How 5G Puts Airplanes at Risk – An Electrical Engineer Explains

Yves here. This is a deliberately very simple explanation of the 5G/airlines controversy that NC readers have debated hotly in comments.

As I’ve indicated, one factor the FCC/cell operator defenders keep minimizing, to their peril, is that passenger and cargo airlines are fault intolerant. Having planes fall out of the sky, or merely go splat instead of landing smoothly, is extremely bad for business, as well as employee, passenger, and bystander life and limb.

This article contends (in effect) that the FCC made a big mistake in not allowing for a bigger buffer between the US 5G C band and air navigation frequencies. It puts the possible blame on old or too sensitive altimeters. But consumer grade hardware isn’t made to the finest tolerances. It’s not crazy to think that faulty cell towers or even phones could transmit a bit wide of their authorized spectrum.

And as we pointed out, upgrading altimeters is likely not just a matter of cost, even though that is a huge issue. Recall Boeing sent a warning out about 5G and its workhorse 777 model. That suggests you can’t just test and fix altimeters one by one. You’d potentially have to upgrade onboard devices, which suggests any fixes would need to take place in a tight timeframe and be closely coordinated. Think the US is capable of anything remotely like that?

By Prasenjit Mitra, Professor of Information Sciences and Technology, Penn State. Originally published at The Conversation

New high-speed cellphone services have raised concerns of interference with aircraft operations, particularly as aircraft are landing at airports. The Federal Aviation Administration has assured Americans that most commercial aircraft are safe, and AT&T and Verizon have agreed to hold off on installing their new cellphone antennas near airports for six months. But the problem has not been entirely resolved.

Concerns began when the U.S. government auctioned part of the C-band spectrum to wireless carriers in 2021 for US$81 billion. The carriers are using C-band spectrum to provide 5G service at full speed, 10 times the speed of 4G networks.

The C-band spectrum is close to the frequencies used by key electronics that aircraft rely on to land safely. Here’s why that can be a problem.

Keeping Order on the Spectrum

Wireless signals are carried by radio waves. The radio spectrum ranges from 3 hertz to 3,000 gigahertz and is part of the electromagnetic spectrum. The portion of the radio spectrum that carries the signals from your phone and other wireless devices is 20 kilohertz to 300 gigahertz.

If two wireless signals in the same area use the same frequency, you get garbled noise. You hear this when you are midway between two radio stations using the same or similar frequency bands to send their information. The signals get garbled and sometimes you hear one station, at other times the other, all mixed with a healthy dose of noise.

Therefore, in the U.S., the use of these frequency bands is tightly regulated by the Federal Communications Commission to ensure that radio stations, wireless carriers and other organizations are assigned “lanes,” or frequency spectra, to use in an orderly fashion.

Bouncing Radio Waves Off the Ground

Modern airplanes use altimeters, which calculate the time it takes for a signal to bounce back from the ground to determine a plane’s altitude. These altimeters are a vital part of automatic landing systems that are especially useful in cases where there is low visibility.

So, if an altimeter interprets a signal from a wireless carrier as the rebounded signal from the ground, it may think that the ground is closer than it is and prematurely try to lower the landing gear and do the other maneuvers that are needed to land an aircraft. If interference with wireless carrier signals corrupts and garbles the altimeter’s radio signals, the altimeter may not recognize the rebounded signal and thus be unable to figure out how close to the ground the plane is.

The portions of the radio frequency spectrum used by airplanes and cellphone carriers are different. The problem is that airplane altimeters use the 4.2 to 4.4 gigahertz range, while the recently sold – and previously unused – C-band spectrum for wireless carriers ranges from 3.7 to 3.98 gigahertz. It turns out the 0.22 gigahertz difference between the signals may not be quite enough to be absolutely sure that a cellphone carrier signal will not be mistaken for or corrupt an altimeter’s signal.

Steering Clear of Trouble – For Now

The telecommunication industry has argued that the gap of 0.22 gigahertz is enough and there will be no interference. The airline industry has been more cautious. Even if the risk is very small, I believe the consequences of a plane crash are enormous.

Who is correct? The chances of such interference are very small, but the truth is that there isn’t much data to say that such interference will never happen. Whether there will be interference depends on the receivers in the altimeters and their sensitivity. In my view, there is no way to ensure that such stray interfering signals will never reach altimeters.

