AirFiber Laser Networks: 622mbps

shinar writes: "In the LA Times, AirFiber announced the July/August avaliability of a new last-mile replacement for fiber: a wireless laser system delivering up to 622 mbps. It's a short-range system, mainly viewed as a delivery system within cities. It's being tested in Dallas, Tokyo, and Madrid, and projected for first commercial release in July or August. Not huge or revolutionary-a few other companies are working on the same sort of thing-but it might be a blessing for people in high-population density areas. "

Comments

interference

[blackdefiance]

would snow, hail, or a hard rain disrupt this?

Laser downside

[Da Unicorn]

As I see it there are a lot of things that could make this unstable. Precipitation in any quantity would attenuate the signal, bugs, birds or dust likewise. Certain health risks if exposed to any high level of coherent light exist. A mirror placed in the path would have interesting effects as well as a crystal.

Neat idea however.

Da

Re:Laser downside

[Stonehand]

Readers of RISKS may remember this piece about lasers, spiders and networking.

Old Tech?

[The Dodger]

This reminds me of a system I tested a few years back when I was working with a TV station. We tried out an infra-red laser system which allowed us to beam a TV feed between two places which had line of sight.

It was quite a while ago, but I remember that the range wasn't very impressive, that it was a bastard to get set up (trying to align a laser over a kilometre distance isn't the easiest thing in the world to do) and that heavy rain, fog, waving your hand in front of it, etc. did have an adverse affect.

D.

Cut the wires!

[zpengo]

I can just see a bunch of punk kids going to the local geek compound and cutting the wires, then spraying lasers them all over the place.

"Luke, I am your fahther...."

Can it kill Darth Maul?

[Ukab the Great]

Or do you have to use something as clumsy as a Cat 5 cable.

Re:Sounds great, but...

[Morbid Curiosity]

how'll they keep pigeons out of the laser path?

Any wireless system needs a heck of a lot of error-correction at the link level. It's been possible to get fairly good bandwidth out of microwave beam transmitters already for at least a decade or so, so I guess similar principles would be in effect.

Besides, isn't it just another parroty error? :-)

Re:Weather?

[arivanov]

Yes it does. So it does with radio, satelite, etc. And this is the reason you back them up with something less speedy but more reliable. Like old good 64K ISDN for example. Unless you want to try your luck of course ;-)

Network interference

[Kaa]

I see major problems with carrier-pigeon network interfering with the laser one.

Initially, as the flights of pigeons would be blocking the laser's line-of-sight, the laser network will suffer: "I am sorry, sir, the packets can't get through -- there are pigeons in the way!".

Eventually I would expect the laser network to deal with this problem technologically, that is, upgrade their lasers to military power. At this point, the carrier-pigeon network will begin to suffer casualties, and notices like "Caution! Dead pigeons falling out of the sky!" will have to be posted around the laser network sites.

Later we may expect carrier pigeons to be fitted with mirror-shielded armor and the network lasers to be outfitted with tracking devices, but that's going to take some time...

"Packets? What packets? There is a pigeon shoot going on!"

Kaa

low latency is cool; bandwidth just increases

[Nemesys]

There's nothing particularly interesting about
how many bits we can get from A to B these
days. Sure some of the tech might be cool
(e.g., the optical switching trickery in the
Vint Cerf interview), but what'll impress me
is when they get latency down to the point
where I can't as a human detect the difference
between the latency across the Pond and the
latency to the other nodes at my LAN party

Smog warning

[Ralph Wiggam]

AirFiber technology information is currently restricted to business partners and current/potential customers.

All I wanted to know was what color laser they are using. If they use red, smoggy or foggy days would result in this very cool "Blade Runner" grid of lasers floating at rooftop level.
Serious note: This technology is only meant for large cities with high population densities. Those cities would also have the most smog. As the smog levels of a city rise, wouldn't the connection get slower and slower? Something for the folks in LA to think about before they drop money on this kind of system.

-B

Security Concerns?

[wesman]

In a perfect world all data transfers would be encrypted. In this world they aren't. Whats to stop someone from splitting off part of the beam and using it in a packet sniffing sort of way. I know that lazers are coherent light so you have to be in the same plane as the beam to pick it up but that plane could easily be a foot or two off a roof top. At least with hardwired networks you need to gain physical access.

Does anyone have any insight?

New York had similar

[NMerriam]

When I was living in NYC a few years ago, there was a laser connection similar to this that was available to anyone with line-of-sight to the Empire State Building. It was about a T1 worth of bandwidth, but it's a lot harder to get line-pf-sight to the ESB than you'd think.

