MikroTik

Hi,

I'm modifying my WISP network topology, so if someone has similar scenarios I'd be glad to hear your experiences. I searched the forum and collected some experiences, but not so clear ...

I'm setting up a Radius server to authenticate all my users ... and give them their IP addresses through DHCP. I want to put some redundancy on my network that's why I'm choosing Radius, DHCP and I'd like to setup my APs with the same SSID. I currently have two towers (different points) and one of those towers have three setorial antennas.

My concerns are (some of them arose after reading some opinions on this forum):

- Setting the same SSID, on the setorial antennas will make users associate back and forth between the cards ? Actually, my question would be ... how much strength difference between channels a wireless card decide to associate to another (average) ?

- I'm currently using 802.11B, but I was planning to use only-G to give them more bandwidth, but I was reading some people on this forum reporting that in outdoor environments B performs better ... By the RFCs, G would be exact upgrade, but in reality ... I don't know ...

So, thanks in advance for all who take some time to answer those questions ....

marcelocbf -

- Setting the same SSID, on the setorial antennas will make users associate back and forth between the cards ? Actually, my question would be ... how much strength difference between channels a wireless card decide to associate to another (average) ?

Typically the client will associate with the first one it sees, and then it goes to the stongest signal...this can cause some oscillation between APs for the client. I do not know what kind of client you are using...some have the ability (like MT) to set a scan list of frequencies, that should help in the oscillations as the client will use the list to scan for an AP in sequence and attach itself to the first one it finds (hopefully your primary for that client). If you see a particular client that is having this issue with this you can do a few a things to stop it...restrict it's access to the 'secondary' AP, etc...

I'm currently using 802.11B, but I was planning to use only-G to give them more bandwidth, but I was reading some people on this forum reporting that in outdoor environments B performs better ... By the RFCs, G would be exact upgrade, but in reality ... I don't know ...

G works fine in an outdoor environment - a quiet one - and not too long of a distance. When things get noisy or there are a lot of consumer grade home wireless APs then 'B' works better every time. You could always setup two APs - one 'B' for long distance and a 'G' for close in.... Actual experience in the area would be required to figure out if it works well for you or not....

marcelocbf - everyone likes definitive answers - unfortunately with wireless theory only takes you so far...as an RF engineer for more than 30 years I can tell you that wireless communications (not just computer wireless) is about 50% theory and 50% 'black magic'. I have seen things that should not work - work great - things that should perform well - perform poorly.... In the end, after some study you figure out why in each case - but it takes having the equipment there and operating to figure it out....

R/

I do not know what kind of client you are using...some have the ability (like MT) to set a scan list of frequencies, that should help in the oscillations as the client will use the list to scan for an AP in sequence and attach itself to the first one it finds (hopefully your primary for that client). If you see a particular client that is having this issue with this you can do a few a things to stop it...restrict it's access to the 'secondary' AP, etc...
R/

That is one idea I hadn't thought about this ... Restricting the access on the second AP ... thanks ...

G works fine in an outdoor environment - a quiet one - and not too long of a distance. When things get noisy or there are a lot of consumer grade home wireless APs then 'B' works better every time. You could always setup two APs - one 'B' for long distance and a 'G' for close in.... Actual experience in the area would be required to figure out if it works well for you or not....
R/

Thanks for the explanation ... Have you ever worked with "A". I currently use "A" for my PtP, but I'd like to know if somebody uses, of think is worth to deploy it as customer multipoint access ...

marcelocbf - everyone likes definitive answers - unfortunately with wireless theory only takes you so far...as an RF engineer for more than 30 years I can tell you that wireless communications (not just computer wireless) is about 50% theory and 50% 'black magic'. I have seen things that should not work - work great - things that should perform well - perform poorly.... In the end, after some study you figure out why in each case - but it takes having the equipment there and operating to figure it out....
R/

Yeah ... I found out this already ....
But like, your explanation (with your experience) about "B" and "G" will help me a lot ... My outdoor environment is not too noisy yet ... I think that helps me ... and I'll give G a try ... Thanks again, my friend ...

marcelocbf -

I use a lot of 5ghz (802.11a) in my metro areas.... 2.4ghz is just too full of noise from consumer gear..... I split the band and use the higher 5ghz for backhauling data to/from the site and then use the lower 5ghz for the clients - works fine for me.

R/

Tom,

Thank you very much for your "contribution" ... they were very helpful ...