Slashdot Mirror
Signal Handoff Could Mean Roaming VoIP over WiFi
wassup writes "According to this article in MIT tech review (and here), researchers at University of California San Diego have developed a technology called SyncScan that will reduce handoff delay in WiFi networks to a few milliseconds. VoIP roaming will be here soon!"
Can you hear me now? Good, I'm wartalking
... until the wireless providers find out. We will finally break the speed of light when all of their lawyers run crying to the FCC and FTC. "OmG, unfair competition!" This, combined with municipal wi-fi, could lead to a much less expensive wireless future for us all. Yay!
I know I'm ignorant. But I'm curious, which is why I'm asking the question. Will this be better than cell phones?
http://nerdfortress.com/
2. duct tape a voip converter box to toolbelt.
3. add a power supply (solar panels or car battery
4. Save money on your mobile voip setup.
5. Profit!
(I did not read ALL TA. #)Fast hand offs between nodes is great, its an interesting article, but basically sync-scan is a smarter protocol for the physical and link layers of wi-fi. From a 'free and open' perspective it still leaves so many questions in my mind over the handling of sessions and authorisation/authentication for those that would provide open wi-fi points. How long will it take before technology such as this hand over method filters out and becomes a standardised free program that anyone can install on a wi-fi enabled linux box and become part of a local public voip service (as a user and provider)?
Last I checked, VOIP uses TCP sockets. When you move between WiFi base stations, you first must discover your new DHCP server, then get a new local IP address, then reconnect to the VOIP server.
This will definitely be an annoying delay.
When the V starts standign for Video rather than VoiceOverIP, then I'll start drooling.
this research wasn't randomly generated
The sync up is interesting.
If you're on one of 11 channels and you spend 10ms every 100ms checking for a beacon on each of the other 10 channels it takes you one second to check 1/10 of the channel-beacon slots. So, after 10 seconds, you've got all slots nailed down to 10ms windows. Once you have all the slots you can update the signal strengths on the active channels once per second and discover any new beacon within 10 seconds.
Yep, pretty cool....
Shouldn't let VoIP go roaming now should you? That'll teach you when it just leaves you for good...
My UID is prime... is yours?
It's just one big LAN and easy pickings for 1337 kiddies with packet sniffers.
And why would you want to join an untrusted network anyway ? So the admin of that network can keep nice juicy logs of everything you are doing ?
Strange.
RTP uses UTP not TCP, which makes sense, because VoIP payload is time sensitive - by the time TCP would get around to requesting/retransmitting, the payload would have become useless.
Why do you assume that different AP's must reside on different subnets? That is not required, and as long as the client remains on the same subnet, no new DHCP lease is required.
"National Security is the chief cause of national insecurity." - Celine's First Law
Got to say, UCSD is getting a very strong group in networking. Savage and Voelker and Snoeren, plus the folks at SDSC, plus CalIT2 (Larry Smarr's latest deal). Watch that space...
--Seen
"I used to be a dilettante. Then I thought I'd try something else for a while."
I'm sure the FCC will step in and protect us from this innovative and helpful new technology with plenty of arbitrary regulations that make little or no sense...
Reading at high threshold levels is group-think.
I will stick with my current approach of having a team of engineers follow me around 24/7 laying cat5 cable for my skype connection.
Do not try to read the dupe, thats impossible. Instead, only try to realize the truth
What truth?
There is no dupe
This idea is obvious in retrospect, as all really useful ideas are. Its basically a modification of the normal behaviour to take into account recent changes in WiFi usage. Instead of intensively hunting for a new AP when the signal has nearly died, the system checks more regularly, but much less intensively, so that it is ready to switch at a moments notice.
I hope they get paid for this.
Of course, this will only work for APs that you have legitimate access to, so if you come within reach of a restricted AP (on a different net maybe) then it can't handover to that, so "roaming" is perhaps too strong a word.
Is about an hour around here, because there's hardly a WiFi hotspot to be found.
But in a closed environment like a school, this technology might be useful for VoIP.
- It's not the Macs I hate. It's Digg users. -
Dear /. editors, from the article, I understands it is software.
