Flexible Broadband Services

Initial Considerations


802.11b wireless radios are rated to operate at 400 metres over open ground, my friends house is about 600 to 800 meters away, so what made me think I could get a link? Well a few things fell into my favour firstly the knowlage that giving the correct circumstances the range of 802.11b could be dramatically extended. Also geograhy was on our side, our road bends around at a 90 degree angle just beofore my friends house and because there are only fields in between I have an unobstructed filed of vision to his house.

Buying the kit

We had a few requirements for the kit firstly it needed to be affordable to nick and I that ment Cisco and 3com were right out the window. Secondly the access point had to be cabable of taking an external ariel as there was no way the signal would work over such a long distance without a bit of a boost.

In the end I decided I decent access point with an external areial coupled with a good pci network card (also with an areial) should do the trick. We worked out the signal decay rates using the formulars from OReillys Building Wireless community Networks. The O'Reilly book is really useful so if you are planning a scheme like this please check it out.

Loss = 20*LOG10(Distance) + 20 * LOG10(Frequency) + 36.6
Frequency for channel 6 2.437GHz
Range (Miles) Natural Decay
0.1 84.33711058
0.2 90.3577105
0.3 93.87953568
0.4 96.37831041
0.5 98.31651067
0.6 99.90013559
0.7 101.2390714
0.8 102.3989103
0.9 103.4219608
1 104.3371106
1.1 105.1649643
1.2 105.9207355
1.3 106.6159776

We estimated the distance to between our houses to be 0.5 miles so therefore the loss to natural signal decay would be 98.32db so our equipment had to have a higher output than about 100db (to account for natural interference and changable weather conditions therefore providing a stable link). Next step was to calculate the output of our various piece of equipment following the formular

Radio - Connector - Cable - Connector - Arresstor - pigtail + Antenna

The proposed equipment for the project would be a Dlink DWL 2000AP+ access point giving an output of 16db for 802.11b, and 15db for 802.11g. The Antenna would be DLINK Cell Type Directional Antenna 8.5db. I decided to use this particular ariel because there is a possibility of another person joining the wireless community and though this is a directional ariel it offers a 70 degree cone of coverage in stead of the very narrow coverage offered by the yagi ariels. At Nicks house we used a Dlink DWL-AG520 a/b/g PCI Access card (its got a small antenna which seems to add about 5db but we used the manufacturers quoated power to calculate)

According to O'Reillys book you lose 0.25 db for each connector, 1.25 for the arrestor and 0.22 per metre of cable,

My House 15 - 0.25 - (3*0.22) - 0.25 - 1.25 -1 + 8.5 = 20.9 total output from my house Nicks House 15

The total power of the system is then Nicks total (15db) added to the total of my radio gear (20.9) giving the system power 35.9db. Next to subtract the path loss from the total

35.9 - 100 = -64.9

So in theory at either end of the link the signal strengh would be -64.9 so the next step was to check the sensitivity of the recievers on the access point according to Dlink at 11Mbps the sensitivity was -85dBm and at 54mbs the sensitivity is -71 dBm. So both well within our operating parameters. Nicks PCI card has reciever sensitivity of -85 db at 11Mbs and -72db at 54Mbs again so the signal of -64.9 should be strong enough to be reiceved at boths ends.

Laying out the Equipment on my bed The equipment layed out on my bed
Picture of the Box the Dlink stuff came in The Boxes Just to prove it is new kit :-)