated.
These assumptions have not aged gracefully. Combined with the framework’s broader methodological flaws, they produce limits that are excessively conservative, waste spectrum capacity, and impose unnecessary constraints on modern LEO constellations.
C. How the EPFD Limits Constrain NGSO Operations
Flawed design assumptions and an unreliable calculation methodology have produced EPFD limits that are far more conservative than modern operating conditions require. To comply, NGSO operators generally must rely on some combination of three responses: avoiding the GSO arc, reducing transmission power, and limiting the number of co-frequency beams.
1. Arc-Avoidance Angles
GSO satellites typically operate with 2 to 3 degrees of separation along the GSO arc. Yet NGSO systems may have to avoid that arc by as much as 18 degrees.
These wide exclusion zones can reduce NGSO coverage by more than 30%, particularly near the equator and at midlatitudes. Operators must then launch additional satellites to compensate for coverage that the rules, rather than any demonstrated interference risk, have taken away.[27]
2. Power Limits
NGSO operators must also reduce transmission power to limit the signal reaching GSO receivers. These restrictions can apply even when an NGSO transmission falls well outside the required arc-avoidance angle.
Lower power means lower data rates and weaker signal performance for NGSO users. The regime thus sacrifices service quality even where the risk of harmful interference may be slight.
3. Number of Co-Frequency Beams
NGSO operators may also limit the number of satellite beams that simultaneously serve a location on the same frequencies, a measure known as “Nco.” This constraint sharply reduces spectrum reuse and system capacity.
Under the current EPFD limits, operators may have to disable beams from satellites that sit well outside the GSO arc but still exceed the prescribed protection criteria. As a result, substantial portions of an NGSO constellation may remain unavailable to serve consumers even when they pose no realistic risk of harmful interference.
Taken together, these constraints underscore the EPFD regime’s growing disconnect from modern satellite operations. The limits rested on questionable assumptions when adopted and have become still less defensible as technology has advanced.
Reform should stop treating each NGSO system as though it holds secondary status in shared satellite bands. It should instead rely on verifiable metrics grounded in measured signal performance, not theoretical worst cases or arbitrary thresholds. The next section outlines a methodology for building that more modern sharing framework.
III. Building a Modern Performance-Based Framework
Since the ITU adopted the EPFD limits in 2000, technological innovation has transformed both the economics and operation of satellite communications. The existing rules belong to an earlier era. A modern framework should reflect today’s multi-orbit satellite market and focus on measurable effects on system performance and consumers.
Reform should begin with metrics grounded in real-world evidence and modern satellite capabilities, including ACM. Three existing benchmarks offer a practical foundation: the Q- and V-band sharing regime, the FCC’s NGSO-NGSO framework, and the interference levels GSO systems already accept from one another. Together, they point toward performance-based standards for both long-term throughput degradation and short-term link unavailability.
The framework must also account for GSO systems that do not use ACM, using I/N thresholds where throughput-based metrics do not fit. And any revised standard must rest on updated reference links, antenna patterns, and simulation methods that reflect how modern constellations actually operate.
This section focuses on single-entry limits, which remain the principal concern of national administrations and the ITU. Aggregate limits raise separate questions that require further study. For now, the more urgent task is to define clear, enforceable rights and obligations for individual constellations so operators can invest and deploy with predictable expectations.
A. Performance-Based Interference Benchmarks
Modernizing the EPFD regime does not require inventing a new framework from scratch. Regulators already use performance-based sharing rules in nearby frequency bands and in analogous NGSO-NGSO and GSO-GSO settings. Each offers a practical benchmark for replacing abstract worst-case assumptions with limits tied to actual system performance.
1. Q- and V-band Sharing Regime
The shortcomings of the existing EPFD limits have been widely discussed in international forums. By 2019, the international community had identified many of the regime’s inefficiencies and counterintuitive results.[28] When the ITU turned to previously undefined FSS frequencies between 37.5 and 51.4 gigahertz, commonly known as the Q- and V-bands, it faced a basic choice: repeat the EPFD model or adopt a framework better suited to modern satellite systems.
The ITU chose the latter. It approved two resolutions that rejected the existing EPFD limits in favor of an NGSO-GSO sharing methodology tied to actual system performance.[29]
That approach reflected the widespread use of ACM, which allows a satellite link to remain connected as signal conditions deteriorate by reducing its data rate. Rather than treating any degradation as a binary loss of service, the new methodology measures how interference affects throughput over time.
More specifically, the framework calculates the time-weighted average amount of data transmitted per hertz of spectrum, using the carrier-to-noise ratio measured across GSO reference links under ITU Recommendation S.2131.[30] That process produced a consensus long-term, single-entry protection criterion of 3% degraded throughput.[31] If interference from an NGSO system reduces a GSO system’s throughput by more than 3%, the NGSO operator must adjust its operations.[32]
The growing use of ACM along the GSO arc strengthens the case for applying a similar throughput-based methodology elsewhere. Although propagation conditions differ somewhat between the Q- and V-bands and the more heavily used Ku- and Ka-bands, the newer framework provides a technically grounded benchmark for reform.
2. NGSO-NGSO Analogy
A second benchmark comes from sharing rules already used within the same Ku- and Ka-band frequencies now governed by EPFD limits.[33] In the United States, the Federal Communications Commission (FCC) has adopted a framework for sharing among NGSO systems with different access priorities based on the order in which their applications were processed.[34]
The FCC’s rules for protecting earlier-round NGSO systems from later entrants draw directly from the ITU’s Q- and V-band methodology.[35] The principal difference concerns short-term interference. After reviewing hundreds of simulations, the FCC concluded that relative short-term metrics perform poorly for ACM-enabled systems. It instead adopted an absolute increase in link unavailability of 0.4%.[36] For long-term interference, it retained the same 3% degraded-throughput threshold.