AFAIK, it never really caught on, but it may still exist. I suspect the line-of-sight is the biggest issue it faced (I might have gotten it if my apt was on the correct side of my building to see the ESB). You'd have to put up a lot of these to solve the last mile in this way. Probably this would be a better solution for getting to the last mile -- ie, mount these on cell towers and then run fiber from them to the area buildings...

What amazing wiretap potential

[dsplat]

Without decent encryption, this practically begs to be wiretapped. I can certainly see some excellent uses for it though. As a data link for sites that want to limit RF interference (radio telescopes, for example) this could be a good thing.

Re:What amazing wiretap potential

[CaseyB]

Without decent encryption, this practically begs to be wiretapped.

Huh? This has FAR more security than any other form of wireless transport. Without compromising the endpoints, it can only be intercepted by putting a reciever in the line of sight, instead of merely being withing broadcast range. This would both be really easy to detect (wtf is that helicopter doing hovering beside my building?) and would be very difficult to do without disrupting the communication.

But for the paranoid, there's certainly no reason not to add encryption to the link.

Bingo! That's exactly the problem...

[ATKeiper]

You nailed it right on the head. The article mentions that this technology will be weather dependent and need to be adjusted not only for a hard rain, but even for fog. And what if a bird flies in the way of your transmitting laser? Ooops, there goes my download.

It's an interesting idea for people who can't use other technologies because of their location, but I doubt it'll have any wide consumer market, when there are so many other established (and more reliable) technologies.

See this related article: Science-fiction staple new entry in high-speed Net (C|Net News.com, 22 Mar 00)

A. Keiper
The Center for the Study of Technology and Society
Washington, D.C.

Some Ideas...

[Spazmoid]

Well.. without know what wavelenght laser they are using, it is hard to predict what envornmental issues might effect it, but here is an idea.

You have several laser guns/recievers on the same building all pointing in different directions... say within a 30-50 degree differenc of each other. Thos signals are all the same and subject to differing amounts of interference. Those signals are all picked up and rebroadcast at different angles to their destination. Adds a little bit of path redundancy.

I could also see being on the roof of a building in the middle and putting up a beam splitter and a couple mirrors and sniffing the media.....

Re:low latency is cool; but impossible

[jmv]

Unless you mean low-latency in the city, forget about it. Lights only travels 300 km/ms and goes down to 200 km/ms in fiber (glass index of refraction ~1.5). Crossing the Atlantic means already 30 ms and the other side of the globe means 100 ms. Sure routers add some latency, but these numbers are the minimun period... unless you run your fiber through a wormhole.

Not an insoluble problem

[Tau Zero]

As you go into the infrared, the influence of fog drops; once the wavelength of the light gets to be much larger than the droplet size, they stop scattering very effectively. If you use a larger lens (say, a fresnel lens instead of a small glass lens) the effect of blocking a small area diminishes. Besides, I'm sure your router is capable of handling the effects of a brief dropout in the signal, and that's all a bird is going to give you.

Free-space optical transmission has distinct advantages over radio and microwave, notably a lack of licensing restrictions and extremely high transmission rates. Would you put up with 2 Mbps when you could have 100 Mbps or more? Neither would I. If I ran an ISP in an area where fog was infrequent (say, from Los Angeles down to San Diego and east to Austin TX or so) this would be a great way to roll out digital services at speeds DSL cannot even dream of, while still being able to make money off a sparsely distributed subscriber population. This will drive the demand curve which will then make it possible to get fiber all the way to the service entrance. It's not the end-all, but it's a great foot in the door.
--

No problems

[jbert]

1) Fog slowing download time.

This is digital. Which means that if the quantity of light getting through drops from 100 Foobars to 30 Foobars you are still getting full bandwidth as long as digital 1 == "more than 20 Foobars". Of course if it drops below (or too near...) that you lose.

2) Beam interruptions

Yup, IP is 'best effort' delivery. You lose packets on the internet all the time (mostly due to congestion). As long as whatever is blocking moves 'soon' its all OK.

3) Pigeons

Break out the old technology. Someone patent the scarecrow quick.

further concerns

[ATKeiper]

Thanks. I think you're quite right when you say "it's a great foot in the door." Just for a bit of technical clarification, can't even slight amounts of scattering seriously distort the data being transmitted?

Moreover, I'm rather curious about upload vs. download speeds.