If the altimeters can register the stray signals as noise and filter them out, then they can function correctly. Upgrading aircraft altimeters is a costly proposition, however, and it’s not clear who would pay the cost.

The FAA has been testing altimeters and clearing ones that can be relied on in the near future. AT&T and Verizon have agreed to not put up 5G transmitters and receivers near the 50 largest airports for six months while a solution is being worked out. This has averted a major crisis in the near term, but it isn’t a permanent solution.

Moreover, regional airlines and rural airports remain at risk of interference.

Print Friendly, PDF & Email


  1. Synoia

    Designing C Bank frequencies equipment is as much and art than engineering. In addition the number of equipment engineers with C band design skilled and experience is probably in the few dozen range, with many of the few dozen in China,

    And reflecting on the world we live on, the possibility of turning a cell phone into a weapon is a little disconcerting.

    1. Watt4Bob

      Years ago I met a radio engineer on a plane headed for Nashville. He was returning from China where he had been teaching Chinese engineers.

      This man was old, he had to be at least 80, and he mentioned that there were only a handful of people who were capable of teaching his subject.

      Sad situation where corporations with technological conflicts get to argue life and death issues in a environment almost devoid of judges qualified to draw conclusions.

      1. WobblyTelomeres

        Not so. We have plenty of engineers with applicable experience. Unfortunately, they are working for the DoD.

        Designing filters isn’t that complicated. The older altimeters may have used simple filters, such as bypass capacitors, with lazy attenuation slopes as there was nothing else in those frequencies. And they worked fine until 5G showed up. But, that is why they are having to test all the older altimeters.

        1. Watt4Bob

          My sources tell me that the reason this wasn’t figured out back when 5G was first proposed is that the Telcom folks argued successfully that it couldn’t be “proved” dangerous because the systems didn’t yet exist.

          Who was testifying for the airline industry?

          It should be clear that it was someone who couldn’t credibly convey the danger.

          My guess, one of those engineers you site, answered one too many questions with “I don’t know?”

          So tell me, how do you convince the decision makers of the necessity of caution, when “You can’t prove it.” is sufficient to get the go ahead?

          OTOH, it could be that Telcom, by enabling the surveillance state, gets what they want most of the time.

          1. WobblyTelomeres

            “I don’t know” can be a valid response. “I can’t know until we test” is often the only correct response.

            An altimeter designed by an engineer who is now dead, who worked at a company that no longer exists, and is documented by, wait, who am I kidding? Documentation? Hahaha.

            I can easily imagine a simple notch filter built out of some capacitors feeding into a simple level detector, said detector being timed by a high resolution clock, the computed time reflecting (ha) the distance from the ground. Just like bats, dolphins, radar, sonar, and lidar.

            If the older altimeters used this method or something equally simple, and if the attenuation slopes were not very steep, then, yeah, there is no way to know if they misbehave in the presence of nearby frequencies without testing them. For that reason, I think the 6 month truce should be sufficient. In the end, someone will have to pony up for newer altimeters with better filters.

            Who was testifying for the airline industry? I don’t know. Obviously, not a dead electrical engineer. Or a live one. Probably a team of worse-than-useless MBAs who were coached by a team of worse-than-useless JDs (is my bias showing?). As I have been told many times, companies are loathe to let engineers testify as they will inevitably blurt out the truth.

            But, you know this.

        2. fajensen

          Designing filters isn’t that complicated.

          Getting avionics equipment through the flight qualification process *is* complicated, even if everything goes right it might take a while and cost serious money.

          It is of course possible that the FAA’s conditioned response to the 737 MAX fiasco is to just wave everything through. Nobody respects them so they might just take the bribes and live happy lives in comfort.

      2. drumlin woodchuckles

        “Teaching Chinese engineers” . . .

        . . . Because the International Free Trade Conspiracy and its Clintonite agents had already long since committed itself to exterminating this kind of engineering from existence within the US.