Using NGSO-NGSO sharing rules as a reference point for GSO protection rests on a straightforward premise: NGSO systems no longer warrant secondary treatment. If a performance threshold allows NGSO systems to share spectrum with one another, it offers a natural starting point for NGSO-GSO sharing as well. Receivers, after all, do not care where interference originates. A signal remains interference whether it comes from a satellite at 1,400 kilometers or 35,786 kilometers. A performance-based threshold therefore should not vary merely because the interfering system occupies a different orbit.
Nearly all satellite operators participating in the FCC proceeding agreed that the Ku- and Ka-bands should at least use the same 3% long-term throughput-degradation criterion.[37] Unless GSO receivers are materially more sensitive than NGSO receivers, the NGSO-NGSO standard offers a reasonable baseline. If GSO equipment does require exceptional protection, GSO operators should bear the burden of justifying that exception rather than shifting its cost across the NGSO ecosystem.[38]
3. GSO-GSO Analogy
The NGSO-NGSO comparison is not exact. GSO systems often use relatively static spot beams with large coverage areas, while NGSO systems use dynamic beams over much smaller footprints.[39] A third approach therefore looks to how GSO systems share spectrum with one another.
GSO systems typically account for interference from their first- and second-order neighbors, extending roughly 6 degrees in each direction along the GSO arc. Beyond that range, interference is generally considered negligible, and operators usually need not coordinate.
If satellite systems sharing the same bands should face comparable obligations, those accepted GSO-GSO conditions can help define when interference protection is no longer necessary. They can also serve as a basis for deriving appropriate NGSO obligations.
Recent studies in the ITU’s Working Party 4A have taken this approach.[40] The studies model a hypothetical GSO satellite 6.5 degrees from a victim GSO link and estimate the interference produced when both systems use the same frequency at a common terrestrial location. They then calculate the resulting short-term unavailability and long-term throughput degradation.
Those values provide de facto benchmarks for acceptable interference because GSO operators generally do not consider coordination necessary at that separation.[41] Put differently, if GSO systems routinely tolerate a given level of interference from one another, they should be able to tolerate the same level from NGSO systems using the same spectrum.
The contrast with the current EPFD regime is stark. As the studies explain, “non-GSO systems must protect GSO networks to a level nearly 30 dB more stringent than what GSO networks require from each other. This disparity becomes even more striking when we consider that both thresholds are meant to define similar regulatory concepts—the point at which interference becomes unacceptable[.]”[42] A more equitable regime would extend to all satellite systems the “interference environments that GSO networks regularly accept from other GSO networks,” regardless of orbital altitude.[43]
B. Selecting Modern Protection Metrics
The available benchmarks point toward a modern interference regime built around two performance measures for ACM-enabled systems: a long-term limit based on degraded throughput and a short-term limit based on increased link unavailability. The United States recently adopted a 3% threshold for time-weighted average throughput degradation and a 0.1% absolute increase in link unavailability.[44]
The precise values remain open to debate. But the FCC’s choices deserve close attention because they reflect technical work across the world’s largest satellite market and align closely with the benchmarks discussed above.
1. Long-Term Degraded Throughput
The long-term EPFD limits offer an obvious starting point for reform because they lack a sound technical foundation. For links using ACM, degraded throughput provides a well-established measure of persistent interference. The remaining question is where to set the threshold for unreasonable degradation.
Across the leading benchmarks, the same answer emerges: 3%.
The Q- and V-band rules now incorporated into Article 22 use a 3% long-term throughput-degradation criterion. Technical studies within Working Party 4A concluded that this threshold would preserve adequate data rates and long-term system performance in those bands.[45]
The same measure appears in the Ku- and Ka-bands. In its NGSO-NGSO proceeding, the FCC adopted a 3% threshold after reviewing technical studies of modern systems and operating conditions.[46] Those studies included 123 dynamic interference simulations.[47] In 91% of all cases—and 100% of downlink cases—long-term degradation measured 3.12% or less.[48] The FCC ultimately selected a 3% limit.[49] Working Party 4A studies applying the GSO-GSO analogy also identify 3% as a reasonable long-term threshold.[50] That convergence supports the FCC’s choice and suggests that the same measure could work beyond U.S. operations.
2. Short-Term Link Unavailability
The appropriate short-term threshold commands less consensus. Resolution 770,[51] later incorporated into Article 22 of the ITU Radio Regulations,[52] uses a 3% relative increase in instances when a link cannot meet its minimum performance requirement.
That metric performs poorly for modern ACM-enabled systems. As the FCC explained in both its NGSO-NGSO proceeding[53] and its recent NGSO-GSO order,[54] relative measures produce distorted results and invite strategic manipulation. Modern satellite systems often maintain link availability above 99%. When a link already fails less than 1% of the time under natural conditions, even a trivial increase in downtime can produce a large percentage increase relative to that small baseline. The apparent change looks dramatic even when the practical effect is negligible. The FCC therefore adopted an absolute increase in link unavailability of 0.4% for NGSO-NGSO sharing.[55] It derived that figure from hundreds of simulated operating scenarios.[56]
Studies applying the GSO-GSO analogy reach a similar result.[57] Depending on the assumed received power level, simulations of a GSO interferer operating 6.5 degrees away have produced an absolute increase in link unavailability of 0.11% or less.[58] Because GSO systems generally do not require coordination at that separation, the resulting performance offers a practical benchmark for NGSO operations.
