UNIX for Advanced & Expert Users

Ok, I currently own a server which has 16 external ip's connecting to it. The ip's are assigned to specific types of accounts... Ie: email, dns, ircd, webpage, ssh, streaming radio.

Currently I have my pf config rules set up to only allow access to the ports that these ip's are to be used for. At this point, everything is working as I wish.

Ie: If someone is using irc, they can only connect to that ip through the ports I allow under the protocol I allow. If a connection is attempted through a protocol/port I do not allow, the final firewall rule just blocks the attempt.

What I would like to do at this point is to start to control the bandwidth. From what I have read with queuing, which is what I think I have to use, is that it controls bandwidth in total usage at any given time. Ie, 10,000 bot's pound an ircd ip using more bandwidth then I allow for would be handled by queueing. But, total bandwidth for a spefic ip used within a month that goes above my allowed limit would not.

Is this correct???

With my understanding as a base here... Can someone tell me:
1). Can I redirect within a queue if bandwidth is exceeded?
- If so, can i redirect to null?
2). What happens when the bandwidth is over the queue limit if nothing else is set up?
3). Can I control bandwidth over a period of time(1 month) with regard to goving over a specified amount. Or would I have to use a 3rd party application(perl program that updates rules) which use's pf statistics as its source for total bandwidth used?

Any and all help is much appriceated.