Researchers Test WiFi Access From Moving Vehicles

Julie188 writes "Researchers from Microsoft and the University of Massachusetts have been working on a technology that would let mobile phones and other 3G devices automatically switch to public WiFi even while the device is traveling in a vehicle. The technology is dubbed Wiffler and earlier this year its creators took it for a test drive with some interesting results. Although the researchers determined that a reliable public WiFi hotspot would be available to their test vehicles only 11% of the time, the Wiffler protocol was able to offload almost 50% of the data from 3G to WiFi."

Comments

call it what it is

[Thud457]

I prefer the OSS term for this technology, "autoleech".

Re:call it what it is

[WrongSizeGlass]

I thought this was called "How Google Got In Trouble (aka 'HoGGIT')"

not gonna work

[alex_guy_CA]

I had my phone setup to auto connect to wifi, but there is a lot of wifi out there that looks open and free to my phone, only it takes you to a page where you have to log in. Peets coffe, most hotels.

When I hit one of these, it sort of grinds everything to a halt, as the phone thinks it has a wi-fi connection but does not.

Re:not gonna work

[Anonymous Coward]

The problem with that approach is that it takes a long time, so it isn't suitable for mobile applications (not even if you're just walking). First you have to scan the available frequencies for beacon frames, then you have to send a frame to associate, then you have to receive an acknowledgment, then you have to send a DHCP broadcast, then the DHCP server has to give you an IP address, then you have to send a ping (echo request) to a host on the internet, then you have to get an echo reply back and only then can you start using your application protocol. If at any step you don't get a response (how long do you wait?), you can either retry or restart the whole process (including rescanning, because you may have moved out of range of a previously promising open network).

Grandparent is right: There needs to be a standard way for wireless access points to declare automatic public internet access. Furthermore, the handshakes (association and address negotiation) should be reduced or eliminated.

Re:not gonna work

[choongiri]

This whole tech, though, is designed for applications where slight delays in sending the data don't matter. It's about offloading 3G usage onto wifi where possible. Every step you mentioned has to happen anyway, and a ping takes what, an extra 50ms? Could it be done more efficiently if you were building up a system from scratch? Sure, but this is about offloading data use onto existing networks.

Define "Public"

[Dancindan84]

There's been a fair number of stories recently of people getting in trouble for "stealing" bandwidth from unsecured wireless routers, and not just when using it for illegal purposes. I don't agree with this. I think it should be the owners responsibility to secure their network, but the possibility for legal ramifications exists.

Re:Define "Public"

[gstoddart]

I don't agree with this. I think it should be the owners responsibility to secure their network, but the possibility for legal ramifications exists.

So, if I have an electrical outlet outside of my house and I don't "secure it", should people be able to plug into my electricity with impunity? How about my garden hose? If I don't physically bar someone from parking in my driveway, that's OK? Is it OK to help yourself to my garden? How about siphoning the gas out of my car?

There's loads of things in the physical world that aren't necessarily secured, but that you don't have a reasonable expectation of being able to use.

I don't agree in any way that just because the wireless isn't 100% locked down that you should get a free pass to just use it. You know you're using a network that isn't yours -- just because you can connect to it doesn't mean you have carte blanch.

Re:Define "Public"

[rotide]

I see where you're going with that line of thinking and I agree to an extent. However, all of those analogies require you to physically go out and take/plug in/steal something that clearly isn't yours and shouldn't be.

Logging onto an unsecured WiFi connection can be done incredibly easy while I'm in my pajamas in the middle of a blizzard. It can also be done innocently and unknowingly. "Wait, there are 4 "linksys" networks, which was mine again?".

While I don't agree with torrenting or otherwise saturating someones connection, leaving it wide open and then being pissed when someone logs onto it is almost as ridiculous as yelling to your neighbor across the street and getting mad when another neighbors listens in and potentially adds their two cents. If you're not going to take the time to secure your broadcast transmissions, don't get pissed at those who listen/use it.

