FCC Opens Unlicensed Direct-to-Device Spectrum to Satellites: What the Proposal Means for Business

New direct-to-device spectrum has become one of the most expensive assets in the space economy, with operators paying billions for licensed frequencies that can reach phones from orbit. On 6 August 2026 the FCC took a different route. It adopted a Notice of Proposed Rulemaking, FCC 26-51 in ET Docket No. 26-169, asking whether satellites could connect with ordinary unlicensed devices in bands that nobody owns. This article explains what was proposed, who stands to gain and what businesses should do during the comment period. It is part of the Kurums Space Economy hub.

What did the FCC decide on direct-to-device spectrum in August 2026?

The FCC adopted a Notice of Proposed Rulemaking on 6 August 2026 and released it the next day. It proposes a general framework for direct-to-device links between satellites and devices that operate under the agency’s Part 15 unlicensed rules.

The law firm Davis Wright Tremaine describes it as the agency’s first comprehensive effort to enable satellite communications for unlicensed consumer and Internet of Things devices at scale. The core proposals are new mobile-satellite service allocations in the named bands on an unprotected, non-interference basis, a clarification that Part 15 devices may operate aboard authorised spacecraft, and an exploration of satellite downlinks with safeguards for coexistence.

Chairman Brendan Carr framed the idea when he circulated the draft in July: “This new effort brings together two of the most cutting-edge areas of wireless technology in ways that help bring connectivity to ‘dead zones.'” The vote was unanimous, according to Broadband Breakfast. The proposal is billed as a complement to licensed direct-to-device services, not a competitor to them, as Light Reading reported.

Which bands are covered and how much spectrum is involved?

The adopted item covers 2400 to 2483.5 MHz and 5725 to 5850 MHz, which together amount to 208.5 MHz. Both are proposed for Earth-to-space uplinks, and the FCC is also asking about satellite downlinks in the 5.8 GHz band.

These are the frequencies used by Wi-Fi, Bluetooth and many industrial and consumer devices. The 2.4 GHz band is globally harmonised and already present in billions of products, and it sits next to the spectrum Globalstar uses for its own direct-to-device service. The 5.8 GHz band has shorter wavelengths, which allow smaller and more directional antennas. The item references the existing Part 15 power ceiling of 4 watts of equivalent isotropically radiated power, and trade coverage indicates the framework would rely on existing power levels.

Uplinks and downlinks raise different problems. A small device transmitting upward at low power is unlikely to disturb its neighbours more than it already does. A satellite transmitting downward illuminates a very large area at once, which is why Converge Digest noted that downlinks present greater coexistence challenges. This asymmetry will shape the business models that emerge: one-way sensor reporting is an easier regulatory case than two-way service.

Why was the 900 MHz band left out?

The draft released in July included 902 to 928 MHz, but the adopted version excluded it. FCC engineering staff said the band is significantly more complex, with several types of incumbent services that have differing levels of protection.

That explanation came from Andrew Hendrickson, chief of the FCC’s Office of Engineering and Technology, as reported by Broadband Breakfast. The band was contested before the vote. NextNav, which holds licences there and wants to build a terrestrial positioning, navigation and timing network, raised interference concerns and asked for the band to be excluded. Rural internet providers also voiced concerns about interference.

Not everyone agreed with the exclusion. A commentary published by Truth on the Market the day before the vote argued that every band should stay in the record so that commenters could test coexistence claims, and pointed to the lower band’s propagation and building penetration, which suit IoT sensors. For satellite IoT companies the loss is real. Low-band signals travel further and need less power, so the most battery-friendly option is off the table for now.

How would the authorisation framework work?

The FCC is asking for comment on three models for terrestrial devices and three for satellite systems. The options include licence by rule, blanket licensing and Part 25 satellite licensing along the lines of the existing Supplemental Coverage from Space framework.

The choice matters commercially. Under licence by rule, a device that already holds Part 15 certification could communicate with an authorised satellite without an individual earth-station licence. Converge Digest reported that the Commission proposes this approach, modelled on Supplemental Coverage from Space. That would mean billions of existing chips are potentially addressable with a firmware change, subject to link budgets. A heavier regime with per-system licensing would slow entry and favour companies with regulatory staff.

The item also touches less obvious uses. It asks about Part 15 equipment inside spacecraft, between spacecraft and during extravehicular activity. It also raises international questions, since unlicensed bands are not allocated to satellite services in the same way worldwide and coordination runs through the International Telecommunication Union. Any US rule would cover US territory; global service would need other administrations to follow.

Who stands to gain commercially?

The clearest beneficiaries are satellite IoT operators that already work with unlicensed technology, device and chip makers, and large constellations that could add a low-cost tier. Users in remote areas gain coverage without buying specialised hardware.

