Slashdot Mirror
Possible Big Boost in WiFi Range
goger writes "An article in the New York Times this morning (insert obligatory note about free registration here, and don't forget the yada's) talks about a startup company in CA that says it will announce WiFi antenna technology today that can give a 2000 ft. range indoors (and up to 4 mi. outdoors). This would be awesome if they really deliver, of course (and if it doesn't require me to set up something the size of a rooftop TV antenna next to my laptop in the coffee shop...)."
They have indeed decided to forfeit width of coverage for distance, by using a self adjusting array of antennas.
, 00 .html
http://www.wired.com/news/wireless/0,1382,56166
The tech is a phased array antenna, there was a good article about using it with 802.11 (notice there is no b) in the IEEE spectrum a while ago.
Consider it a sort of software antenna, you have a series of antenna that you can bias towards a particular direction. You then listen for incoming signals and use a processor to calculate environmental multipath (RF signals bouncing off buildings, etc.) and then fire off your signal so that the main signal and multipath reflections arrive at the reciever at the same time. Instant gain.
I'm skeptical on the reported max range but they should get a good amount. If you're sitting in the middle of a parabolic dish and so is your target, sure I expect that kind of increase in range, but in the real world...
Google Link
Your statements are true for a passive antenna design. There's only so much effeciency you can squeeze out of a chunk of metal.
However, this antena is not a passive antenna. It's a fully active antenna, specifically a phased array antenna. Because it's an active antenna, and it's probably got quite a few DSP's in it to sort out all the antenna signals, I'll bet this antenna is going to cost signifigantly more than a plain access point. I really do think that what they're claiming is possible. It's not like it boosts the S/N ratio in the passive antenna sense. It's able to tease out and build a better signal by watching a great many tiny antennas. It's an awful lot of DSP horsepower to get that done, though.
I'll venture it'll cost about $800, which includes an integral access point. And they'll say "to get the same coverage, you'd need 8 regular access points, or 8 * $200."
This, if it's for real, will make mesh topologies a real threat to the existing landline-based ISPs. Effective inter-node range has always been the biggest barrier to the potential acceptance of mesh systems. With this kind of range, only a few people in a community would need to run mesh routers in order to provide access to the whole community.
Nothing new here. All it is is a phased array. By definition a phased array is directive, meaning it only covers a portion of the spherical space around the antenna. The reason this is good, is because directivity means gain. So yes, this new antenna will increase range, but only in a given direction. There are antenna schemes for 802.11b that get 10's of miles in range by using very directive antennas. There have been some articles on slashdot about this. The only interesting thing here is that the directivity is achieved by a phased array so it should be stearable. They may scan the 3 dimensional space (but this will cut availabitlity and thus bandwidth at any point). Or they may be able to stear multiple beams to point at areas with a concentration of devices. They may even be able to do it on the fly as demand changes, but there is no big breakthrough here. This technology is older than I am and probably older than most of you too.
I'm suprised the posting didn't link to the company website. The tech info doesn't really offer a whole lot in the way of useful information though. Anyone know a place with better details? (/. seems slow for me this morning too)
Yeah, but those are point to point links right? Aeronet wireless bridges claim out of the box point to point bridging in excess of 25 miles. No big deal.
This is different though, this is more like a roaming point to point connection. It monitors all of the users within the area, then establishes and monitors multiple point to point connections using a phased array antennae.
Basically they get both an omnidirectional AND directional signal (not really, but that's the effect).
This is nothing new - there is a lot of antennas which extend the coverage already on the market. A friend of friend of mine got 27km link with parabolic ones.
Also there is a lot of omnidirectional antennas avalaible. But as always:
a) longer range is alwyas at the expense of angular
coverage (unless there is an amplifier) - it is
a law of physics - conservation of energy
b) stronger antennas may violate FCC regulations
Kubus
The problem here, is that the FCC intentionally limits the usefulness of Part 15 devices. Sure, you can build yourself a gigantic super-duper gain antenna and a kW amplifier for your signal, but it won't be legal. Did you ever wonder why you don't see access points or PCMCIA cards with convenient BNC or N-type connectors on them? That's because they can't get type approval to sell them if they are 'easily modifyable' by the consumer.
The FCC limits not only power output, but also the gain of the antennas matched with the power output. (ERP) If you have a 1W transmitter and a 12dB antenna attatched to it, your effective radiated power would be 16W, which would not fly under part 15. (I believe the maximum ERP for a part 15 device is something like 1W, and the maximum power output at the device is 100mW) That's not saying a manufacturer couldn't make a superb antenna that was fixed to a tremendous coax run, so that the coax loss was recovered by the antenna.
Basically, the point is, Uncle Sam isn't going to let anything too amazing happen with the tech until he gets some auctioned spectrum money from us. The only thing we can do is work on better reception (pre-amps / low loss coax / DSP) methods, or get a license to operate under a different part.
I apparently forgot that sig != uptime...
Navini makes a phased-array wireless solution that is supposedly being tested by NTelos, a regional ISP here in Charlottesville VA. I saw some preliminary documents for the trial: one that showed the coverage area, a roughly circular area with two "pie" slices taken out of it -- essentially shadows cast by mountains near the antenna location. The other document was a map giving signal strength throughought the coverage area produced by a program called "antenna explorer" or something like that. The coverage area was still considerably larger than the DSL coverage area from the same provider. I have not met anyone who is actially on the service. The trial documents indicated that they would add more coverage areas if the trial went well.
burnt sig