The important word is test. The order does not suddenly make 5G a legal command-and-control link for every Part 107 aircraft. It creates a nationwide, time-limited experiment designed to collect the evidence regulators need before making broader spectrum decisions.
What the FCC actually authorized
On 11 September, the FCC's Wireless Telecommunications Bureau and Office of Engineering and Technology granted a conditional waiver for the Department of Transportation's Mobile Network Aviation Assessment Program, or MNAAP. The waiver runs until 1 October 2029.
The program can involve up to 2,000 DOT-authorized general-aviation pilots using an app on ordinary wireless devices and up to 200 unmanned aircraft at a time carrying dedicated MNAAP modules.
That scale matters. Previous cellular-drone work has often been tied to a particular operator, network or experimental authorization. MNAAP is intended to measure commercial network performance across the contiguous United States and across a broad collection of licensed bands.
Why a waiver is needed in the first place
A mobile network looks attractive for drones because coverage already exists, modems are inexpensive, encryption and authentication are mature, and operators know how to run large networks. The regulatory picture is less convenient.
FCC rules prohibit airborne use in several flexible-use bands. The Commission has specifically pointed to restrictions affecting Cellular Radiotelephone Service and CBRS, while the Table of Frequency Allocations limits aeronautical mobile use in additional bands.
Those rules were not written around today's idea of thousands of low-altitude autonomous aircraft using commercial networks. An airborne transmitter also behaves differently from a phone at street level. At altitude it may see many more base stations, creating interference and handover conditions that terrestrial network planning did not optimize for.
MNAAP gives regulators a bounded environment in which to measure those effects instead of assuming that ground-network performance translates directly into the sky.
The test spans much more than one flavor of 5G
The waiver covers a wide spectrum range. Published summaries of the order include 600, 700 and 800 MHz spectrum; AWS and PCS; WCS and BRS; 3.45 GHz, CBRS and 3.7 GHz mid-band spectrum; and millimeter-wave bands including 24, 28, 37 and 47 GHz.
That breadth is useful because "cellular for drones" is not one engineering problem.
Lower frequencies generally offer better propagation and coverage. Mid-band spectrum offers a different balance of capacity and reach. Millimeter-wave can provide substantial bandwidth but is far more sensitive to geometry and blockage. An aircraft moving through altitude, terrain and antenna patterns can expose behavior that a ground drive test will miss.
For drone designers, this argues against treating a modem's advertised network support as evidence that the link is suitable for an aviation function. Coverage, latency, handover behavior, antenna placement, network congestion and failure modes all matter.
Cellular is attractive because BVLOS needs infrastructure
The commercial value becomes clearer when the aircraft leaves visual range.
A short-range RC link is excellent for many piloted and recreational aircraft. A docked inspection drone, delivery aircraft or long linear-infrastructure mission has a different problem. The operator needs communications beyond the local launch site, often across territory where building a private radio network is uneconomic.
Commercial mobile networks already solve a similar geographic problem for billions of terrestrial devices.
Recent BVLOS demonstrations on North Dakota's Vantis network show the other side of the same trend: scalable drone operations increasingly depend on reusable communications, surveillance and airspace infrastructure rather than a bespoke stack built for each mission.
But a cellular link is not automatically an aviation-grade link
A phone losing data service for several seconds is annoying. A command-and-control system has to treat the same event as an expected failure mode.
That changes the questions engineers should ask.
How does the aircraft behave when the modem loses service? What happens during a handover between cells? Can two independent links fail for the same geographic reason? Does the autopilot have a deterministic lost-link procedure? Is the cellular connection carrying primary C2, telemetry, payload data or merely a redundant path? How is the network's health exposed to the flight controller and operator?
These are system questions, not modem specifications.
The FCC order also requires participants to avoid harmful interference and places controls on the test equipment. The MNAAP modules must be properly authorized, approved through provider certification processes, non-transferable and decommissioned when the waiver ends.
That is a useful reminder for DIY and commercial builders: the experiment is controlled precisely because airborne use changes the RF environment.
What this means for builders today
For an ordinary U.S. Part 107 pilot, the immediate answer is deliberately boring: this waiver is not permission to attach a 5G modem and declare the aircraft's C2 system compliant.
It is still worth following closely.
A builder can already design software so that communications are transport-independent. Telemetry, mission updates and health reporting should not be unnecessarily coupled to one radio technology. Autopilots should have explicit lost-link behavior. A cellular path can be evaluated where legally permitted without making it the only thing keeping the aircraft safe.
Developers can also watch what MNAAP measures. The most valuable output will not be a headline saying that 5G "works for drones." It will be data about where it works, how performance changes with altitude, what handovers look like, which bands behave well, and what failure patterns appear.
That evidence can influence both future FCC spectrum rules and the architectures vendors choose for BVLOS aircraft.
The larger regulatory direction
The FCC is already examining broader changes to UAS spectrum policy. Its 2026 proceeding asks whether restrictions on airborne operation in flexible-use terrestrial bands should be relaxed and what interference protections would be necessary.
MNAAP gives that policy process something it badly needs: measurements.
This does not guarantee that regulators will open all of the tested bands to general drone operations. Network operators have terrestrial customers to protect, aviation links have demanding reliability requirements, and spectrum rights become more complicated when transmitters move vertically through a network designed largely around users near the ground.
The experiment can still narrow the uncertainty.
What to watch through 2029
The interesting result is not whether a drone can connect to a cell tower. We already know it can.
Watch for evidence about reliability at operational altitudes, interference into terrestrial networks, transitions between bands and cells, geographic dead zones, provider differences, and whether commercial networks can support safety functions without pretending they are deterministic private links.
If those results are strong, cellular connectivity could become one layer of the infrastructure that makes routine BVLOS less bespoke. If the results expose difficult interference or reliability limits, that is equally useful. It tells aircraft designers where cellular belongs in a redundant architecture rather than asking a consumer network to behave like something it was never designed to be.
For civilian drones, that distinction is more valuable than another connectivity marketing claim.