Britain's CAA Considers Laptop Ban on Commercial Aircraft

hayb writes "An article in Britain's The Register claims that NASA and United Airlines have conducted tests on various aircraft and have found that ultra-wideband (UWB) devices "knocked out" collision-avoidance systems and impaired instrument landing systems. It states that the blanket ban on all devices in necessary because flight crews do not have the knowledge to differentiate between standard notebooks and ones with UWB devices."

Comments

Err, then UWB is at fault

[Anonymous Coward]

The whole point of UWB is that it is spread-spectrum technology.

So, because it is using a huge bandwidth at low power, it is supposed to be able to share it with other narrowband and high-power transmissions.

The point is, though, if it is causing interference with *any* narrowband communications, something is very wrong. Personally, I don't believe that UWB is going to be practical anyway, because it's just increasing the noise-floor, which is fine for a few devices here and there, but once every cell phone in the world has a UWB connection to it's hands-free kit, the noise-floor is going to be really high.

Infact, a good comment would be, 'Imagine a beowulf cluster of those!', because, ironically, they would all interfere with each other, (in the real world - in an ideal world they wouldn't, but do you really believe that manufacturors are going to implement the full specs, etc? No, they will just implement the least possible subsystem, which will be rubbish).

Exactly WHAT were they using?

[Raetsel]

Okay, let me get this straight...

1. FCC approves UWB devices for testing at power levels an order of magnitude less than is commonly believed to cause ANY interference,

   AND

2. UWB devices have been tested, and found to interfere with the #1 topic guaranteed to scare large populations?

What device did they test? Where'd they get these things? How can I know they didn't just hook up a 30KV spark-gap transmitter and go "See??? Interference!" (Booga booga booga!!)

Oh, great. "UWB will cause a 747 to crash into the White House, curdle your milk, kidnap your virgin daughter and sell her to the Hells Angels, molest your wife, and defraud every company you've ever invested in!"

Great, sure. The airline industry (like any industry) hates to spend money unless it's absolutely necessary. Look at the current state of US air traffic control. (Yike!) Heck, look up the state of aviation radios, even! There's a simple little thing called "heterodyne detection" that isn't present! (People have died as a result!) Yes, there are fancy computers, and GPS, and "glass cockpits" -- but there are some extremely basic technologies of aviation that haven't changed in 50 years simply because nobody has said "That's dangerous and idiotic, we've had better tech for a generation! Do it right!!!"

On second thought... this is probably a good thing. It'll return air travel to its' proper place -- an enforced, several-hour vacation! Relax, look out the window, marvel at the world you live in. No phones, no computers, but lots of distractions. God forbid, you might even talk to your neighbor. (I wonder how many people even remember how to work with a pen and a piece of paper..?)

Re:Exactly WHAT were they using?

[wowbagger]

May I ask what you think you mean by "heterodyne detection"? Because all the aviation band radios I've ever designed test equipment for were superheterodyne receivers, just like any other modern radio.

The problem with UWB is that simple fact that it occupies the same spectrum as everything else, by design - as a result it acts to raise the noise floor of all signals.

This isn't a problem when you are dealing with a signal that is tens of decibels above the noise floor, but if you are dealing with a weak signal, like GPS, LORAN, or another aircraft's ATCRBS transponder where you only have a few dB headroom, you don't have any leeway for the noise floor to rise. It doesn't matter what supercalofragilistic sheilding you put on anything to keep unwanted frequencies out, because UWB occupies the wanted frequencies too!.

And as for the tech in aviation not changing - tell it to the trial lawyers who pounce on any excuse to sue aviation manufacturers. If anybody introduces something new, and the plane it is in crashes, then the manufacturer will be sued for putting in "New, Experimental, untested technology!" - even if the reason the plane crashed was that the pilot was drunk, stoned, and inexperenced. So neither the manufactures nor the FAA will approve any new tech without giving it a multi-decade probe.

Let me run the numbers for you on interference.

Assume you have a UWB device at 10 mW output. Assume the bandwidth is 3 GHz, centered at 2 GHz. Assume the spectral shaping is rectangular. Thus, the energy is evenly spread from .5 GHz to 3.5 GHz.

First observation: the signal overlaps the GPS frequency allocation.

OK, now what is the power density? 10 mW over 3 GHz is 3.3E-12 W/Hz.

Now, consider GPS. GPS signals are about 20.46MHz wide. That means our UWB signal will be producing 3.3E-12 W/Hz * 20.46 MHz = 6.8E-5 Watts of signal, or -11.6 dBm of signal. The signal you get from the birds is less than about -90 dBm. You UWB signal is over 80 dB HIGHER than the GPS signal. Even with the coding gain you get from the fact that GPS is spread-spectrum you are still 30 dB under the noise floor. That means you could reduce the UWB signal by 30 dB (1/1000 the power) and STILL swamp the GPS signal.

