Receiver resilience becomes an airworthiness requirement on 29 September

A floor for receivers, not a limit on transmitters

On 29 September 2026 the United States makes interference tolerance a condition of airworthiness. The Federal Aviation Administration's final rule, Requirements for Interference-Tolerant Radio Altimeter Systems, takes effect that day ^[1]. It writes an Interference Tolerance Mask into 14 CFR 91.220, specifying “the maximum amount of radio frequency interference (RFI) that the RA system must safely tolerate at the surface of the RA antenna”, with a safety margin above the expected environment ^[1]. The Federal Communications Commission's Upper C-band Report and Order, which licenses the transmitters that mask was modelled against, takes effect on the same day ^[2].

That symmetry is the finding. Spectrum coexistence disputes are normally settled on the transmitter side, through power limits, guard bands and exclusion zones. This one has been settled on the receiver side. The equipment that suffers the interference must now be built to survive it, on a schedule, at a cost the operator carries.

What the rule requires

The mask applies below 500 ft above ground level, across the 48 contiguous states and the District of Columbia ^[1]. It was evaluated against the FCC's technical limits for C-band base stations: 65 dBm/MHz maximum equivalent isotropically radiated power, a maximum antenna height of 450 ft above ground level, and out-of-band emissions into the altimeter band no greater than -28.4 dBm/MHz EIRP or -46 dBm/MHz conducted ^[1][2]. The FAA applied a six decibel safety margin above that modelled environment, and reduced the aggregation assumption used in the proposal from six decibels to five ^[1].

Two compliance dates apply. Aircraft operating under parts 121 and 129 with 30 or more passenger seats, or more than 7,500 lb payload capacity, must comply by 30 December 2030. Everything else, including parts 91, 125, 133, 135, 136, 137 and 194 and the remaining part 129 operators, has until 31 October 2034 ^[1].

Non-compliance is managed rather than prohibited. The rule reserves the power to supersede existing airworthiness directives “to impose operating limitations on the use of RAs that do not meet the performance requirements”, and allows the Administrator to authorise continued operation after a compliance date subject to limitations ^[1]. The enforcement instrument is the one used between 2022 and 2024: restrict the approach.

The population and the bill

The FAA's regulatory evaluation covers roughly 58,500 radio altimeters across about 40,700 civil aircraft. Of those, 27,200 units on 12,500 part 121 and 129 aircraft fall under the 2030 deadline ^[7]. Airways Magazine, reading the signed rule, gives 58,514 units on 40,730 aircraft ^[8].

Cost is disputed in a way that matters. The final rule estimates $4.8bn to $7.1bn undiscounted, at $80,000 to $120,000 per altimeter and about $40,000 for rotorcraft units ^[3][8]. The January 2026 proposal estimated $4.49bn ^[4], and some coverage has continued to quote that superseded figure. Business aviation told the FAA that real replacement cost sits at the upper bound rather than the lower ^[18].

Nothing fitted today complies

FreeFlight Systems, an altimeter manufacturer that participated in the rulemaking, states that no radio altimeter in service meets the new interference tolerance mask, and describes both a service-bulletin upgrade for fielded units and a next-generation unit still to be certified ^[9]. The industry standard intended as the means of compliance, RTCA DO-155A with its EUROCAE equivalent ED-30A, is scheduled for publication in March 2027 ^[10]. The rule does not mandate that standard. It sets a performance mask and permits legacy equipment to meet it if it can ^[1].

The sequence is therefore: the mask becomes law this month, the transmitters are licensed this month, the compliance standard arrives in eighteen months, the first deadline falls in four years and the second in eight. Upper C-band service is expected to begin in December 2030 in the 75 largest United States markets and in July 2031 elsewhere ^[6][20].

Why the receivers are exposed

The mechanism is ordinary receiver behaviour, documented for NTIA by MITRE. Strong out-of-band energy drives a receiver front end into compression, reducing sensitivity and distorting the wanted signal, and receivers are “most sensitive to interference that is in adjacent bands, i.e., those frequencies just beyond the passband where the receiver filtering has not yet been able to provide high attenuation” ^[11]. The same study is direct about cause: altimeters fitted before the 2020 FCC order “were engineered to operate in the spectrum allocation environment that prevailed at that time” ^[11]. The equipment is not defective. The environment moved.

The arithmetic explains the urgency. Radio altimeters occupy 4.2-4.4 GHz. Lower C-band 5G sits at 3.7-3.98 GHz. The Upper C-band order licenses flexible use to 4.14 GHz, with a 20 MHz guard band to 4.16 GHz and fixed satellite service retained to 4.2 GHz ^[2][6]. Separation between the highest licensed mobile carrier and the altimeter band falls to tens of megahertz, against 220 MHz under the Lower C-band arrangement.

The evidence base is thinner than the invoice

Sources disagree on how much interference has actually occurred, and the disagreement is worth publishing rather than resolving. NASA's Aviation Safety Reporting System carries crew narratives describing a captain's altimeter frozen at the on-ground indication of -4 ft after rotation at Phoenix, and uncommanded speedbrake deployment at roughly 3 ft on landing at San Francisco. Both are crew-attributed and unconfirmed ^[12]. FlightGlobal reported approximately 80 instances of aircraft system interference sourced to the FAA ^[19].

Against that, a study of Prague Airport examined 229,476 movements between September 2021 and November 2023 and found a single candidate event, concluding that no failure was shown to have been caused by C-band interference ^[13]. CTIA's own laboratory programme found that with bandpass filters fitted, tested altimeters showed “no breakpoint” against an adjacent-channel signal of +25 dBm at the receiver port ^[14]. That is the wireless industry testing the wireless industry's case. The rule rests on a modelled worst case, not on an observed failure rate.

The regional reading

Malaysia is not exposed on the transmitter side. MCMC stated in 2022 that Malaysian C-band 5G occupies 3.4-3.6 GHz, describing it as “much further down on the C-band with a broader guard band” and comparable to European and Singaporean practice ^[15]. That leaves roughly 600 MHz of separation from the altimeter band. EASA has recorded no confirmed case of 5G interference with radio altimeters in Europe, while noting that European Commission plans for 3.8-4.2 GHz will probably require retrofit in due course ^[16]. The problem is being exported on a delay, not avoided.

Regional exposure is on the operator side. Carriers serving the United States under part 129 are captured by the 2030 deadline, and the FCC's altimeter rebate programme is restricted to aircraft on the United States registry ^[2]. IATA petitioned the D.C. Circuit on 31 August 2026 against that exclusion, arguing the Commission's reasoning is unsupported by the record ^[17]. Regional carriers should assume they are paying for their own retrofits.

What to do with this

• Write receiver selectivity into spectrum-adjacent procurement as a stated performance requirement rather than an assumed property ^[1].
• Operators flying into the United States under part 129 should build fleet equipage plans this year; manufacturers advise part 121 operators to file theirs by the end of 2026 ^[9].
• Record and report interference events. The arguments above were decided on a contested record assembled from operator reports ^[12][13].
• Watch 1 January 2028, when the voluntary mitigations agreed with United States carriers around 50 airports expire, two years before the first deadline ^[5].

Domain close

Receiver resilience has been a design preference for as long as spectrum has been shared. In one jurisdiction and one band it is now a legal minimum with a date attached. Any organisation operating a fielded receiver population should read the schedule rather than the headline. Six years elapsed between the FCC allocation that created the problem and an enforceable receiver standard, and that standard arrives eighteen months after the emitters are licensed. The interval, not the mask, is the part that will repeat.