Finally, you mention L.A. as a possible city where fog might not be a problem. That's probably true, and smog probably wouldn't be a problem there these days (as it would have been a few years ago). But mightn't minor tremors (in an area which, like much of California, is prone to seismic activity) inspire a need to perpetually realign the trasmitting and receiving gear? Over long distances, even the tiniest vibration could hugely affect the angle of the beam.

Are these realistic concerns?

A. Keiper

Re:further concerns

[Tau Zero]

can't even slight amounts of scattering seriously distort the data being transmitted?

They'll attenuate it, but distortion isn't much of a factor. Visualize it this way: the transmitter emits a train of pulses, which travel at the speed of light. For the pulses to be distorted, something has to either block stronger or weaker parts preferentially (a non-linearity), or delay part of the pulse so that it starts interfering with the next pulse. For this to happen with light going through fog, a substantial part of the pulse would have to follow a longer path. This requires at least two (and possibly three) scattering events: one off the line of sight and one back toward the receiver (assuming the receiver is not highly directional and only taking signal directly from the transmitter, in which case you need a bounce back onto the line of sight and then a third bounce to send it to the receiver). As you can see just from hand-waving, the probability of a photon coming in with a delay is vastly smaller than it being lost. Thus, distortion isn't a big worry.

I'm rather curious about upload vs. download speeds.

That's going to depend on the exact technology used, and the tradeoffs chosen by the designers.

But mightn't minor tremors (in an area which, like much of California, is prone to seismic activity) inspire a need to perpetually realign the trasmitting and receiving gear?

Sure. If you need accurate pointing to an arc-minute, anything that jiggles the optics by that much or more is going to screw up your system. (This includes birds landing on it, squirrels climbing on it, the wind blowing it around...) On the other hand, the same technology which we use to make CD players which focus themselves and follow tracks wherever they waver ought to be able to make a self-aligning laser transceiver. If CD drive prices are any guide, they'll be cheap in quantity.
--

Re:More tricks and gimmicks.

[Detritus]

The problem in many areas is the telephone company. They control the right-of-ways and own the copper and fiber. It wouldn't be a problem if they leased the copper or fiber at a reasonable price, but they want to sell you a T1 or T3 for an outrageous price that has no relationship to the cost of providing the service.

You don't need to be fancy to tap (DIY experiment)

[orpheus]

I really think the 'laser wiretap' is overblown. After all microwaves, etc. are quite a bit easier to tap thna a laner, since lasers are more tightly collimated.

*HOWEVER* I'd like to point out that (on a purely
technical level) tapping a technical signal is easier than you might think, because the intruder can amplify the signal at the point of interception, and therefore needs to divert only a fraction of the signal.

Even the tricky issue of 'the optical edge' (inserting the optical element without the boundary of the optical element causing a detectable interruption) is easily handled by using "edgeless" optical 'diverters' like smoke or steam

Get out your pocket pointer laser, and a phototransistor. Point the phototransistor at a point along the laser beam, and insert some of the following in the beam at that point:

Smoke/Steam/mist - if you see the beam, so will the phototransistor; no edge to temporarily interrupt the beam; even if you accidentally trigger the gadget that automatically photographs beam blocking objects, then photo won't show anything very incriminating (re: your identity)

For silly hacker/cracker movie scripts:

a transparent helium balloon (they 'light up' when the beam hits them)

a sheet of glass (monitor the edge of the sheet)

talcum powder (or powder of choice)

__________

Fog interference vs. San Francisco

[billstewart]

OC12 line-o-sight is great. Unfortunately, the Center Of The Internet is the San Francisco Bay Area, aka Fog City, so one of the places it would be most useful you can't use it :-)

In practice, San Jose doesn't have anywhere near as much fog problem, and a 500-meter distance could connect the PacBell/ATT POP with the AboveNet/UUNet/MFS POP, though there are bundles of OC48 fiber under the street that do the job just fine.

Lasers, Broken-window fade and Plywood fade

[billstewart]

A friend of mine was at the University of Colorado back in the late 70s. They had an infrared laser connecting two of their computer buildings - it went out the window of one to the rooftop of the other, and worked quite reliably. There was a bit of rain and snow fade, but it was reasonably tolerant, and it's on a relatively dry side of the mountains. Then one day the window broke. The laser system survived this... until the workman came and put a piece of plywood over the broken window :-)

The computer system, being the fine piece of 70s mainframe gear it was, crashed hard, and the computer center folks had trouble explaining to the guy that he couldn't put that plywood there because there was an invisible laser beam going right where he was standing....