UWB isn't magic - it doesn't magically pull bandwidth from nowhere, and it WILL interfer with other signals. You want to park the signal in a band nobody else is using, like up in the THz band, great! But don't put it down with everybody else, because contrary to what its proponents say, it does not play well with others.

Re:Exactly WHAT were they using?

[Raetsel]

I know exactly what I mean by "heterodyne detection." I refer to the capability of a circuit to detect when two transmitters attempt to operate on the same frequency simultaneously. The 1977 Tenerife airport crash of two 747s (KLM & Pan Am) is frequently used as an example of (1) a heterodyne happening (it was recorded on the cockpit tapes), and (2) the need for this feature so all parties are alerted to the event.

Geez, you couldn't even type "aviation radio heterodyne" into Google to see what I was talking about. Everything I've posted comes up on the first page!

• Advanced Aviation Technology Ltd. makes a device for this purpose, their sales pitch (section 3) describes the problem. (How nice of them.)

• Salon posted an article about the problem on March 28th. Same example -- Tenerife airport.

Over 500 people died in that fireball, and we still haven't standardized a solution. It's been twenty-five-and-a-half years! That's enough time to come up with a lawyer-resistant solution!

Not going to bother bashing lawyers here -- this is Slashdot, feelings on that subject are well known.

• 
  "Let me run the numbers for you on interference."

Um... no. Your numbers are way off the mark. Assumptions are dangerous, you have an internet connection, why didn't you use it? Google for "FCC UWB limits" -- the first link is a whole set of info on power levels and spectrum allocation. Digging a bit deeper, you'll find:

• "...For now, UWB communications devices will be restricted to intentional operation only between 3.1 and 10.6 GHz; through-wall imaging and surveillance systems restricted between 1.99 and 10.6 GHz (and used only for law enforcement, fire and rescue, and other designated organizations) and automotive radars restricted to frequencies above 24.075 GHz."

  Further, maximum output -- anywhere in the spectrum -- must be under limits set by part 15 (for now). That's -41.25 dBm/MHz.

Further, since you seem particularly worried about interference with GPS, I point you to this PowerPoint presentation from IEEE, or Google's rather poor HTML translation of the same thing. (It loses the graphics) The point? UWB is specifically regulated to stay out of the GPS bands.

This will not work

[david_e_v]

The thing is that, if you really want your laptop to emit these UWB, you can do it (you don't have to power off your laptop even if you have it in its case, you know).
If there is really a case against the USE of laptops within the airplanes, there is an absolute need for some kind of screening system (we should be forced to put our laptops in special cases). If not, then this is just another case of false sense of security, and all this discussion is nonsense.

Re:heh, way to go

[sql*kitten]

(Business travelers are apparently the highest margin passenger class becuase they tend to book nicer seats and fly on shorter notices so they're higher up the essentially exponentail cost function correlating time-to-flight-from-ticket-booking and ticket price.)

The reason business travellers are willing to pay for better seats is that they have the room to work on board the plane! There are other advantages too (fully flexible ticket, greater cabin allowance so you don't have to check luggage and can avoid reclaim queues, etc) but really, what matters is being able to work on board. You can get an awful lot done with no distractions between London and San Francisco, especially if you can time the flight to coincide with a working day - that's one of the reasons that companies foot the bill for business class. Work that you can do on board these days pretty much demands a computer. Ban laptops and two things will happen: business people who really need to be there will be sent coach, and everyone else will invest in videoconferencing.

The one airline that's smart enough to train its cabin crew in what is acceptable or not is going to own the market.

Not that I care though. If it's good for safety it's beyond question.

What's "good for safety" is the plane never taking off. There is always a compromise between expediency and safety.

And honestly, if you don't have your work done by the time you catch the plane to your distant meeting, the chances of you being ready are slim-to-none anyway

That's not true either. Ever done business travel? It's common to get a template done at the office, get the very latest figures as assumptions that morning, and work them into your document on the plane. If it's just a transatlantic flight, you'll probably have to deliver that document or presentation as soon as you can get to the client's office from the airport.

I'm not scared of flying..

[CoderByBirth]

...but I'm still a bit amazed at how lightly people take issues like this.

Your sitting in a metal crate with two giant combustion engines delivering an insane amount of power to get you off ground.
A plane consists of several thousand electronic, mechanical, and electromechanical systems, a zillion bolts and hundreds of tonnes of lightweight metal. And any single part of this giant system might fail at any time.

The fact that accidents don't happen more often than they actually do must be considered an engineering miracle.

So, you can't smoke and sip a gin&tonic while writing some shitty design document nobody cares about and which you might as well write when you get there?
Boo-fucking-hooo

Read a book.

TCAS and ILS and terrorists with telephones

[tlambert]

If this were really an issue, we would be seeing terrorists with small devices built into cell phone cases that were built using a switch, a battery, a capacitor, a coil, an electromechanical relay, and a large antenna loop: a spark gap generator, of the type one makes from Radio Shack project kits.

