Interesting [ local copy ] but I agree with alioth.
I have known a number of people who work deep in the “3G” business but have never managed to get any of them to talk to me, let alone openly, about what they do. It all seems to be rather confidential, and one can see why. Some are (or have been) on EuroGA but don’t post.
If I was in that business, my biggest concern would be an airliner taking off with say 300 people, of whom say 250 will have their phones on, and I will suddenly get 250 phones up there, climbing to some altitude and trying to negotiate connections to 1000s of towers. Multiply this by say 10-30 for any given country’s network, plus all those arriving, in holding patterns. And this process isn’t just the phone connecting to the one tower, or multiple towers. It involves communication between the towers – normally this is done with underground (often fibre) links but in the countryside it is done with tower-to-tower microwave links. These links can be easily saturated and this can cap the bandwidth. I recall reading about this a few years ago, by one specialist in Usenet. It actually gets really complicated because the system maintains very dynamic databases of where each phone is so that incoming calls and sms can be routed to it.
So, if I was doing this, I would put in some really aggressive software which simply kills all phones trying to connect to too many towers, for some time – say a few mins. Nobody will actually complain because everybody is used to phones sometimes not working.
And I see some manifestation of this when flying. When entering some country, I get connectivity briefly (and a strong signal), but after a few minutes the connectivity vanishes totally (but the signal is still as strong as before). It’s kind of pretty obvious what they did… and it is exactly what I would have done too. This happens regularly e.g. entering France from the UK and it remains the case (with sporadic connectivity every say 30 mins) until I am approaching the Alps and then (as expected) connectivity returns, and again this is predictable because anybody skiing on top of some mountain will create the “airliner problem” but this time they do expect their phone to work… plus there is the public policy requirement for emergency usage. Everybody knows about this issue and everybody can think of the way to solve it. And nobody on an airliner expects their phone to work, anyway… And different countries work this differently. For example I get instant and total zero connectivity over Belgium (Vodafone) and it never returns, even briefly… note that I almost never land there and overfly at FL090 minimum.
So I suspect the answer is simpler than some technology issue. I think the system is hacked to protect itself 
4G works a lot better but it seems to be known that they had to replace most of the ground system for that too. But 4G is still killed in the same places.
I have to imagine that doppler effect is taken into account (at least to some extent) or is not significant enough for 4G networks since I could get a perfectly good and usable 4G connection on the AVE between Zaragoza and Barcelona at >300km/hr (pretty much Bonanza cruising speed). Given that most European countries have ~300km/hr high speed rail these days I would have to imagine it’s a design consideration.
I wonder whether there are external antennas for mobile phones such as the Samsung lines. Have not found anything useable so far but I could imagine that with an antenna on the underside of the plane, you might get much better connections.
DavidC wrote:
Have you noticed that those horrible pulsing noises (rat-a-tat-tat) are absent these days
Unfortunately, no!
There seems to still be plenty of places where you only get 2G (I wish I had an option to disable 2G on my phone, there’s only an option to disable 4G)
Also 4G causes audio interference (I thought those days would be gone), I’ve noticed audio noise being caused by my phone with a different characteristic to the GSM noises if I leave the phone near my mixer or microphone cables or speakers (with the phone showing a solid 4G connection). I still have to put the phone in airplane mode (with WiFi on) if it’s near that stuff.
I don’t find voice works well at all, but obviously you have the other problem in that unless you can feed it via the headset, it’s a bit useless in the cockpit…
(Connected to Headset via Bluetooth!)
Have you noticed that those horrible pulsing noises (rat-a-tat-tat) are absent these days. The prevalence of 3G and 4G means that 2G is rarely used. Since the later generations broadcast RF continuously (CDMA and OFDMA), rather than using timeslots (TDMA), there is no pulsing of RF transmissions and hence much less interference.
I am no expert on GSM but I think the reason is different.
About 15 years ago mobile phones would radiate continuously while you were making a voice call. This was when concerns were raised over the cancer risk, because 400mW or so next to your head would in theory heat up the brain a little bit. But this mode also resulted in a poor battery life.
Then came DTX (discontinuous transmission) whereby the data got packaged and sent out with a low duty cycle.
Then the networks started stretching how often a phone “logs onto” the system. Nowadays a phone sits there silently listening, and goes active for a second or two every 10 mins to log on, report it’s position to the system, etc. The power output is variable in AFAIK 255 steps; the weaker the incoming signal, the more power it sets for the outgoing. These measures makes it possible to make a smartphone which more or less works for, ahem, 1 day 
But the old ones e.g. the Nokia 6310i did 2 weeks between charges.