The FCC recently adopted an even more conservative threshold for NGSO-GSO sharing: a 0.1% absolute increase in link unavailability.[59] The agency explained that this value approximates “the maximum short-term interference a GSO satellite operator in Ku-band would expect from another GSO satellite operating 6.5 degrees away on the GSO arc, a distance at which no coordination between the GSO operators would be required under the ITU Radio Regulations.”[60] Applying the same threshold to NGSO systems would “maintain a [GSO] link availability near 99.9% in the presence of an operational NGSO system.”[61]
The requirement also applies to every link in a GSO network. The most sensitive link therefore determines the NGSO operator’s system parameters and mitigation measures. Most other links will experience substantially less than a 0.1% increase in unavailability.[62]
Recent ITU studies support that conclusion. One U.S. study modeled interference from a 30,000-satellite NGSO constellation across 230 GSO reference links operating over the United States.[63] None of the customer-terminal links reached the 0.1% threshold, and 90% experienced an increase of 0.00125% or less.[64] The effects on gateway links were smaller still. The median gateway experienced an increase in unavailability of 0.0000003%, while 90% experienced an increase of 0.0000178% or less.[65]
C. Protection Metrics for Non-ACM Systems
Some GSO systems still do not use ACM for their communications. This is especially common in the Ku-band, where point-to-multipoint video services, including broadcast-satellite service (BSS) systems, remain in operation. For those links, degraded throughput is the wrong measure because the system cannot lower its data rate to preserve the connection.
Here, the GSO-GSO analogy offers a better benchmark. A single NGSO system produces roughly the same interference as two GSO satellites operating 6.5 degrees away along the GSO arc when measured against an I/N threshold of -10.5 decibels for 80% of the time. ITU coordination rules generally treat interference between GSO satellites separated by more than 6 degrees as negligible and do not require formal coordination. The same logic should apply to NGSO systems. If an I/N threshold of -10.5 decibels produces effects comparable to those already accepted among GSO operators, NGSO networks should receive the same allowance. That threshold has already been tested in several settings, including studies involving international mobile telecommunications (IMT) networks in the Ka-band.[66] More recent real-world measurement campaigns and technical studies submitted to Working Party 4A provide further support,[67] as discussed below.
D. Implementing and Updating the New Framework
Any protection metric must be practical to implement. New operators need a reliable way to demonstrate compliance, and regulators need a credible basis for enforcement. That requires simulation parameters that reflect modern satellite operations as closely as possible.
A revised NGSO-GSO sharing framework should therefore correct the flawed assumptions identified in Section II. In particular, it should:
- Update the GSO reference links to account for ACM.
- Revise GSO receive-antenna patterns to reflect the equipment used in common operating scenarios.
- Adopt simulation methods that capture how modern constellations operate, including dynamic traffic shifting that avoids worst-case geometries.
The framework should also include a process for periodically reviewing and, when warranted, updating these inputs as technologies, network architectures, and operating conditions evolve.
IV. Real-World Validation
The central flaw in the current EPFD regime is its reliance on theoretical assumptions that became obsolete decades ago—or never reflected operating systems in the first place. Reform should not repeat that mistake. Any new protection framework must account for modern conditions and demonstrate that its metrics work outside simulations and laboratory settings.
SpaceX has led several field-testing campaigns with GSO operators around the world. Taken together, these tests show that degraded throughput can serve as a workable interference metric for modern satellite systems. They also suggest that protection levels near those adopted by the FCC can expand NGSO spectrum use without materially degrading GSO service.
A. Romania
The first measurement campaign began in Romania in 2024 and examined interference to active FSS links using the Intelsat 39 satellite.[68] The tests measured desensitization caused by NGSO beams and converted those results into long-term throughput degradation under Recommendation S.2131.[69]
The campaign found that NGSO satellites could operate within 2 degrees of the GSO arc—an 89% reduction from the existing EPFD requirement—while using eight co-frequency spot beams, a 700% increase over current Nco restrictions. Even under those conditions, peak long-term spectral-efficiency loss reached only 0.7%, while the average loss was about 0.25%.[70]
B. Colombia
In Colombia, SpaceX worked with DirecTV Colombia to measure interference to a consumer dish receiving FSS signals from Intelsat 30.[71]
The tests found that an NGSO system operating with a 4-degree arc-avoidance angle—a 78% reduction from current EPFD requirements—and eight simultaneous co-frequency spot beams caused negligible long-term throughput degradation. The absolute increase in short-term link unavailability was about 0.05%.[72]
C. Nigeria
Field tests overseen by Nigeria’s telecommunications regulator examined NGSO interference to a consumer dish receiving BSS signals from Eutelsat 36.[73]
The results showed that a 7,500-satellite NGSO system could satisfy the -10.5 decibel long-term I/N threshold for non-ACM systems while using eight co-frequency spot beams and operating within 3 degrees of the GSO arc.[74] The same configuration also met a 0.1% short-term unavailability threshold.[75]
The study also considered aggregate interference. It found that a constellation with twice as many satellites operating about 2.75 degrees from the GSO arc would still remain below the NGSO-NGSO unavailability threshold and satisfy a -6 decibel long-term protection criterion.[76]
D. Botswana
Testing in Botswana similarly examined expanded NGSO operations near BSS links using Intelsat 20.[77]
The results showed that NGSO satellites operating within 4.5 degrees of the GSO arc and using eight co-frequency spot beams remained well below the -10.5 decibel long-term I/N threshold for non-ACM systems.[78] The resulting short-term link unavailability was about 0.0005%.[79]
E. Jordan
The Jordan campaign differed from the others by testing interference across four GSO satellites providing BSS coverage rather than a single GSO link.[80]
The results showed that a 7,500-satellite NGSO system could satisfy both a 0.1% short-term increase in link unavailability and the -10.5 decibel long-term I/N threshold for non-ACM systems.[81] It met those criteria under either of two configurations: a 3-degree arc-avoidance angle with six co-frequency spot beams, or a 4-degree angle with eight beams.[82]
V. The Economic and Social Gains from EPFD Reform
Modernizing the EPFD limits would correct decades of inefficient spectrum allocation. Real-world testing shows that NGSO systems can use substantially more capacity while preserving GSO service quality. Reform would therefore unlock valuable spectrum, lower the cost of satellite broadband, and reduce the infrastructure needed to compete.
