I've read that it requires perfect line of sight, which makes it sometimes tricky. Thanks, -Drew -----Original Message----- From: Jared Mauch [mailto:jared@puck.nether.net] Sent: Thursday, March 29, 2012 12:45 PM To: Eugen Leitl Cc: NANOG list Subject: Re: airFiber On Thu, Mar 29, 2012 at 06:34:21PM +0200, Eugen Leitl wrote:

Claim: 1.4 GBit/s over up to 13 km, 24 GHZ, @3 kUSD/link price point.

http://www.ubnt.com/airfiber

Yeah, I got this note the other day. I am very interested in hearing about folks experience with this hardware once it ships. I almost posted it in the last-mile thread. Even compared to other hardware in the space the price-performance of it for the bitrate is amazing. I also recommend watching the video they posted: http://www.ubnt.com/themes/ubiquiti/air-fiber-video.html You are leaving out that it's an unlicensed band, so you can use this to have a decent backhaul to your house just by rigging it yourself on each end. - Jared -- Jared Mauch | pgp key available via finger from jared@puck.nether.net clue++; | http://puck.nether.net/~jared/ My statements are only mine.

They are taking pre-orders now for a (hopefully) June delivery. I'm at a conference now and got the rundown yesterday from Ubiquiti. This product was designed completely from the ground up by the former Motorola Canopy 100 team. It -should- deliver ~700mbit in both directions @ full duplex. Note that 24ghz is very susceptible to "rain fade" and should be used in caution in certain climates, especially at longer distances approaching 10+km. Anyhow, check the video out on ubnt.com for an introduction and technical overview - it's worth watching. Josh On Thu, Mar 29, 2012 at 1:30 PM, Phil Regnauld <regnauld@nsrc.org> wrote:

Drew Weaver (drew.weaver) writes:

...

I've read that it requires perfect line of sight, which makes it sometimes tricky.

Thanks, -Drew

       Define perfect line of sight ? How is this different from any other wireless        link and the associated Fresnel zone ?

       http://en.wikipedia.org/wiki/Fresnel_zone

       Even 100 Mbit/s wireless equipment (which ubqt also happens to make great        gear for, at 800 USD / link) will need unobstructed view of the remote        point - and it's not all or nothing, the performance will degrade.

       Cheers,        Phil

Respectfully, the claim isn't a "decline in the cost of backhaul bandwidth between 10 and 100 times", the claim is "Operators will be able to get 10 to 100 times more data throughput for the same dollar." which granted is a very good thing, but it does not imply how much more money one would have to spend with a competitor to reach that bandwidth level. It is only an assumption that you would have to buy between 10 and 100 of the competitor's products and put them in parallel (not feasible anyway) to get the same performance thereby costing between 10 and 100 times a much. Logically it's possible that the competitor's product which matches AirFiber is only penny more, which it's not, but that's all one could logically conclude from UBNT's statement - for the same price you get a lot more bandwidth _not_ how much more you'd have to spend to get that performance level from a competitor. Ubiquiti gear is shattering price barriers, but I believe the difference in cost between their product and their competition's which can offer the same bandwidth is less than 10:1 and certainly not 100:1. AirFiber is reported to be $3000 a pair (both ends of the link). 100:1 would mean the competitor's cost is $300,000. I don't believe anyone else's 24 GHz UNLICENSED gear is in that price range. Also keep in mind this is unlicensed gear (think unprotected airspace). Nothing stops everyone else in town from throwing one up and soon you're drowning in a high noise floor and it goes slow or doesn't work at all. Like what's happened to 2.4GHz and 5.8GHz in a lot of places. There's few urban or semi-urban places where you still can use those frequencies for backhaul. The reason why people pay the big bucks for licenses and gear for licensed frequencies is you're buying insurance it's going to work in the future. Greg On Mar 29, 2012, at 1:53 PM, Gordon Cook wrote:

On Mar 29, 2012, at 1:58 PM, Josh Baird wrote:

...

Anyhow, check the video out on ubnt.com for an introduction and technical overview - it's worth watching.

The claim is a huge decline in the cost of backhaul bandwidth for wisps between 10 and 100 times. I have just finished the preparation of an extensive article on a nebraska wisp whose network is backhaul radios on towers about 5 miles apart. he is on over 100 towers across a space of 150 miles by roughly 40 miles

here is the text of the video which indeed is very good

Robert Pera, CEO Ubiquity: Ubiquity had a lot of strength. We had hardware design software design, mechanical design, antenna design. We had firmware and protocol design but the one thing that we were missing was really our own radio design at our old modem design.

