The Folsom
Kindly Shut Your Noise Hole.
Premium
join:2003-01-31
Yucaipa, CA
 ·Verizon FIOS
| reply to GeekGirl1
Re: Works great if you only drive on a test track
said by GeekGirl1  :"Tarzan swinging through the jungle on a vine attached to one branch. If either breaks, Tarzan's forward motion is abruptly interrupted. If multiple vines and branches are readily at hand, however, he easily can keep moving forward..." RF multipath puts so many branches in the way that Tarzan picks the wrong one and falls off the tree.  More like "George of the Jungle"...
»www.youtube.com/watch?v=1fB0tv31-ys
♫"Watch out for that treecar..."♫
--
Civil disobedience is still disobedience.
|
|
TheMG
join:2007-09-04
Edmonton, AB
 ·TELUS
| reply to GeekGirl1
Doesn't this system also rely on having plenty of open Wifi APs along the way? I can see there would be very few places where this would work reliably. Ok, maybe it would stand a chance in a few of the major US cities, but anywhere else? Probably not. |
|
GeekGirl1
The Medium IS the Message
Premium
join:2007-01-28
Morrisville, PA
4 edits | Did anyone with RF experience at MS review the white paper? »www.cs.umass.edu/~arunab/paper/s···vifi.pdf Very good documentation of the model assumptions, protocols, experimental observations. I think these guys did some really nice work- it seems to present both positive and negative aspects objectively.
However, this is an attempt to optimize performance to a single model. Everything was based on an open test track. The paper is very clear on that point and a lot of work needs to be done in a realistic RF environment.
A more realistic environment using buildings, trees, and other objects found on the highway (tractor trailers, buses, etc.) will make or break this project. I'm thinking that there are enough differences to significantly change the model. IOW, all the parameters have to change according to the environment.
(Why were directional antennas even considered for this study? It's not appropriate.)
PC World referenced an MS team member using an analogy of "Tarzan swinging through the jungle on a vine attached to one branch. If either breaks, Tarzan's forward motion is abruptly interrupted. If multiple vines and branches are readily at hand, however, he easily can keep moving forward..." RF multipath puts so many branches in the way that Tarzan picks the wrong one and falls off the tree.
I don't agree with MS spokesman in the PC World article that cell plans will go away if Wi-Fi is available. The cellphone standards are optimized for urban multipath environments. How about a cellphone vs. NIC + software smackdown? Location: downtown Manhattan. I'll place my bet on the cellphone. |
|
