5G versus commercial aircraft Radio Altimeters in the US, wtf?

[ve7xen]

At first, I wondered why radio altimeters were such a big issue.  You have both a barometric altimeter, which is usually set to the local pressure, and on approach you have at least glideslope.  During the time I flew general aviation (US), two of my airplanes had radio altimeters, and I never used them except for amusement.

One problem is that when flying into airports in the SW and mountain states, the surrounding terrain can be well below the airport's elevation.  That might be bad, if one descended to 200 feet AGL before reaching the threshold.  Then I looked at the AD notices.  Apparently, in modern airliners, the radio altimeter is linked to important landing functions, such as reverse thrust and automatic braking.  The two "passengers" in the front seats cannot control those independently.

Once obvious solution seems to be ignored.  Why not have the pilots make those decisions?

I'm not a pilot, but I am pretty sure spoilers and autobrakes are typically triggered by the weight-on-wheels sensors, not the RA, and reverse is typically manual (though gated by weight-on-wheels), though there are probably some aircraft out there with different logic. These systems are armed manually or possibly by the aircraft during a managed approach (which might use RA to decide, but could easily use the INS altitude or some other signal), but activated when the aircraft touches down. If there's one lesson to take from the decades of aviation safety improvement, it's that humans are fallible, slow, and easily distracted, so reducing workload with automation is almost always a safety win. Having pilots do things manually would increase the amount of errors, both in the tasks themselves, as well as other errors caused by the increased workload during an already high workload phase of flight.

The biggest user of RA is one that can't be replaced with something else - EGPWS, which is a safety system designed to save the plane when the pilots have lost situational awareness. Obviously, given its purpose, you can't replace this with pilots, and it depends pretty heavily on RA.

It also provides altitude callouts (50-40-30-20-RETARD) which aid pilots in timing the flare, reducing human factors on landing performance, and the higher altitude ones provide procedural checkpoints. Also rather obviously essential for autoland. Of course you can get away without it, but you'll have sloppier landings and likely need increased minimum landing distances to account for that.

[jpanhalt]

From the AD lint provided in POst #17:

The FAA determined anomalies on Boeing Model 787-8, 787-9, and 787-10 airplanes due to 5G C-Band interference
which may affect multiple airplane systems using radio altimeter data, regardless of the approach type or weather.
These anomalies may not be evident until the airplane is at low altitude during approach. Impacted systems include,
but are not limited to: autopilot flight director system; autothrottle system; engines; thrust reversers; flight controls;
flight instruments; traffic alert and collision avoidance system (TCAS); ground proximity warning system (GPWS);
and configuration warnings.
During landing, this interference could prevent proper transition from AIR to GROUND mode, which may have
multiple effects. As a result, lack of thrust reverser and speedbrake deployment and increased idle thrust may occur;
and brakes may be the only means to slow the airplane. Therefore, the presence of 5G C-Band interference can result
in degraded deceleration performance, increased landing distance, and runway excursion.

[Halcyon]

Some reading released by the Australia Civil Aviation Safety Authority:

https://www.casa.gov.au/search-centre/airworthiness-bulletins?search_api_fulltext=5G&field_dt_effective%5Bmin%5D=&field_dt_effective%5Bmax%5D=&sort_by=title&sort_order=ASC

https://www.casa.gov.au/no-sign-5g-interference-australia

Essentially, they have said: While CASA and the Australian Transport Safety Bureau (ATSB) have urged pilots to report any anomalies with radio altimeters near 5G towers, they have yet to see any. In fact, the ATSB says there have been no reports of radio altimeter incidents linked to 5G since the telecommunications technology rolled out 2 years ago. One reason for this is that Australian 5G transmissions currently do not extend into the part of the spectrum worrying the U.S. aviation industry.

Radio altimeters operate in 4.2-4.4Ghz range and the 5G transmissions subject to the interference debate are in the adjacent 3.7-4.2GHz spectrum. Australian 5G transmissions currently top out at 3.7GHz, well below the radio altimeter frequencies.

[LaserSteve]

It gets more interesting.   To meet some modern reliability standards,  on some aircraft the decision  logic  needs  concurrence of. 2x WOW and 3x radar altimeter.

It seems WOW has a lower reliability  then radalt.

Steve

[ve7xen]

From the AD lint provided in POst #17:

I stand corrected, though I really wish they would go into more detail on the logic.

Doesn't really change the conclusion that you can't replace RA with pilots, though.

It gets more interesting.   To meet some modern reliability standards,  on some aircraft the decision  logic  needs  concurrence of. 2x WOW and 3x radar altimeter.

