I’ve spent the week here at the Airline Passenger Experience Association (APEX) conference, and it’s been a great show. This group has its roots in traditional inflight entertainment offerings, but this year, it’s connectivity that’s been getting the headlines. Wifi providers are all trying to figure out how to speed things up, add more capacity, and reduce costs. Sounds like a piece of cake.
Wifi on airplanes has been something of a frustrating experience. We’ve been promised that it’ll work (which isn’t always the case), but even when it does it’s not uncommon for it to be really slow. Nobody has been close to offering broadband speeds like you’d see on the ground. That doesn’t mean the wifi providers aren’t trying, but it’s not an easy thing to do. At least they’re making some progress.
Possibly the biggest news on that front this week came from Gogo, provider of wifi to most of the domestic US airlines. Gogo today uses an air-to-ground system and that has some real limitations. If you’ve ever been on a flight with a lot of people using the service, you know it can barely crawl at times. They made a change previously that increased speeds (called ATG-4) but it wasn’t enough. You were still looking at less than 10 mbps to the entire airplane.
So now Gogo is rolling out GTO. The idea here is that it will still use air-to-ground transmission from the airplane. But a new antenna will be installed up top that will allow it to receive data back via Ku band satellite. According to Gogo, this will increase the pipe to 60 mbps on the airplane.
Why not just use satellite for everything? After all, if you have to put on a new antenna anyway, what’s the point of using the air-to-ground stuff? Well, air-to-ground is a lot less expensive, so Gogo wants to use that as much as it can. But by doing all upload via the ground and download via satellite, Gogo can use a smaller, lighter antenna to communicate with the satellite than it would otherwise. I guess receiving data is a lot easier than transmitting it.
Now, if you have a ton of people using the system, it’s still going to be slow, but this is way more capacity than you see today so it should make a big difference. The launch customer? Virgin America, of course. It should be up and running by the back half of 2014.
I’ll be curious to hear how many airlines end up installing this. It’ll cost money, you have to pull the airplane out of service for a few days to install the antenna, and that new antenna adds weight and creates drag, albeit less than a standard antenna would. But people want fast internet, and this is one way to achieve that.
But Gogo isn’t the only one speeding things up. JetBlue has been talking about its new wifi offering via subsidiary LiveTV for some time. It just got government approval and the A320s will be getting internet starting shortly. Next up, United’s 737-900s followed by 737-800s. JetBlue’s Embraers will follow next year. This is promised to be very fast, but I’ll reserve judgment until I can try it out for myself.
Row 44, meanwhile, the company that provides inflight internet most notably to Southwest, is busy adding capacity to its network. Some of it is meant to increase speeds but it’s also going to expand the footprint. It just finalized coverage over the Atlantic, for example, and Icelandair is going to begin installing the system.
Overall, demand for wifi just continues to grow. As providers race to grow capacity, we still wait to see if there’s actually a viable business model. The economics of inflight wifi today are tough, but it seems that airlines see this now as a cost of doing business. That’s good for travelers – it means we’ll keep getting more access. In the meantime, you can expect airlines to continue to push hard on suppliers to speed things up while bringing costs down. That’s good news for everyone.
The basic problem for terrestrial based providers is that the available spectrum is pretty limited (a few Mhz). The Shannon Equation tells you how much capacity you can get based upon available bandwidth and signal to noise ratio. Suffices to say that with only a few Mhz of available bandwidth shared among multiple aircraft, the demand for bandwidth can vastly outstrip the supply.
By contrast the spectrum available for Sat based services is quite large (literally Ghz), and the traffic tends to be highly assymetric, i.e. downlink traffic tends to dwarf uplink (which is why most broadband home internet connections have far more downlink than uplink capacity). So if you want really high capacity downlink, you are kind of forced to go the Sat route. Unfortunately that also means substantial weight (and drag) penalties for the operator. HOwever if you are willing to pay the price, you can in fact have enormous bandwidth available. It will be interesting to see exactly how much capacity the carriers are willing to pay for.
You need substantial gain in the antenna to provide reasonable capacity when the other end is 22,500 miles above you. The other problem with the Sat link is latency. If you have watched Sat based interviews on Television, you have probably noticed the delays between questions and any sort of response.
Sat Based Internet users will tell you that services that require short latency such as interactive gaming and VOIP services just aren’t very responsive with a Sat link. The propagation delay on the Sat link alone is about
250 milliseconds (several times what it is on the ground) and that still doesn’t count propagation delay in the ground based portion of the network. However this isn’t a problem for streaming video. The other potential negative for sat based (K,Ku and Ka band) systems is heavy rain (as any DirecTV® customer will tell you). Fortunately most of the time, when passengers can use the service, the aircraft will be above most of the weather.