Engineer 1: The group of guys who are here have been working together for about 20 years. we collectively have a lot of experience in the wireless data world - probably more so than any other company. This team of people originally were all hired into Motorola, some of us go back to the late 1980s. We actually worked on a program called altair. Altair was one of the 1st attempts at doing in building wireless networking. It was the 1st wireless local area network product ever. It was actually the 1st time that I am aware of that anyone had actually built a broadband wireless networking product.

What we did on altair continued on through Motorola and eventually became a product called canopy. Canopy is a very popular product now. It is a wireless Internet distribution system used to provide high-speed Internet people in houses where there typically is no access to cable or to DSL

Gary Schulz: we had kind of run the canopy product through its maturity and did not see a lot of additional room for growth there. When the ubiquity management approached us, we were looking for the opportunity to continue to build new stuff and that's what made it very interesting to come over and work for Ubiquity Because their focus is on the new stuff. It is on working on high speed and low cost.

The freedom to design at our level was just go and do it. What are you going to do? it was like start with a clean sheet of paper. start with nothing. We could build and design this product in any way we saw fit. The idea was just to be the best we could. air fiber is the start of the new product line within Ubiquity. It is the 1st of several products that are highly efficient, high data rate, wireless broadband products.

Greg Bedian: Our design is something that is a little bit crazy. We are trying to build a 0 IF radio at 24 GHz and do this for a 100 MHz bandwidth which is something that I am not sure anyone else has been crazy enough to try.

Chuck Macenski: As fast as you can send a packet on an ethernet wire we can receive it and transmit with no limitations.

Air fiber is designed to be mounted in a reasonably high location. It is a point to point network where the 2 antennas see each other. this is a system that under certain circumstances can work up to 10 miles. It is going to be very easy to deploy and align. It is a product that is going to require only one person to carry it up the tower and install it. There is a display on the bottom that tells you what sort of power is being received as well as a very comprehensive web interface.

We designed all aspects of it. The modem, the radio, the mechanical housing. This is a completely designed from scratch, purpose built solution just to deliver backhaul. So it is not based on wi-fi or anybody else's standards. As a result it does not suffer from any of the other overhead normally associated with that.

Built for speed -- if you want to compare the data rates of existing products to our product, other products on the market today would give you the expected data rate of the flow of water through a garden hose. Our product will provide the flow rate of a firehose. This product will provide 1.4 Gb per second of data flow which is 300 times faster than you would normally be able to get from your own home Internet service provider.

Operators will be able to get 10 to 100 times more data throughput for the same dollar. That is the big impact that this product is going to have.

Rick Keniuk: we looked at 24 GHz. We actually wanted to do something up in high frequency and that happens to be the next unlicensed band beyond six gigahertz. You can put it out anywhere. You don't have to do anything. No special paperwork. No license fees. Nobody to go get permission from to operate the radio. The nice part is that it him allows anyone to operate the product and started up without any issues of having to get licenses or jump through certain hoops of where you can place the product. It is a freedom thing.

Inside the air Fiber Design -- As far as I know no one builds a modem with this level of sophistication. Most people when the building modem commit to custom silicon. But doing it this way is very expensive very time-consuming. It is rigid in its architecture. If you make mistake, you cannot reprogram it. If someone wants to change a feature, it's locked in stone and too late, once it is committed. We call this a modem but there may be times that we can actually change the identity of it by loading new software into it on the fly. This programmable. It is flexible. And it can basically do whatever function you want to do.

With most systems, the farther you get away, the longer the amount of time that you have to wait for the packet to actually get there. we actually have a patent pending that allows us to synchronously send packets in between radios. So that packets transmitted from both ends of the link and actually meet in space halfway in between. It does not have to wait before it transmits. In this case they are both synchronize through global positioning And they can send packets simultaneously

[This next paragraph is a summary] They point out at the end that in the developing world there are many people who given the high scrap value of copper are motivated to dig up copper cables between transmission centers in order to sell the copper. And furthermore that in many cases they go looking for cables and do not understand the difference between a fiber-optic cable copper cable. When they find the cable, they cut in order to extract it. And when they see it's not fiber, they just leave it alone. The nice thing about our solution is that other than the radios themselves there is nothing you have to protect in between the point-to-point links. [End summary]

When you are given an opportunity to try to create something new and do something differently than anyone else has done, as an engineer, that's always very exciting. Ubiquiti has a reputation for being very disruptive in the market place and we found hat very attractive. We like to think about products differently than anyone else. It is going to be a whole lot less costly and much higher performance than anything else that is out there right now.