Or, they would just have cell phones, since they are also supposedly a source of interference with something other than AirFone revenues ;^)).

In reality, this article is _mostly_ bogus.

The ILS (Instrument Landing System) is vulnerable to electronic interference, mostly because it is an incredibly ancient implementation, and has not yet been replaced with anything designed in the last two decades.

The antique ILS in even the most modern aircraft is why you can't use electronic equipment during takeoff and landing (landing is obvious; so's takeoff, if you realize that it might have to be aborted, in which case it turns into a landing).

Most airports, however, are in urban areas, with a high telephone cell density. If this were ever a real issue, we would see aircraft dropping out of the sky as they flew over any urban area. SFO, PHX, and SLC tend to have a higher than average instrument requirement (the first for fog, the second two for temperature inversion based wind shear; want to vomit? Fly Tucson to Phoneix. SLC also has snow visibility issues in winter). For most airports, the systems are largely ignored. SLC has an upgraded system that ~60% of modern planes can use, actually; it's a deployment issue.

The TCAS (Traffic Collision Avoidance System) is actually based on paired receivers. It's succeptable to powerful broad-band interferences; "powerful", in this case, means "orders of magnitude higher than the those currently permitted for use in UWB devices".

The failure you would see (and you would probably need a specially manufactured transmitter to see it) would be a 180 degree polar flip (i.e. if the transponder you cared about were 23 degrees down and 17 degrees right, it would read as 157 degrees up and 163 degrees left). This actually happens a lot, and the hardware is built to automatically compensate through multiple samples (i.e. sustained interference is required).

The fix for this is to go to trios instead of pairs of receivers.

As we saw just the other week, though, TCAS itself is generally ignored in favor of ground instructions, we lost two planes in a collision in Germany specifically because TCAS was ignored.

Given that TCAS is almost never used, anyway, because the controllers keep the planes far enough apart, the interference is isn't likely to be an issue.

In any case, I think the overall concern is a result of the fear of out-of-spec devices, which met emissions at the time of manufacture, and have since, for whatever reason, ended if with a much higher signal strength.

Personally, I think they are worried over nothing: it's just an uncommenly slow news day, what with most of the U.S. shut down for Labor Day...

-- Terry

Re:Stability And Security

[KC7GR]

Dan Kaminsky writes...

Now, I'm not afraid of gameboys. See, I've *met* Boeing safety engineers. Hell, I've quoted em, learned a bit from em. Paranoid doesn't begin to describe them. These guys imagine everything, and implying that they didn't budget for even a miniscule amount of shielding and noise resistance...it's almost insulting.

-=-=-=-=-=-=-=-=-

I can go you one better than that. I've SEEN the innards and design of lots of the "Black Boxes" that make up the core of modern avionics. I've also seen how the wiring harnesses are put together, and what's being used material-wise.

If this UWB test scrambled something in the CAS or ILS, then either the test itself was seriously flawed or the UWB unit was spewing spurious signals. The avionics "black boxes" themselves are heavily shielded, and the cabling going to the antennas is a type that has two layers of shield braid. Don't even get me started on the grounding systems.

Banning laptops entirely would be far too extreme a measure, one that (as others have pointed out) would tend to piss off an airline's most critical customers. Assuming this test was actually valid, I would say train the flight crews to check for UWB devices and be done with it. Radio transmitters (and some types of receivers) are already prohibited for inflight use. Let's leave it at that.

ATC Tech

[Raetsel]

• "...the fact that the U.S. ATC relies on "dated" technology may be the reason it's so successful."

Oh God, thank you. I needed a laugh! "US ATC" and "reliable" in the same sentence, with a straight face even!

To the regular person, I suppose ATC could be looked at as 'reliable' -- but go talk to a controller sometime; the people who have to present the aura of reliability when something fails. Ask him (or her) how often their radio breaks. Or how hard it is to get vacuum tubes for some of their equipment. Perhaps you could visit the vampires -- the people who sit in an almost completely dark room dealing with everything IFR (and VFR in controlled airspace). Everything is voice and paper -- it's a sobering sight. Yes, there is a lot of computerization, but the interaction goes

• Pilot (flight plan) -> computer -> piece of paper -> controller <-> pilot!
  (Note the heads on the arrows.)

It's a wonder these people stay sane sometimes.

Canada privatized their ATC system, and (to an outsider) it has worked quite well. Communications systems are much better. The controllers don't have to keep track of planes on slips of paper, they can actually interact with the computer. One has to consider, however, that Canada doesn't deal with nearly the same daily volume of aircraft that the United States does, so their successes may not scale the way we'd need.

I must admit that the last time I was in an ATC facility was before the whole Y2K thing, and a lot of money was spent to upgrade things for that particular scare. Perhaps things are better now, but ATC doesn't live on internet time -- so I doubt it.