It seems WOW has a lower reliability  then radalt.

Steve

It's hard to find authoritative information on this stuff, since the FCOMs and repair manuals aren't typically publicly available. Not too surprising that RA is more reliable than WOW, because moving parts suck. I'm really surprised that safety inhibits like reverser lockout would not 'fail open', and in fact that they make them even more likely to fail inhibit by adding more tests. I guess that's what the safety analysis concluded was correct, but it's surprising to me. Touching down and not being able to use spoilers or reversers is a pretty serious problem, but I suppose you just balk.

[LaserSteve]

Sorting the wheat from the chaff is difficult on other forumns,  but here goes: Dig hard enough in this thread you'll find links to European government   studies and tests,  and then some calculations by a poster named "satcom_guru" on pprune and the tw/itter platform.

FAA ADs re 5G interference
Started: 9th Dec 2021 by WillowRun 6-3
On PPRUNE forumn:

https://www.pprune.org/rumours-news/644134-faa-ads-re-5g-interference.html

All I can tell is Faa/Fcc/CTIA are keeping tech data closely held. CEPT in Europe is not. Band in question is 4.0 to 4.4 Ghz. Cellular band edge in the US is 3.8 Ghz. No one seems to be talking about potential intermod or mixing products either. Ie Aviation Band + 5G = mixing products? Made all the more interesting by the fact that the US ERP per tower sector is twice that of the rest of the world in 5G bands. Most other nations left a larger guardband. Also a "third harmonic" product from the lower edge of 5G has been mentioned.

Landing capability most seriously lost  under the current US   directives is visibility 1/2 mile or less.

Steve

[jmelson]

Sorting the wheat from the chaff is difficult on other forumns,  but here goes: Dig hard enough in this thread you'll find links to European government   studies and tests,  and then some calculations by a poster named "satcom_guru" on pprune and the tw/itter platform.

It seems to me (non pilot) that all it takes is a guy with a Cessna (single engine light plane) equipped with a radioaltimeter to shoot a bunch of approaches to places that have 5G sites running, and the answer will be obvious.  Either it interferes or it doesn't.  I've seen so much crazy stuff online.  There was one article stating that all 5G sites have their antennas pointing straight UP!  How is that going to work for phones?  I know that radioaltimeters are affected by all sorts of stuff, and give momentarily fluctuating readings, and the planes need to deal with that already.  If you fly directly over another large aircraft, it will momentarily show the distance between the two planes.  It seems that airlines and aircraft makers have NOT done these tests, and are predicting that aircraft will be falling out of the skies.  And, the FCC seems to think they have done enough tests to know there will not be an issue.  5G sites run at lower power than 4G and are closer to their users, and transmit intermittently, so it seems aircraft zipping quickly overhead will only see momentary fluctuations, if any.
Well, I know nothing, but it seems SOMEBODY doesn't have real data.
Jon

[ve7xen]

Sorting the wheat from the chaff is difficult on other forumns,  but here goes: Dig hard enough in this thread you'll find links to European government   studies and tests,  and then some calculations by a poster named "satcom_guru" on pprune and the tw/itter platform.

It seems to me (non pilot) that all it takes is a guy with a Cessna (single engine light plane) equipped with a radioaltimeter to shoot a bunch of approaches to places that have 5G sites running, and the answer will be obvious.  Either it interferes or it doesn't.  I've seen so much crazy stuff online.  There was one article stating that all 5G sites have their antennas pointing straight UP!  How is that going to work for phones?  I know that radioaltimeters are affected by all sorts of stuff, and give momentarily fluctuating readings, and the planes need to deal with that already.  If you fly directly over another large aircraft, it will momentarily show the distance between the two planes.  It seems that airlines and aircraft makers have NOT done these tests, and are predicting that aircraft will be falling out of the skies.  And, the FCC seems to think they have done enough tests to know there will not be an issue.  5G sites run at lower power than 4G and are closer to their users, and transmit intermittently, so it seems aircraft zipping quickly overhead will only see momentary fluctuations, if any.
Well, I know nothing, but it seems SOMEBODY doesn't have real data.
Jon

This kind of empirical data is maybe good enough for a startup elevator pitch or something, but it's not how you do a safety analysis, it doesn't and can't really cover the gamut. They do, after all, still ban cell phone transmitters on aircraft even after all these years of the public completely ignoring the requests to turn their phones off for 10 minutes during takeoff. Now that those transmissions are moving much closer to an important aviation band, you think some small-n of test approaches is going to convince them it's not a problem?