Those gains would extend well beyond the satellite industry. Greater capacity and lower deployment costs would strengthen competition across satellite and terrestrial broadband markets. They would also make high-speed service more viable in rural and remote communities that conventional networks have struggled to reach.
Better connectivity, in turn, expands access to employment, education, health care, information, and civic life. EPFD reform is therefore not merely a technical adjustment. It is an opportunity to replace idle spectrum and regulatory scarcity with broader economic participation and social connection.
A. The Economic Case for EPFD Reform
The costs of the current EPFD regime extend beyond reduced NGSO coverage and capacity. By leaving usable spectrum idle and raising the infrastructure needed to compete, the rules suppress output, increase costs, and reinforce barriers to entry. Modernized limits would unlock substantial spectrum capacity while imposing little measurable harm on GSO systems.
1. Opportunity Costs
The full cost of the EPFD rules includes not only their direct constraints on NGSO operations, but also the opportunity costs—the benefits forgone under an inefficient spectrum regime.[83] Those costs are clearest in the large GSO-arc avoidance angles that NGSO systems must observe. If, as nearly every relevant technical study suggests, double-digit avoidance angles are unnecessary to protect GSO receivers, then the rules leave large portions of usable spectrum capacity dormant.
This problem is neither new nor unique to satellites. Early wireless regulation treated interference as a contaminant to eliminate[84] rather than an unavoidable feature of shared spectrum that can be managed.[85] That same instinct migrated from broadcasting and other terrestrial services to satellite communications.
When the ITU developed rules for co-frequency NGSO-GSO operations, it adopted an exceptionally cautious approach. The resulting protection masks sought to prevent even low-power NGSO signals from reaching GSO receivers.[86] That gave then-dominant GSO networks predictable protection, but it did so through blunt and technically inefficient limits that ignored the costs of foreclosing productive spectrum use.
The proper objective is not to eliminate interference, but to maximize joint efficiency. Nobel laureate Ronald Coase once made the point about pollution: “I am sure that pollution exists, I know that much; what I do not know is whether we have enough of it.”[87]
Radio interference presents the same basic tradeoff. It is a byproduct of socially valuable activity—in this case, satellite communications. Sound policy should not eliminate productive activity merely because it creates some interference. It should compare the costs of that interference with the benefits generated by greater spectrum use.[88]
Technical simulations and real-world measurement campaigns show that GSO-arc avoidance angles can fall by more than 80%, to as little as 3 to 4 degrees, without compromising reasonable protection. That change would open frequencies at geometries where NGSO transmissions are now barred.[89]
Economic studies estimate that this additional use could increase total spectrum capacity by 74% to 180%, depending on the band.[90] The average cost per unit of capacity could fall by 43% to 64%.[91] By expanding broadband access, increasing capacity, and lowering prices, reform could generate between $10 billion and $100 billion in economic gains.[92] The corresponding reduction in GSO spectral efficiency would remain below 2%.[93]
The Coasean trade is lopsided. Modernized limits would allow a rapidly growing NGSO sector to improve performance, lower prices, and generate billions of dollars in economic value. The resulting effects on GSO systems would remain largely de minimis.
2. Enhancing Competition
Lower launch and operating costs have fueled a second revolution in LEO, allowing NGSO systems to challenge established broadband providers and drive broader technological progress. EPFD reform could accelerate that competition.
Consider Nco. Under the modernized criteria examined in technical studies, the current limit of one active co-frequency satellite over a location could rise to as many as eight—a 700% increase in system capacity.[94] Other research finds that a hypothetical NGSO constellation requiring 462 satellites to provide global coverage under current EPFD rules could deliver the same performance with 360 satellites under updated limits, reducing required infrastructure by 28%.[95]
Those savings matter because satellite communications require enormous upfront investment and present formidable barriers to entry.[96] Reducing the number of satellites needed for global service lowers launch costs, eases the capital demands of large-scale manufacturing, and weakens the economies of scale that favor established operators.
Smaller constellations could then enter more readily and compete on their merits. They could target specialized markets, offer differentiated services, and compete through quality, features, and price rather than sheer fleet size. Satellite competition would become less of a contest in volume and more of a contest in value.
The gains would extend beyond competition among NGSO systems. Broadband markets are converging.[97] A single constellation can offer fixed and mobile services to consumers and businesses while also providing backhaul between local networks and the broader internet.
As those functions converge, traditional market boundaries become less meaningful. New satellite systems can compete not only with other constellations, but also with GSO providers and terrestrial broadband networks. Stronger competition across technologies could generate billions of dollars in additional economic value and consumer benefits.
B. Closing the Digital Divide
Modernizing the EPFD rules would produce benefits far beyond existing satellite markets. Most notably, expanded NGSO deployment offers one of the most powerful tools for closing the digital divide.
For millions of Americans—and billions of people worldwide—who lack meaningful broadband choice, satellite service may be the most viable option. That is especially true in rural areas, where rugged terrain, sparse populations, and weak commercial returns make extensive terrestrial infrastructure prohibitively expensive. Satellite networks face no comparable last-mile construction problem. Once deployed, a constellation can serve remote and densely populated areas through much of the same infrastructure.
EPFD reform would make that access easier to deliver by removing outdated constraints that limit satellite coverage and capacity. In some areas, those constraints do not merely raise costs. They can prevent viable markets from emerging at all.
