Slashdot Mirror
5 km Range Commercial Wi-Fi Available
prostoalex writes "Japanese company Maspro Denkoh will start selling transmission systems, enabling WiFi signal over 5 km, Network World magazine reports. From the article: "There are two types of antenna: one is a tube-shaped model about 40 centimeters long, and the other is a much shorter and square-shaped model. Combining two of the tubular antenna -- one on each end of the link -- will result in a transmission distance of about 5 kilometers while one of each antenna will work on distances up to 2 kilometers and two of the compact antenna will be fine for up to a kilometer, the company said.""
This'd be great for home, too, (if it's reasonably priced) because it would guarantee the highest bandwidth is available throughout the house, without any dead-spots, i.e. my wireless won't have to step down to 1mb/s when I'm at the other end of the house.
Perhaps the limits on transmitter antenna gain are less strict ... IN JAPAN!
Has the FCC prosecuted WiFi users for using non-stock, non-approved antennas?
They will if use becomes widespread.
Actually, since it won't scale to become widespread, they probably wouldn't have to.
Dense areas can't support this kind of thing for thousands of people in a square mile.
resigned
There's no need to proescute. Just the implied threat is good enough.
Eg.
FCC recieves complaints of wireless interference from someone.
FCC says "Well, that band is unlicensed, tough."
FCC recieves complaints of interference of licensed band from wireless source.
FCC sends a fellow out with directional equipment, locates your AP, knocks on your door and says "Hi, I'm from the FCC. It looks like your running a illegal high-powered transmitter there. Care to turn that off? It's interfering with the hospital 3 blocks away."
And that's pretty much the end of it.
You are in a twisty maze of processor lines, all alike.
There is a lot of hype here.
You are correct that latency is an issue. For 802.11 especially, since each packet is acknowledged individually before the next one is sent (imagine TCP with a fixed window size of 1), you add 1 round trip time to the time to send each packet (at 11mbps, a 1500 byte packet takes about a millisecond and a half, and the acknowledgement takes a few hundred microseconds, so this isn't really a big performance killer).
Another effect, though, is with transmission timeouts. Suppose station A sends a packet, and begins waiting for a reply. Station B sends an ACK. Normally, station A would have priority to send the next packet, since it only has to wait 10 microseconds, versus station B which must wait 50 + 10*R (where R is a random number from 0 to 31). Unfortunately, with a large propogation delay and a small value of R, station B might assume the channel is free and send a packet (a TCP ACK, perhaps) before the second packet from A arrives, resulting in a collision. But that only happens at distances over about 5 miles I believe. Caveat: I haven't tested this effect myself, so I could be wrong in my interpretation of the spec.
Even worse is if the stations are so far apart that a transmission times out before an ACK comes back, but that requires much larger distances (I think it's somewhere around 60 miles, but I could be off by quite a bit on that one).