What they might do is do the analysis, figure out what the worst case scenarios are, and then fly those scenarios with test articles.

If you're flying over another aircraft while on low approach, something has gone seriously procedurally wrong, the airspace under approach paths is controlled.

FCC seems to have sided with the mobile carriers who coincidentally just gave them an $81bn payday, shocker.

I think the magnitude of the problem is likely overblown, but I also totally understand the pushback here. It's happening far too fast for the massively bureaucratic and risk-averse aviation industry to properly deal with, and is getting rammed down their throats despite it being likely to cost them many millions.

[Someone]

I think the magnitude of the problem is likely overblown, but I also totally understand the pushback here. It's happening far too fast for the massively bureaucratic and risk-averse aviation industry to properly deal with, and is getting rammed down their throats despite it being likely to cost them many millions.

Airport wants control over what happens in its approach path, its simple, they can buy the land! (yes not all airports have the luxury, but at the same time many airports have sold off/leased that same land they want control over).

[coppercone2]

I think its a good thing that they are taking their time to check stuff. People often claim the FCC is incompetent (hence the cross checking is good.. there are many stories on HAM forums and stuff showing that you might want to double check FCC anything)... All I see is people worrying about really important systems?

You need to be really sure about anything airplane related, given all the problems lately.

They probobly grounded a buncha clunker planes 

[ejeffrey]

Rule #1. Follow the money.

     I wonder how much the US cell phone carriers expect to make from the 5G market?  And what would it cost them out of pocket to replace it with something else if the 5G system was completely shutdown?

Well they (AT&T and Verizon) paid the FCC $69bn for the rights to the C band frequencies, so they obviously expect to make a fair bit.

It does feel as if the FCC were squeezing the pips a bit... South Korea use 3.5GHz, Europe use 3.7GHz and the US, 3.98GHz.

Eventually but for the first year at least they will only be using 3.7-3.8.  they will eventually have to make do with "only" a  220 MHz guard band but right now it is 400 MHz.  I get the political and historical baggage bit technically there is absolutely no reason this should be a problem.  Even a 220 MHz guard band is far more than is needed and 400 MHz is frankly obscene.

The prior operators in that spectrum are satellite video services.  The reshuffling moved all their operation into the upper half of the band, so they now operate in the 4.0-4.2 band between the 5G midband and the radar altimeters.  All those terminals are receiving extremely weak satellite signals and were designed to operate in the 3.7-4.0 band so deliberately had no filters.  Every operational system in the country had a filter installed.  They need to operate with only a 20 MHz guard band.

It's a lot more effort to get safety systems updated but there is absolutely no technical reason why a radar altimeter should have trouble filtering these signals.  I'm assuming at least some of the devices are ancient designs but if someone designed a safety critical radio receiver with that poor out of band rejection in this century I would fire them and decertify the device. 

The FAA has looked at radar altimeter interference before.  The #1 source of interference is radar altimeters.  They don't shut them off on the ground so flying over a busy airport you get reflections from dozens of aircraft taxiing around or just sitting.  It can be pretty strong too since the source is so close to the ground.  Probably not as much as a worst case 5G signal but it is also right in band and the same type of signal.  This is a real world problem that while has a low probability of causing problems it's not zero but the FAA (controlled by the airlines and Boeing...) decided it wasn't worth the cost to upgrade.  So I also don't have a lot of sympathy for their current concern.  And what do you know, now that everyone is looking at the FAA they are mandating testing and in 1 week they have tested the altimeters in 60% of aircraft and they all work.  If they had done that 2 years ago or 2 months ago they would be done by now with either clearance to fly or an actual problem.

[ejeffrey]

Sorting the wheat from the chaff is difficult on other forumns,  but here goes: Dig hard enough in this thread you'll find links to European government   studies and tests,  and then some calculations by a poster named "satcom_guru" on pprune and the tw/itter platform.

FAA ADs re 5G interference
Started: 9th Dec 2021 by WillowRun 6-3
On PPRUNE forumn:

https://www.pprune.org/rumours-news/644134-faa-ads-re-5g-interference.html

All I can tell is Faa/Fcc/CTIA are keeping tech data closely held. CEPT in Europe is not. Band in question is 4.0 to 4.4 Ghz. Cellular band edge in the US is 3.8 Ghz. No one seems to be talking about potential intermod or mixing products either. Ie Aviation Band + 5G = mixing products? Made all the more interesting by the fact that the US ERP per tower sector is twice that of the rest of the world in 5G bands. Most other nations left a larger guardband. Also a "third harmonic" product from the lower edge of 5G has been mentioned.

