Direct to cell (D2C) satellite services are about to change the RF noise environment across Australia.
Not incrementally. Structurally.
Systems like Starlink (Optus) and AST SpaceMobile (TPG) are transmitting from Low Earth orbit (LEO) using the same spectrum as terrestrial mobile networks.
That is intentional. It is called supplemental coverage from space (SCS).
D2C systems operate within ITU power flux density (PFD) masks and under Australian Communications and Media Authority licensing frameworks, which is why they are permitted. Those limits do not account for continuous aggregate interference at the receiver level.
The problem is simple.
A weak signal that is always present is not the same as thermal noise.
D2C does not just extend coverage. It raises the baseline RF noise floor, nationwide.
D2C turns spectrum from a mostly quiet resource into a continuously occupied one.
If you design or operate:
then your link budgets, interference assumptions and availability targets may already be outdated.
D2C effectively introduces a nationwide, time-varying interference floor that most terrestrial systems were never designed to tolerate.
Direct to cell (D2C) satellite systems introduce a persistent, low-level interference source across terrestrial mobile spectrum. Unlike traditional interference, this signal is continuous, spatially dynamic and nationwide. The result is a measurable increase in effective noise floor (around 1 to 3 dB), reducing link margin and availability for systems not designed with external interference in mind.
Supplemental coverage from space (SCS) is the regulatory framework that allows a satellite operator to transmit on a terrestrial mobile carrier’s licensed spectrum, in areas where that carrier has no terrestrial coverage. The phone in your pocket sees the satellite as just another cell. No special hardware, no separate antenna, no dish.
In Australia the two visible deployments are:
Both rely on the satellite emitting a downlink that looks, to a handset, indistinguishable from a terrestrial base station. The satellite’s beam footprint on the ground is large, on the order of tens of kilometres across, and it sweeps as the bird passes overhead.
That last point matters. A terrestrial cell’s interference contour is fixed. A D2C cell’s footprint moves at roughly seven kilometres per second.
Satellite to terrestrial coexistence is not new. C band, Ku band and the various FSS bands have been shared for decades through coordination zones, exclusion contours and antenna pattern envelopes.
What is new about D2C is three things.
They were never planned for satellite downlinks. Adjacent and in-band victims include fixed links in the 1.4 to 2.7 GHz range, PMR users, telemetry, SCADA backhaul, and a long list of apparatus licences that have always assumed the only emitters in-band were known, fixed and licensed.
A geostationary bird illuminates a fixed footprint. A D2C constellation illuminates everywhere, all the time, on a duty cycle that depends on user demand and orbital geometry. There is no quiet zone you can plan a sensitive receiver into.
D2C emissions sit below ITU recommended PFD masks for the bands in question. That is how they got approved. Below the mask is not the same as below your receiver’s sensitivity. A weak signal that is one hundred percent of the time present in-band is not the same noise environment as a thermal floor with occasional bursts.
This is the heart of the direct to cell interference problem. It is not a single strong emitter. It is a permanent, low-level rise in the in-band noise floor.
It behaves like noise, but it is not random, and it is not intermittent.
The remote site noise floor in Australia has historically been close to thermal in the bands above about a gigahertz, particularly outside towns. That is one of the main reasons regional fixed links and rural IoT work as well as they do at low transmit powers.
Once D2C is operating overhead, the in-band noise floor at any given site is no longer purely thermal. It is thermal plus the integral of D2C downlink PFD over the satellite’s beam, weighted by your antenna pattern in the direction of the visible constellation.
For a fixed link with a moderately directional antenna pointed at the horizon, the increase is small. For a low gain omni receiver, particularly one looking up, it is not.
The practical consequence: link budgets that assumed kT B noise plus a receiver noise figure now need a margin for an external interference floor that varies with elevation and time. For high availability links, that margin needs to cover the worst case satellite pass geometry, not the average.
If you are designing to ITU-R P.530 availability targets, the rain fade and multipath calculations do not change. The receiver’s effective sensitivity does.
A two or three decibel rise in the in-band noise floor, sustained, eats directly into your fade margin. On a link that was already marginal in the worst rain hour of the year, that can drop you below the availability target.
The fix is the same as it always is. Higher gain antennas, lower system noise figure, or shorter hops. The point is that you have to know to apply it.