  2. tegnost

    I am not an engineer, electrical or otherwise, but it seems to me that one possible explanation could be the standard neoliberal framework in operation. Two gigantic industries want profit, but not cost. Typically this is achieved by getting ,gov to cover costs. Alex Morfesis offered a reasonable theory a week or so ago…ATT/Verizon et. al. didn’t/don’t want to upgrade signal towers…too expensive (paraphrasing). So never let a crisis go to waste, the answer will of course be society (socialism for the rich donchaknow) will pay for the upgrade (hoocoodanode!) and voila’! Profit!
    Oh and by the way, ten times the speed means ten times the throttling, so Moar Profit!
    Isn’t the modern world just the absolute best thing, ever?

    1. Stephen V.

      Similar to Big Pharma, the TeleComms, as you say, have a primary mission to seek profit (aka “shareholder value”) rather than service to customers. The sick part is that these profits are er, re-invested into lobbying govt agencies: FDA for Pharma and FCC for TeleComms. These agencies are demonstrably captured. I don’t know if this explains the lack of media coverage, but there is a stunning vacuum when it comes to litigation against the FCC on the part of municipalities and public interest groups. Another externalized cost aside from signal towers is human health! FCC wireless safety standards have not been updated since…1996. Think about that. More here:

      1. p fitzsimon

        Couldn’t you say the same about the airlines. The airlines or their suppliers (Boeing? or whoever supplied Boeing) should have known about the altimeter frequency interference long before this became a last minute issue. Also I don’t think anyone can guarantee there wont be interference from broken equipment no matter how much band gap exists. If the altimeter is that critical shouldn’t there be a backup of some type?

    2. Jonathan Holland Becnel

      Sounds exactly like New Orlean’s power company, Entergy, that is very good at playing the pauper while the CEO and the board make out like Fat Cats.

  3. Sam F

    It is surprising that the altimeter band 4.2 – 4.4 GHz was not tested for compatibility with the C-band transmitters as part of approval process for the 5G band devices. FAA and FCC could have required that the new devices have sharper filters to avoid interference.

    But the precipitous declines in product reliability over recent decades show that the USG runs on bribes, and puts zero value on the public interest until they get caught. The problem is that the top levels of executive agencies are political party crooks installed by politicians, and the rest are afraid to lose their careers under the present or future administrations, if they investigate any political party racketeering.

    The DOJ and FBI completely ignored my pleas that they investigate political racketeering by Repub politicians in Florida, even after mailing full packages of evidence and legal argument to their local, DC, and OIG offices several times over three years, under both Repub and Dem administrations. I am now suing them directly as a branch of the racketeering operation, and the judiciary are faking up excuses not to request that they investigate the politicians. But of course we have no way to sue the judiciary, a major defect in the Constitution.

    But the judicial branch is also completely corrupt, and is obstructing even that case by simply lying.

    1. Oh

      Your points are well made. I too have noticed how the judiciary keep using their own interpretations of the law or simply refuse to hear cases to help their buddies.

      Unless you have plenty of money and a team of sharp lawyers (liers?) you have no chance in the US courts. Our constitution is just “a piece of paper” like Bush or Cheney said.

      I never unstood why justices are appointed by political parties and why they run for re-appointment.

  4. Questa Nota

    Airplanes fly right over my house daily. The 5G rollout raises an added concern. I may want to relocate somewhere with less radio frequency traffic, so will research the pro and con aspects.

    Is there a radio frequency analog to Dark Skies?

  5. Susan the other

    Thanks for this explanation. It is quite clear. Making it even more incomprehensible why it wasn’t explained by the media long before this. It is a straightforward problem – with an expensive solution. If airlines and cell phone companies were both utilities, operating on a non-profit basis, this would be already solved by now.

  6. redleg

    The author somehow overlooked, or maybe the editor cut, the concept that the transmitters should be tested to ensure that they aren’t transmitting into the altimeter spectra. It shouldn’t be the altimeter’s problem since the 5G spectra is new.
    This shouldn’t be that complicated- if you cause a problem, you need to fix it. Alas, that’s not how this f’d up country works.

    1. ambrit

      It’s a matter of scale. There are what, tens of thousands of aircraft altimiters, versus millions of 5G repeater stations.
      Second, those 5G repeaters are generally sited near the tops of telephone and electric poles, so, physically working on them will require either ladders or bucket trucks. Probably cheaper to replace the units with “upgraded” units. Working on so many units will take time, time we do not have if we want both fast mobile data rates and flying aircraft.
      There are so many ‘things’ that could go wrong with 5G communications. 4G is good enough. This is a case where upgrading for profit’s sake is hitting the proverbial brick wall at speed.