The United States’ long-running efforts to connect unserved communities illustrate the problem. Across several versions of the FCC’s high-cost subsidy programs, the government has struggled to persuade providers to accept available funding and build infrastructure. High deployment costs and low expected revenues often made participation uneconomical.[98]
Phase I of the Connect America Fund awarded only $115 million of the $300 million available.[99] Phase II awarded $1.49 billion of $2.15 billion.[100] The Tribal Mobility Fund distributed just $16.6 million of its $50 million allocation,[101] while the Rural Digital Opportunity Fund awarded $9.2 billion of the $16 billion initially set aside.[102] Provider demand repeatedly fell short of available subsidies because the economics of deployment remained forbidding.[103]
Satellite broadband has since reached a level of quality that makes it a credible alternative to terrestrial service and, in many places, the only practical broadband option. Modernizing NGSO-GSO sharing rules would remove technical barriers just as reusable launch systems, vertically integrated manufacturing, and falling input costs are reducing the physical and financial barriers to deployment.
Greater spectrum efficiency would allow operators to expand capacity, increase speeds, and improve reliability. The federal government has begun to recognize that shift by revising its policies for the $45 billion Broadband Equity, Access, and Deployment program to treat NGSO service as a potential source of high-speed connectivity in unserved areas.[104] The next step is to ensure that outdated interference rules do not prevent those systems from delivering the promised service. The United States’ domestic EPFD reforms move in that direction.
The welfare gains extend well beyond faster internet access. Broadband can connect workers in isolated communities to national labor markets through remote work. It can bring specialist care to areas with few medical providers through telehealth. It can expand the resources available through schools, libraries, and community institutions. On-demand instruction can make educational opportunities less dependent on geography, while online communities can give dispersed minority groups new avenues for association and expression.
Each of these gains reflects the same basic point: better connectivity creates opportunities for economic, intellectual, and social exchange that distance once foreclosed. Modernizing the NGSO-GSO sharing framework would help turn those possibilities into practical options for communities that terrestrial networks have struggled to reach.
VI. From Technical Consensus to International Reform
WRC-23 directed the ITU Radiocommunication Sector (ITU-R) to study the EPFD limits in Article 22 and report its findings to WRC-27.[105] Since then, technical studies and field tests have produced a substantial record showing that modernized protections can preserve GSO service while allowing far more efficient NGSO operations.
The United States has already acted on that evidence through domestic reform. The next task is to convert successful implementation into broader international support.[106]
That will require two things. First, the United States and other administrations should use real-world operations to demonstrate that the new framework works, then carry that evidence into WRC-27 rather than wait for another study cycle. Second, policymakers should reject narrow compromises that smooth out a few limits while leaving the EPFD regime’s obsolete assumptions and built-in hierarchy intact.
A. Prove the Model, Then Build Consensus
Procedure and politics now appear to be the main barriers to what should be a straightforward technical improvement. Critics argue that the United States has inverted the usual ITU process by testing reforms through domestic rules and real-world operations before securing international agreement.
There is little precedent for amending the Radio Regulations without a formal WRC agenda item. The closest analogue may be the late addition of global flight tracking to the WRC-15 agenda after the Malaysia Airlines disaster.[107] But that objection says more about the ITU’s process than about the merits of reform.
The ITU’s combination of preliminary studies, regional preparations, conference cycles, and implementation periods can stretch to eight years before a new rule takes effect. In a fast-moving satellite market, that timetable is less a deliberative process than a pause button.
Other areas of internet and space governance offer a more practical model: demonstrate that a system works, then formalize it.
Google followed that path with QUIC, a low-latency internet-transport protocol designed to combine the speed of the User Datagram Protocol (UDP) with reliability features traditionally associated with the slower Transmission Control Protocol (TCP).[108] Google deployed QUIC across Chrome, Search, and YouTube and studied its performance for years before submitting it to the Internet Engineering Task Force (IETF).[109]
By the time the IETF took up the proposal, QUIC’s technical viability had already been established. Commercial adoption followed quickly.[110] The protocol carried a majority of internet traffic before the IETF published its first formal Requests for Comments.[111]
International space governance has followed a similar pattern. Rather than wait for a comprehensive multilateral treaty, the United States and its partners have increasingly relied on nonbinding political commitments and state practice. The Artemis Accords, led by the National Aeronautics and Space Administration, now include 68 signatory states and seek to establish norms that may eventually harden into customary international law.[112]
These examples offer both a precedent and a playbook. Technical studies and operational experience since WRC-23 already provide substantial evidence for reform at WRC-27. Waiting for another study cycle and WRC-31 would add four years of delay without a clear technical justification.
Over the next year, the United States should encourage commercial operations under its new domestic framework. It should also press other administrations to adopt similar rules, building a record of successful implementation and a broader coalition before WRC-27 in Shanghai.
B. Reject Cosmetic Reform
As administrations debate the future of EPFD, incumbent GSO operators developing NGSO capabilities of their own have advanced a narrower proposal. It would modestly relax the most restrictive limits, particularly in the upper Ka-band, and bring greater consistency across frequency bands.[113] That is a worthwhile objective.
The problem lies in what the proposal leaves untouched. It would preserve the outdated reference links, propagation models, antenna assumptions, and improvised long-term protection criteria embedded in the current regime. It also would retain the mistaken premise that NGSO systems should receive secondary treatment in bands where they hold co-primary status.
The proposal would therefore adjust the margins while preserving the framework’s central defects. Satellite technology is moving quickly. Regulation should keep pace rather than remain tethered to assumptions made 25 years ago.
VII. Conclusion
The evidence now points in one direction. The current EPFD limits rest on obsolete design assumptions, an incomplete methodology, and political compromises that no longer reflect the satellite market. Since WRC-23, technical studies, domestic proceedings, and real-world measurement campaigns have reached the same conclusion: modern satellite systems can share spectrum far more efficiently without compromising GSO service.
The question is no longer whether reform is technically feasible. It is whether regulators will continue imposing unnecessary costs on one of the world’s fastest-growing communications technologies. The current regime leaves valuable spectrum capacity idle, raises NGSO deployment costs, weakens competition, and limits broadband access in communities that terrestrial networks struggle to reach.