There is precedent. The FCC has previously authorised Hubble Network to receive Bluetooth Low Energy signals by satellite, and trade coverage also cites a waiver for Spire Global. A general rule would replace one-off permissions with a predictable path, which lowers the cost of capital for small operators. Readers following young companies in this segment can find more in our space startup news archive.

Large operators are positioned too. SpaceX already offers direct-to-device service; Converge Digest reported that Starlink Mobile counted 7.4 million unique devices a month across roughly 30 countries as of 31 March. Amazon is acquiring Globalstar and its Leo unit is developing a direct-to-device system for 2028, according to Light Reading. For these companies unlicensed access would be an addition to licensed spectrum, useful for trackers, wearables and sensors that do not justify a cellular modem. Our Starlink business story covers how the broadband base supports such extensions.

Who objects, and what are the interference concerns?

The main concerns come from existing users of the bands. Davis Wright Tremaine lists interference risks to Wi-Fi and IoT operations, national security and cyber threats, and international coordination as issues the FCC itself flags in the item.

Unlicensed spectrum works because everyone accepts interference from everyone else at low power. Adding satellites changes the geometry. A receiver in orbit sees the combined noise from an enormous number of devices on the ground, which makes reliable reception hard. A transmitter in orbit could raise the noise floor for Wi-Fi over a wide region. The proposal answers the second concern by making satellite operations unprotected and non-interfering: if they cause harm they must stop, and they cannot claim protection from others.

That condition is also the commercial weakness. A service with no interference protection cannot guarantee availability, which limits it to applications that tolerate delay and loss. Businesses that need assured links, for example safety systems or financial transactions, will still buy licensed capacity.

How does this fit the wider race for licensed spectrum?

It arrives after a year of very large licensed spectrum deals. Carr has said the agency’s decisions “spurred nearly $50 billion in D2D spectrum deals over the last year alone.” Unlicensed access offers a route into the market for companies that cannot pay those prices.

The transactions he refers to include SpaceX acquiring 65 MHz of spectrum from EchoStar for $19.6 billion, Amazon purchasing Globalstar for $11.6 billion and Rocket Lab buying Iridium and its spectrum for about $8 billion, as listed by Broadband Breakfast. AST SpaceMobile takes another route, providing service over terrestrial spectrum held by AT&T and Verizon. We examine the Iridium transaction in our analysis of the Rocket Lab Iridium acquisition, and the strategy behind it in our Rocket Lab company story.

The two tracks reinforce each other. Licensed spectrum supports premium, protected services such as voice and broadband to phones. Unlicensed spectrum could support a high-volume, low-margin layer of connected objects. Capacity in orbit is growing at the same time: the 26 Starlink V3 satellites deployed in September, covered in our report on the first Starship orbital flight, show how quickly supply can expand once heavy launch is available. More satellites with more sensitive antennas make weak unlicensed signals easier to receive.

What happens next, and what should businesses do?

Comments are due on 9 November 2026 and reply comments on 7 December 2026. After that the FCC reviews the record and may draft final rules, which would need another vote. No date for final action has been set.

A realistic expectation is that final rules, if adopted, come some time in 2027 or later. Rulemakings with contested interference questions often take longer than their sponsors hope. The FCC also plans other spectrum actions on a similar horizon; Broadband Breakfast noted auction preparations aimed at 2028. Companies should treat this as an option that may open, not a certainty to build revenue forecasts around.

Three practical steps follow. Device makers should check whether their existing radios could close a link to low Earth orbit at Part 15 power, since that decides whether the opportunity is real for their products. Satellite operators should model unprotected service honestly, with availability figures that reflect shared spectrum. And enterprise buyers of asset tracking or remote monitoring should ask suppliers which bands they rely on and what happens to service levels if the rules land differently from the proposal.

Frequently Asked Questions

Sources

• Davis Wright Tremaine: FCC D2D satellite spectrum NPRM
• Broadband Breakfast: FCC approves proposal on unlicensed spectrum for direct-to-device, without 900 MHz
• Broadband Breakfast: FCC to vote on proposal for more direct-to-device spectrum in August
• Covington Inside Global Tech: FCC to vote on unlocking spectrum for direct-to-device operation
• Light Reading: FCC weighs D2D for unlicensed Wi-Fi and Bluetooth devices
• Converge Digest: FCC opens unlicensed spectrum frontier for direct-to-device satellite services
• Truth on the Market: The FCC is right to keep every band on the table for direct-to-device
• Exterra: FCC unanimously adopts proposal to open unlicensed spectrum for satellite direct-to-device links