============================================================= The COOK Report on Internet Protocol, (PSTN) 609 882-2572 Back Issues: http://www.cookreport.com/index.php?option=com_docman&task=cat_view&gid=37&Itemid=61 Cook's Collaborative Edge Blog http://gordoncook.net/wp/ Subscription info: http://www.cookreport.com/index.php?option=com_content&view=article&id=54&Itemid=65 =============================================================

On Mar 29, 2012, at 1:58 PM, Josh Baird wrote:

...

Anyhow, check the video out on ubnt.com for an introduction and technical overview - it's worth watching.

Probably it will be a good alternate to FSO based laswer links for backhual. Probably cheaper & more reliable solution then hanging lasers between towers for backhaul? On Fri, Mar 30, 2012 at 1:03 AM, Oliver Garraux <oliver@g.garraux.net>wrote:

...

Also keep in mind this is unlicensed gear (think unprotected airspace). Nothing stops everyone else in town from throwing one up and soon you're drowning in a high noise floor and it goes slow or doesn't work at all. Like what's happened to 2.4GHz and 5.8GHz in a lot of places. There's few urban or semi-urban places where you still can use those frequencies for backhaul. The reason why people pay the big bucks for licenses and gear for licensed frequencies is you're buying insurance it's going to work in the future.

Greg

I was at Ubiquiti's conference. I don't disagree with what you're saying. Ubiquiti's take on it seemed to be that 24 Ghz would likely never be used to the extent that 2.4 / 5.8 is. They are seeing 24 Ghz as only for backhaul - no connections to end users. I guess point-to-multipoint connections aren't permitted by the FCC for 24 Ghz. AirFiber appears to be fairly highly directional. It needs to be though, as each link uses 100 Mhz, and there's only 250 Mhz available @ 24 Ghz.

It also sounded like there was a decent possibility of supporting licensed 21 / 25 Ghz spectrum with AirFiber in the future.

Oliver

-- Anurag Bhatia anuragbhatia.com or simply - http://[2600:3c01:e000:1::5] if you are on IPv6 connected network! Twitter: @anurag_bhatia <https://twitter.com/#!/anurag_bhatia> Linkedin: http://linkedin.anuragbhatia.com

On 3/29/12 21:53 , Jonathan Lassoff wrote:

On Thu, Mar 29, 2012 at 12:33 PM, Oliver Garraux <oliver@g.garraux.net> wrote:

...

I was at Ubiquiti's conference. I don't disagree with what you're saying. Ubiquiti's take on it seemed to be that 24 Ghz would likely never be used to the extent that 2.4 / 5.8 is. They are seeing 24 Ghz as only for backhaul - no connections to end users.

I suspect this is just due to cost and practicality. ISPs, nor users will want to pay 3k USD, nor widely utilize a service that requires near-direct LOS. I could see this working well in rural or sparse areas that might not mind the transceiver.

Cost will continue to drop, fact of the matter is the beam width is rather narrow and they attenuate rather well so you can have a fair number of them deployed without co-channel interference. if you pack a tower full of them you're going to have issues.

...

I guess point-to-multipoint connections aren't permitted by the FCC for 24 Ghz.

The whole point of these unlicensed bands is that their usage is not tightly controlled. I imagine hardware for use still should comply with FCC's part 15 rules though.

...

AirFiber appears to be fairly highly directional. It needs to be though, as each link uses 100 Mhz, and there's only 250 Mhz available @ 24 Ghz.

Being so directional, I'm not sure that cross-talk will as much of an issue, except for dense hub-like sites. It sounds like there's some novel application of using GPS timing to make the radios spectrally orthogonal -- that's pretty cool. If they can somehow coordinate timing across point-to-point links, that would be great for sites that co-locate multiple link terminations.

Overall, this looks like a pretty cool product!

--j

On Mar 29, 2012, at 12:33 PM, Oliver Garraux wrote:

...

Also keep in mind this is unlicensed gear (think unprotected airspace). Nothing stops everyone else in town from throwing one up and soon you're drowning in a high noise floor and it goes slow or doesn't work at all. Like what's happened to 2.4GHz and 5.8GHz in a lot of places. There's few urban or semi-urban places where you still can use those frequencies for backhaul. The reason why people pay the big bucks for licenses and gear for licensed frequencies is you're buying insurance it's going to work in the future.