Administrations should therefore pursue comprehensive reform of Article 22 rather than modest adjustments to legacy EPFD masks. Updated rules should use performance-based criteria tied to actual network effects, including degraded-throughput measures for ACM-enabled systems and validated I/N thresholds for non-ACM services. They should also update reference links, antenna patterns, and simulation methods—and establish a process for revisiting those inputs as technology evolves.
Just as important, future rules should reflect the co-primary status of NGSO and GSO systems. The goal should not be to preserve a hierarchy built for the satellite industry of 2000. It should be to define reasonable interference obligations that protect service while allowing all operators to make productive use of scarce spectrum.
Over the coming year, administrations should expand domestic implementation, continue coordinated field testing, and use that operational record to build consensus before WRC-27. The ITU need not choose between protecting incumbent networks and enabling the next generation of satellite broadband. The evidence shows it can do both. WRC-27 should replace precautionary guesswork with demonstrated performance—and turn Article 22 from a brake on innovation into a workable foundation for a multi-orbit future.
[1] See Int’l Telecomm. Union, Minutes of the Eleventh Plenary Meeting for WRC-23, Doc. 526-E, at 4–5 (Jan. 15, 2024).
[2] See Int’l Telecomm. Union, Radio Regulations No. 22.2. National regulators later incorporated these limits into domestic rules. See, e.g., 47 C.F.R. §§ 25.146, 25.289 (2025).
[3] Press Release, Euroconsult, Non-Geostationary Orbit Constellations Redefining the High Throughput Satellites Market Landscape (Apr. 25, 2024).
[4] Jonathan McDowell, Space Activities in 2025, Version 1.4, Jonathan’s Space Report (Feb. 4, 2026), https://planet4589.org/space/papers/space25.pdf.
[5] See, e.g., Mike Dano, 2025 Global Satellite Broadband Performance Report, Ookla (Feb. 4, 2026), https://www.ookla.com/articles/2025-global-satellite-broadband-performance-report; Sue Marek, Latency Is the Achilles’ Heel for HughesNet, Viasat, Ookla (July 15, 2025), https://www.ookla.com/articles/hughesnet-viasat-performance-2025.
[6] The simulations and assumptions underlying the equivalent power flux-density limits date to 1997–1999. See United States of America Contribution to Chairman’s Report, SkyBridge System Parameters Needed for Simulations of Interference Between NGSO and GSO Systems, Doc. 4-9-11/192-E (June 29, 1998).
[7] United States of America Contribution to Working Party 4A, Working Document Towards a Preliminary Draft New Report [Article 22 EPFD Limit Studies], Doc. 4A/84-E, at 1–2 (Apr. 19, 2024) [hereinafter U.S. WP 4A Proposal].
[8] Int’l Telecomm. Union Radiocommunication Bureau Contribution to Working Party 4A, GSO FSS/BSS Reference Links Used in ITU-R Studies in 1995–2000 to Derive the Existing RR Article 22 EPFD Limits, Doc. 4A/251-E (Oct. 9, 2024).
[9] For an overview of these shortcomings, see U.S. WP 4A Proposal, supra note 7, at 3–4.
[10] United States of America Contribution to Working Party 4A, Proposed Updates to Technical Studies in Response to WRC-23 Minutes on Article 22 EPFD Limits, Doc. 4A/789-E, at 13–14 (Oct. 17, 2025).
[11] United States of America Contribution to Working Party 4A, Working Document Containing Technical Work Relating to the GSO Earth Station Gain Patterns Used by Recommendation ITU-R S.1503, Doc. 4A/792-E, at 1–2 (Oct. 20, 2025).
[12] See Comments of Kuiper Systems LLC, SB Docket No. 25-157, at 6–7 (filed July 28, 2025).
[13] Id.; see also Reply Comments of Kuiper Systems LLC, SB Docket No. 25-157, at 9 (filed Aug. 27, 2025).
[14] Int’l Telecomm. Union Radiocommunication Sector, Resolution 76, Protection of GSO FSS and GSO BSS Networks from Maximum Equivalent Power Flux Density Produced by Multiple Non-GSO FSS Systems in Frequency Bands Where EPFD Limits Have Been Adopted (WRC-2000).
[15] See Claude E. Shannon, Communication in the Presence of Noise, 37 Proc. Inst. Radio Eng’rs 10, 16–18 (1949) (demonstrating how background noise affects signal strength).
[16] Int’l Telecomm. Union Radiocommunication Sector, Recommendation S.1323-0 (1997), superseded by Recommendation S.1323-2 (approved Sept. 2002).
[17] Chairman, Joint Task Group 4-9-11, Report of the Third Meeting of JTG 4-9-11, Doc. 4-9-11/367-E, at 16 (Feb. 5, 1999).
[18] See, e.g., Revising Spectrum Sharing Rules for Non-Geostationary Orbit, Fixed-Satellite Service Systems, Second Report and Order and Order on Reconsideration, FCC 24-117, IB Docket No. 21-456 (rel. Nov. 15, 2024) [hereinafter NGSO-NGSO Order].
[19] Id. ¶¶ 40–47.
[20] Id.
[21] Id.
[22] See Contribution of Tonga (Kingdom of) and Ecuador, Working Document Towards a Draft New Report on Radio Regulation Article 22 EPFD Limits Issues, Doc. 4A/971-E (June 21, 2023).
[23] Id. at 5 fig. 4.
[24] Although ITU participants know this compromise well, no formal record explains why the parties selected that figure. The record also fails to explain what real-world operations would constitute a “0.5” system or how to assess departures from that assumption.
[25] See Int’l Telecomm. Union Radiocommunication Sector, Recommendation S.1323-0, supra note 16.
[26] See United States of America, SkyBridge System Parameters Needed for Simulations, supra note 6.
[27] See U.S. WP 4A Proposal, supra note 7, at 9–10.
[28] Int’l Telecomm. Union Radiocommunication Sector, Report S.2462-0, Sharing Between 50/40 GHz Geostationary Networks and Non-Geostationary Systems (July 2019).