In the aircraft, one gets the noises in the headset when the phone is logging on to the system – every 10 mins or so. That’s if it is getting a GSM signal. If it isn’t (which is the case most of the time at altitude) it won’t be transmitting anything at all.
One reason for the improved coverage is the deployment of lower frequency spectrum, made available from the switch off of analogue TV. Bands at 700 and 800MHz provide longer range than the higher frequencies, so are good for rural and less densely populated areas. When combined with the superior technical performance of 3G or 4G, signal range and penetration (as well as speed) are much improved.
I don’t know the answer but are they using the 700MHz band for 4G (LTE) data? I would have thought that the data rate on 4G (10-20mbits/sec possible in some places e.g. Germany) would need a higher frequency carrier.
Have you noticed that those horrible pulsing noises (rat-a-tat-tat) are absent these days. The prevalence of 3G and 4G means that 2G is rarely used. Since the later generations broadcast RF continuously (CDMA and OFDMA), rather than using timeslots (TDMA), there is no pulsing of RF transmissions and hence much less interference.
One reason for the improved coverage is the deployment of lower frequency spectrum, made available from the switch off of analogue TV. Bands at 700 and 800MHz provide longer range than the higher frequencies, so are good for rural and less densely populated areas. When combined with the superior technical performance of 3G or 4G, signal range and penetration (as well as speed) are much improved.
Dealing with a single device flying slowly across the countryside (say at 100 knots, not that much faster than automobile traffic) would likely be easier than handling a large block of co-located devices hurtling at several hundred knots. You can imagine that a high speed train crossing between cell sectors places a lot of strain on the system to handover all sessions quickly. Likewise, a planeload of 200 active phones.
Disruption to the cell network on the ground from fast moving airborne planes did used to be a problem and one reason why cellphones weren’t used onboard. I’d assume that technical improvements have been made to circumvent that.
Peter wrote:
I don’t find voice works well at all, but obviously you have the other problem in that unless you can feed it via the headset, it’s a bit useless in the cockpit
I have been able to call an airfield in Germany from around FL080 over northern France when we realized that we would arrive just after closing time. They agreed to stay open a bit longer. That was with a bluetooth headset, and the whole call took around 1 minute and was completely clear and stable.
allowing aerial connections to cell towers will cause the entire planet to implode upon itself or something. Despite this, good 3G connections can be had at altitude throughout most of the US, New Zealand and now apparently Australia. Despite this flagrant disregard for the future of human civilisation, none of these places seem to have imploded just yet.
Well, yes, it’s pretty obvious that if that was true, it would take only a few people on an airliner who forgot to turn their phones off to sabotage the entire cellular network(s) down below – which obviously is not happening.
So, obviously, there is software in the cellular system which prevents this happening, and it must have always been there. So all those dire warnings must have always been rubbish, because airliners have been around a lot longer than cellular networks.
Whether you actually get a usable connection (for sms, voice, data) from high-up is a different matter.
It seems pretty evident that the more brief a connection is needed for the job, the more likely you are to succeed.
Hence, SMS, with its short packets, has always worked fairly well, at all GA altitudes. It used to work better on the old phones which could retry say every 10 seconds; modern phones try a few times and then abandon it (unless you hack the firmware). SMS needs only a basic GSM connection, no internet.
I don’t find voice works well at all, but obviously you have the other problem in that unless you can feed it via the headset, it’s a bit useless in the cockpit…
As for data, 4G is finally making things work sort-of usefully because establishing the connection (DHCP etc) is about 5x faster than on the older 3G / 3G+HSPA. It seems to happen in under 1 second so you have a pretty good chance of getting something done. Whatsapp, Telegram, even small emails can work. With 3G the setup would take 5 seconds and at 150kt you often lost the connection soon after that.
But whether the towers point high up must be country dependent. In the UK they don’t seem to, but in mountainous areas (S. France for example) they always have done and I was running dial-up connections there, for minutes at a time, in 2004 at FL080.
in Australia the national Telecom Telstra have enabled their 3G/4G cell towers to point upwards so that you can get a good signal at say 35,000ft
They must have also done something to facilitate rapid handovers between towers. I don’t know if any 3G/4G specialist will have the details (I know several of them read EuroGA and have previously posted) but I recall the 4G system deals with this issue very differently to 3G (and GPRS, etc).