5G allocation in the US ends at 3.98.  RA band is 4.2-4.4 (worldwide).  The worst case 3rd order intermod product is 4.42.  Furthermore, a given RA doesn't use the whole band at once.  The standard setup is a swept frequency, and they are measuring delta-f between the transmit and receive due to time-of-flight.  The normal frequency difference is in the 10-100 kHz range.  So to get a 3rd order product at 4.42 you would need to be transmitting at the low end of the band and your IF filter should easily be able to block out the 220 MHz distortion product.  So inter-modulation shouldn't be a problem.  The only realistic problem is saturating the first stage LNA. 

[LaserSteve]

Case Study from Estonia:

https://www.icao.int/safety/FSMP/MeetingDocs/FSMP%20WG11/IP/FSMP-WG11-IP06_RA_interference.pdf

Only 1200 units of the RADALT model estimated to still be in service. However cost and timeframe of the redesign for a given aircraft install estimated at 3 years to obtain needed approvals.

Steve

[ejeffrey]

Wow, thanks.  All of that seems to be pre-5G with transmitters between 800 MHz and 2.1 GHz which kind of boggles the mind. They say there have been 2300 cases since 2017. 

Honestly I can't see how such an instrument can even be considered usable in the modern world.  According to the FAA (https://www.faa.gov/air_traffic/technology/asr-11/) the primary radar system at an airport has 25 kW peak power at 2.7-2.9 GHz.  It's also deliberately designed to project towards approaching aircraft.  A device that can be disrupted by a cell tower transmitting <100 watt at 2.1 GHz is not a device that should be operating under any circumstances.

Since the FAA mandated manufacturers to test altimeters, in the last few weeks they have approved altimeters on 90% of commercial aircraft for use in low visibility landings near C band 5G towers: https://www.reuters.com/business/aerospace-defense/us-faa-approves-90-planes-low-visibility-landings-near-5g-airports-2022-01-25/

[coppercone2]

well keep in mind how real life microwave filters end up looking like on a transfer function.. there could be peaks, modeing in waveguide, etc. I see alot of graphs that look like a cutoff then it bounces like a ball falling down a hill. I bet microstrip filters don't help too much with that.

[LaserSteve]

I attend a modest sized church with a multi-carrier cell tower on an adjacent property.

When the Digital Television switchover occurred,  our wireless microphones were now illegal. So my spectrum analyser attended church fairly often as we [actually I] decided  what band and type of modulation to switch to.


I've had really bad experiences with VHF wireless mics. 

900 Mhz was studded  with data bursts from gas and electric meters.

2.4 Ghz was out of the question with 150-200 cell phones attending service pinging for Bluetooth/ WiFi.  2.4 Ghz WISP was on the tower.

That left "T band" and up. We did not qualify for licensed spectrum and frequency allocation under the new rules.

I reassigned all our critical command and control to 5.6 ghz WIFI and the band's audio transport to hard wired ethernet.

Large amounts of whitespace WIFI was emanating from the high school across the street.


Four frequency agile decent spread spectrum wireless mics cost the church 2700$ due to spectrum auctions and unfunded mandates.  When the TV repack occurred, the cell companies already had their new gear on the tower. I was not planning on attending that Sunday, but got a call to bring the spec an as two mics went down that Sunday from jamming.

The frequency scanning software in the new receivers had a bad habit of placing the suggested  channel about 50 kHz
from the edge of the  nearest  ATSC spectral  mask it can find.. So manual channel selection is a must. We have the software to plot the mic receiver's scan on a pc. I am amazed by just how packed 450 to 950 Mhz is in Northern Ohio.

I have a WA5VJB Log Periodic on the A input of each wireless mic receiver  and the factory omni on the B input.
The LP does a good job of "looking under" the tower and most of the time the voter in the receiver prefers A as it looks at signal to noise ratios. Signal strength is about the same on A and B as I have the antennas about 60 feet from the alter. I have bandpass filters on everything now.

A chance meeting with a contract installer confirmed remotely controlled dynamic channel selection on nearly everything on the tower and auto adjustment of downtilt and other parameters. Out of band spurs are all over the place from the tower.

My point being when a modest church out in the country needs a spectral coordinator  due to a tower 350 feet away, things are a bit out of hand. I wonder how much they would have had to spend correcting all this if they did not have a well equipped Ham in the congregation. I at least gave leadership a one year heads up about the need for new
gear. I follow the FCC craziness religiously, as my Ham habit focuses on the microwave bands.