I don't know about the name. Phone companies have always worked on the basis that they had something we needed - a network of transmitters maintained 24/7 and connected to the general phone system. Local calls in cities don't need to touch the phone system, or even the internet, just switch on some cheap routers and let them create a city wide network at practically no cost - it would be like one big cordless phone, sure it would probably be patchy, but people would live with it for most calls - which in the city go something like:
"hey where are you? im outside x"
"oh im like 1 minute away from x, stay there"
and text messaging would work fine. If there was congestion or you wanted to call a land line then your phone just switches to your usual network and you pay for the call. Personally i think this would be good for everyone including the networks - that push-to-talk bullshit is a lesser version of this.
This comment does not represent the views or opinions of the user.
Cell operators like Verizon spend BILLIONS on proprietary "3G" networks. Their networks require lots of towers, yet have poor coverage and lots of "signal shadows". WiMax access points have ranges from 30-50 MILES and don't have the same signal shadow problems. WiMax phone networks will steamroll cell operators with cheap networks yet better coverage and service.
cpeterso
wifi hotspots can cover remote areas like deserts, mountains, etc. given that WiFi has such a short range and it's basically impractical to build wifi networks in anywhere besides highly dense areas, it looks like there is a still a use for cell phones.
however, given that certain cities are deploy WiFi hardward throughout the city giving everyone free wireless internet, VOIP WiFi is an excellent idea to start testing there. i personally make 90% of my calls within the city. sometimes i make calls from the middle of nowhere, but that is far and few between.
HD Trailers
Airespace (www.airespace.com), recently acquired by Cisco, has had this functionality in it's wireless products for some time now. They claim roaming between their access points does not adversely interfere with VoIP call clarity. Obviously, this is designed to work within a single network that has twenty or thirty APs connected to the same network that cover an entire building, not your neighborhood which has a few APs connected to diverse ISPs which are scattered around the block. ;-)
Will we be able to connect to several networks at once?
:(
The really badass cisco servers have been able to do this for like 40 years.
But it never filtered down, it would have been awsome to have 2 Modems and phone lines or Use Both my and My neighbors Cable while he used both his and mine.
I guess they are making good use of the wifi channels but this could still be better
when is the industry going to catch up to this technology designed almost 10 years ago?
and still hasnt come to market.
the routing/communications protocol called L2R has been doing this, even with video, for almost a decade
wake me up with something actually useful happens will ya? yawn..........
I just watched a presentation last week where some grad students were looking to replace Mobility IP with their protocol IPMN. It only has a delay of 100 -> 200 ms. when switching networks, and no data packets are dropped. To test it they made VOIP calls to texas and VA.
Actually VOIP over WiFi is more likely to be useful in deserts and other remote areas because those who care can setup their own network. It might not be worthwhile for a cell phone company to put up a cell tower, but a farmer can put a WiFi station on his silo and get pretty good coverage of his ranch. Sure it won't have a large coverage areas, but it covers his needs.
"VoIP roaming will be here soon!"
Already here. It's called a cell phone.
Vote for Pedro
NetMotion Wireless has products that have been doing this for years.
By the way, that is precisely why roaming between Wi-Fi access points is important. WiFi access points are much less expensive than 3G towers, and you can get very high bandwidth if you deploy a dense set of access point. However, if you have many access points, you will roam very frequently, and you need to be able to roam very fast.
What has this story got to do with Apple? I don't see any iPod/G5 references anywhere.
Tsk. tsk... moderators. Pay more attention and remember the three golden rules of (recent) Slashdot...
1) Stories must mention Apple
2) Stories must promote Apple
3) failing that, post some Roland P crap.
VoIPoWiFi ? Kinda catchy.
yes yes yes, for all the ?hippies?, smoke signals/firephones will still be available/useable if wireless VoIP fails. they are not regulated/oppressed by the FCC, but probably are by the USFS, so check the ?legality? of your hippie ways, lest fees/?jailtime? insue.
some of US should consider ourselves very fortunate to be among those scheduled to survive after the big flash/implementation of the creators' wwwildly popular planet/population rescue initiative/mandate.
of course.. how could I not have seen it before..
From what little the article said, it looks like just another 3rd tier professor claiming to solve a major problem, but not really understanding all the issues involved.
As others have pointed out, this doesn't solve the issue of AP's being on different networks. But that's not an issue on many campus's. (Educational & Industrial)
It also doesn't address the increased power needs that such an algorithm would create, nor the performance problems that might be created as well.
On the whole, either the articles are flawed or the research is.