Some D2C deployments will be in-bands directly adjacent to spectrum used for fixed services or PMR. Older receiver front ends with relaxed selectivity will see the satellite emission as a blocking signal even if it is technically out of band.
If you have legacy equipment in adjacent bands, particularly anything more than about ten years old, the assumption that adjacent band emissions are negligible is worth checking against the new D2C PFDs.
Apparatus licensees have always had a coordination right against new entrants. The mechanics of coordinating against a moving satellite footprint are not the same as coordinating against a fixed transmitter.
ACMA’s position, broadly, is that D2C operates within the host MNO’s existing licence and within agreed PFD masks, so there is no per pass coordination obligation. What there is, however, is the ability to log and report interference if you can demonstrate it. That requires baseline measurements taken before D2C reaches full operational density. If you have sensitive receivers in-band, taking those measurements now is cheap insurance.
The headline use case for D2C is exactly the part of Australia where terrestrial RF engineering has always been hardest. Mining, agriculture, emergency services, regional ISPs and remote government sites all operate in-bands that are now or will soon be shared with D2C downlinks.
Three categories of system are most at risk from direct to cell interference.
Rural fixed wireless access. WISP class deployments using sub 6 GHz mobile adjacent bands rely on a quiet noise floor for cell edge performance. D2C raises that floor.
Telemetry and SCADA over PMR. Narrowband systems with low link margins, often running on legacy equipment, have the least headroom to absorb a noise floor change.
Long range IoT. LoRa, Sigfox and similar systems operate in ISM bands that are not directly affected by D2C, but adjacent band emission considerations still apply, particularly for receivers without strong front end filtering.
For most other users, D2C is a net positive. A handset that worked nowhere now works somewhere. The trade is that the spectrum itself is becoming busier, and the design margins that used to be generous are not anymore.
The 2025 LIPD class licence updates and the broader spectrum reform work both touch on satellite to terrestrial coexistence, but neither directly regulates D2C beyond confirming that supplemental coverage from space operates within the host MNO’s licence.
That is the crux of the issue. The regulatory mechanism treats D2C as if it were just more terrestrial cells. The physics does not.
This is one of the areas where ACMA’s pivot toward dynamic and data driven spectrum management is going to matter. Static coordination contours do not capture a moving interference source. Real time measurement, reporting and coordination tools do.
If you are designing or operating RF systems in-bands the MNOs are using for D2C, or in adjacent bands, three practical steps.
D2C will expand coverage where terrestrial RF cannot economically reach.
It will also permanently change the RF noise environment in the bands it uses.
A weak signal that is always present is not the same as thermal noise.
The engineers who account for satellite to cell interference now will avoid the interference problems everyone else will be diagnosing later.
Not in the traditional sense of discrete interference events. It raises the baseline noise floor in the bands it uses, which reduces link margin for terrestrial receivers operating in or adjacent to that spectrum.
Low margin systems, legacy PMR and SCADA, rural WISP networks and omnidirectional receivers looking partly upward. Highly directional fixed links pointed at the horizon are least affected by satellite to cell interference.
Early estimates suggest around 1 to 3 dB depending on antenna pattern, elevation angle and orbital geometry. Worst case satellite passes can be higher than the time averaged figure.
D2C in Australia uses the host MNO’s existing terrestrial mobile spectrum under the supplemental coverage from space (SCS) framework. Starlink operates on Optus’ holdings and AST SpaceMobile operates on TPG’s holdings, primarily in sub 2 GHz bands at first.
No per pass coordination is required because D2C operates within the host MNO’s existing licence and within agreed PFD masks. However, apparatus licensees retain the right to log and report interference, which makes baseline noise floor measurements valuable now.
Related Insights
A working engineer's guide to ACMA's 900 MHz replan: the new band edges, what happens to legacy PMR, STL and telemetry users, how LIPD class licensing changes in 915 to 928 MHz, and what LoRaWAN, NB IoT, LTE M and private LTE deployments actually have to recalculate this year.
May 13, 2026
Regional Australia is facing a growing spectrum and connectivity challenge driven by sparse infrastructure, satellite dependence, and rising demand for wireless services. Here’s what’s really happening.
Apr 1, 2026
RF interference is becoming more common as wireless systems multiply and spectrum becomes increasingly congested. Learn what causes it, why it’s worsening, and what engineers can do to mitigate it.
Mar 25, 2026