  It is far worse for touring concert sound technicians who can easily need to find sixty channels at each venue.

   I found hash and spurs all over the place from our neighbor. The tower tech mentioned corrosion can act as a non-linear mixer close in, resulting in  interesting effects in the very near field.

Couple that with gear without "Brick Wall" input filters and your bound to end up with problems.

Primary long distance search radar in the US, around the coasts runs around 1200-1300 mhz.

So I'm not surprised that there are a few locations where odd occurances happen.

  I'm an instrumentation technician for a research oriented
medium sized university. The spec an has had to visit campus a few times as well, for similar problems.

Steve

[ejeffrey]

well keep in mind how real life microwave filters end up looking like on a transfer function.. there could be peaks, modeing in waveguide, etc. I see alot of graphs that look like a cutoff then it bounces like a ball falling down a hill. I bet microstrip filters don't help too much with that.

I will say it's possible that the problems noted in that document are from non-compliant cell towers, specifically those in the 2.1 GHz band generating a lot of 2nd harmonic distortion.  If so, obviously the non-compliant hardware should just be fixed or disabled.  I don't know how likely that is, although if it happened in the US and the FAA reported interference from a tower operating outside it's assigned band you can bet that there would be a response.

But if you take that report at face value there is no need to make excuses for them.  A radar altimeter that is screwed up by a LTE tower operating 2-3 GHz away is just hot garbage.  Yes, you *can* make a bandpass filter that has an unwanted transmission peak there.  That doesn't excuse putting it into a supposedly safety critical device.  There is no way that is acceptable, not now, not even in 1960.  It shouldn't be flying.

When the Digital Television switchover occurred,  our wireless microphones were now illegal

Assuming you are talking about the 700 MHz band: that spectrum was allocated for UHF TV channels but little used in most areas.  It was never assigned for unlicensed use or licensed for wireless microphone use.  It is extremely debatable at best that the use was ever "legal" although it was certainly ignored as long as it didn't stomp on an active UHF channel at that specific location.  This is of course the whole reason why it was popular for wireless microphones: analog microphones are extremely susceptible to interference and since those bands were reserved for TV but unused they were largely clear of the interference that ISM bands see.  But the risk of squatting on spectrum that you don't own is that it can get reassigned from under you. Probably the FCC should have assigned some spectrum specifically for analog wireless microphones, but I guess they thought that 900 MHz was good enough?

[LaserSteve]

The final decision from the FCC left roughly four Mhz for wireless mics in a cellular guard band. To obtain a license
and co-ordination you must need at least 64 devices for your
event.  Otherwise it is a free for all on picking channels.

The final rule making came down less then six months from the switch over.  The lawyers and the lobbyists were probably quite busy deciding who would purchase what and when. The entertainment industry was quite busy trying to band together and come up with money to lobby.  Rumors
of just where the proposed allocation would be were flying around the media.

Wireless mic manufacturers were offering rebates and trade ins in an attempt to get as much gear as possible out of 700.

Fcc is lead by lawyers, all of which come from the cellular industry. I note that a certain commissioner, when she first
joined FCC, was talking about all these Uhf and Vhf spectrum bands that were under utilized and should be auctioned off. I remember her "I know a band or two we should sell" quote quite well. She was hinting about selling UHF ham / biz band and the ~ 1 Ghz Navigation/Radar band.

Now she's the Chair of the Commission. She's calling for Congess to re-establish the in house,  independent,  non-partisan  science advising agency it disbanded twenty years ago, to prevent this sort of thing occurring again.  That is to her credit for figuring out she presides over quite a mess.

That is quite a turn around.  It took her what, four years to learn this?

There were jokes flowing around various email list-serves about what that spectrum was actually quietly used for by the military. It didn't take long for that offer to disappear.

Yeah, those Radalts should have never made it through approval.  However when the approval process is a self submitted packet, one wonders who checks it.

The point of this post is less lawyers, less lobbyists. More Engineers and planners are needed. More technical enforcement is needed.

NTIA advises Congress on spectrum. NTIA and FCC are not under the same roof, so to speak. Throw in the international allocation process every few years and you have considerable Chaos.

What woke me up this morning is realizing that four years ago I flew into Vilnius LT on a very similar aircraft with an IFR approach in marginal weather.

I'm a big fan of Capitalism. But there are a few things that should be carefully managed by independent agencies. . Spectrum is one of them. National standards are another.
Just think what a mess privately managed time and frequency standards would be.

Steve