The Rugby Pitch test

So we got our nice new 802.11n USB wifi cards and I must say they are pretty awesome. They have a nice little management application for setting frequencies (2.4Ghz/5.8Ghz), transmission power, beacon timings, pretty much everything we're going to need for the Howth-Bray link. Even better is the cards ability to go into AP mode, this will definitely come in handy when we plan a Dublin wide link. All we're missing now is a DHCP/DNS server on the primary laptop, to have a proper infrastructure.

Myself, Michael and Jeff decided to break open the boxes and fire them up, the first test being a simple laptop to laptop with the antennas provided by the manufacturer. We were only getting about 1700kb/s throughput between them initially but it wasn't until we realised that the cards were probably too close to each other (which they definitely were) that we got anywhere closer to the 150Mbit as promised by the manufacturer. We got a steady 8-9MB/s between them once we spaced them about 10 feet apart and turned down the transmission power. Happy with our results, we figured it was time to give them and the rest of the equipment a proper test.

Upon grabbing the antennas and assembling them and in the process praying for sunshine, we attempted a 120 metre Rugby Pitch link between Jeff's laptop and mine. The fantastic thing about these little USB adapters is that we don't need an external power source to run them, as there is more than enough power on the USB. The antennas are such a high gain that they only require small amount of power to go the much longer distances as the beam is more focused. We used a 700MB test file and set up a web server on Jeff's laptop and I downloaded it to mine. There was absolutely no problem, we connected fine and managed a maximum throughput of 12MB/s.

Anyway, here's a couple of pics for now, I'll try to upload more at a later date:

Antenna pics

Well myself, Michael and Jeff finally got a chance to crack open the antennas for enough time to take pictures. They are a lot bigger than I had originally thought. They seemed very light, but the box apparently weighed 15kg. Anyway, they're here now. You can see the full set of pics here.

'The Urban Cloud' project page launched!

I've just launched a new project page detailing progress of my dissertation. You can access it at http://urbancloud.colinlyons.com or on the bar to the right.

Enjoy!

Theoretically speaking, this will work.

After about an entire day of myself and Michael breaking out the text books and trying to figure out theoretically if this Howth-Bray link will work I stumbled across this calculator which confirmed exactly what we figured out. So go us anyway!

Things are looking very good in terms of us being able to establish a full 300Mbit link, as to do so requires about -71dB between the two nodes and we have -49dB. One problem though, it turns out the USB dongles Jeff has purchased from eBay will only transmit at 150Mbit, even though it can recieve at 300Mbit, sneaky Chinese manufacturers had it in the small print. But it will do for the moment, we can always buy the DIR-825 routers at a later stage.

We decided to take the highest legal transmission power (which is 2W EIRP) into account in the example, although in reality we'll probably be using a lot less for the tests. It's just good to know for now that we have a lot of wiggle room.

As for the inital test with the equipment we've decided to use a distance that is long enough that the units wont pick up the signal by themselves but short enough to ensure there is absolutely no dedgredation of the signal. We've decided to use the rugby pitch grounds here in Trinity. The distance is about 400 feet (which is over 120 metres).

Long USB cable

Bad news - Jeff just tried making a butchered USB extension cable, with a 14M ethernet cable to lengthen. The loss is too large, and the device he tried wasn't recognised. So we're going to need to think up some other way to get the data from the roof down to wherever.

Something like this maybe? (if 10m would be enough). I think it will be.

Obviously for anyone who ends up with an ethernet router running DD-WRT this is not really an issue.

Legal Issues, TCP throughput and such

Just an Update.

I just found out that in order for us to use 5.8Ghz ISM band we must register (and thank christ it's free) with Comreg. I guess this is something we'll have to do at a later stage, but it's good to know now. The equipment falls within the ISM 5.725-5.875Ghz range (which is 4 x 40Mhz channels), so we're not breaking any laws there.

We also need to figure out how much power we're going to need in order to create the Howth-Bray link to see if we're breaking the 2w EIRP maximum comreg has.

I also found out the TCP throughput that we'll get in the best case senario is around 130Mbit, so if realistically we get 100Mbit, that would be awesome!

Another thing we need to verify is that if there is any difference between 802.11n and 802.11a at 5.8Ghz on the physical layer. I'm pretty sure that there is no difference in the physical layer (as they both use OFDM for modulation) besides having a channel size difference (20Mhz vs 40Mhz).

Antennas have arrived

Okay wow... the antennae have already arrived. Pretty damn good service, Jeff ordered them less than 24 hours ago. Hopefully the USB dongles will be here soon enough.
I'll try my best to post pics when I get to see them.