[29] Int’l Telecomm. Union Radiocommunication Sector, Resolution 769 (WRC-19); id., Resolution 770 (rev. WRC-23).
[30] Int’l Telecomm. Union Radiocommunication Sector, Recommendation S.2131-1, Method for the Determination of Performance Objectives for Satellite Hypothetical Reference Digital Paths Using Adaptive Coding and Modulation (Jan. 2022).
[31] The long-term protection criterion measures the annual reduction in time-weighted average spectral efficiency for generic GSO reference links using adaptive coding and modulation. Int’l Telecomm. Union, Radio Regulations, supra note 2, art. 22, Nos. 22.5L, 22.5M.
[32] Int’l Telecomm. Union, Radio Regulations, supra note 2, art. 22, No. 22.5L.
[33] NGSO-NGSO Order, supra note 18.
[34] See 47 C.F.R. § 25.261.
[35] NGSO-NGSO Order, supra note 18.
[36] Because baseline availability for systems using adaptive coding and modulation often exceeds 99%, even a slight change in link conditions can produce a large relative increase in unavailability—and, in turn, in the derived carrier-to-noise ratio. See NGSO-NGSO Order, supra note 18, ¶ 28; Letter from Jayson L. Cohen to Marlene H. Dortch, IB Docket No. 21-456, at 2, S-6–S-9 (July 25, 2024) [hereinafter Cohen Letter].
[37] NGSO-NGSO Order, supra note 18, ¶¶ 11–16.
[38] See Principles for Promoting Efficient Use of Spectrum and Opportunities for New Services, Policy Statement, FCC 23-27, ET Docket No. 23-122, ¶¶ 20–22, 33–35 (rel. Apr. 21, 2023) (recognizing that transmitting and receiving systems share responsibility for adapting to a changing RF environment and that operators should develop error-tolerant systems where technically feasible).
[39] Many modern GSO satellites nevertheless use steerable or configurable beams that largely replicate dynamic beamforming techniques.
[40] United States of America, Proposed Updates to Technical Studies, supra note 10, at 19–22; see also Contribution of Tonga (Kingdom of) and Ecuador, supra note 22.
[41] United States of America Contribution to Working Party 4A, supra note 10, at 15–18.
[42] Id. at 21.
[43] Id.
[44] See Modernizing Spectrum Sharing for Satellite Broadband, Report and Order, FCC 26-26, SB Docket No. 25-157 (rel. May 1, 2026) [hereinafter NGSO-GSO Order].
[45] Int’l Telecomm. Union Radiocommunication Sector, Resolution 770, supra note 29.
[46] NGSO-NGSO Order, supra note 18, ¶ 19.
[47] Id. ¶ 12.
[48] Cohen Letter, supra note 36, at 2, S-6–S-9.
[49] 47 C.F.R. § 25.261.
[50] United States of America, Proposed Updates to Technical Studies, supra note 10, at 15–18.
[51] Int’l Telecomm. Union Radiocommunication Sector, Resolution 770, supra note 29. WRC-23 made minor changes to the methodology for calculating interference. See Int’l Telecomm. Union Radiocommunication Sector, Recommendation S.2157-0 (WRC-23). For an assessment of those changes, see John Pahl, Resolution 770 After WRC-23, Transfinite Sys. (Apr. 26, 2024), https://www.transfinite.com/content/Resolution_770_After_WRC_23.
[52] Int’l Telecomm. Union, Radio Regulations, supra note 2, art. 22, No. 22.5L.
[53] NGSO-NGSO Order, supra note 18, ¶¶ 19–21.
[54] NGSO-GSO Order, supra note 44, ¶¶ 54–56.
[55] Id. ¶ 28.
[56] Cohen Letter, supra note 36, at 2, S-6–S-9.
[57] United States of America, Proposed Updates to Technical Studies, supra note 10, at 19–22.
[58] Id. at 20.
[59] NGSO-GSO Order, supra note 44, ¶ 54.
[60] Id. ¶ 55.
[61] Id. ¶ 56.
[62] Id. ¶ 55.
[63] See United States of America Contribution to Working Party 4A, Proposed Updates to Technical Studies in Response to WRC-23 Minutes on Article 22 EPFD Limits, Doc. 4A/1030-E, at 2–4 (Apr. 24, 2026).
[64] Id. at 4.
[65] Id.
[66] See Reply Liaison Statement from Working Party 4A to Task Group 5/1, WRC-19 Agenda Item 1.13 (IMT), Doc. TG 5/1-411 (WRC-19).
[67] See United States of America Contribution, Proposed Updates to Technical Studies, supra note 63, at 15–23.
[68] See SpaceX Contribution to Working Party 4A, Real-World Interference Measurements in Support of Studies on EPFD Limits, Doc. 4A/542-E (Apr. 23, 2025).
[69] Int’l Telecomm. Union Radiocommunication Sector, Recommendation S.2131-1, supra note 30.
[70] Id. at 11-13.
[71] See Interference Measurement Campaign on EPFD Limits in Colombia, 46th Meeting of Permanent Consultative Committee II, Inter-Am. Telecomm. Comm’n, Doc. No. 6309/25 (Aug. 21, 2025), attached to Reply Comments of Space Exploration Holdings, SB Docket No. 25-157 (Aug. 27, 2025). During the five-month campaign, researchers first measured the signal received by the GSO terminal under free-space conditions. They then measured the effects of NGSO test beams activated at 15-minute intervals under various configurations.
[72] Id.
[73] See Article 22 EPFD Studies, Info. Doc. 16 to APM27-2 (Aug. 8, 2025), attached to Reply Comments of Space Exploration Holdings, SB Docket No. 25-157 (Aug. 27, 2025).
[74] Id.
[75] Id.
[76] Id. at 9-11.