      1. CJ

        It’s not 5G per se, it’s the frequency. I don’t think anyone ever made a 4G/3G base station with radios for that frequency (and those technologies wouldn’t benefit like 5G does from the higher frequencies). But if they did run older standards on this particular band we’d have the same issue. It’s just that 5G was the current technology when they bought the licenses, and the press is equating 5G and this band.
        5G has been running on lots of other frequencies (600MHz, 850MHz, 1.9GHz, etc.) for quite a while. 850 is the original band you would have used on your phone 25 years ago. “Ultra 5G” or whatever marketing term they use is just 5G at a higher frequency, which can be faster because more Hz can carry more bits.

        1. Jessica

          So long since I studied this at school, but does a higher frequency for the carrier wave increase the amount of data that the signal can contain?

          1. CJ

            Yes. That’s why the fastest 5G uses millimeter-wave frequencies, say 28GHz and above. But it doesn’t go far at all, and struggles to get through walls.

            Regardless of frequency, 5G is more spectrally efficient than 4G, which is much better than 3G, etc. That means more bits per hertz. For the carriers it means they can simultaneously serve more customers (or give each of them more data) in a given area with the same spectrum (chunk of bandwidth).

          2. Alex V

            No. It does not. The reason higher frequencies carry more data in practice is that they are less crowded so more bandwidth is available. Bandwidth is the primary determinant of the amount of data that can be transmitted.

            1. CJ

              You are correct, I was being too simplistic. I was just trying to say that, for example if you owned the 25-30 GHz band you could theoretically carry 10x more data than 2.5-3.0 GHz. Sorry about that.

            2. Durans

              Actually, to some degree it does. Digital data is often transmitted by alternating two rather close frequencies. You need to see enough of a particular cycle to differentiate them. Since higher frequencies are shorter in time that means you need less time to make the determination. In this way higher frequencies can transmit more data by making the “1s and 0s” shorter and therefore transmitting at a higher rate. As to whether or not current cellphone technologies are actually pushing against these limits I have no idea.

              1. Alex V

                Frequency shift keying is not used by any widespread high bandwidth application, and is primarily used at much lower frequencies than 5G. Maybe there are some secret military applications not in the public domain. Cellphones have never used FSK for anything but basic signaling functions. More modern modulations far outperform FSK. I hedged my statement with “primary determinant”, since in engineering nothing is every absolute.




      2. Oh

        Electric utility and telephone crews are working on equipment installed high on poles. I don’t think this is any problem. Replacing units may take ime but that’s the cost of doing business.

        They want to pack (cram) more bandwidth using 5G so they can get people to pay more for playing games and streaming. I’m hoping the 5G game will go bust. They’re trying like hell to get people to buy 5G phones by offering big discounts (which they’ll recoup in a short time). “we” don’t need 5G.

    2. Duke of Prunes

      And even if you test new transmitters, it doesn’t mean they will not fall out of compliance over time. I mentioned this on earlier thread, but I was told by an RF engineer about a situation where a small and relatively weak paging system transmitter was failing in a way that caused it to take out some of a major city’s police radio system. The two systems coexisted just fine for some number of years until they didn’t. There needs to be multiple layers of safety built into critical systems.

    3. Jeff Hails

      That is a definite issue. Ham radio operators had always been responsible to prove innocence and provide corrections for any report of interference. The capabilities exist to assure strict compliance but the real world throws in conditions that can’t always be predicted or accounted for. I worked in electronics for better than 30 years. Just a dumb grunt technician. I had no fear of unemployment because things never work as well as the engineers would like or the executives will allow.

  7. Matthew G. Saroff

    The subtext to all of this is that the FAA has had years to test and address this, and sat on their hands.

    The problem is with the FAA.

    This has not been an issue in Europe, but the gap between 5G and the radar altimeters is (slightly) larger.

    I would note that we’ve seen this sort of issue, without safety issues, with a number of spectra adjacent to military uses as well.

    IMHO, the risks are minuscule, but I am a Mechanical Engineer, not an RF engineer.

  8. MichaelC

    It’s unlikely there are millions of 5G transmitters close enough to airports to potentially interfere w altimeters.