As Matt noted, latency will preclude most interactive apps and protocols. It will also impact web surfing, as there are many round trips on the network (PC->Web Server->PC) to get a web page. So, one web page round trip at home which might be 60ms or so will turn into more than half a second.
Ask anyone who uses satellite (e.g., Hughes) to describe the real-world experience.
Still, it’s better than nothing.
Agreed, latency is going to be the big issue, and there is only so much they can do about it.
That being said one of the good things about Gogo’s hybrid ground/satellite system is it cuts out two round trips to the satellite.
In a purely satellite internet a single web request has four round trips to the satellite:
1. The request goes from the plane to the satellite
2. From the satellite to the ground.
3. From the ground to the satellite.
4. From the satellite to the plane.
At the theoretical maximum that takes half a second to happen, which is an insanely long time on fixed line internet. Replacing half of those legs with communication to the ground should speed things up.
To be clear by theoretical maximum I’m saying that is the theoretical fastest speed. (Light speed to a geostationary satellite.) After it goes through all the equipment I’d expect it to take a bit longer.
“That?s good news for everyone. ”
Except for those of us who really, really, really valued being disconnected from the office while flying so we could actually catch up on work, reading, sleep, etc… now my company expects you to be connected and responsive when flying. It will only be a matter of time before telephone conversations on flights become as common as a city bus.
Setting limits is a good thing.
I know I’m not your average road warrior, but I find many people’s inability to set limits between work and personal time to be disheartening.
CF – How much talk has there been about storing movies, TV, etc onboard the plane and streaming it to passengers devices?
I find this is a really neat idea. I’m also really curious on the technicals of Row44’s/WN’s in flight live TV.. I should’ve asked that last week.. (damn.) (e.g. Is the video transcoded onboard before being sent to the devices? Is it a special stream that Row44 prepares for the airlines? Does the plane just pick up DirectTV/Dish and pass the bitstream onto the passenger’s devices and have software deal with it there.)
Nick – Oh sure, there’s plenty of talk about that and it’s happening today. That’s what GogoVision is and other providers are doing it as well. It’s a great idea for sure. The only problem is that all the “early window” content — movies that haven’t been released on DVD but are on airplanes — won’t get put into anything that streams to a passenger device. And of course, you can’t watch anything below 10,000 feet either.
As for Row 44, the TV stuff is awesome and cheap. Internet is tough because everyone is requesting something different. But with TV, they can just blast the same signal everywhere. So they use broad beams to send it to airplanes all throughout a region. Row 44 uses Ku for this as well.
Flying WiFi is a great idea – for some, but It is also a technology apparently a few years ahead of its time. In the ten y ours that it has been around, how many different Tx/Rx systems have been tried? How many work well for more than one or two connections per aircraft? And I doubt that they have many satisfied customers. In my view the several players should concentrate their investment capital on research, and a test fleet, not entire fleets as I think one has for JetBlue. It costs too much. When they need real world traffic to test the next new system, gift the service or charge a minimal fee to generate some results. Several providers seem to have very deep pockets, but even those have limits and a couple of the current players are likely to fold before they get it right. I’ve tried flying WiFi twice, on different systems. Both were horrible and I’ve sworn it off probably for several years and I look forward to the disconnected time when flying.
I don’t get the big deal about wi-fi in the air. I connect at home before leaving, check the latest dispatches at most airports which have connections, enjoy the silence in the air, and check in at the Golden Arches or my hotel at the other end. The only real advantage would be in a trans-Pacific flight which can last up to fourteen hours. That would make it worth paying for a connection.
I agree that in air wi-fi only really appeals on long flights. If I’m doing a quick 2 hour less trip Delta doesn’t charge me less for the service, and I usually can wait the 2 hours. That’s why I haven’t grasped the rush to put wi-fi on all these planes when many still don’t have any IFE options. Hopping on a Delta MD-90 early tomorrow and sure would love to watch some morning news but no….only wi-fi. The foreign carriers I’ve flown lately seem to have their priorities straight, let the Americans sort out the wi-fi thing while it’s teething and give the customers something today that does work like good seatback IFE options.
Airline Passeger Experience. Checking the fares on the website and seeing all, or nearly all at $10 or $5 roundtrip (around 2pm, Thurs. Sept. 12, UA) HNL, ANC, LAS, you name it domestically, now that’s an experience I will take over having or not having WiFi anytime. (Bookings seemingly got confirmed–confirmation number and confirmation email–for the half hour or so I sat there looking on in amazement until the system went dead.)
Perhaps a future post commenting on how a website with its fare listings could go so amiss might be forthcoming?
Sorry for the interuption!