Greg

I was at Ubiquiti's conference. I don't disagree with what you're saying. Ubiquiti's take on it seemed to be that 24 Ghz would likely never be used to the extent that 2.4 / 5.8 is. They are seeing 24 Ghz as only for backhaul - no connections to end users. I guess point-to-multipoint connections aren't permitted by the FCC for 24 Ghz. AirFiber appears to be fairly highly directional. It needs to be though, as each link uses 100 Mhz, and there's only 250 Mhz available @ 24 Ghz.

It also sounded like there was a decent possibility of supporting licensed 21 / 25 Ghz spectrum with AirFiber in the future.

Oliver

I don't think it's an FCC issue so much as 24Ghz has so much fade tendency with atmospheric moisture that an omnidirectional antenna is about as effective as a resistor coupled to ground (i.e. dummy load). The only way you can get a signal to go any real distance at that frequency is to use a highly directional high-gain antenna at both ends. Owen

On Mar 30, 2012, at 6:01 PM, Dylan Bouterse wrote:

A couple of thoughts. First, it's not fair to compare 24GHz to 2.4 or even 5Gig range due to the wave length. You will get 2.4GHz bleed through walls, windows, etc. VERY close to a 5GHz transmitter you may get some bleed through walls but not reliably. 24GHz will not propagate through objects as it's millimeter wavelength. That coupled with the fact it is a directional PTP product, you will be able to get a good amount of density of 24GHz PTP links using the same frequency in a small area (downtown for instance).

The comparison isn't on wavelength, it's on the unlicensed-ness of it. Think CB vs Ham Radio. Where 2.4GHz and 5.8GHz are congested people have no where to go but up. You may not be alone up there. Guys already running 24GHz links might look at the sudden availability of cheap 24GHz gear in a different light. Granted there's many things in AirFiber's favor regarding congestion being less of a problem. The short range and high directivity, high cost, etc, but remember this isn't the only 24GHz product out there. In the kind of places where one of these links might be needed, others might have the same need. If you're thinking about the implications of possible congestion/interference when you're thinking about a link between the main office and the warehouse at a plant to give the guys in the warehouse internet that's not mission critical that's one thing. If it's key infrastructure for your ISP business then things start to look different. The licensed links start looking better regarding reliability down the road because you have a protected frequency. For ISPs out in farm country this is less of an issue, but in the more urban areas it is a concern. You start getting interference to your backhaul and you've got serious issues. You possibly have downgraded service or no service at many towers involving lots of customers.

Another point, the GPS on the airFiber will also allow for frequency reuse to a point. I would like to see smaller channel sizes though. I hear it will be a software upgrade down the road. I'm shocked the old Canopy guys didn't code that into the first release to be honest.

The GPS/reuse thing is for transmitters that are synced, that is transmitters belonging to the same system. Someone else's system won't be synced with yours and you won't see that benefit. So if you're thinking that's going to help between competitors it won't. Greg

Dylan

-----Original Message----- From: Owen DeLong [mailto:owen@delong.com] Sent: Thursday, March 29, 2012 7:18 PM To: Oliver Garraux Cc: NANOG list Subject: Re: airFiber (text of the 8 minute video)

On Mar 29, 2012, at 12:33 PM, Oliver Garraux wrote:

...

...

Also keep in mind this is unlicensed gear (think unprotected airspace). Nothing stops everyone else in town from throwing one up and soon you're drowning in a high noise floor and it goes slow or doesn't work at all. Like what's happened to 2.4GHz and 5.8GHz in a lot of places. There's few urban or semi-urban places where you still can use those frequencies for backhaul. The reason why people pay the big bucks for licenses and gear for licensed frequencies is you're buying insurance it's going to work in the future.

Greg

I was at Ubiquiti's conference. I don't disagree with what you're saying. Ubiquiti's take on it seemed to be that 24 Ghz would likely never be used to the extent that 2.4 / 5.8 is. They are seeing 24 Ghz as only for backhaul - no connections to end users. I guess point-to-multipoint connections aren't permitted by the FCC for 24 Ghz. AirFiber appears to be fairly highly directional. It needs to be though, as each link uses 100 Mhz, and there's only 250 Mhz available @ 24 Ghz.

It also sounded like there was a decent possibility of supporting licensed 21 / 25 Ghz spectrum with AirFiber in the future.

Oliver

I don't think it's an FCC issue so much as 24Ghz has so much fade tendency with atmospheric moisture that an omnidirectional antenna is about as effective as a resistor coupled to ground (i.e. dummy load).

The only way you can get a signal to go any real distance at that frequency is to use a highly directional high-gain antenna at both ends.

Owen