Re:Define "Public"

[Dancindan84]

All of the things you've described have 2 things that make them very different from using unsecured wireless:

1. All of those things you've described require someone to trespass on your properly to get access to. Wireless could be available from as far away as across the street or in a completely different building.
2. It's very easy to secure a wireless network, whereas securing those physical things would be rather awkward.
3. There's no easy way to know if the network is public, or someone's private unsecured network. Sometimes it's obvious, but often it isn't.

• Last year when I went to Myrtle Beach for a vacation, the beach had public wifi set up for anyone to use. There were also some private wifi from the nearby condos that were unsecured. If I hadn't been told about the public wifi on the beach by the condo company I rented from, I wouldn't have known which was which. The SIDs weren't named to be obviously private or public.)

Re:Define "Public"

[chill]

And if you leave your front porch light on, should I be able to stand on the public sidewalk and read by it?

Or, if you leave your blinds open and your big screen T.V. on, should I be able to stand on the public sidewalk and watch?

Your cases are different because there are per-usage charges for the items you mention: water and electricity. If you paid a flat-rate for either, regardless of usage, it would be an interesting question. Especially because neither of them are "yours", you are just paying for usage from a utility.

A different case for your driveway, garden or gasoline. They are finite resources that use by another deprives your of their use. That is one of the basic issues behind theft.

Internet bandwidth, if used such that it didn't interfere with your usage, is a different animal. As long as you aren't saturating it, you aren't suffering a loss either of money or resource. With out a loss of product or service, how can you justify calling it a crime?

Re:Define "Public"

[gstoddart]

Logging onto an unsecured WiFi connection can be done incredibly easy while I'm in my pajamas in the middle of a blizzard. It can also be done innocently and unknowingly. "Wait, there are 4 "linksys" networks, which was mine again?".

*laugh* For one, there is no "it was so easy I did it in my underpants" defense. Ease doesn't equate with right -- stealing candy from babies is trivial, for instance. ;-)

There's also a huge difference between inadvertently using the wrong wifi, and intentionally looking for unsecured wifi.

Sadly, wifi routers are so cheap and easy to get, that lots of people just fire it up, go through the setup wizard, and never think of it again.

For much of the consumer public, these things are treated like toasters. Turn 'em on and go. They just don't realize there's more to it. The availability of the tech has outstripped the knowledge of the people using it.

Right or wrong, these have become consumer devices used as black boxes -- the companies making them should make them a little more secure, and try to steer you into having some protections on it -- having them all have the same SSID and passwords is bad. Unfortunately, that would likely lead to more support issues for them as people call and say "teh wireless doesn't work" as people get lost in the instructions or lose their passwords.

Theres never been a problem using wifi in a car...

[Viol8]

... so long as its not moving. If you're a passenger in a car doing 70mph you're going to be in and out of range of a wifi hotspot in a matter of seconds so what exactly is the point of this research? To prevent people getting bored in traffic jams in towns?

Re:Yo moron

[WrongSizeGlass]

But how would a city bus line offering Wi-Fi negotiate carriage with every AP on its routes?

And would a bus using this technology in the Netherlands have to register as an ISP?

only 50% - must not have been on 128 at rush hours

[Locutus]

with the typical AP having only a 300m range in open air and traveling at 55+ MPH, they would be in and out of the AP quite quickly. But, if they were sitting in traffic then that would be another story. I've been quite disappointed with how many of the Android apps rely on 100% data connectivity instead of intermittent connectivity. Even the facebook app just dumps a notification and does not continue with the post or upload unless the user interacts with the notification. I found no setting in the maps/navigation app to cache the route but must rely on me manually scrolling through the entire route to cache it and then hit the road. Believe it or not, there are still dead xG spots out there and wifi-only is currently not an option.

Maybe this study will wake up the apps developers to intermittent connectivity and make the device much easier to use.

LoB