    I wonder how difficult would it be to hack airport adjacent transmitters to intentionally interfere with altimeters.
    Seems to me the practical solution would be to remove 5G transmitters to a distance from airports suffucient to mitigate the risk of intentional or unintentional interference.

    And that cost should be borne by the 5G operators.

    I live in LGAs neighborhood. I’m not aware of any community pushback on the rollout plan, but I’ll keep my ears open

    Thanks for the summary.

  9. Duke of Prunes

    This reminds me of an expression a boss of mine used: “FAA Management”. As in, you wait until the plane crashes to figure out why. He used this when he got pushback when trying to proactively seek solutions to potential problems. “Hey, let’s not use FAA management on this one… the plane hasn’t crashed, maybe we can still save it”.

  10. Copeland

    Wasn’t there another 5G link in NC a few days ago pointing out how the rest of the flying world rolled out 5G with a bigger gap between mobile phone and airline frequencies? In addition, the rest of the world requires mobile transmitters near airports to be carefully aimed lower to prevent any problems, and a few other tweaks that I don’t recall.

    USA: nothing matters and nobody cares, its the wild west, and it always has been.

    1. marku52

      “USA: nothing matters and nobody cares, its the wild west, and it always has been.”

      Not true. Under GWB, Enron malfeasers were actually put in jail and Arthur Anderson was de-corporation-ized.

      We used to take this stuff seriously. Now it’s just who has the biggest bribe, er, contribution.

    2. Jeff Hails

      I don’t recall seeing it here but two big differences were distance buffer and power level. The distance and power levels provided a much higher safety zone. The radiation pattern is also a big deal but that can’t eliminate the possibility of surface reflections also creating a problem. The cell carrier can’t control who builds what where on land around the cell site.

  11. John

    Am I correct that the cell services did there best to ignore this “little problem?” Is fast cell service so important that the lives must be put in jeopardy or is it the two rules of Neo-liberalism: “Because markets” and “Go die?”

  12. VT Digger

    Can we just go back to 2G and call it good?
    Nobody needs 5G…oh…except the denizens of the metaverse prison…

  13. marku52

    “wireless devices is 20 kilohertz to 300 gigahertz.”

    There must be a misprint. 20kHz is way too low for commercial use. The transmitting antenna would be huge. A quarter wave antenna, for example would be 3.75Km long. Try fitting that in your phone.

      1. Zamfir

        To add: the Russians even have an even crazier 82Hz ( Hz, really) system. The “antenna” here consist of running hundreds of amps through low-conductivity rock formations.

        1. T_Reg

          Meanwhile, forget 5G, there are reports of birds and insects being harmed by current EMR. Hard to know whether they’re accurate, but there seem to be a lot of lethal coincidences. I think that should be as big a priority for analysis as the altimeter mess.

  14. Cynical Engineer

    I am not an RF engineer, but I work with RF technology a lot. A few observations about this problem:

    Radio receivers are not frequency-specific, but instead rely on bandpass filters to attenuate the signals outside of the frequency range of interest. You will also virtually never hear an RF engineer talk about “shielding”, but rather in terms of attenuation. The circuitry or shielding on the radio receiver can provide a certain amount of attenuation of undesirable signals. The stronger the out-of-band signal, the harder it is to keep it out of your receiver. The more sensitive your receiver, the bigger the problem.

    I ran into this when I was working on a box with both a 900MHz radio and a 2.4GHz radio on the same circuit board. When the 1w 900MHz transmitted, it swamped the receiver on the 2.4GHz radio even though it was a completely different frequency.

    I have a great deal of sympathy for the aircraft operators who are afraid of having a brand new high-power intentional emitter operating within a few hundred kilohertz of their frequency band. In my opinion, the FCC never should have considered licensing cellular operations within 500MHz of the aircraft altimeter frequencies. Between having low-power handsets operating IN the passenger aircraft, and the power levels of the ground towers, this has trouble written all over it.

  15. Scott1

    Who are the heroes? Who are the villains? I forget the name of the FCC Trump Administration villain. The FAA and the Airlines operating passenger airliners are the heroes? It is they who say since their mission is sure enough to make a profit they can neither guarantee safety or profitability if cockpit instruments are caused not to work. We have heard nothing but bad noise when it has come to 5G technology. Huawei equipment was said to be set up so it would spy on US communications, and that embedded in it was malware the Communists could use during cyberwar. Many operators appeared to say “We need it, so we’re buying it.” Next thing we know about 5G is that its implementation could cause chaos. Canada had freed the CEO.