[77] See SpaceX Contribution to Working Party 4A, Interference Measurement Campaign on EPFD Limits in Botswana, Doc. 4A/624-E (Oct. 7, 2025).
[78] Id. at 13.
[79] Id.
[80] See SpaceX Contribution to Working Party 4A, Interference Measurement Campaign on EPFD Limits in Jordan, Doc. 4A/604-E (Oct. 3, 2025).
[81] Id.
[82] Id. at 12.
[83] Eric Fruits & Kristian Stout, Comments of the International Center for Law & Economics Regarding Modernizing Spectrum Sharing for Satellite Broadband, SB Docket No. 25-157, at 7–8 (July 28, 2025), https://laweconcenter.org/resources/icle-comments-re-modernizing-spectrum-sharing-for-satellite-broadband [hereinafter ICLE Comments].
[84] See Henry E. Smith, Property as the Law of Things, 125 Harv. L. Rev. 1691, 1711 (2012).
[85] See Philip J. Weiser & Dale N. Hatfield, Spectrum Policy Reform and the Next Frontier of Property Rights, 15 Geo. Mason L. Rev. 549, 558–62 (2008).
[86] U.S. WP 4A Proposal, supra note 7, at 1–2.
[87] Bruce Lehman attributed this statement to Ronald H. Coase during an April 2003 presentation at the Federal Reserve Bank of Atlanta’s Business Method Patents and Financial Services Conference. See Federal Reserve Bank of Atlanta, Business Method Patents and Financial Services Conference Program, http://www.frbatlanta.org/invoke.cfm?objectid=A6BDAC9C-384A-4C59-9096A079D324A9B7&method=display.
[88] See Ronald H. Coase, The Federal Communications Commission, 2 J.L. & Econ. 1, 28–29 (1959).
[89] See supra Section IV.
[90] See Harold Furchtgott-Roth, The Economic Benefits of Updating Regulations That Unnecessarily Limit Non-Geostationary Satellite Orbit Systems 7 (Furchtgott-Roth Econ. Enters. Aug. 11, 2023).
[91] Id. at 8 app. A.
[92] Id.
[93] Id. at 7 app. B.
[94] See supra Section IV.
[95] Furchtgott-Roth, supra note 90, at 8.
[96] See LEO Policy Working Group, Low Earth Orbit Satellites: Policies to Promote Spectrum Sharing, Foster Competition, and Close Digital Divides (Oct. 30, 2025).
[97] Id.
[98] Low expected revenue often reflected both a limited customer base and implicit rate regulation through soft price caps. See Daniel A. Lyons, Narrowing the Digital Divide: A Better Broadband Universal Service Program, 52 U.C. Davis L. Rev. 803, 834–38 (2018).
[99] See, e.g., John Pender et al., Three USDA Rural Broadband Programs: Areas and Populations Served 3–5 (U.S. Dep’t of Agric. Econ. Rsch. Serv. Oct. 2023), https://www.ers.usda.gov/media/9071/eib-258.pdf?v=78922.
[100] Id.
[101] See Lennard G. Kruger, Tribal Broadband: Status of Deployment and Federal Funding Programs 7–8 (Cong. Rsch. Serv., updated July 17, 2018), https://www.congress.gov/crs_external_products/R/PDF/R44416/R44416.10.pdf.
[102] See Rural Digital Opportunity Fund, Universal Serv. Admin. Co., https://www.usac.org/high-cost/funds/rural-digital-opportunity-fund (last visited Aug. 10, 2026).
[103] See Gregory L. Rosston & Scott Wallsten, Overhauling the Universal Service Fund: Aligning Policy with Economic Reality, Tech. Pol’y Inst. (Aug. 28, 2024), https://techpolicyinstitute.org/publications/broadband/overhauling-the-universal-service-fund-aligning-policy-with-economic-reality.
[104] U.S. Dep’t of Com., Broadband Equity, Access, and Deployment (BEAD) Program: BEAD Restructuring Policy Notice 8–11 (June 6, 2025), https://www.ntia.gov/sites/default/files/2025-06/bead-restructuring-policy-notice.pdf.
[105] Int’l Telecomm. Union, Minutes of the Eleventh Plenary Meeting for WRC-23, supra note 1, at 4–5.
[106] Modernizing Spectrum Sharing for Satellite Broadband, Notice of Proposed Rulemaking, FCC 25-23, SB Docket No. 25-157, ¶ 41 (rel. Apr. 29, 2025).
[107] See, e.g., Peter B. de Selding, In Busan, Regulators Put Global Flight Tracking on WRC-15 Agenda, SpaceNews (Oct. 31, 2014), https://spacenews.com/42385in-busan-regulators-put-global-flight-tracking-on-wrc-15-agenda.
[108] See Frederic Lardinois, Google Wants to Speed Up the Web with Its QUIC Protocol, TechCrunch (Apr. 18, 2015), https://techcrunch.com/2015/04/18/google-wants-to-speed-up-the-web-with-its-quic-protocol.
[109] See Mark Nottingham, What’s Happening with QUIC, Internet Eng’g Task Force (Oct. 29, 2018), https://www.ietf.org/blog/whats-happening-quic.
[110] See F5, QUIC Will Eat the Internet (Feb. 22, 2021), https://www.f5.com/company/blog/quic-will-eat-the-internet.
[111] See, e.g., Matt Joras & Yang Chi, How Facebook Is Bringing QUIC to Billions, Meta (Oct. 21, 2020), https://engineering.fb.com/2020/10/21/networking-traffic/how-facebook-is-bringing-quic-to-billions.
[112] See NASA, The Artemis Accords (last updated June 25, 2026), https://www.nasa.gov/artemis-accords.
[113] See, e.g., Letter from George v. John to Marlene Dortch, Modernizing Spectrum Sharing for Satellite Broadband, SB Docket No. 25-157 (filed Mar. 2, 2026).