  16. chuck_m

    I am an RF/Microwave engineer with 30 years of experience and I can assure you that C-band technology is very mature. If all that is needed are better filters than that can be done as 220 MHz of separation seems enough to me.

    By the way, 3.7 to 3.98 GHz is S-band not C-band.

    L-band: 1 to 2 GHz
    S-band: 2 to 4 GHz
    C-band: 4 to 8 GHz
    X-band: 8 to 12.4 GHz
    and so on….

    1. Cynical Engineer

      But do all of the existing radio altimeters have good enough filters?

      Are you volunteering to be on the plane as they test this out?

      1. chuck_m

        I never said they do. I am just saying that the problem is solvable assuming inadequate filtering is the problem . But I wouldn’t put the filters on the altimeters but rather on the transmitters at the cell site.

        I’m surprised this is an issue. It seems they should have considered this when writing the 5G standard.

        1. BillS

          The other problem is, of course, spurious adjacent channel power emitted from 5G transmitters. Base stations are likely to be less of a problem here because they (should) have high quality ceramic or cavity filters to block spurious emission into the C band radar channel. Handsets and IoT (internet of sh*t) devices are likely to be the major spurious emitters here. Consumer electronics manufacturers will not likely jump to design in expensive transmitter filters into a handheld device. Filters on altimeters will not correct this problem.

          Given that many altimeters use a frequency modulated signal, as long as the interference level is below the point where it desensitizes the altimeter receiver, probability of interference is pretty low. Pulse-based radar altimeters may be more adversely affected because they do not have the benefit of the spectrum spreading effect of the FM radars.

          I understand that aircraft operators are loathe to take on the added risk, however small, because the penalty for failure is so high (Risk = Failure Probability * Cost of Failure). This is all very similar to the Light Squared debacle of some years ago, whose L-band channels fell very close to the GNSS bands with the predictable outcry from the FAA and DoD. The mixture of regulatory capture, incompetence, agency competition and big money spectrum sales created a perfect storm.

          All this to have Netflix, YouPorn and Cat pix at super speed. We live in a Brave New World.

          1. Mikeyjoe

            Does transmitter power increase as the number of devices increase?
            The reason I ask this is for example a Boeing 777 with between 300 to 350 passengers has hundreds of devices. Their demand in this frequency is much greater than the altimeter. Precise altimeter readings are not just important on final approach, of course, but also on the flyways.

  17. cyclist

    I recall reading a link about the French vs. US approach to this problem. Apparently, European 5G operates on a frequency band more widely separated from the altimeter band. Nevertheless, the French actually used some aircraft to actually test the potential for interference and, from their findings, have limited both the power output and minimum separation at airports. The antennae were also required to aim downwards. It was also mooted that the US carriers want to use higher output powers so that they can get away with funding fewer towers.

    1. Synoia

      There are two set of frequency standards. One for N America, and one for the rest of the world.

      Why? Because the US is exceptional. Exceptional has twp meanings. I am beginning to believe only one of the meanings is correct when used with US Jurisprudence.

      Not agreement capable.

    1. drumlin woodchuckles

      We can’t have the promised Ubiquinet of SpySh*t if we don’t have the 5G to support it. If that means that air travel is abolished so that the 5G can roll, well . . . sacrifices must be made.

      Duty Now for the Future!

      Perhaps we should rename it “5 Eyes G” to clarify the issues at stake. Or maybe ” 5EyesG” if that scans better.

  18. Alex Vaivars

    More technical background regarding radar altimeters, which overwhelmingly use FMCW as the radar technique for ranging:

    The radar is continuously changing frequency so the odds of interference covering the entire frequency band of a sweep is exceedingly low, unless a transmitter is completely out of spec and transmitting spurious emissions over a broad spectrum at extremely high power.

  19. drumlin woodchuckles

    So it looks like everyone who flies on an airplane into or out of any airport ” too close” to 5G radiation emission sources is taking a gamble. They could be said to be making a Darwin bet.

    There may be a lot of Darwin Awards getting handed out around all the major airports.

Comments are closed.